JP2012128669A - Vehicle color determining device, computer program and vehicle color determining method - Google Patents

Abstract

A vehicle color determination apparatus, a computer program, and a vehicle color determination method that can determine the color of a vehicle without performing advanced image processing.
A plate recognition unit recognizes a license plate based on a vehicle image detected on a captured image. The plate position calculation unit 15 calculates the plate position. The classification number detection unit 16 detects the classification number on the license plate based on the plate image output from the plate recognition unit 14. The vehicle color determination area specifying unit 17 specifies the shape of the vehicle color determination area and the position with respect to the plate based on the plate position calculated by the plate position calculation unit 15 and the classification number detected by the classification number detection unit 16. The vehicle color determination unit 18 determines the color of the vehicle based on the pixel value of the vehicle color determination region specified by the vehicle color determination region specification unit 17.
[Selection] Figure 1

Description

  The present invention relates to a vehicle color determination device that determines a color of a vehicle based on a captured image obtained by imaging a vehicle traveling on a road, a computer program for realizing the vehicle color determination device, and a vehicle color determination method. .

  In recent years, in the field of traffic monitoring and the like, a technique for determining a color (vehicle color) of a vehicle has been developed as one technique for extracting a specific vehicle. For example, a characteristic part of a vehicle (for example, a license plate, a side mirror, a front mirror, a headlamp, etc.) on a color image acquired by a camera is detected, and a specific part of the vehicle (for example, , A bonnet area, a roof area, a vehicle side and door area, a trunk area, etc.), and a specific part extraction program that can accurately extract a part where the color of the vehicle should be discriminated regardless of environmental conditions is disclosed (See Patent Document 1).

JP 2006-164193 A

  However, in the technique of Patent Document 1, for example, when a bonnet region is extracted as a part where the color of the vehicle is to be determined, the positions of the two headlamps and the two headlamps are determined based on the positional information of the headlamp and the side mirror. The position of the side mirror is regarded as a vertex of the rectangle, and the intersection of the diagonal lines of the rectangle is detected. Then, the bonnet region must be specified by extracting a region where the attribute information (for example, the photometric quantity or the color attribute) does not change from the detected intersection. For this reason, advanced image processing is required to exclude an unnecessary area when determining the color of the vehicle, and a technique for determining the color of the vehicle with simple image processing has been desired.

  The present invention has been made in view of such circumstances, and is a vehicle color determination device capable of determining the color of a vehicle without performing advanced image processing, and a computer program for realizing the vehicle color determination device It is another object of the present invention to provide a vehicle color determination method.

  A vehicle color determination device according to a first aspect of the present invention is a vehicle color determination device that determines a color of a vehicle based on a captured image obtained by capturing an image of a vehicle traveling on a road with an imaging unit. A classification number detecting means for detecting a classification number on the number sign, and a vehicle for specifying the shape of the vehicle color determination area on the captured image and the position with respect to the number mark based on the classification number detected by the classification number detection means A color determination area specifying unit; and a vehicle color determination unit that determines a color of the vehicle based on a color component of each pixel of the vehicle color determination area specified by the vehicle color determination area specifying unit.

  According to a second aspect of the present invention, in the first aspect, the vehicle color determination area specifying unit corresponds to a small vehicle or a normal vehicle whose classification number detected by the classification number detection unit is equal to or smaller than a predetermined passenger capacity. In this case, a rectangular area located above the number mark on the captured image, where the vertical side of the rectangular shape forms a predetermined angle with respect to the horizontal axis on the captured image. It is configured to be specified as a color determination area.

  According to a third aspect of the present invention, there is provided the vehicle color determination apparatus according to the second aspect, wherein the vehicle color determination region specifying means has a classification number detected by the classification number detection means that is a small vehicle for carrying a person, a predetermined passenger capacity or less. In the case of a normal vehicle or a small vehicle for special use, the vehicle is characterized in that a vertically long rectangular region is specified as a vehicle color determination region.

  The vehicle color determination device according to a fourth aspect of the present invention is the vehicle color determination device according to the second aspect, wherein the vehicle color determination region specifying means is a horizontally long rectangle when the classification number detected by the classification number detection means is a small automobile for transporting cargo. The region of the shape is configured to be specified as the vehicle color determination region.

  In the vehicle color determination device according to a fifth aspect of the present invention, in any one of the second to fourth aspects, the predetermined angle is an angle formed by a horizontal axis on the captured image and a road extending direction. It is characterized by.

  The vehicle color determination device according to a sixth aspect of the present invention is the vehicle color determination device according to any one of the second to fifth aspects, wherein the vehicle color determination region specifying means has a classification number detected by the classification number detection means as a small vehicle or a predetermined vehicle. When the vehicle corresponds to an ordinary vehicle having a passenger capacity or less, a vehicle color determination region is specified at a position along the road extending direction from the number sign on the captured image.

  The vehicle color determination device according to a seventh aspect of the present invention is the vehicle color determination area according to the first aspect, wherein the vehicle color determination region specifying means is configured to perform the imaging when the classification number detected by the classification number detection means corresponds to a normal vehicle for transporting cargo. A substantially rectangular shape above the number mark on the image, and a vertical side of the rectangular shape forms a predetermined angle with respect to a horizontal axis on the captured image, and a lower side of the substantially rectangular shape An area where the center is cut out in a concave shape is specified as a vehicle color determination area.

  The vehicle color determination device according to an eighth aspect of the present invention is the vehicle color determination area specifying means according to the first aspect, wherein the vehicle color determination area specifying means corresponds to a normal automobile having a classification number detected by the classification number detection means of a predetermined number of passengers or more. A substantially rectangular shape above the number mark on the picked-up image, and the vertical side of the rectangle forms a predetermined angle with respect to the horizontal axis on the picked-up image. A region having a convex region at the center in the side is specified as a vehicle color determination region.

  The vehicle color determination apparatus according to a ninth aspect of the present invention is the vehicle color determination area specifying means according to the first aspect, wherein the vehicle color determination area specifying means corresponds to a special purpose ordinary vehicle or a large special vehicle when the classification number detected by the classification number detection means is Two rectangular shapes separated from each other by an appropriate length on the captured image with the upper position of the number mark interposed therebetween, and the vertical sides of the rectangular shape form a predetermined angle with respect to the horizontal axis on the captured image. An area to be formed is specified as a vehicle color determination area.

