JP5500388B2 - Shooting position specifying system, shooting position specifying program, and shooting position specifying method - Google Patents

Description

  The present invention relates to a shooting position specifying system, a shooting position specifying program, and a shooting position specifying method for specifying a shooting position of a real landscape image that is a landscape image shot by a camera.

  As a conventional example of the imaging position specifying system as described above, for example, there is a technique described in Patent Document 1 below. This Patent Document 1 describes a configuration in which a camera is mounted on a moving body. Specifically, landscape information generated from a landscape image obtained by photographing the traveling direction of the moving body is prepared in advance for a plurality of positions. In addition, a configuration is described in which a feature portion is matched between an actual image obtained by photographing with a camera mounted on a moving body and landscape information. Then, based on the matching result and the shooting position corresponding to the landscape information used for the matching, the shooting position of the actual image, that is, the position of the moving body at the time of shooting the actual image is specified.

  By the way, the actual landscape image (the actual landscape image in the example of Patent Document 1) can vary greatly in image characteristics such as brightness, contrast, and hue depending on the weather conditions at the time of shooting. Therefore, the weather conditions at the time of photographing a real landscape image are the weather conditions at the time of photographing a landscape image for generating reference data (landscape information in the example of Patent Document 1) referred to for specifying the photographing position. If they are different, even if the shooting position of the actual landscape image and the shooting position corresponding to the reference data are the same, the degree of coincidence obtained by matching between the actual landscape image and the reference data is May be lower. That is, in such a case, the accuracy of matching is lowered, and there is a possibility that erroneous recognition may occur when specifying the shooting position of the actual landscape image.

  In order to solve the above problem, reference data corresponding to a plurality of different weather conditions is prepared in advance, and the reference data to be matched is selected in consideration of the weather conditions at the time of shooting a real landscape image; It is possible to do. However, Patent Document 1 does not mention the weather conditions at the time of shooting a real landscape image. Therefore, as a matter of course, Patent Document 1 describes a problem that may occur when the reference data to be matched is selected in consideration of the weather conditions at the time of shooting a real landscape image, and a problem for solving the problem There is no indication of the means.

JP 2010-151658 A

  Therefore, it is desired to realize a technique that can specify the shooting position of the actual landscape image well regardless of the weather conditions at the time of shooting the actual landscape image.

  The characteristic configuration of the shooting position specifying system for specifying the shooting position of a real landscape image that is a landscape image shot by a camera according to the present invention is that a plurality of landscape images shot from different shooting conditions from the same shooting position. Reference data composed of a group of unit reference data, which is data generated based on each of them and is associated with the weather condition at the time when the landscape image of the generation source is captured, is referred to as the reference data The reference data storage unit that stores the unit reference data that is generated in association with the shooting position of the landscape image, and the actual landscape data generated based on the actual landscape image and the reference data are matched, A shooting position specifying unit that specifies a shooting position of the actual landscape image based on a shooting position associated with the reference data that has been successfully matched; When matching the actual scenery data and the reference data by the position specifying unit, for each of a plurality of divided areas obtained by dividing the shooting area of the real scenery image, the position of the divided area in the shooting area; Based on current weather condition information representing current weather conditions and past weather condition information representing past weather conditions more than a predetermined time before the present, from the group of unit reference data constituting the reference data, A unit reference data selection unit that selects the unit reference data corresponding to each divided region.

  According to the above feature configuration, the reference data is data generated based on each of a plurality of landscape images taken under different weather conditions from the same shooting position, and the generation source landscape image is taken It is composed of a group of unit reference data, which is data associated with the meteorological conditions at the time. Then, when matching the actual scenery data generated based on the actual scenery image and the reference data, unit reference data used for the matching is selected from the group of unit reference data constituting the reference data. Therefore, it is possible to select unit reference data based on a landscape image having image characteristics such as brightness, contrast, and hue similar to those of an actual landscape image, and it is possible to suppress the possibility of erroneous recognition.

  By the way, it is considered that the unit reference data corresponding to the current weather condition is selected in order to select the unit reference data based on the landscape image similar in image characteristics such as luminance, contrast, or hue to the actual landscape image. It is done. However, as a result of diligent research, the inventors of the present application can vary the followability of the change in the image characteristics with respect to the change in the weather conditions depending on the subject. Therefore, by simply selecting the unit reference data corresponding to the current weather conditions, It has been found that there is a possibility that the possibility of misrecognition cannot be sufficiently suppressed when the weather conditions have changed greatly in the past within a predetermined time from the present.

  In the present invention, based on the above knowledge, the unit reference data is selected for each of a plurality of divided areas obtained by dividing the shooting area of the actual landscape image. It is assumed that the conditions are considered. As a result, the unit reference data corresponding to the past weather conditions for the divided areas that have low followability to the change of the weather conditions and the image characteristic is high in the proportion of the subject that is not in the current state but in the same state as the past weather condition. Can be selected. With such a configuration, even if the weather conditions have changed greatly in the past within a predetermined time from the present, if the corresponding shooting positions are the same or close to each other, the actual scenery data and reference It is possible to increase the degree of coincidence obtained by matching with the business data. That is, according to the above-described feature configuration, even when the weather conditions have changed greatly in the past within a predetermined time from the present time, the possibility of misrecognition can be sufficiently suppressed. It becomes possible to specify the shooting position of the actual scenery image well regardless of the weather conditions.

  Further, in the present invention, if the viewing angle and the inclination angle with respect to the horizontal plane in the photographing direction are the same, the subject having low followability of the change in the image characteristic with respect to the change in the weather condition is determined as the photographing region regardless of the photographing position of the actual landscape image. Based on the knowledge obtained by the inventors of the present invention that they tend to exist in the same portion, the unit reference data for the divided region is selected in consideration of the position of the divided region in the imaging region. As a result, it is possible to appropriately select unit reference data without identifying subjects included in the divided areas, and the system configuration can be simplified.

  Here, a first area and a second area that do not overlap each other are set in the imaging area, and the unit reference data selection unit includes the current weather condition information for the divided areas included in the first area. The unit reference data corresponding to the weather condition to be represented is selected, and the divided area included in the second area is changed from the weather condition represented by the past weather condition information to the weather condition represented by the current weather condition information. Based on the weather condition represented by the current weather condition information and the weather condition represented by the past weather condition information, and selecting the unit reference data corresponding to the selected weather condition, It is preferable.

  According to this configuration, the unit reference data is set for the first area by setting the second area to include the divided area in which the ratio of the subject having a low followability of the change in the image characteristic to the change in the weather condition is high. In the second region, image characteristics depending on the direction of the change in weather conditions are considered by considering the change in weather conditions from the past to the present while reducing the factors considered when selecting The unit reference data can be selected in consideration of the difference in the followability of the change.

  As described above, in the configuration in which the second area is set in the photographing area, the actual landscape image is a landscape image including a road surface, and the second area is set to include at least a road surface. It is preferable to adopt the configuration.

  According to this configuration, the followability of changes in image characteristics with respect to changes in weather conditions, such as the time it takes for the road surface wet with rain to dry and the time it takes for snow accumulated on the road surface to melt, The unit reference data can be appropriately selected for each divided region, taking into account the fact that generally it tends to be low for the sky and buildings.

  As described above, in the configuration in which the second region is set so as to include at least the road surface of the road, the camera is mounted on the moving body, and estimation for acquiring estimated position information representing the estimated position of the moving body A position information acquisition unit; and a road width information acquisition unit that acquires road width information that is information related to the road width of the road around the position indicated by the estimated position information, wherein the second region includes the road width information acquisition unit. It is preferable that the configuration is variably set according to the acquired road width information.

  According to this configuration, it is possible to appropriately set the second area so that all or most of the road surface is included in the second area without detecting the road width by image recognition.

  Further, as described above, in the configuration in which the second region is set so as to include at least a road surface, the camera is mounted on a moving body and acquires estimated position information indicating the estimated position of the moving body. An estimated position information acquiring unit, and a region attribute information acquiring unit for acquiring region attribute information including at least a city part and a suburb as attribute items about the position indicated by the estimated position information, It is also preferable that the area is variably set according to the area attribute information acquired by the area attribute information acquisition unit.

  According to this configuration, in the suburbs, on the basis of the relationship between the road width and the region attribute information, such as a road width that is generally narrower than that in an urban area, it is possible to obtain all or large road surfaces without directly acquiring road width information. The second region can be appropriately set so that the portion is included in the second region.

  Further, in each of the above configurations in which the second area is set in the imaging area, the unit reference data selection unit includes precipitation data for the past weather condition information about the divided area included in the second area. In the case where the present weather condition information represents a weather condition not accompanied by precipitation, the weather condition represented by the past weather condition information is selected, and the past weather condition information represents a weather condition not accompanied by precipitation. When the current weather condition information represents a weather condition accompanied by precipitation, it is preferable that the weather condition represented by the current weather condition information is selected and the unit reference data corresponding to the selected weather condition is selected. It is.

  According to this configuration, when the weather condition changes from a condition with precipitation such as “rain” or “snow” to a condition without precipitation such as “sunny”, some subjects (for example, the road surface) On the other hand, the followability of changes in image characteristics with respect to changes in weather conditions decreases, while the weather conditions change from conditions that do not involve precipitation such as “sunny” to conditions that involve precipitation such as “rain” and “snow” In such a case, for all or almost all subjects, the unit reference data for the divided areas included in the second area is appropriately taken into consideration that the followability of the change in the image characteristics with respect to the change in the weather condition is appropriately considered. Can be selected appropriately.

  Further, in the shooting position specifying system having each configuration described above, the shooting position specifying unit derives the degree of coincidence between the real scene data and the unit reference data for each divided area, and the each of the divided areas It is preferable that the overall coincidence degree is derived for the entire photographing area based on the weighting factor set according to the position in the photographing area and the coincidence degree, and the success or failure of the matching is determined based on the overall coincidence degree. It is.

  According to this configuration, it is possible to set a high importance (weighting coefficient) for a divided region including a subject that is important for matching, and it is possible to improve matching accuracy. At this time, it is also preferable that the weighting factor for the divided region is variably set according to external factors such as weather conditions and time zone conditions.