  In the vehicle color determination device according to a tenth aspect of the present invention, in any one of the seventh to ninth aspects, the predetermined angle is 80 ° to 100 ° with respect to a horizontal axis direction on the captured image. It is characterized by.

  According to an eleventh aspect of the present invention, in the vehicle color determination device according to any one of the seventh to tenth aspects, the vehicle color determination area specifying means is configured such that the classification number detected by the classification number detection means is for freight transportation. When it corresponds to a normal car, a normal car with a predetermined passenger capacity or more, a special car for a special purpose, or a large special car, an angle of 80 ° to 100 ° is formed with respect to the horizontal axis direction on the captured image from the number sign. The vehicle color determination area is specified at a position along the direction.

  A vehicle color determination device according to a twelfth aspect of the present invention is the vehicle color determination apparatus according to any one of the second aspect to the eleventh aspect, wherein the number mark determination means for determining the size of the number mark of the vehicle and the number determined by the number mark determination means And a special-purpose vehicle determination unit for determining whether the vehicle is a special-purpose ordinary vehicle or a special-purpose small vehicle based on the size of the sign and the classification number detected by the classification number detection unit.

  A computer program according to a thirteenth invention is a computer program for causing a computer to execute a step of determining a color of a vehicle based on a captured image obtained by imaging a vehicle traveling on a road by an imaging unit. Detecting a classification number on a vehicle number sign based on the captured image; identifying a shape of a vehicle color determination area on the captured image and a position with respect to the number sign based on the detected classification number; And determining the color of the vehicle based on the color component of each pixel in the specified vehicle color determination area.

  A vehicle color determination method according to a fourteenth aspect of the present invention is the vehicle color determination method by the vehicle color determination device that determines the color of the vehicle based on the captured image obtained by imaging the vehicle traveling on the road by the imaging unit. Detecting a classification number on a vehicle number sign based on the vehicle, identifying a shape of a vehicle color determination area on the captured image and a position with respect to the number sign based on the detected classification number, and specifying Determining the color of the vehicle based on the color component of each pixel in the vehicle color determination area.

  In the first invention, the thirteenth invention, and the fourteenth invention, the classification number on the vehicle number mark (number plate) is detected based on the captured image. The classification number is displayed on the right side of the place name of the license plate, and classifies the size (for example, small, normal, large, etc.) and the use (for example, passenger, cargo, special purpose, large special, etc.) of the automobile. It is a number. The number sign is, for example, a vehicle is detected based on a difference between an image when no vehicle is present and an image when the vehicle is present, and a rectangular number sign is cut out based on the detected vehicle image. It can be detected by performing processing. The classification number can be detected by a general character recognition process.

  Based on the detected classification number, the shape of the vehicle color determination area on the captured image and the position with respect to the number mark are specified. For each classification number, the shape of the vehicle color determination area on the captured image and the position (for example, distance, direction, etc.) with respect to the number mark are determined in advance, and the vehicle color determination area corresponding to the detected classification number is selected. be able to. And the color of a vehicle is determined based on the color component of each pixel of the specified vehicle color determination area. For example, the color component of each pixel in the vehicle color determination area is converted from RGB to HSI (H: hue, S: saturation, I: luminance), and the color of the vehicle is determined based on the hue H and / or luminance I. be able to.

  Since the position and shape of the vehicle color determination area are specified based on the detected classification number, advanced image processing for determining the determination area for determining the color of the vehicle based on the attribute information is performed as in the past. The color of the vehicle can be determined. Also, since the vehicle color determination area is specified based on the classification number classified according to the size and application of the vehicle, the vehicle color determination area can be specified according to various types of vehicles, and the vehicle color determination The vehicle color can be accurately determined without including unnecessary portions.

  In the second invention, when the detected classification number corresponds to a small car or a normal car having a predetermined passenger capacity (for example, 10 persons) or less, it is located above the number mark (number plate) on the captured image. A rectangular area in which the vertical side of the rectangle forms a predetermined angle with respect to the horizontal axis on the captured image is specified as a vehicle color determination area. The rectangular area on the captured image is, for example, a parallelogram. In the case of a small vehicle or a regular vehicle with a predetermined number of passengers or less, a rectangular (square or rectangular) bonnet area exists above the license plate, so the detected classification number is a small vehicle or a regular vehicle with a predetermined number of passengers or less. In the case of corresponding to, a rectangular area above the number mark (number plate) is specified as the vehicle color determination area. In addition, an area in which a rectangular side in the vertical direction forms a predetermined angle with respect to the horizontal axis on the captured image is defined as a vehicle color determination area when a vehicle is imaged obliquely from the front. Is inclined from the vertical direction on the captured image by an angle corresponding to the installation state of the imaging unit. Since the vehicle color determination area is specified based on the classification number, the color of the vehicle can be determined without performing advanced image processing.

  In the third invention, when the detected classification number is a small vehicle for carrying a person, a normal vehicle having a predetermined passenger capacity (for example, 10 persons) or less, or a small vehicle for a special purpose, a vertically long rectangular shape. Is specified as the vehicle color determination area. In the case of a small car for transporting people, a normal car with a specified passenger capacity or less, or a small car for special purposes, a bonnet area with a vertically long rectangular shape (vertical rectangular shape) exists above the license plate. If the classification number is a small car for transporting people, a normal car with a certain number of passengers or less, or a small car for special purposes, a vertically long rectangular area above the number sign (number plate) The vehicle color determination area is specified. Since the vehicle color determination area is specified based on the classification number, the color of the vehicle can be determined without performing advanced image processing.

  In the fourth invention, when the detected classification number is a small automobile for transporting cargo, a horizontally long rectangular area is specified as a vehicle color determination area. In the case of small vehicles for cargo transportation, there is a horizontally long bonnet area above the license plate, so if the detected classification number is small automatic for cargo transportation, A horizontally long rectangular area above the number mark (number plate) is specified as a vehicle color determination area. Since the vehicle color determination area is specified based on the classification number, the color of the vehicle can be determined without performing advanced image processing.

  In the fifth invention, the predetermined angle is an angle formed by the horizontal axis on the captured image and the extending direction of the road. Depending on the installation conditions of the imaging unit, when the vehicle is imaged from the front, the vehicle color determination area on the captured image is not in the vertical direction but inclined in the direction of the road extension with respect to the horizontal axis. By tilting the vertical side of the region with respect to the horizontal axis only in the direction in which the road extends, it is possible to prevent unnecessary portions from being included in the vehicle color determination region.

  In the sixth aspect of the invention, when the detected classification number corresponds to a small vehicle or a normal vehicle having a predetermined passenger capacity or less, a vehicle color determination region is provided at a position along the road extending direction from the number sign on the captured image. Identify. Depending on the installation conditions of the imaging unit, if the vehicle is imaged obliquely from the front, the vehicle color determination area on the captured image is not in the vertical direction, but in the direction of the road extension with respect to the horizontal axis. By specifying the vehicle color case area at a position along the road extending direction, it is possible to prevent unnecessary portions from being included in the vehicle color determination area.

  In the seventh invention, when the detected classification number corresponds to a normal vehicle for transporting cargo, it is a substantially rectangular shape at a position above the number sign on the captured image, and the vertical direction of the rectangular shape is A region in which the side forms a predetermined angle with respect to the horizontal axis on the captured image and the lower center of the substantially rectangular shape is cut out in a concave shape is specified as the vehicle color determination region. In the case of ordinary vehicles for cargo transportation, there is a rectangular front grill above the license plate, and there are bonnet areas on both sides and the upper side of the front grill, so the detected classification number is the ordinary vehicle for cargo transportation. Is a substantially rectangular shape above the number mark (number plate), and the vertical side of the rectangular shape forms a predetermined angle with respect to the horizontal axis on the captured image, A region obtained by cutting out the lower center of the substantially rectangular shape into a concave shape is specified as the vehicle color determination region. In addition, an area in which a rectangular side in the vertical direction forms a predetermined angle with respect to the horizontal axis on the captured image is defined as a vehicle color determination area when a vehicle is imaged obliquely from the front. Is inclined from the vertical direction on the captured image by an angle corresponding to the installation state of the imaging unit. Since the vehicle color determination area is specified based on the classification number, the color of the vehicle can be determined without performing advanced image processing.

  In the eighth invention, when the detected classification number corresponds to an ordinary car having a predetermined number of passengers (for example, 11 persons) or more, it is a substantially rectangular shape located above the number sign on the captured image, The rectangular vertical side forms a predetermined angle with respect to the horizontal axis on the captured image, and a region having a convex region at the lower center of the substantially rectangular shape is specified as the vehicle color determination region. In the case of an ordinary vehicle having a predetermined number of passengers or more, there is a rectangular bonnet region above the license plate, and a rectangular bonnet region with a short height is located above the two headlights. Therefore, when the detected classification number corresponds to an ordinary vehicle having a predetermined number of passengers or more, it has a substantially rectangular shape above the number mark (number plate) and has a vertical side of the rectangular shape. Forms a predetermined angle with respect to the horizontal axis on the captured image, and specifies a region having a convex region at the lower center of the substantially rectangular shape as a vehicle color determination region. In addition, an area in which a rectangular side in the vertical direction forms a predetermined angle with respect to the horizontal axis on the captured image is defined as a vehicle color determination area when a vehicle is imaged obliquely from the front. Is inclined from the vertical direction on the captured image by an angle corresponding to the installation state of the imaging unit. Since the vehicle color determination area is specified based on the classification number, the color of the vehicle can be determined without performing advanced image processing.

  In the ninth invention, when the detected classification number corresponds to an ordinary vehicle or a large special vehicle for special purposes, it is two rectangular shapes separated by an appropriate length with the upper position of the number sign in between on the captured image. Thus, an area in which the rectangular side in the vertical direction forms a predetermined angle with respect to the horizontal axis on the captured image is specified as the vehicle color determination area. In the case of special-purpose ordinary vehicles or large special vehicles, there is a rectangular bonnet area on both sides of the front of the vehicle, so when the detected classification number corresponds to special-purpose ordinary vehicles or large special vehicles Are two rectangular shapes separated by an appropriate length with the upper position of the number mark (number plate) in between, and the vertical sides of the rectangular shape form a predetermined angle with respect to the horizontal axis on the captured image. The area is specified as the vehicle color determination area. In addition, an area in which a rectangular side in the vertical direction forms a predetermined angle with respect to the horizontal axis on the captured image is defined as a vehicle color determination area when a vehicle is imaged obliquely from the front. Is inclined from the vertical direction on the captured image by an angle corresponding to the installation state of the imaging unit. Since the vehicle color determination area is specified based on the classification number, the color of the vehicle can be determined without performing advanced image processing.

  In the tenth invention, the predetermined angle is 80 ° to 100 ° with respect to the horizontal axis direction on the captured image. In the case of ordinary cars for transporting cargo, ordinary cars with a certain passenger capacity (for example, 11 people) or more, special purpose ordinary cars or large special cars, the front of the vehicle is almost vertical compared to small cars And a relatively large area. For this reason, depending on the installation conditions of the imaging unit, even when the vehicle is imaged from the front rather than in front of the vehicle, the vehicle color determination area is substantially vertical on the captured image. By setting the angle 80 ° to 100 ° with respect to the horizontal axis, it is possible to prevent unnecessary portions from being included in the vehicle color determination region.

  In the eleventh aspect of the invention, when the detected classification number corresponds to an ordinary vehicle for transporting cargo, an ordinary vehicle having a predetermined passenger capacity or more, an ordinary vehicle for special purposes, or a large special vehicle, the number sign indicates The vehicle color determination area is specified at a position along a direction that forms an angle of 80 ° to 100 ° with respect to the horizontal axis direction of the vehicle. In the case of ordinary cars for transporting cargo, ordinary cars with a certain passenger capacity (for example, 11 people) or more, special purpose ordinary cars or large special cars, the front of the vehicle is almost vertical compared to small cars And a relatively large area. For this reason, depending on the installation conditions of the imaging unit, even when the vehicle is imaged from the front rather than directly in front, the vehicle color determination area is substantially vertical on the captured image. By specifying the vehicle color determination region at a position along a direction that forms an angle of 80 ° to 100 ° with respect to the direction, it is possible to prevent unnecessary portions from being included in the vehicle color determination region.