  The technical features of the shooting position specifying system according to the present invention having the above-described configurations can be applied to a shooting position specifying program and a shooting position specifying method. Therefore, the present invention also includes such a program and method. Can be subject to rights.

  In this case, the characteristic configuration of the shooting position specifying program for specifying the shooting position of the real landscape image that is a landscape image shot by the camera is that a plurality of landscape images shot under different weather conditions from the same shooting position. Reference data for acquiring reference data composed of a group of unit reference data, which is data generated based on each of the data and associated with weather conditions at the time when the landscape image of the generation source is captured The acquisition function, the actual scenery data acquisition function for acquiring the actual scenery data generated based on the actual scenery image, the matching between the actual scenery data and the reference data, and the reference data successfully matched The information of the shooting position of the landscape image that is the generation source of the unit reference data to be configured is acquired, and the actual landscape data is based on the information of the shooting position A shooting area specifying function for specifying the shooting position of the real scenery image that is the generation source, and the matching of the real scenery data and the reference data by the shooting position specifying function, the shooting area of the real scenery image is divided. For each of the plurality of divided areas, the position of the divided area in the imaging area, current weather condition information indicating current weather conditions, and past weather condition information indicating past weather conditions more than a predetermined time before the current time And a unit reference data selection function for selecting the unit reference data corresponding to each divided region from the group of unit reference data constituting the reference data.

  The characteristic configuration of the shooting position specifying method for specifying the shooting position of a real landscape image that is a landscape image shot by a camera is a real landscape data acquisition step of acquiring real landscape data generated based on the real landscape image; , Data generated based on each of a plurality of landscape images taken from the same shooting position under different weather conditions, and data associated with the weather conditions at the time when the generation source landscape image was shot The reference data acquisition step of acquiring reference data composed of a group of unit reference data, and the shooting of the divided area for each of a plurality of divided areas obtained by dividing the shooting area of the real landscape image Based on the location in the region, current weather condition information representing the current weather conditions, and past weather condition information representing past weather conditions more than a predetermined time before the present. The unit reference data selection step for selecting the unit reference data corresponding to each divided region from the group of unit reference data constituting the reference data, and the unit reference data selection step selected by the unit reference data selection step Using the unit reference data, the actual landscape data and the reference data are matched, and information on the shooting position of the landscape image from which the unit reference data constituting the reference data that has been successfully matched is acquired. And a shooting position specifying step of specifying the shooting position of the real scenery image from which the real scenery data is generated based on the information of the shooting position.

  Naturally, these shooting position specifying programs and shooting position specifying methods can also obtain the effects of the above-described shooting position specifying system, and further incorporate some additional techniques mentioned as examples of suitable configurations thereof. It is possible.

It is a mimetic diagram showing a schematic structure of a navigation device concerning an embodiment of the present invention. It is explanatory drawing of the structure of the data for reference which concerns on embodiment of this invention. It is explanatory drawing of the imaging area of the real scenery image which concerns on embodiment of this invention, and the division area formed by dividing the said imaging area. It is a figure which shows the selection conditions of the unit reference data which concern on embodiment of this invention. It is a flowchart which shows the whole procedure of the imaging position specific process which concerns on embodiment of this invention. It is a flowchart which shows the procedure of the matching process which concerns on embodiment of this invention.

  Embodiments of the present invention will be described with reference to the drawings. Here, the case where the imaging position specifying system according to the present invention is applied to an in-vehicle navigation device will be described as an example. The navigation apparatus 1 according to the present embodiment includes a shooting position specifying unit 2 (see FIG. 1), and the vehicle (not shown in the figure, hereinafter sometimes referred to as “own vehicle”) has the shooting position specifying unit 2 as a core. A shooting position specifying system for specifying a shooting position of a landscape image (hereinafter referred to as “real landscape image SP”) shot by a camera 90 mounted on the camera has been constructed. That is, in this embodiment, the camera 90 is provided in a vehicle (an example of a moving body). Hereinafter, the imaging position specifying system (navigation apparatus 1) according to the present embodiment will be described in the order of "schematic configuration of navigation apparatus", "configuration of imaging position specifying unit", and "operation processing procedure". In the present embodiment, the vehicle corresponds to the “moving body” in the present invention.

1. Schematic Configuration of Navigation Device First, a schematic configuration of the navigation device 1 according to the present embodiment will be described. As shown in FIG. 1, the navigation device 1 includes a navigation control unit 12, a shooting position specifying unit 2, an estimated position calculation unit 16, an actual shot image processing unit 17, a reference data storage unit 11, and map information storage. Part 10. The navigation device 1 has a function of calculating an estimated position of the own vehicle (hereinafter sometimes referred to as “estimated own vehicle position”) based on a GPS signal from a GPS (Global Positioning System) satellite. A function of correcting the estimated vehicle position based on the shooting position of the actual landscape image SP specified by the position specifying unit 2 is provided.

  As will be described in detail later, as shown in FIG. 1, the navigation control unit 12 and the shooting position specifying unit 2 include a plurality of functional units. Each functional unit (except for the map information storage unit 10 and the reference data storage unit 11, which is the same hereinafter) included in the navigation device 1 is input with an arithmetic processing unit such as a common or independent CPU as a core member. A functional unit for performing various processes on the data is implemented by hardware or software (program) or both. Each of these functional units is configured to be able to exchange information with each other via a communication line such as a digital transfer bus, and also receives data from the map information storage unit 10 and the reference data storage unit 11. It is configured to be extractable. Here, when each function part is comprised with software (program), the said software is memorize | stored in memory | storage means, such as RAM and ROM which an arithmetic processing unit can refer.

  The map information storage unit 10 and the reference data storage unit 11 are databases configured by a storage device. This storage device includes, as a hardware configuration, a recording medium capable of storing and rewriting information, such as a hard disk drive and a flash memory. The map information storage unit 10 and the reference data storage unit 11 may have independent hardware, or may be provided in common hardware.

  The navigation control unit 12 includes a route setting unit 13, a route search unit 14, and a route guide unit 15. The route setting unit 13 is a functional unit that sets, for example, a departure place such as the position of the host vehicle, an input destination, a passing point, and traveling conditions (whether or not an expressway is used). The route search unit 14 is a functional unit that performs calculation processing (route search processing) for searching for a guidance route from the departure point to the destination based on the conditions set by the route setting unit 13. The route guidance unit 15 operates according to the guidance from the display screen of a monitor (not shown) or voice guidance from a speaker (not shown) according to the route from the departure place to the destination searched by the route search unit 14. It is a functional part which performs the calculation process (route guidance process) for performing suitable route guidance with respect to a person.

  The map information storage unit 10 stores (stores) road map data that is referred to when performing the above-described route search processing and route guidance processing, or when executing map display processing on a monitor. Although illustration is omitted, in this embodiment, the road map data includes information on a number of nodes having position information on the map expressed by latitude and longitude, and a route corresponding to the road by connecting two nodes. Information of a number of links to be configured. Nodes are set at road intersections, turning points, branch points, and the like, for example.

  The road map data includes information such as road type information (types of expressways, toll roads, national roads, prefectural roads) and link length information as link information (link attributes) of each link. The road map data also includes information on shape interpolation points that are arranged between two nodes (on the link) and indicate the detailed shape of the link, and road width information RW that is information on the road width of the road. The road width information RW is, for example, information that directly represents the road width with a numerical value, or information that indirectly represents the road width with the number of lanes.

  Further, in the present embodiment, the road map data includes regional attribute information in which each area is classified as “urban area”, “suburb”, or “mountainous area” according to the tendency such as the average height or the density of buildings. RA is also included. That is, in this example, the regional attribute information RA includes “urban area”, “suburb”, and “mountainous area” as attribute items in order from the side where the average height of buildings and the tendency of congestion are high.

  Then, as will be described later, the road width information acquisition unit 34 provided in the photographing position specifying unit 2 refers to the road map data stored in the map information storage unit 10 to obtain road width information RW around the estimated vehicle position. get. Further, as will be described later, the region attribute information acquisition unit 35 provided in the photographing position specifying unit 2 refers to the road map data stored in the map information storage unit 10 and the region attribute around the estimated vehicle position. Information RA is acquired.

  The estimated position calculation unit 16 is a functional unit that calculates the estimated own vehicle position based on the outputs of the GPS measurement unit 91, the distance sensor 92, and the direction sensor 93 provided in the vehicle. The GPS measurement unit 91 receives a GPS signal from a GPS satellite and outputs the GPS signal to the estimated position calculation unit 16. The distance sensor 92 detects the vehicle speed and movement distance of the vehicle 100 and outputs information on the vehicle speed and movement distance as a detection result to the estimated position calculation unit 16. The azimuth sensor 93 includes, for example, a gyro sensor, a geomagnetic sensor (electronic compass), and the like, and outputs azimuth information as a detection result to the estimated position calculation unit 16.

  The estimated position calculation unit 16 analyzes the GPS signal, calculates the current position (latitude and longitude) of the host vehicle, and obtains GPS position data. Further, the estimated position calculation unit 16 calculates a dead reckoning position based on the moving distance information and the azimuth information sent every moment to obtain dead reckoning position data. And the estimated position calculation part 16 performs the calculation which calculates the estimated own vehicle position which is an estimated position of the own vehicle by a well-known method from GPS position data and dead reckoning position data. Note that the estimated position calculation unit 16 may be configured to calculate the estimated own vehicle position using only one of the GPS position data and the dead reckoning position data.

  Note that the estimated vehicle position calculated as described above is information including measurement errors, and may be off the road in some cases. In this embodiment, in such a case, based on the road map data stored in the map information storage unit 10, the estimated position calculation unit 16 corrects the vehicle position on the road indicated on the road map ( Map matching process). Then, information on the estimated own vehicle position (estimated position information EP) calculated by the estimated position calculation unit 16 is output to the estimated position information acquisition unit 31 included in the imaging position specifying unit 2. The estimated position calculation unit 16 may be configured not to perform the map matching process.

  The reference data storage unit 11 converts the reference data RD to the shooting position (more precisely, shooting position information) of the landscape image from which unit reference data RDo (to be described later) constituting the reference data RD is generated. It is a storage unit that stores (stores) in association with each other. As will be described later, the reference data RD is data referred to for specifying the shooting position of the real landscape image SP. The reference data RD is composed of a group of unit reference data RDo generated based on each of a plurality of landscape images shot under different weather conditions from the same shooting position. In other words, the reference data RD is composed of a plurality of unit reference data RDo, and in the plurality of unit reference data RDo, a plurality of landscape images captured from the same shooting position under different weather conditions are included. The unit reference data RDo generated based on each is included. Each unit reference data RDo is associated with a weather condition at the time when the generation source landscape image is captured. In the present embodiment, the reference data storage unit 11 stores a plurality of reference data RD.

  In this embodiment, as shown in FIG. 2, the reference data RD is composed of a group of four unit reference data RDo. That is, in this example, one reference data RD is constituted by a total of four unit reference data RDo generated based on each of four landscape images from the same shooting position (in this example, a predetermined position on the road). Has been. The landscape image for generating the reference data RD is, for example, a captured image of a landscape by a data collection vehicle that travels for the purpose of generating the reference data RD. In addition, the shooting position associated with the reference data RD is determined with higher accuracy than the estimated vehicle position calculated by the estimated position calculation unit 16. Apart from this, it is also possible to adopt a configuration in which the reference data RD is maintained by learning based on a landscape image taken while the host vehicle or another vehicle (a normal vehicle different from the data collection vehicle) is running. It is.

  In the present embodiment, the landscape image from which the unit reference data RDo constituting the same reference data RD is generated includes a plurality of landscape images taken under different weather conditions and has different times. It contains multiple landscape images taken under belt conditions. In other words, in the present embodiment, the group of unit reference data RDo constituting the same reference data RD was shot from the same shooting position under the same weather conditions and different time zone conditions. Unit reference data RDo generated based on each of a plurality of landscape images is also included. Specifically, in the present embodiment, the weather condition is divided into “sunny” and “rain”, and the time zone condition is divided into “day” and “night”. And the same reference data based on a total of four landscape images taken under four sets of conditions that are combined with the time zone conditions (at least one of the weather conditions and the time zone conditions are different from each other) Four unit reference data RDo constituting the RD are generated. That is, in this embodiment, as the unit reference data RDo corresponding to the same shooting position, four unit reference data RDo associated with the time zone condition in addition to the weather condition are prepared.

  Regarding weather conditions, “sunny” is an example of weather conditions that do not involve precipitation, and “rain” is an example of weather conditions that involve precipitation. As for time zone conditions, “daytime” represents a time zone from sunrise to sunset, and “night” represents a time zone from sunset to next day sunrise. For example, a landscape image taken at 12:00 (noon) can be used as a landscape image for generating the unit reference data RDo corresponding to the time zone condition “daytime”. Further, as a landscape image for generating the unit reference data RDo corresponding to the time zone condition “night”, for example, a landscape image taken at 0:00 (Masako) can be used.

  In the following description, the unit reference data RDo based on the landscape image under the condition that the time zone condition is “daytime” and the weather condition is “sunny” is “first condition unit reference data RDo1”, and the time zone condition is The unit reference data RDo based on the landscape image under the condition of “daytime” and the weather condition “rain” is “second condition unit reference data RDo2”, the time zone condition is “night”, and the weather condition is “ The unit reference data RDo based on the landscape image under the condition of “sunny” is set as the “third condition unit reference data RDo3”, the scenery under the condition that the time zone condition is “night” and the weather condition is “rain”. The unit reference data RDo based on the image is referred to as “fourth condition unit reference data RDo4”.

  By the way, the unit reference data RDo is data to be matched (pattern matching) with actual landscape data AD described later, and is data representing image features suitable for landscape image recognition. In this embodiment, the unit reference data RDo, as conceptually shown in FIG. 2, is a collection of image feature points (contour information of features (for example, buildings, road markings, etc.) extracted from the source landscape image). Data of edge point images (feature point images) represented as feature point groups). Such an edge point image can be generated by performing binarization processing, edge detection processing, or the like on the generation source landscape image. In the binarization process and the edge detection process, a process based on a luminance difference (density difference), a saturation difference, or a hue difference in a landscape image can be performed.

  In the present embodiment, each of a plurality (four in this example) of unit reference data RDo constituting the same reference data RD is basically generated at the same recognition level when detecting (extracting) contour information. It is assumed. Therefore, as conceptually shown in FIG. 2, when the weather condition is “rain”, the detection is more effective than the case where the weather condition is “sunny” due to the reflection of light by the water film on the road surface. The possible contour information can be different. In FIGS. 2B and 2D, when the weather condition is “rain”, recognition of road markings (see FIG. 3A) provided on the road surface due to irregular reflection by a water film on the road surface, etc. As an example, the rate decreases and the feature points corresponding to the road marking decrease compared to when the weather condition is “sunny”. Depending on the position of the sun with respect to the vehicle and the color or material of the road marking, the contrast of the wet road marking may increase, and the strength of the edge representing the outline of the road marking may increase. In any case, the contents of the unit reference data RDo (in this example, the distribution of edge points in the edge point image) generally tend to be different between the case where the weather condition is “sunny” and the case where it is “rain”. is there.

  In addition, as illustrated in FIG. 2, when the time zone condition is “night”, the recognition rate of the features in the low luminance region other than the region where the luminance is appropriately secured by the lighting device is reduced, The feature points corresponding to the features in the low-luminance region tend to decrease compared to when the time zone condition is “daytime”. In this example, the camera 90 is a front camera that captures a landscape in front of the host vehicle, and the low-luminance region is illustrated in FIG. 2C and FIG. The upper part in the landscape image, which is outside the illumination range of the headlight (an example of a lighting device) that is not secured and the brightness is not secured, is the center.

  The actual captured image processing unit 17 is a functional unit that acquires the actual landscape image SP captured by the camera 90 and generates actual landscape data AD based on the actual landscape image SP. In the present embodiment, the real landscape image SP is a landscape image including a road surface as illustrated in FIG. In the present embodiment, the actual captured image processing unit 17 extracts image features from the actual landscape image SP, and the actual landscape data AD that represents the image features with the same information (method) as the unit reference data RDo described above. Is generated. As described above, in the present embodiment, the unit reference data RDo is data of edge point images, and the actual captured image processing unit 17 performs binarization processing, edge detection processing, and the like on the actual landscape image SP. As a result, contour information of the features included in the actual landscape image SP is extracted, and actual landscape data AD that is data of edge point images (feature point images) is generated. Note that the actual captured image processing unit 17 basically extracts feature outline information from the actual landscape image SP at the same recognition level as that used when generating the unit reference data RDo.

2. Configuration of Shooting Position Specifying Unit Next, the configuration of the shooting position specifying unit 2 that is a main part of the present invention will be described. As shown in FIG. 2, the shooting position specifying unit 2 includes an estimated position information acquisition unit 31, a real landscape data acquisition unit 32, a reference data acquisition unit 33, a road width information acquisition unit 34, a region attribute information acquisition unit 35, a weather condition The information acquisition part 36, the imaging | photography position specific | specification part 40, the unit reference data selection part 42, the division area adjustment part 43, and the own vehicle position determination part 44 are provided.

2-1. Configuration of Estimated Position Information Acquisition Unit The estimated position information acquisition unit 31 is a functional unit that acquires estimated position information EP representing the estimated position of the own vehicle (estimated own vehicle position). In the present embodiment, the estimated position information EP calculated by the estimated position calculator 16 is input to the estimated position information acquisition unit 31 so that the estimated position information acquisition unit 31 acquires the estimated position information EP. Yes.

2-2. Configuration of Real Landscape Data Acquisition Unit The real landscape data acquisition unit 32 is a functional unit that acquires actual landscape data AD generated based on the actual landscape image SP. In the present embodiment, the actual scene data acquisition unit 32 inputs the actual scene data AD generated based on the actual scene image SP by the actual captured image processing unit 17 to the actual scene data acquisition unit 32. Is configured to get.

  In the present embodiment, the photographing of the actual landscape image SP by the camera 90 is repeatedly executed at a predetermined timing. For example, the predetermined timing is set every predetermined time interval (for example, every 1 second or 10 seconds), or set every time the vehicle travels a predetermined distance (for example, 1 meter or 10 meters). can do. Further, the predetermined timing can be set at a time when the host vehicle approaches a specific point (for example, an intersection or a pedestrian crossing). In the present embodiment, the actual captured image processing unit 17 generates the actual scenery data AD from the new actual scenery image SP each time the camera 90 acquires the new actual scenery image SP, and the generated actual scenery. The data AD is output to the real landscape data acquisition unit 32. That is, in the present embodiment, the actual landscape data acquisition unit 32 acquires the actual landscape data AD from the actual captured image processing unit 17 at the same frequency as when the actual landscape image SP is captured by the camera 90.

2-3. Configuration of Reference Data Acquisition Unit The reference data acquisition unit 33 is a functional unit that acquires reference data RD. In the present embodiment, the reference data acquisition unit 33 refers to the reference data storage unit 11 and includes one or a plurality of reference data RD that are candidates for matching processing (pattern matching processing) by the imaging position specifying unit 40 described later. To get.

  Specifically, as described above, each reference data RD is stored in the reference data storage unit 11 in association with the shooting position of the generation-source landscape image. Then, the reference data acquisition unit 33 obtains the estimated position information EP representing the estimated vehicle position at the time of shooting the actual landscape image SP from the estimated position information acquisition unit 31, and shoots in an area near the estimated vehicle position. One or a plurality of reference data RD associated with the position are extracted as reference data RD for matching candidates.

  Here, the region in the vicinity of the estimated host vehicle position can be set based on the maximum value of errors that can be included in the estimated position information EP. For example, the reference data acquisition unit 33 sets the circular area whose radius is equal to the maximum value of the error as the area near the estimated own vehicle position, with the estimated own vehicle position indicated in the estimated position information EP as the center. The configuration may be such that the reference data RD in which the shooting position of the source landscape image is located inside the circular area is extracted.