  In the twelfth invention, the size of the number sign of the vehicle is determined, and based on the determined number sign size and the detected classification number, either the special use ordinary car or the special use small car is selected. judge. The size of the number mark can be determined, for example, by the size of the character (serial number) of the car number, which is plate information other than the classification number. For example, the size of the number mark can be classified into a large number mark, a medium number mark, and a small number mark. And if the classification number is a special purpose small car and a special purpose ordinary car, if the number sign is a large number sign, it can be determined that it is a special purpose ordinary car, and the number sign is a medium size number If it is a standard, it can be determined that the vehicle is a special purpose small car. Thereby, even when the classification numbers are the same, the vehicle color determination area can be correctly specified.

  According to the present invention, since the position and shape of the vehicle color determination area are specified based on the classification number, as in the past, an advanced image for determining the determination area for determining the color of the vehicle based on the attribute information The color of the vehicle can be determined without performing processing.

It is a block diagram which shows an example of a structure of the vehicle color determination apparatus of this Embodiment. It is explanatory drawing which shows an example of the picked-up image for plates imaged with the plate recognition camera. It is a schematic diagram which shows an example of a license plate. It is explanatory drawing which shows the relationship between a classification number, the classification of a motor vehicle, and a use. It is explanatory drawing which shows the outline | summary of the identification method of a vehicle color determination area | region. It is explanatory drawing which shows the 1st specific method of the vehicle color determination area | region by the vehicle color determination apparatus of this Embodiment. It is a schematic diagram which shows an example of the vehicle color determination area | region identified by the 1st identification method. It is explanatory drawing which shows the 2nd specific method of the vehicle color determination area | region by the vehicle color determination apparatus of this Embodiment. It is a schematic diagram which shows an example of the vehicle color determination area | region identified by the 2nd identification method. It is explanatory drawing which shows the 3rd identification method of the vehicle color determination area | region by the vehicle color determination apparatus of this Embodiment. It is a schematic diagram which shows an example of the vehicle color determination area | region identified by the 3rd identification method. It is explanatory drawing which shows the 4th identification method of the vehicle color determination area | region by the vehicle color determination apparatus of this Embodiment. It is a schematic diagram which shows an example of the vehicle color determination area | region identified by the 4th identification method. It is explanatory drawing which shows the 5th specific method of the vehicle color determination area | region by the vehicle color determination apparatus of this Embodiment. It is a schematic diagram which shows an example of the vehicle color determination area | region identified by the 5th identification method. It is a flowchart which shows the process sequence of the vehicle color determination apparatus of this Embodiment. It is a flowchart which shows the process sequence of the vehicle color determination by the vehicle color determination apparatus of this Embodiment.

  Hereinafter, the present invention will be described with reference to the drawings illustrating embodiments. FIG. 1 is a block diagram illustrating an example of a configuration of a vehicle color determination device 100 according to the present embodiment. An imaging device 200 and a central device 300 are connected to the vehicle color determination device 100. The imaging device 200 may be configured to be built in the vehicle color determination device 100.

  The imaging apparatus 200 includes an illumination unit 210, a plate recognition camera 220, a car color determination camera 230, and the like. The illumination unit 210 includes a plurality of LEDs, and irradiates the vehicle traveling on the road with near-infrared light at the timing (for example, every 33 ms) when the plate recognition camera 220 images the vehicle. Thereby, for example, even at night, the registration number can be detected by capturing the vehicle with the plate recognition camera 220 to recognize the license plate.

  The plate recognition camera 220 images a vehicle traveling on the road, and outputs image data obtained by the imaging to the vehicle color determination device 100. The plate captured image (first captured image) obtained by capturing with the plate recognition camera 220 is, for example, a black and white image. Here, the black-and-white image is a grayscale image composed of white, black, gray in between, and the like.

  The vehicle color determination camera 230 images a vehicle traveling on the road and outputs image data obtained by the imaging to the vehicle color determination device 100. The color determination captured image (second captured image) obtained by capturing with the vehicle color determination camera 230 is a color image.

  The plate recognition camera 220 and the car color determination camera 230 are built in the imaging apparatus 200, that is, are incorporated in the same housing, and the mounting position of each camera, the optical axis of the lens, and the like are determined when the imaging apparatus 200 is manufactured. Fixed in place. Thus, when the imaging device 200 is installed on the side of the road, it is not necessary to adjust the mounting positions of the plate recognition camera 220 and the vehicle color determination camera 230.

  FIG. 2 is an explanatory diagram showing an example of a plate-captured image captured by the plate recognition camera 220. As shown in FIG. 2, the plate captured image has a high resolution, and the horizontal axis is represented by xp and the vertical axis is represented by yp with the upper left of the captured image being the center of coordinates.

  The vehicle color determination apparatus 100 includes a control unit 10 that controls each unit in the apparatus, a plate recognition image input unit 11, a vehicle color determination image input unit 12, a vehicle detection unit 13, a plate recognition unit 14, and a plate position calculation unit 15. A classification number detection unit 16, a vehicle color determination area specifying unit 17, a vehicle color determination unit 18, a storage unit 19, a day / night determination unit 20, a clock 21, and the like.

  The plate recognition image input unit 11 acquires image data obtained by imaging with the plate recognition camera 220, and stores the acquired image data in the storage unit 19 in association with the imaging time for each frame.

  The vehicle color determination image input unit 12 acquires image data obtained by imaging with the vehicle color determination camera 230, and stores the acquired image data in the storage unit 19 in association with the imaging time for each frame.

  The vehicle detection unit 13 takes out image data obtained by imaging with the plate recognition camera 220 from the storage unit 19 and detects a vehicle (including a part of the vehicle) based on the taken-out image data. As the vehicle detection process, for example, the vehicle can be detected based on the difference between the plate-captured image when the vehicle is not present and the plate-captured image when the vehicle is present. The vehicle detection unit 13 outputs the detected vehicle image to the plate recognition unit 14. Note that the vehicle image may be an image of only the vehicle portion obtained by extracting the vehicle portion from the plate-captured image, or may specify the range of the vehicle portion in the plate-captured image.

  Based on the detected vehicle image, the plate recognition unit 14 recognizes the license plate by performing a cutting process of the license plate portion as a vehicle number mark. The plate recognition unit 14 outputs the plate image of the recognized license plate, the coordinates of the license plate on the picked-up image for the plate, and the like to the plate position calculation unit 15 and the classification number detection unit 16.

  Moreover, the plate recognition part 14 determines the magnitude | size of the cut out license plate. The size of the license plate can be determined by, for example, the size of the character (serial number) of the car number, which is plate information other than the classification number. For example, the size of the license plate can be classified into a large number mark, a medium number mark, and a small number mark.

  The size of the large number mark is, for example, 60 mm x 120 mm. The large number mark is used for ordinary cars with a gross vehicle weight of 8 tons or more, a maximum loading capacity of 5 tons or more, or a passenger capacity of 30 or more. It is done. Moreover, the size of the character of a medium size number mark is 40 mm x 80 mm, for example, and a medium size number mark is used for a normal car, a small car, and a large special car which do not correspond to a large number mark and a small number mark. Moreover, the size of the character of a small number mark is 35 mm x 60 mm, for example, and a small number mark is used for a part of small two-wheeled motor vehicles and light vehicles.

  The plate position calculation unit 15 includes an image memory therein, and a predetermined position of the license plate unit (for example, a rectangular license plate unit) from the coordinates on the plate image of the plate image output from the plate recognition unit 14. (Hereinafter referred to as “plate position”), and the calculated coordinates (xp, yp) of the plate position and the imaging time of the captured image for the plate are stored in the image memory in association with each other. In addition, a plate position is not limited to the above-mentioned position, The other location of a license plate part may be sufficient.

  The classification number detection unit 16 detects the classification number on the license plate based on the plate image output from the plate recognition unit 14. The classification number detection unit 16 can also detect the vehicle number (excluding the classification number) on the license plate based on the plate image.

  FIG. 3 is a schematic diagram showing an example of a license plate. As shown in FIG. 3, the classification number is displayed on the right side of the place name of the license plate, and the size (for example, small, normal or large) of the automobile and the application (for example, passenger, cargo, special purpose, large special, etc.) ) Is a 1 to 3 digit number (in the example of FIG. 3, it is displayed as OO).

  The classification number detection unit 16 uses, for example, a matched filter that receives a binarized image obtained by binarizing a plate image and a reference image that is a car number template to determine the similarity between the binarized image and the reference image. The correlation distribution shown can be created, and the number indicated by the maximum value of the correlation distribution can be detected for each number of the classification number.

  FIG. 4 is an explanatory diagram showing the relationship between the classification number, the type of automobile, and the application. In FIG. 4, the classification number identification code is set for convenience, and the classification number identification code is used for the type and application of the vehicle according to the classification number and the size of the license plate (for example, the size of the letters of the serial number). Based on this, the method for identifying the vehicle color determination area is classified. In other words, the method for specifying the vehicle color determination area differs depending on the classification number identification code. Rather than simply specifying the vehicle color determination area according to the vehicle type (for example, small, normal, or large), the vehicle color determination area is specified based on the classification number of the automobile and the size of the license plate. It is possible to extract an optimal region for color determination according to various forms of an automobile without performing advanced image processing.

  As shown in FIG. 4, when the classification numbers are 5, 7, 50 to 59, 70 to 79, 500 to 599, and 700 to 799, the type and use of the vehicle are “a small vehicle for transportation of people” Is. Further, when the classification numbers are 3, 30 to 39, and 300 to 399, the type and use of the automobile are “ordinary automobiles with a boarding capacity of 10 or less for transportation of people”. In addition, when the classification numbers are 8, 80 to 89, and 800 to 899 and the size of the license plate is a medium size (medium size number mark), the type and use of the vehicle is “small vehicle for special use” It is. The above three types and uses are summarized and the classification number identification code is A.

  When the classification number is 4, 6, 40 to 49, 60 to 69, 400 to 499, and 600 to 699, the type and use of the automobile are “small cars for freight transportation”, and the classification number is identified. The sign is B.

  When the classification numbers are 8, 80 to 89, and 800 to 899, and the number plate is large (large number mark), the type and use of the car is “ordinary car for special use”. . When the classification numbers are 0, 00 to 09, 000 to 099, 9, 90 to 99, and 900 to 999, the type and application of the automobile are “large special automobiles”. The above-mentioned two types and uses are summarized and the classification number identification code is C.

  When the classification numbers are 1, 10 to 19, and 100 to 199, the type and use of the automobile are “ordinary automobiles for freight transportation”, and the classification number identification code is D.

  When the classification numbers are 2, 20 to 29, and 200 to 299, the type and use of the vehicle are “ordinary vehicles with a passenger capacity of 11 or more for transportation of people”, and the classification number identification code is E. To do. Mini cars are included as part of small cars.

  The vehicle color determination area specifying unit 17 specifies the plate position coordinates (xs, ys) on the color determination captured image corresponding to the plate position coordinates (xp, yp) calculated by the plate position calculation unit 15. Based on the classification number detected by the classification number detecting unit 16, the shape of the vehicle color determination area on the color determination captured image and the position with respect to the coordinates (xs, ys) of the plate position are specified.

  In addition, the vehicle color determination area specifying unit 17 determines either a special-purpose ordinary vehicle or a special-purpose small vehicle based on the size of the license plate determined by the plate recognition unit 14 and the detected classification number. Thereby, even when the classification numbers are the same, the vehicle color determination area can be correctly specified.

  FIG. 5 is an explanatory diagram showing an outline of a method for specifying the vehicle color determination area. As shown in FIG. 5, the vehicle color determination area is specified by detecting the distance and direction from the coordinates (xs, ys) of the plate position according to the detected classification number and the position of the vehicle color determination area. The shape and size of the vehicle color determination area are specified according to the classification number.