  In addition, the above “at the time of shooting the actual landscape image SP” is an arbitrary time point from the time when the camera 90 actually captures the actual landscape image SP to the time when the actual landscape data acquisition unit 32 acquires the actual landscape data AD. It can be.

2-4. Configuration of Road Width Information Acquisition Unit The road width information acquisition unit 34 is a functional unit that acquires road width information RW of a road around the position indicated by the estimated position information EP (hereinafter sometimes referred to as “current position”). As described above, in this embodiment, the road map data stored in the map information storage unit 10 includes the road width information RW, and the road width information acquisition unit 34 stores the road stored in the map information storage unit 10. The road width information RW of the road around the current position is acquired with reference to the map data. Then, the road width information RW acquired by the road width information acquisition unit 34 is output to the divided region adjustment unit 43 described later. It should be noted that the range of the position from which the road width information RW is acquired is preferably a range including a road that is included in a size capable of appropriately extracting the contour information in the actual landscape image SP.

2-5. Configuration of Regional Attribute Information Acquisition Unit The regional attribute information acquisition unit 35 is a functional unit that acquires regional attribute information RA including at least a city area and a suburb as attribute items about the position indicated by the estimated position information EP. As described above, in this embodiment, the region attribute information RA is included in the road map data stored in the map information storage unit 10, and the region attribute information acquisition unit 35 is stored in the map information storage unit 10. The regional attribute information RA around the current position is acquired with reference to the road map data.

  In the present embodiment, the regional attribute information RA includes “mountain area” as an attribute item in addition to “urban area” and “suburb”. For this reason, the region attribute information acquisition unit 35 acquires information indicating which region of the “urban area”, “suburb”, and “mountainous region” the position surrounding indicated by the estimated position information EP corresponds to. Then, the region attribute information RA acquired by the region attribute information acquisition unit 35 is output to the divided region adjustment unit 43 described later.

2-6. Configuration of Division Area Adjustment Unit The division area adjustment unit 43 is a functional unit that adjusts the size of the division area DA set in the shooting area WA of the real landscape image SP. As will be described later, in the present invention, when matching the actual scenery data AD and the reference data RD, each of the plurality of divided areas DA obtained by dividing the shooting area WA of the actual scenery image SP is used as a reference for matching. Unit reference data RDo used for the matching is selected from a group of unit reference data RDo constituting data RD. Then, the divided area adjustment unit 43 considers the type and arrangement area of the subject included in the real scenery image SP and increases the size of the divided area DA in order to improve the matching accuracy between the real scenery data AD and the reference data RD. It is possible to adjust the height. As will be described later, in this embodiment, since a weighting factor is set for each of the divided areas DA, the divided area DA divides areas having different selection conditions for the unit reference data RDo and is important for matching. It is set to delimit areas of different degrees.

  In the present embodiment, as shown in FIG. 3, as the divided area DA set in the imaging area WA, there are three divided areas of a first type divided area DA1, a second type divided area DA2, and a third type divided area DA3. An area is set. Here, the first type divided area DA1 is an area that is set so that the proportion of the sky in the divided area is high. The second type division area DA2 is an area that is set so that the proportion of buildings in the division area is high. The third type division area DA3 is an area that is set so that the proportion of the road surface in the division area is high.

  As shown in FIG. 3B, in this embodiment, a first area A1 and a second area A2 that do not overlap each other are set in the imaging area WA. In this example, the first area A1 and the second area A2 are set so that the added area is equal to the imaging area WA. Each of the first area A1 and the second area A2 is set so as to include only the divided area DA in which the selection conditions of the unit reference data RDo by the unit reference data selection unit 42 described later are the same.

  In the present embodiment, the first area A1 is set to include the first type divided area DA1 and the second type divided area DA2, and the second area A2 is set to include the third type divided area DA3. . Specifically, the first area A1 is set to be equal to the area obtained by adding the first type divided area DA1 and the second type divided area DA2, and the second area A2 is set to be the third type divided area DA3. Is set to be equal to Thus, in the present embodiment, the second area A2 is set so as to include the third type divided area DA3 that is set so that the ratio of the road surface in the divided area is increased, so that at least the road It is set to include the road surface. Then, the divided area adjustment unit 43 appropriately adjusts the size of each divided area DA set in advance at a predetermined position and range without identifying the subject included in the real landscape image SP as described below. It is configured to be able to. That is, in this embodiment, the size of each divided area DA is configured to be variably set.

  In the following description, the direction (lateral direction in FIG. 3) corresponding to the direction (horizontal direction of the vehicle) orthogonal to both the traveling direction and the vertical direction of the vehicle in the imaging area WA is referred to as “left and right in the imaging area WA. The direction (vertical direction in FIG. 3) corresponding to the direction along the vertical direction (vertical direction of the vehicle) is referred to as “vertical direction in the imaging area WA”.

  In this embodiment, the divided area adjustment unit 43 variably sets the third type divided area DA3 based on the road width information RW of the road around the position indicated by the estimated position information EP acquired by the road width information acquisition unit 34. The second type divided area DA2 is variably set based on the area attribute information RA about the position indicated by the estimated position information EP acquired by the area attribute information acquisition unit 35.

  Specifically, the divided area adjustment unit 43 continuously or stepwise sets the lateral width of the third type divided area DA3 (the width in the left-right direction in the imaging area WA) as the road width of the road represented by the road width information RW increases. Set to be larger. In the example shown in FIG. 3, the third type divided area DA3 is set as a polygonal area, and increasing the horizontal width of the third type divided area DA3 is inclined with respect to the left-right direction in the imaging area WA. This corresponds to reducing the angle of intersection of the sides of the polygon extending in the direction with the horizontal direction. Accordingly, the third type divided area DA3 can be set so that most of the road surface is included in the divided area without performing image processing on the actual scenery image SP and detecting the road width. In the present embodiment, when the divided region adjustment unit 43 repeatedly acquires the road width information RW at a predetermined timing and the road width indicated by the road width information RW changes significantly, the size of the third type divided region DA3 is increased. adjust.

  In this example, since the second area A2 is set to be equal to the third type divided area DA3, the second area A2 is also changed by changing the third type divided area DA3 as described above. Is done. In other words, in this example, the divided region adjustment unit 43 sets the second region A2 variably according to the road width information RW acquired by the road width information acquisition unit 34. At this time, the size of the first area A1 is changed in accordance with the changed size of the second area A2 so that the sum of the first area A1 and the second area A2 becomes equal to the imaging area WA. At the same time, the position and size of the first type divided area DA1 and the second type divided area DA2 set in the first area A1 are also adjusted.

  In addition, the divided area adjustment unit 43 sets the vertical width of the second type divided area DA2 (the vertical width in the imaging area WA) when the area attribute information RA represents “urban area” and the area attribute information RA indicates “suburb”. So that the vertical width of the second-type divided area DA2 in the case of "" and the vertical width of the second-type divided area DA2 in the case where the area attribute information RA represents "mountain" are reduced in the order of description. The second type divided area DA2 is set according to the attribute information RA. Accordingly, the second type divided area DA2 can be set so that most of the building is included in the divided area without performing image processing on the real landscape image SP and detecting the building. In the present embodiment, when the divided region adjustment unit 43 repeatedly acquires the region attribute information RA at a predetermined timing and the content of the attribute item indicated in the region attribute information RA changes, the second type divided region DA2 Adjust the size.

  In this example, the first-type divided area DA1 and the second-type divided area DA2 are set so that the added area is equal to the first area A1, and thus the second-type divided area as described above. The first type divided area DA1 is also changed in accordance with the change of DA2.

2-7. Configuration of Weather Condition Information Acquisition Unit The weather condition information acquisition unit 36 represents information indicating the current weather conditions around the position indicated in the estimated position information EP (hereinafter referred to as “current weather condition information CW”), and the present. It is a functional unit that acquires information indicating past weather conditions before a predetermined time (hereinafter referred to as “past weather condition information PW”). Although illustration is omitted, in the present embodiment, the navigation device 1 is provided with a communication device that transmits and receives data by wireless communication via a wireless base station or the like, and the weather condition information acquisition unit 36 includes the communication device. Then, current weather condition information CW and past weather condition information PW around the current position are acquired from a weather information providing server or the like that provides weather condition information. The current weather condition information CW and the past weather condition information PW acquired by the weather condition information acquisition unit 36 are output to the unit reference data selection unit 42.

  In this example, as described above, the weather condition information is divided into two, “sunny” and “rain”, and the weather conditions represented by each of the current weather condition information CW and the past weather condition information PW are: It is either “Sunny” or “Rain”. Here, although the case where the weather condition information acquisition unit 36 acquires the information of the weather conditions classified in advance as “sunny” and “rain” is shown as an example, the weather condition information acquisition unit 36 It is configured to be able to acquire information on weather conditions other than “clear” and “rain” such as “cloudy” and “snow”, and the weather condition information acquisition unit 36 sets the acquired weather conditions to “sunny” and “rain”. It can also be set as the structure classified into these. In such a case, when the weather condition is “cloudy”, it may be classified as “sunny”, and when the weather condition is “snow”, it may be classified as “rain”.

  In addition, the time point when the past weather condition information PW indicates the weather condition is a time point that is a predetermined time or more before the present time, and this predetermined time can be set based on, for example, the time until the rainy road surface dries. it can. For example, the predetermined time can be 30 minutes, 1 hour, 2 hours, or the like. In addition, the predetermined time may be variably set based on information on roads around the current position (information such as general paved roads, drainage paved roads, and unpaved roads).

  As a method of wireless communication by the communication device, for example, a known communication network such as a mobile phone network or a wireless local area network (LAN) can be used. In addition, wireless communication is configured using VICS (registered trademark; Vehicle Information and Communication System), or between roads and vehicles that communicate with communication devices installed on the road side. It is good also as a structure performed using communication.