  For each classification number, information defining the shape and size of the vehicle color determination area on the color determination captured image and the position (for example, distance, direction, etc.) with respect to the number sign is stored in the storage unit 19 in advance. The vehicle color determination area corresponding to the detected classification number can be selected. The details of the vehicle color determination area specifying method will be described later.

  The vehicle color determination unit 18 converts the pixel value of the vehicle color determination region specified by the vehicle color determination region specification unit 17 from, for example, RGB to HSI (H: hue, S: saturation, I: luminance), According to the value of the saturation S, it is determined whether the color is chromatic or achromatic, and the color of the vehicle is determined based on the hue H and / or the luminance I. The type of color to be determined can be determined as appropriate. For example, according to the value of hue H, it can be classified into red, yellow, green, water, blue, purple, and the value of luminance I. Depending on the situation, it can be divided into white, gray, black, etc.

  More specifically, the vehicle color determination unit 18 obtains the hue H and the luminance I for all the pixels in the vehicle color determination region, determines the color for each pixel, and totals the number of pixels of each color. The vehicle color determination unit 18 determines the color of the vehicle body based on the total number of pixels for each color. For example, the color with the largest number of pixels can be the color of the vehicle body. Or it is good also as the 1st candidate of the color of a vehicle body, and a 2nd candidate in order with the number of pixels more than a predetermined threshold. Or you may determine the color which mixed each color of the pixel number more than a predetermined threshold value as the color of a vehicle body.

  The vehicle color determination unit 18 performs color determination regardless of chromatic color or achromatic color during daytime, for example, according to the determination result from the day / night determination unit 20 described later, and only achromatic color at night. Judgment of color is performed.

  The day / night determination unit 20 determines whether the day / night is different (for example, whether the mode is the day mode) according to the output (time) from the clock 21, and outputs the determination result to the vehicle color determination unit 18.

  The vehicle color determination device 100 outputs information in which the determined vehicle color is associated with identification information (for example, vehicle number) for identifying the vehicle to the central device 300.

  As described above, since the position and shape of the vehicle color determination area are specified based on the detected classification number, the determination area for determining the color of the vehicle is determined based on the attribute information as in the prior art. The color of the vehicle can be determined without performing image processing. Also, since the vehicle color determination area is specified based on the classification number classified according to the size and application of the vehicle, the vehicle color determination area can be specified according to various types of vehicles, and the vehicle color determination The vehicle color can be accurately determined without including unnecessary portions.

  Next, a specific example of a method for specifying the vehicle color determination area according to the classification number will be described. FIG. 6 is an explanatory diagram showing a first method for specifying a vehicle color determination region by the vehicle color determination device 100 of the present embodiment, and FIG. 7 shows an example of a vehicle color determination region specified by the first specification method. It is a schematic diagram shown. The vehicle color determination region illustrated in FIGS. 6 and 7 corresponds to the case where the classification number identification code of FIG.

  In the case where the detected classification number corresponds to classification number identification code A, that is, the type and use of the vehicle are “a small vehicle for carrying people, a predetermined passenger capacity for carrying people (for example, 10 people) In the case of the following "normal vehicle or small vehicle for special use", a vertically long rectangular area is specified as the vehicle color determination area.

  More specifically, as shown in FIG. 6, on the picked-up image for vehicle color determination, the length is set at a distance s1 in a direction that forms an angle θ with respect to the horizontal axis from the plate position (xs, ys). A parallelogram having the midpoint of the base of a1, the length b1 of the other side and the diagonal α is specified as the vehicle color determination region. In this case, the length a1 of one opposite side and the length b1 of the other opposite side have a relationship of a1 <b1, and the vehicle color determination region is a vertically long parallelogram (vertically long rectangular shape).

  As shown in FIG. 7, in the case of a small car for transporting people, a normal car having a predetermined passenger capacity or a small car for special purposes, a vertically long rectangular (long rectangular) bonnet area above the license plate Therefore, if the detected classification number is a small car for transporting people, a regular car with a specified passenger capacity or less, or a small car for special purposes, it is located above the number sign (number plate). A vertically long rectangular (parallelogram) area is specified as the vehicle color determination area. Since the vehicle color determination area is specified based on the classification number, the color of the vehicle can be determined without performing advanced image processing.

  In FIG. 6, the diagonal α (predetermined angle) of the parallelogram can be an angle formed by the horizontal axis on the vehicle color determination captured image and the road extending direction. That is, an area in which a rectangular vertical side forms a predetermined angle with respect to the horizontal axis on the captured image is specified as a vehicle color determination area. Depending on the installation conditions of the imaging device 200, when the vehicle is imaged from the front, the vehicle color determination area is not in the vertical direction on the vehicle color determination captured image, but is inclined in the road extending direction with respect to the horizontal axis. Therefore, by inclining the vertical side of the car color case area by the road extending direction with respect to the horizontal axis (by making the car color judgment area a parallelogram), the car color judgment area includes an unnecessary part. Can be prevented.

  In FIG. 6, the angle θ can be an angle formed by the horizontal axis on the vehicle color determination captured image and the road extending direction. Depending on the installation conditions of the imaging device 200, when the vehicle is imaged from the front, the vehicle color determination area is not in the vertical direction on the vehicle color determination captured image, but is inclined in the road extending direction with respect to the horizontal axis. Therefore, by specifying the vehicle color case area from the plate position to the position along the road extending direction, it is possible to prevent unnecessary portions from being included in the vehicle color determination area.

  Note that the angles θ, α, lengths (distances) s1, a1, b1, and the like shown in FIG. 6 are empirically calculated in advance as parameters for specifying the position, shape, and size of the vehicle color determination region. , And can be stored in the storage unit 19 in association with the classification number. Note that the angles θ and α can be set as appropriate according to the installation state of the imaging apparatus 200, and may be 0 °. In this case, the shape of the vehicle color determination area can be a rectangle or a square instead of the parallelogram.

  FIG. 8 is an explanatory diagram showing a second method for specifying the vehicle color determination area by the vehicle color determination device 100 of the present embodiment, and FIG. 9 shows an example of the vehicle color determination area specified by the second specification method. It is a schematic diagram shown. The vehicle color determination region illustrated in FIGS. 8 and 9 corresponds to the case where the classification number identification code in FIG.