2-8. Configuration of Unit Reference Data Selection Unit The unit reference data selection unit 42 forms unit reference data constituting the reference data RD to be matched when matching the actual scene data AD and the reference data RD by the shooting position specifying unit 40 described later. This is a functional unit that selects unit reference data RDo used for matching from a group of RDo. As described above, in the present embodiment, the reference data acquisition unit 33 acquires one or a plurality of reference data RD that are candidates for matching with the actual landscape data AD. Then, the unit reference data selection unit 42 performs a unit reference data RDo selection process for the reference data RD selected as a matching target from the matching candidate reference data RD.

  The unit reference data selection unit 42, for each of a plurality of divided areas DA obtained by dividing the shooting area WA of the real landscape image SP, the position of the divided area DA in the shooting area WA, the current weather condition information CW, and the past The unit reference data RDo corresponding to each divided area DA is selected based on the weather condition information PW and further information on the current time zone condition in this example. As described above, the current weather condition information CW and the past weather condition information PW are input from the weather condition information acquisition unit 36 to the unit reference data selection unit 42.

  By the way, in the present embodiment, the weather condition is divided into “sunny” and “rain”. Therefore, in the present embodiment, the weather condition represented by the current weather condition information CW (hereinafter referred to as “current weather condition”) from the weather condition represented by the past weather condition information PW (hereinafter sometimes referred to as “past weather condition”). There is a case of “no change” in which the current weather condition and the past weather condition are the same, and a case of “with change” in which the current weather condition and the past weather condition are different. “There is a change” has two directions, a change from “sunny” to “rain” and a change from “rain” to “sunny”. When the weather condition changes from “sunny” to “rain”, image characteristics such as brightness, contrast, and hue (hereinafter, simply “ "Image characteristics") changes following a change in weather conditions without a large delay. On the other hand, when the weather condition changes from “rain” to “sunny”, some subjects included in the actual scenery image SP (hereinafter referred to as “the rainy road surface” take time). , “Low tracking subject”), the image characteristics follow with a relatively large delay with respect to changes in weather conditions.

  In view of the above, in the present embodiment, not only the present, but also the second area A2 including the third-type divided area DA3 that is set so that the proportion of the low-tracking subject (the road surface in this example) is increased. The structure takes into account past weather conditions. Specifically, the unit reference data selection unit 42 determines the current weather condition from the weather condition represented by the past weather condition information PW for the divided area DA (in this example, the third type divided area DA3) included in the second area A2. Based on the change to the weather condition represented by the information CW, select either the weather condition represented by the current weather condition information CW or the weather condition represented by the past weather condition information PW, and correspond to the selected weather condition The unit reference data RDo to be selected is selected.

  More specifically, in this embodiment, the unit reference data selection unit 42 selects unit reference data RDo for the divided area DA (in this example, the third type divided area DA3) included in the second area A2. When the past weather condition information PW represents “rain” and the current weather condition information CW represents “sunny”, “rain” that is the weather condition represented by the past weather condition information PW is selected, and the selected weather Unit reference data RDo corresponding to the condition (ie, “rain”) is selected. Further, the unit reference data selection unit 42, when the past weather condition information PW represents “sunny” and the current weather condition information CW represents “rain”, the “rain” is the weather condition represented by the current weather condition information CW. And the unit reference data RDo corresponding to the selected weather condition (ie, “rain”) is selected. Of course, when the past weather condition information PW represents “sunny” and the current weather condition information CW represents “sunny”, the unit reference data RDo corresponding to “sunny” is selected, and the past weather condition information When the PW represents “rain” and the current weather condition information CW represents “rain”, the unit reference data RDo corresponding to “rain” is selected.

  On the other hand, for the first area A1 including the first-type divided area DA1 and the second-type divided area DA2 that are set so that the ratio of subjects other than the low-tracking subject (sky or building in this example) occupies is high, Only current weather conditions are considered. Specifically, the unit reference data selection unit 42 stores the current weather condition information CW for the divided area DA (in this example, the first type divided area DA1 and the second type divided area DA2) included in the first area A1. A weather condition to be expressed is selected, and unit reference data RDo corresponding to the selected weather condition is selected.

  More specifically, in the present embodiment, the unit reference data selection unit 42 uses the unit for the divided area DA (in this example, the first type divided area DA1 and the second type divided area DA2) included in the first area A1. When selecting the reference data RDo, if the current weather condition information CW represents “sunny”, the unit reference data RDo corresponding to “sunny” is selected, and if the current weather condition information CW represents “rain” Unit reference data RDo corresponding to “rain” is selected.

  As described above, in the present embodiment, the divided area DA (the third-type divided area DA3 in this example) included in the second area A2 set in the lower portion in the up-down direction in the imaging area WA is not only present. In consideration of past weather conditions, the divided area DA (in this example, the first type divided area DA1 and the second type included in the first area A1 set on the upper side in the up-down direction in the imaging area WA with respect to the second area A2 The divided area DA2) is configured so that only the current weather conditions are taken into account, so that the position of the divided area DA in the imaging area WA is taken into account when selecting the unit reference data RDo.

  In the present embodiment, the time zone condition is divided into “daytime” and “night”, and the unit reference data RDo is prepared in association with the time zone condition in addition to the weather condition. Yes. Then, for all or almost all subjects included in the actual landscape image SP, the followability of changes in image characteristics with respect to changes in time zone conditions is generally high. Therefore, in the present embodiment, the unit reference data selection unit 42 performs both of the divided area DA included in the first area A1 and the divided area DA included in the second area A2, that is, the first type divided area DA1, the first area. For all of the second-type divided area DA2 and the third-type divided area DA3, a current time zone condition that is a time zone condition at the time of shooting the real landscape image SP is selected, and the selected time zone condition (that is, the current time zone) Unit reference data RDo corresponding to (condition) is selected.

  More specifically, in the present embodiment, when selecting the unit reference data RDo, the unit reference data selection unit 42 selects “daytime” when the current time zone condition is “daytime”, and the selection. The unit reference data RDo corresponding to the set time zone condition (ie, “daytime”) is selected. The unit reference data selection unit 42 selects “night” when the current time zone condition is “night”, and sets the unit reference data RDo corresponding to the selected time zone condition (ie, “night”). select.

  Although illustration is omitted, the navigation device 1 is provided with a time acquisition unit that acquires the current time, and the unit reference data selection unit 42 acquires the current time zone condition via this time acquisition unit. . At this time, the navigation device 1 is configured to be able to acquire information on the sunrise and sunset times of each day and information on the current date, and the current time is the sunrise time and sunset time corresponding to the current date. If the current time zone condition is “daytime” and the current time is the time before sunrise or the time after sunset corresponding to the current date, The time zone condition can be determined to be “night”. Alternatively, the navigation device 1 includes a lighting state acquisition unit that acquires the lighting state of the headlight provided in the host vehicle, and when the headlight is in a non-lighting state, the current time zone condition is “daytime”. If the headlamp is in the lighting state, the current time zone condition can be determined to be “night”.

  In summary, the unit reference data RDo selection conditions by the unit reference data selection unit 42 are as shown in the table of FIG. As shown in this table, each divided area DA, except for the divided area DA included in the second area A2 when the current weather condition is “sunny” and the past weather condition information is “rain”, Unit reference data RDo corresponding to the current weather condition and the current time zone condition is selected. On the other hand, for the divided area DA included in the second area A2 when the current weather condition is “sunny” and the past weather condition is “rain”, the past weather condition and the current time zone condition are exceptionally different. Corresponding unit reference data RDo is selected.

  From another viewpoint, when the past weather condition is “sunny”, the unit reference data RDo corresponding to the current weather condition is selected for all the divided areas DA. When the current weather condition is “rain”, the unit reference data RDo corresponding to “rain” is selected for all the divided areas DA. When the past weather condition is “rain” or the current weather condition is “sunny”, the unit reference data corresponding to the current weather condition is uniformly obtained for the divided area DA included in the first area A1. Whereas RDo is selected, the divided area DA included in the second area A2 is based on both the current weather condition and the past weather condition, that is, based on the change from the past weather condition to the current weather condition. Unit reference data RDo corresponding to weather conditions or past weather conditions is selected. Note that the selection condition for the second region A2 shown in FIG. 4 is determined in advance in consideration of a change from the past weather condition to the current weather condition (direction of change). By selecting the unit reference data RDo based on the selection condition, the unit reference data RDo is selected based on the change from the past weather condition to the current weather condition.

2-9. Configuration of Shooting Position Specifying Unit The shooting position specifying unit 40 performs matching between the actual scenery data AD generated based on the actual scenery image SP and the reference data RD, and the shooting associated with the reference data RD that has been successfully matched. This is a functional unit that identifies the shooting position of the actual landscape image SP based on the position.

  The photographing position specifying unit 40 includes a matching execution unit 41 as shown in FIG. 1 as a functional unit for performing matching between the actual scenery data AD and the reference data RD. Then, the shooting position specifying unit 40 acquires information on the reference data RD that is determined to have been successfully matched by the matching execution unit 41, and acquires information on the shooting position associated with the reference data RD. Then, the shooting position specifying unit 40 specifies the acquired shooting position as the shooting position of the actual scenery image SP, and outputs information on the shooting position of the specified actual scenery image SP to the vehicle position determination unit 44 described later. .

  The matching execution unit 41 is a functional unit that performs matching (pattern matching) between the actual scene data AD and the reference data RD. As described above, in the present embodiment, the reference data acquisition unit 33 acquires one or a plurality of reference data RD that are candidates for matching with the actual landscape data AD. Then, the matching execution unit 41 sequentially selects one reference data RD from the matching candidate reference data RD as the matching target reference data RD, the matching target reference data RD, and the actual scenery Processing necessary for matching (pattern matching) with the actual scenery data AD acquired by the data acquisition unit 32 is executed.

  In the present invention, one unit reference data RDo is selected by the unit reference data selection unit 42 for each of a plurality of divided areas DA obtained by dividing the shooting area WA of the real landscape image SP. Then, for each of the divided areas DA, the matching execution unit 41 stores the data representing the image characteristics in the divided area DA in the divided area DA included in the unit reference data RDo selected for the divided area DA. Matching between the actual scenery data AD and the reference data RD is executed so as to be compared with the data representing the image feature in the corresponding region. That is, the matching execution unit 41 uses the unit reference data RDo selected by the unit reference data selection unit 42 to execute matching between the real scene data AD and the reference data RD. In this example, since the data representing the image feature is the data of the edge point image (feature point image), the arrangement of the feature point group in the divided area DA is the unit reference data selected for the divided area DA. Matching is executed by comparing with the arrangement of feature points in the corresponding region of RDo.