  When the detected classification number corresponds to the classification number identification code B, that is, when the type and use of the automobile is “a small automobile for freight transportation”, the horizontally long rectangular area is defined as a vehicle color determination area. As specified.

  More specifically, as shown in FIG. 8, on the picked-up image for vehicle color determination, the length is set at a distance s2 in a direction that forms an angle θ with respect to the horizontal axis from the plate position (xs, ys). A parallelogram having the midpoint of the bottom side of a2, the length b2 of the other side and the diagonal α is specified as the vehicle color determination region. In this case, the length a2 of one side and the length b2 of the other side have a relationship of a2> b2, and the vehicle color determination region is a horizontally long parallelogram (vertically long rectangular shape).

  As shown in FIG. 9, in the case of a small vehicle for transporting cargo, a horizontally long bonnet region (horizontal rectangle) exists above the license plate, so the detected classification number is a small size for transporting cargo. In the case of an automobile, a horizontally long rectangular (parallelogram) area above the number mark (number plate) is specified as a vehicle color determination area. Since the vehicle color determination area is specified based on the classification number, the color of the vehicle can be determined without performing advanced image processing.

  In FIG. 8, the diagonal α (predetermined angle) of the parallelogram can be an angle formed by the horizontal axis on the vehicle color determination captured image and the road extending direction. That is, an area in which a rectangular vertical side forms a predetermined angle with respect to the horizontal axis on the captured image is specified as a vehicle color determination area. Depending on the installation conditions of the imaging device 200, when the vehicle is imaged from the front, the vehicle color determination area is not in the vertical direction on the vehicle color determination captured image, but is inclined in the road extending direction with respect to the horizontal axis. Therefore, by inclining the vertical side of the car color case area by the road extending direction with respect to the horizontal axis (by making the car color judgment area a parallelogram), the car color judgment area includes an unnecessary part. Can be prevented.

  In FIG. 8, the angle θ can be an angle formed by the horizontal axis on the vehicle color determination captured image and the road extending direction. Depending on the installation conditions of the imaging device 200, when the vehicle is imaged from the front, the vehicle color determination area is not in the vertical direction on the vehicle color determination captured image, but is inclined in the road extending direction with respect to the horizontal axis. Therefore, by specifying the vehicle color case area from the plate position to the position along the road extending direction, it is possible to prevent unnecessary portions from being included in the vehicle color determination area.

  Note that the angles θ, α, lengths (distances) s2, a2, b2, and the like shown in FIG. 8 are empirically calculated in advance as parameters for specifying the position, shape, and size of the vehicle color determination region. , And can be stored in the storage unit 19 in association with the classification number. Note that the angles θ and α can be set as appropriate according to the installation state of the imaging apparatus 200, and may be 0 °. In this case, the shape of the vehicle color determination area can be a rectangle or a square instead of the parallelogram.

  As shown in FIGS. 6 to 9 described above, when the detected classification number corresponds to a small car or a normal car having a predetermined passenger capacity (for example, 10 persons) or less, a number sign ( A region that is a rectangular region above the license plate and in which the vertical side of the rectangular portion forms a predetermined angle with respect to the horizontal axis on the captured image is specified as a vehicle color determination region. As described above, the rectangular area on the vehicle color determination captured image is, for example, a parallelogram. In the case of a small vehicle or a regular vehicle with a predetermined number of passengers or less, a rectangular (square or rectangular) bonnet area exists above the license plate, so the detected classification number is a small vehicle or a regular vehicle with a predetermined number of passengers or less. In the case of corresponding to, a rectangular area above the number mark (number plate) is specified as the vehicle color determination area. In addition, the region where the vertical side of the rectangular shape forms a predetermined angle α with respect to the horizontal axis on the captured image is the vehicle color determination region. This is considered because the region is inclined by an angle corresponding to the installation state of the imaging device 200 from the vertical direction on the vehicle color determination captured image. Since the vehicle color determination area is specified based on the classification number, the color of the vehicle can be determined without performing advanced image processing.

  FIG. 10 is an explanatory diagram showing a third method for specifying the vehicle color determination region by the vehicle color determination device 100 of the present embodiment, and FIG. 11 shows an example of the vehicle color determination region specified by the third specification method. It is a schematic diagram shown. The vehicle color determination area illustrated in FIGS. 10 and 11 corresponds to the case where the classification number identification code in FIG.

  When the detected classification number corresponds to the classification number identification code C, that is, when the type and application of the automobile are “ordinary automobile for use in special use, large special automobile”, the number is displayed on the picked-up image for vehicle color determination. Two rectangular shapes separated by an appropriate length with the upper position of the sign (number plate) in between, and a region in which the vertical sides of the rectangle form a predetermined angle with respect to the horizontal axis on the captured image It is specified as a color determination area.

  More specifically, as shown in FIG. 10, from the position at a distance s3 in the direction that forms an angle θ with respect to the horizontal axis from the plate position (xs, ys) on the vehicle color determination captured image, Two parallelograms having a vertex at a position separated by a distance S4 in the lateral direction and the left lateral direction and having a length a3 on one side, a length b3 on the other side, and a diagonal α are used as vehicle color determination regions. Identify.

  As shown in FIG. 11, in the case of a special-purpose ordinary vehicle or large special vehicle, a front grille or a special structure exists on the front surface of the vehicle, and a rectangular bonnet region exists on both sides of the front surface of the vehicle. Therefore, avoiding these front grills or special structures, two rectangular shapes separated by an appropriate length with the upper position of the number mark (number plate) in between, and the vertical sides of the rectangular shape are captured images. A region that forms a predetermined angle with respect to the upper horizontal axis is specified as a vehicle color determination region. In addition, an area in which a rectangular side in the vertical direction forms a predetermined angle with respect to the horizontal axis on the captured image is defined as a vehicle color determination area when a vehicle is imaged obliquely from the front. Is inclined by an angle corresponding to the installation state of the imaging device 200 from the vertical direction on the vehicle color determination captured image. Since the vehicle color determination area is specified based on the classification number, the color of the vehicle can be determined without performing advanced image processing.