  In the present embodiment, the matching execution unit 41 derives the degree of coincidence between the real scene data AD and the unit reference data RDo (reference data RD) for each divided area DA, and in each shooting area WA of the divided area DA. Based on the weighting coefficient set according to the position and the degree of coincidence, the overall degree of coincidence in the entire imaging area WA is derived. That is, in this example, by performing a partial matching process for each divided area DA, an overall matching degree that is a matching degree in the entire photographing area WA between the actual scene data AD and the reference data RD is derived. It is configured.

More specifically, in the present embodiment, the degree of coincidence between the real scene data AD and the reference data RD in the N-th type division region (N is a natural number equal to or less than M, and M = 3 in this example) is K [ N], and the weight coefficient set in the N-th type division region is L [N], and the overall matching degree is derived as the sum of all N of products of K [N] and L [N]. . In this example, since M = 3, the total coincidence is expressed as follows.
Total coincidence = K [1] · L [1] + K [2] · L [2] + K [3] · L [3]

  The “matching degree” is an index indicating the degree of matching (similarity) between two data, and can be derived based on various known methods in the field of pattern matching. In this example, since each of the real scene data AD and the unit reference data RDo is an edge point image (feature point image), a feature point group that coincides with the arrangement of the feature point group included in the divided area DA. The degree of coincidence is derived according to the ratio included in the corresponding area of the unit reference data RDo.

  The weighting coefficient set for the divided area DA is set according to the position of the divided area DA in the imaging area WA. In this example, the sum of the weighting coefficients for each divided area DA is “1”. Is set. Specifically, in the present embodiment, the first-type divided area DA1 that is positioned on the upper side in the vertical direction in the imaging area WA and is set so that the proportion occupied by the sky is high includes features that are important for matching. Therefore, a smaller weighting coefficient is set than other divided areas DA. On the other hand, it is located on both sides in the left-and-right direction in the imaging area WA, and is set in the second type division area DA2 that is set so that the ratio occupied by the building is high. For the third type division area DA3 set to be higher, there is a high possibility that important features related to matching are included, and therefore a larger weighting coefficient is set than for the first type division area DA1.

  In the example shown in FIG. 3B, a building (not shown) located at a position away from the traveling direction is difficult to properly extract contour information by image recognition. In order to reduce the importance of the central part in the left-right direction in the imaging area WA, which can be included, the first type divided area DA1 is set to extend downward in the central part. Thus, in the present example, the second type divided area DA2 is an area divided on both the left and right sides in the imaging area WA. In the present invention, even if the areas are separated in the photographing area WA in this way, the same type (in this example, any of the first type, the second type, and the third type) of the divided areas DA. Assume that one or a plurality of regions constituting one form a single divided region DA as a whole.

  And the matching execution part 41 determines the success or failure of matching based on the whole matching degree derived | led-out as mentioned above. In the present embodiment, the matching execution unit 41 sequentially performs the overall matching degree derivation process for all the reference data RD acquired as matching candidates by the reference data acquisition unit 33. Then, the matching execution unit 41 performs matching when there is reference data RD having a derived overall matching degree equal to or higher than a predetermined determination threshold among the reference data RD acquired as matching candidates. If it is determined that the reference data RD is not found, it is determined that matching has failed. If there is only one reference data RD whose overall matching degree is equal to or higher than the determination threshold value, it is determined that matching with the reference data RD is successful, and a reference whose overall matching level is equal to or higher than the determination threshold value. When there are a plurality of use data RD, it is determined that the matching with the reference data RD having the highest overall matching degree is successful.

  Here, the configuration in which the matching execution unit 41 performs the process of deriving the overall matching degree in order for all the reference data RD extracted as matching candidates by the reference data acquisition unit 33 is shown as an example. A configuration may be adopted in which the matching process is terminated when matching with the reference data RD is successful when the reference data RD having a matching degree equal to or higher than the determination threshold value appears for the first time. In such a configuration, the reference data acquisition unit 33 selects one or a plurality of reference data RD in which a shooting position is associated with an area in the vicinity of the estimated host vehicle position as reference data RD for matching candidates. The acquisition of the reference data RD may be performed every time the overall matching degree derivation process is performed.

  By the way, the weighting coefficient set in the divided area DA can be set variably according to external factors such as weather conditions and time zone conditions. For example, when the current weather condition is “rain”, the value of L [2] / L [3] may be set larger than that when the current weather condition is “sunny”. . For example, when the current time zone condition is “night”, the value of L [2] / L [3] is set to be smaller than that when the current time zone condition is “daytime”. can do.

  Further, the weighting factor set for the divided area DA may be variably set according to the area attribute information RA. For example, when the regional attribute information RA is “mountainous part”, since the density of buildings is generally low, the value of L [2] / L [3] is the same as the regional attribute information RA is “urban part”. Or it can be set as the structure set small compared with the case where it is a "suburb".

2-10. Configuration of the Vehicle Position Determination Unit The vehicle position determination unit 44 is a functional unit that corrects the estimated vehicle position represented by the estimated position information EP acquired by the estimated position information acquisition unit 31 based on the shooting position of the actual landscape image SP. is there. As described above, the information on the shooting position of the actual landscape image SP specified by the shooting position specifying unit 40 is input to the own vehicle position determining unit 44. In this embodiment, the host vehicle position determination unit 44 corrects the estimated host vehicle position by replacing the shooting position input from the shooting position specifying unit 40 with the estimated host vehicle position. Note that the vehicle position determination unit 44 not only simply replaces the shooting position input from the shooting position specification unit 40 with the estimated vehicle position, but also the actual scenery by the shooting position specification unit 40 from the time of shooting the actual landscape image SP. It can also be set as the structure which adjusts the own vehicle position in consideration of the travel distance of the own vehicle until the imaging | photography position of image SP is specified. Information on the travel distance of the host vehicle can be acquired by the distance sensor 92 or the like.

3. Procedure of Operation Process Next, with reference to FIG. 5 and FIG. 6, a procedure of the shooting position specifying process executed in the shooting position specifying unit 2 according to the present embodiment, that is, a shooting position specifying method will be described. The procedure of the imaging position specifying process described below is executed by hardware or software (program) or both constituting each functional unit of the navigation device 1 (imaging position specifying unit 2). When each of the above function units is configured by a program, the arithmetic processing device included in the navigation device 1 operates as a computer that executes the program that configures each of the above function units. FIG. 6 is a flowchart showing the matching processing procedure of step # 06 in FIG.

3-1. Overall Procedure of Shooting Position Specifying Process As shown in FIG. 5, when the real scene data acquisition unit 32 acquires the real scene data AD (step # 01), the estimated position information acquisition unit 31 acquires the estimated position information EP. (Step # 02), the reference data acquisition unit 33 acquires one or a plurality of reference data RD to be matching candidates from the reference data storage unit 11 based on the estimated position information EP (step # 03).

  Next, the weather condition information acquisition unit 36 acquires the current weather condition information CW and the past weather condition information PW (step # 04, step # 05), and the matching process is executed (step # 06). This matching process will be described later with reference to the flowchart of FIG.

  As a result of the matching process, when it is determined that the matching between the real scene data AD and the reference data RD is successful (step # 07: Yes), the photographing associated with the reference data RD that has been successfully matched. The position information is acquired (step # 08), and the shooting position of the actual landscape image SP is specified based on the shooting position information (step # 09). In the present embodiment, the shooting position acquired in step # 08 is set as the shooting position of the real landscape image SP. On the other hand, when it is determined that the matching is not successful (step # 07: No), the process ends.

3-2. Matching Processing Procedure Next, the matching processing procedure in step # 06 in FIG. 5 will be described with reference to FIG. The matching execution unit 41 selects one reference data RD that is a target of the matching process from the matching candidate reference data RD acquired in Step # 03 of FIG. 5 (Step # 11). The unit reference data selection unit 42 then selects one of the plurality of (four in this example) unit reference data RDo constituting the reference data RD selected in step # 11, in the imaging area WA of the divided area DA. Based on the position, current weather condition information CW, past weather condition information PW, and in this example, the current time zone condition, unit reference data RDo is selected for each of the plurality of divided areas DA (step # 12). In the present embodiment, the unit reference data RDo is selected based on the selection conditions shown in the table of FIG. 4 for each of the divided areas DA included in the first area A1 and the divided areas DA included in the second area A2.

  Next, for each of the divided areas DA, the matching execution unit 41 includes data representing the image features in the divided area DA included in the unit reference data RDo selected for the divided area DA. The actual scene data AD and the reference data RD are matched so as to be compared with the data representing the image feature in the area corresponding to, and the overall matching degree is derived (step # 13). In the present embodiment, as described above, the degree of coincidence between the real scenery data AD and the reference data RD is derived for each divided area DA, and according to the position of each divided area DA in the shooting area WA. Based on the weighting coefficient and the degree of coincidence set in this way, the overall degree of coincidence in the entire imaging area WA is derived.

  When the reference data RD that has not been subjected to the process of deriving the overall matching degree is included in the reference data RD acquired as a matching candidate (Step # 14: Yes), the matching execution unit 41 The processing from step # 11 to step # 13 is repeatedly executed until there is no reference data RD for which the matching degree derivation processing is not performed.

  When all the reference data RD acquired as matching candidates are subjected to the overall matching degree derivation process (step # 14: No), a matching success / failure determination is performed (step # 15). In the present embodiment, as described above, in the reference data RD acquired as a matching candidate, when there is reference data RD having a derived overall matching degree equal to or higher than a predetermined determination threshold value, It is determined that matching is successful, and if there is no such reference data RD, it is determined that matching has failed. In this example, when there is only one reference data RD having an overall matching degree equal to or greater than the determination threshold, it is determined that the matching with the reference data RD is successful, and the overall matching degree is determined as described above. When there are a plurality of reference data RD equal to or greater than the threshold, it is determined that matching is successful with the reference data RD having the highest overall matching degree among them.

  Here, the acquisition step of the current weather condition information CW by the weather condition information acquisition unit 36 and the acquisition step of the past weather condition information PW by the weather condition information acquisition unit 36 are the acquisition steps of the matching candidate reference data RD. Although the case where it is executed between (Step # 03) and the step of matching processing (Step # 06) has been described as an example, acquisition of current weather condition information CW and past weather condition information PW by the weather condition information acquisition unit 36 The step may be executed at any timing as long as it is before the unit reference data RDo selection step (step # 12). Further, the step of acquiring the current weather condition information CW and the past weather condition information PW by the weather condition information acquisition unit 36 is not performed every time the actual landscape data AD is acquired, and the unit reference data RDo is obtained using the previously acquired information. It can also be set as the structure which performs a selection process.

4). Other Embodiments Finally, other embodiments according to the present invention will be described. Note that the features disclosed in each of the following embodiments can be used only in that embodiment, and can be applied to other embodiments as long as no contradiction arises.

(1) In the above embodiment, the configuration in which the weighting factor is individually set for each divided area DA has been described as an example. However, the embodiment of the present invention is not limited to this, and the same weighting factor is set for all the divided areas DA included in the first area A1, and all the areas included in the second area A2 are set. The same weighting factor may be set for the divided area DA.

(2) In the above embodiment, the matching execution unit 41 derives the degree of coincidence between the real scene data AD and the reference data RD for each divided area DA, and positions of the divided areas DA in the respective shooting areas WA. The configuration for deriving the overall coincidence in the entire imaging area WA based on the weighting factor and the coincidence set according to the above has been described as an example. However, the embodiment of the present invention is not limited to this, and the data of the areas corresponding to the divided areas DA to be selected for each of the plurality of selected unit reference data RDo are collected, and the entire imaging area WA is collected. It is also possible to generate a set unit reference data corresponding to the above data and perform a matching process on the actual scene data AD and the set unit reference data for the entire shooting area WA to derive an overall matching degree. . In such a configuration, a configuration in which a weighting factor is not set or a configuration in which a weighting factor is set for each region set independently of the divided region DA can be adopted.

(3) In the above-described embodiment, even when there are a plurality of reference data RD having an overall matching degree equal to or greater than the determination threshold, one reference data RD is selected from among the reference data RD, and imaging associated with the reference data RD is performed. The configuration in which the position is specified as the shooting position of the real landscape image SP has been described as an example. However, the embodiment of the present invention is not limited to this, and when there are a plurality of reference data RD whose overall coincidence is equal to or higher than the determination threshold, all of them are selected based on the magnitude of the overall coincidence. A configuration may be adopted in which the shooting position of the real landscape image SP is specified based on the shooting position associated with a plurality of reference data RD that are a part. In such a configuration, for example, an average value of a plurality of shooting positions can be set as the shooting position of the real landscape image SP.

(4) In the above embodiment, the weather condition (current weather condition) represented by the current weather condition information CW and the weather condition (past weather) represented by the past weather condition information PW for the divided area DA included in the second area A2. The configuration in which the unit reference data RDo is selected according to the selection condition shown in FIG. However, the embodiment of the present invention is not limited to this, and when the current weather condition information CW and the past weather condition information PW represent the same weather condition, the weather condition represented by any of the weather condition information Considering that the same unit reference data RDo is selected even if selected, the weather condition is selected based on the following selection criteria for the divided area DA included in the second area A2, and A corresponding unit reference data RDo may be selected. In other words, when the current weather condition is “rain”, the current weather condition is selected, and in other cases (that is, when the current weather condition is “clear”), the past weather condition is selected. If the past weather condition is “sunny”, the current weather condition is selected. In other cases (that is, if the past weather condition is “rain”), the past weather condition is selected. Even in such a configuration, the selection condition of the weather condition takes into account the change from the past weather condition to the current weather condition (direction of change). The unit reference data RDo is selected based on the change to the condition.

(5) In the above-described embodiment, the configuration divided into two of “sunny” and “rain” with respect to the weather conditions has been described as an example. However, the embodiment of the present invention is not limited to this, and “sunny” is classified into a plurality of states based on the amount of clouds, or “rain” is classified into a plurality of states based on the precipitation. It is also possible. For example, the weather condition can be divided into three sections, “sunny”, “weak rain”, and “strong rain”. In this case, it is preferable to prepare data corresponding to each of “weak rain” and “strong rain” in addition to “sunny” as the unit reference data RDo. At this time, for example, regarding the selection condition of the unit reference data RDo shown in FIG. 4, “strong rain” is the same as “rain” in the above embodiment, and “weak rain” is “clear” in the above embodiment. The same can be said. As a result, if the past weather condition is "weak rain", refer to the unit considering that it does not take a long time for the wet road surface to dry or that the road surface condition is "clear". Selection of data RDo becomes possible.

(6) In the above-described embodiment, the configuration divided into two of “sunny” and “rain” regarding the weather condition has been described as an example. However, the embodiment of the present invention is not limited to this, and as a weather condition with precipitation, in addition to `` rain '', a configuration in which `` snow '' is further set, and as a weather condition without precipitation, In addition to “sunny”, “cloudy” may be set. In this case, it is preferable to prepare unit reference data RDo corresponding to each of all weather conditions. At this time, for example, regarding the selection condition of the unit reference data RDo shown in FIG. 4, “snow” is the same as “rain” in the above embodiment, and “cloudy” is the same as “sunny” in the above embodiment. be able to. In such a configuration, as in the above embodiment, the weather condition information acquisition unit 36 may be configured to acquire information on the weather conditions previously classified into any of the conditions, or the weather condition information acquisition unit 36 may classify the acquired weather conditions.

(7) In the above-described embodiment, the configuration divided into two of “daytime” and “night” regarding the time zone condition has been described as an example. However, the embodiment of the present invention is not limited to this, and in addition to “daytime” and “night”, “early morning” and “evening” are set as time zone conditions. You can also. Such a division can be set based on, for example, altitude information with respect to the horizon of the sun and information on sunrise and sunset times. In such a configuration, unit reference data RDo corresponding to each of a plurality of (for example, four) time zone conditions is prepared, and unit reference data RDo corresponding to the current time zone condition is selected. be able to.

(8) In the above embodiment, the unit reference data RDo associated with the time zone condition in addition to the weather condition is described as an example of the unit reference data RDo corresponding to the same shooting position. . However, the embodiment of the present invention is not limited to this, and the unit reference data RDo associated only with the weather conditions may be prepared. In such a configuration, for example, it is preferable that the unit reference data RDo is generated based on a landscape image under the condition that the time zone condition is “daytime”.
In addition to the weather conditions, or in addition to the weather conditions and the time zone conditions, unit reference data RDo is also prepared in association with other conditions. The unit reference data RDo is also based on the other conditions. It is also possible to adopt a configuration in which is selected. For example, the shooting conditions for landscape images are divided into “forward light state” and “backlight state”. If the current shooting condition is “front light state”, the unit reference data RDo associated with “front light state” is used. If the current shooting condition is “backlight state”, the unit reference data RDo associated with the “backlight state” can be acquired. It is also possible to adopt a configuration that takes into account seasonal conditions.

(9) The shapes of the first area A1 and the second area A2 and the shapes of the divided areas DA shown in the above embodiment are examples, and these shapes can be changed as appropriate. For example, the shape of the first region A1 and the second region A2 can be a rectangular shape, or at least a part of the divided region DA can be a rectangular region.

(10) In the above embodiment, the configuration in which the two divided areas DA (first type divided area DA1 and second type divided area DA2) are set in the first area A1 has been described as an example. The number (division number) of the divided areas DA set in the area A1 can be changed as appropriate. For example, one or three divided areas DA can be set in the first area A1. Similarly, in the above-described embodiment, the configuration in which one divided area DA (third type divided area DA3) is set in the second area A2 has been described as an example, but is set in the second area A2. The number of divided areas DA (number of divisions) can be changed as appropriate. For example, two or three divided areas DA can be set in the second area A2.

(11) In the above embodiment, the regional attribute information RA has been described as an example of a configuration including “urban area”, “suburb”, and “mountainous area” as attribute items. However, the attribute item included in the regional attribute information The type and number of these can be changed as appropriate. For example, the regional attribute information has a configuration including only “urban area” and “suburb” as attribute items, or a structure including “urban area” and “suburb” further subdivided (for example, “urban area” is an average of buildings) It is also suitable as a subdivision of “first urban area (high-rise urban area)” and “second urban area (urban area)” in accordance with trends such as height and density. In the above embodiment, the configuration in which the vertical width of the second type division area DA2 is adjusted based on such area attribute information RA is described as an example. However, the second type based on such area attribute information RA is described. A configuration in which the adjustment of the divided area DA2 is not performed may be employed.

(12) In the above embodiment, the road map data stored in the map information storage unit 10 includes road width information RW that is information related to the road width of the road, and the road width information acquisition unit 34 includes the road map data. The configuration for acquiring the road width information RW with reference to FIG. However, the embodiment of the present invention is not limited to this. When the road map data included in the map information storage unit 10 does not include road width information RW of the road around the current position, road width information acquisition is performed. The unit 34 may be configured to acquire the road width information RW as an estimated value based on road type information (type of highway, toll road, national road, prefectural road, etc.) included in the road map data. In such a configuration, for example, the road width information RW can be estimated based on the general relationship between the road type and the road width that the highway is generally wider than the prefectural road.

(13) In the above embodiment, the configuration in which the second region A2 is variably set by the divided region adjustment unit 43 according to the road width information RW acquired by the road width information acquisition unit 34 has been described as an example. However, the embodiment of the present invention is not limited to this, and the second region A2 is variably set by the divided region adjustment unit 43 according to the region attribute information RA acquired by the region attribute information acquisition unit 35. It is good also as a structure. In such a configuration, based on the general tendency that the road width increases in the order of “mountainous part”, “suburb”, and “urban part”, the lateral width of the second area A2 (in the above embodiment, the third-type divided area DA3 The width can be set to be variable. The second area A2 may be a fixed area without being variably set.

(14) In the above embodiment, the first area A1 and the second area A2 that are set so as not to overlap each other in the shooting area WA are set so that the added area is equal to the shooting area WA. The configuration has been described as an example. However, the embodiment of the present invention is not limited to this. Based on conditions different from the selection conditions of the unit reference data RDo set in each of the first area A1 and the second area A2, the unit reference data RDo If there is a divided area DA that is preferably selected, the third area A3 including the divided area DA may be set in addition to the first area A1 and the second area A2. In such a configuration, the area obtained by adding the first area A1, the second area A2, and the third area A3 can be set to be equal to the imaging area WA. For the divided area DA included in the third area A3, for example, the unit reference data RDo corresponding to the weather condition represented by the past weather condition information PW can be selected.

(15) The assignment of the function units described in the above embodiments is merely an example, and a plurality of function units can be combined or one function unit can be further divided.

(16) In the above embodiment, the configuration in which the camera 90 is a front camera that captures a landscape in front of the vehicle has been described as an example. However, the embodiment of the present invention is not limited to this, and the camera 90 may be a camera that captures a landscape behind or side of the vehicle, or obliquely forward or obliquely behind.

(17) In the above embodiment, the case where the imaging position specifying system according to the present invention is applied to the in-vehicle navigation device 1 has been described as an example. However, the embodiment of the present invention is not limited to this, and the imaging position specifying system according to the present invention can also be applied to devices (travel control devices, etc.) other than the navigation device provided in the vehicle. .

(18) In the above embodiment, the configuration in which the photographing position specifying unit 2, the reference data storage unit 11, and the map information storage unit 10 are provided in the vehicle is shown as an example. The functional unit, the reference data storage unit 11 or the map information storage unit 10 may be provided outside the vehicle, and information and signals may be transmitted and received via a communication network such as the Internet.

(19) In the above embodiment, the case where the vehicle corresponds to the “moving body” according to the present invention has been described as an example. However, the embodiment of the present invention is not limited to this, and the present invention can be applied to a moving body other than a vehicle. For example, the present invention can be applied to a moving body (robot or the like) that travels by a mechanism other than wheels (for example, a crawler mechanism or a leg mechanism). In addition, the mobile body targeted by the present invention is not limited to a self-propelled mobile body, and the present invention can also be applied to a mobile body that moves by external force, such as a mobile terminal (for example, a mobile phone with a camera). It is.

(20) In the above embodiment, the configuration in which the estimated position acquired by the estimated position information acquisition unit 31 is corrected based on the shooting position specified by the shooting position specifying unit 2 has been described as an example. When the range is limited, the shooting position specifying unit 2 does not include the estimated position information acquisition unit 31, and the reference data acquisition unit 33 stores all the reference data RD stored in the reference data storage unit 11. Can be selected as reference data RD for matching candidates. In the above-described embodiment, the configuration in which the position of the moving body is determined based on the shooting position of the specified real scenery image SP is shown as an example. However, the shooting position of the specified real scenery image SP is used for other purposes. It can also be used.

(21) In the above embodiment, the configuration in which the camera 90 is mounted on a moving body has been described as an example. However, the embodiment of the present invention is not limited to this, and a configuration in which the camera 90 is not mounted on a moving body, for example, a configuration in which the camera 90 is a single device may be employed. In this case, the position of the camera 90 can be specified based on the shooting position of the specified real scenery image SP.

(22) In the above embodiment, the configuration in which the reference data storage unit 11 stores a plurality of reference data RD has been described as an example. However, the reference data storage unit 11 stores a single reference data RD. It is also possible to do.

(23) Regarding other configurations as well, the embodiments disclosed herein are illustrative in all respects, and the embodiments of the present invention are not limited thereto. That is, as long as the configuration described in the claims of the present application and a configuration equivalent thereto are provided, a configuration obtained by appropriately modifying a part of the configuration not described in the claims is naturally also included in the present invention. Belongs to the technical scope.

  INDUSTRIAL APPLICABILITY The present invention can be suitably used for a shooting position specifying system, a shooting position specifying program, and a shooting position specifying method for specifying a shooting position of a real landscape image that is a landscape image shot by a camera mounted on a moving body. it can.

11: Reference data storage unit 31: Estimated position information acquisition unit 34: Road width information acquisition unit 35: Area attribute information acquisition unit 40: Shooting position specifying unit 42: Unit reference data selection unit 90: Camera A1: First area A2: No. Two areas AD: actual landscape data CW: current weather condition information DA: divided area EP: estimated position information PW: past weather condition information RD: reference data RDo: unit reference data RW: road width information RA: area attribute information SP: actual Landscape image WA: Shooting area

Claims (9)

A shooting position specifying system for specifying a shooting position of a real landscape image that is a landscape image shot by a camera,
Data generated based on each of a plurality of landscape images taken under different weather conditions from the same shooting position, and corresponding to the weather conditions at the time when the source landscape image was taken A reference data storage unit that stores reference data configured by a group of unit reference data in association with a shooting position of a landscape image that is a generation source of the unit reference data that constitutes the reference data;
Photographing that matches the actual scenery data generated based on the actual scenery image and the reference data, and identifies the photographing position of the actual scenery image based on the photographing position associated with the reference data that has been successfully matched A position identification part;
When matching the actual landscape data and the reference data by the shooting position specifying unit, for each of a plurality of divided areas obtained by dividing the shooting area of the real landscape image, the position of the divided area in the shooting area Based on current weather condition information representing current weather conditions and past weather condition information representing past weather conditions more than a predetermined time before the present, among the unit reference data group constituting the reference data And a unit reference data selection unit that selects the unit reference data corresponding to each divided area. A first area and a second area that do not overlap each other are set in the imaging area,
The unit reference data selection unit selects the unit reference data corresponding to the weather condition represented by the current weather condition information for the divided area included in the first area, and the divided area included in the second area For the region, based on the change from the weather condition represented by the past weather condition information to the weather condition represented by the current weather condition information, the weather condition represented by the current weather condition information, and the weather represented by the past weather condition information The imaging position specifying system according to claim 1, wherein any one of conditions is selected, and the unit reference data corresponding to the selected weather condition is selected. The actual landscape image is a landscape image including a road surface,
The imaging position specifying system according to claim 2, wherein the second area is set so as to include at least a road surface. The camera is mounted on a moving body,
An estimated position information acquisition unit that acquires estimated position information representing an estimated position of the mobile body, and a road width information acquisition unit that acquires road width information that is information related to the road width of the road around the position indicated by the estimated position information, In addition,
The imaging position specifying system according to claim 3, wherein the second region is variably set according to the road width information acquired by the road width information acquisition unit. The camera is mounted on a moving body,
An estimated position information acquisition unit that acquires estimated position information representing an estimated position of the mobile object, and a region that acquires area attribute information including at least a city area and a suburb as attribute items about the position indicated by the estimated position information An attribute information acquisition unit,
The imaging position specifying system according to claim 3, wherein the second area is variably set according to the area attribute information acquired by the area attribute information acquisition unit. The unit reference data selection unit, for the divided area included in the second area,
In the case where the past weather condition information represents a weather condition with precipitation and the current weather condition information represents a weather condition without precipitation, the weather condition represented by the past weather condition information is selected,
If the past weather condition information represents a weather condition without precipitation and the current weather condition information represents a weather condition with precipitation, select the weather condition represented by the current weather condition information,
The imaging position specifying system according to any one of claims 2 to 5, wherein the unit reference data corresponding to the selected weather condition is selected.   The shooting position specifying unit derives a degree of coincidence between the real scene data and the unit reference data for each divided area, and a weighting factor set according to the position of each of the divided areas in the shooting area The imaging position specifying system according to any one of claims 1 to 6, wherein an overall matching degree in the entire imaging region is derived based on the matching degree and the success or failure of the matching is determined based on the overall matching degree. . A shooting position specifying program for specifying a shooting position of a real landscape image that is a landscape image shot by a camera,
Data generated based on each of a plurality of landscape images taken under different weather conditions from the same shooting position, and corresponding to the weather conditions at the time when the source landscape image was taken A reference data acquisition function for acquiring reference data composed of a group of unit reference data;
An actual scenery data acquisition function for acquiring actual scenery data generated based on the actual scenery image;
The actual scenery data and the reference data are matched, and information on the shooting position of the scenery image from which the unit reference data constituting the reference data that has been successfully matched is acquired, A shooting position specifying function for specifying a shooting position of the real scenery image from which the real scenery data is generated based on information;
When matching the actual scenery data and the reference data by the photographing position specifying function, for each of a plurality of divided areas obtained by dividing the photographing area of the real scenery image, the position of the divided area in the photographing area Based on current weather condition information representing current weather conditions and past weather condition information representing past weather conditions more than a predetermined time before the present, among the unit reference data group constituting the reference data A unit reference data selection function for selecting the unit reference data corresponding to each divided area;
Shooting position specifying program to make computer realize. A shooting position specifying method for specifying a shooting position of a real landscape image that is a landscape image shot by a camera,
An actual landscape data acquisition step of acquiring actual landscape data generated based on the actual landscape image;
Data generated based on each of a plurality of landscape images taken under different weather conditions from the same shooting position, and corresponding to the weather conditions at the time when the source landscape image was taken A reference data acquisition step for acquiring reference data composed of a group of unit reference data;
For each of a plurality of divided areas obtained by dividing the shooting area of the real landscape image, the position of the divided area in the shooting area, current weather condition information indicating current weather conditions, and a predetermined time or more before the current time A unit reference data selection step for selecting the unit reference data corresponding to each divided region from the group of unit reference data constituting the reference data based on past weather condition information representing past weather conditions of When,
Using the unit reference data selected in the unit reference data selection step, the real scene data and the reference data are matched, and the unit reference data constituting the reference data that has been successfully matched is generated. A shooting position specifying step that acquires information of a shooting position of the original landscape image and specifies a shooting position of the real landscape image from which the real landscape data is generated based on the information of the shooting position; Method.

Images