  In FIG. 10, the diagonal α (predetermined angle) of the parallelogram is 80 ° to 100 ° with respect to the horizontal axis direction on the vehicle color determination captured image. In the case of a special automobile or large special automobile, the front surface of the vehicle is substantially vertical and has a relatively large area compared to a small automobile. For this reason, even if the vehicle is imaged from the front rather than in front of the vehicle according to the installation conditions of the imaging device 200, the vehicle color determination area is substantially vertical on the vehicle color determination captured image. By making the vertical side of 80 ° to 100 ° with respect to the horizontal axis, it is possible to prevent unnecessary portions from being included in the vehicle color determination region.

  Further, in FIG. 10, the angle θ can be set to an angle of 80 ° to 100 ° with respect to the horizontal axis direction on the vehicle color determination captured image from the plate position. In the case of a special automobile or large special automobile, the front surface of the vehicle is substantially vertical and has a relatively large area compared to a small automobile. Therefore, depending on the installation conditions of the imaging device 200, even when the vehicle is imaged from the front rather than in front of the vehicle, the vehicle color determination area is substantially vertical on the vehicle color determination captured image. By specifying the vehicle color determination region at a position along a direction that forms an angle of 80 ° to 100 ° with respect to the horizontal axis direction on the color determination captured image, an unnecessary portion is included in the vehicle color determination region. Can be prevented.

  Note that the angles θ, α, lengths (distances) s3, s4, a3, b3, etc. shown in FIG. 10 are empirically calculated in advance as parameters for specifying the position, shape, and size of the vehicle color determination area. It can be stored in the storage unit 19 in association with the classification number. Note that the angles θ and α can be appropriately set according to the installation state of the imaging device 200.

  FIG. 12 is an explanatory diagram showing a fourth method for specifying the vehicle color determination region by the vehicle color determination device 100 of the present embodiment, and FIG. 13 shows an example of the vehicle color determination region specified by the fourth specification method. It is a schematic diagram shown. The vehicle color determination area illustrated in FIGS. 12 and 13 corresponds to the case where the classification number identification code in FIG.

  When the detected classification number corresponds to the classification number identification code D, that is, when the type and use of the vehicle is “ordinary vehicle for freight transportation”, the plate position of the vehicle color determination image is displayed. An approximately rectangular shape at an upper position, in which the vertical side of the rectangle forms a predetermined angle α with respect to the horizontal axis on the captured image, and the lower center of the approximately rectangular shape is cut out in a concave shape Is specified as the vehicle color determination area.

  More specifically, as shown in FIG. 12, on the picked-up image for vehicle color determination, the length is set at a distance s5 in a direction that forms an angle θ with respect to the horizontal axis from the plate position (xs, ys). There is a midpoint of the base of (a4 + 2 × a5), a parallelogram having a length b5 of the other side and a diagonal α is specified, and the lower center of the parallelogram is a4. The shape cut out by the parallelogram whose other side is b4 is specified as the vehicle color determination area.

  As shown in FIG. 13, in the case of an ordinary vehicle for transporting cargo, a rectangular front grille is provided above the license plate, and bonnet areas are present on both sides and above the front grille. In the case of corresponding to an ordinary vehicle for transporting cargo, it is a substantially rectangular shape above the number sign (number plate), and the vertical side of the rectangular shape is relative to the horizontal axis on the captured image. An area formed by forming a predetermined angle and cutting out the lower center of the substantially rectangular shape in a concave shape is specified as a vehicle color determination area. In addition, an area in which a rectangular side in the vertical direction forms a predetermined angle with respect to the horizontal axis on the captured image is defined as a vehicle color determination area when a vehicle is imaged obliquely from the front. Is inclined by an angle corresponding to the installation state of the imaging device 200 from the vertical direction on the vehicle color determination captured image. Since the vehicle color determination area is specified based on the classification number, the color of the vehicle can be determined without performing advanced image processing.

  In FIG. 12, the diagonal α (predetermined angle) of the parallelogram is 80 ° to 100 ° with respect to the horizontal axis direction on the vehicle color determination captured image. In the case of an ordinary vehicle for transporting cargo, the front surface of the vehicle is almost vertical and has a relatively large area compared to a small vehicle. For this reason, even if the vehicle is imaged from the front rather than in front of the vehicle according to the installation conditions of the imaging device 200, the vehicle color determination area is substantially vertical on the vehicle color determination captured image. By making the vertical side of 80 ° to 100 ° with respect to the horizontal axis, it is possible to prevent unnecessary portions from being included in the vehicle color determination region.

  In FIG. 12, the angle θ can be an angle of 80 ° to 100 ° with respect to the horizontal axis direction on the vehicle color determination captured image from the plate position. In the case of an ordinary vehicle for transporting cargo, the front surface of the vehicle is almost vertical and has a relatively large area compared to a small vehicle. Therefore, depending on the installation conditions of the imaging device 200, even when the vehicle is imaged from the front rather than in front of the vehicle, the vehicle color determination area is substantially vertical on the vehicle color determination captured image. By specifying the vehicle color determination region at a position along a direction that forms an angle of 80 ° to 100 ° with respect to the horizontal axis direction on the color determination captured image, an unnecessary portion is included in the vehicle color determination region. Can be prevented.

  Note that the angles θ, α, lengths (distances) s5, a4, a5, b4, b5, etc. shown in FIG. 12 are previously empirically used as parameters for specifying the position, shape, and size of the vehicle color determination area. It can be calculated and stored in the storage unit 19 in association with the classification number. Note that the angles θ and α can be appropriately set according to the installation state of the imaging device 200.

  FIG. 14 is an explanatory diagram showing a fifth method for specifying a vehicle color determination region by the vehicle color determination device 100 of the present embodiment, and FIG. 15 shows an example of a vehicle color determination region specified by the fifth specification method. It is a schematic diagram shown. The vehicle color determination region illustrated in FIGS. 14 and 15 corresponds to the case where the classification number identification code in FIG.

  When the detected classification number corresponds to the classification number identification code E, that is, when the type and use of the vehicle is “ordinary vehicle with a passenger capacity of 11 or more for transportation of people”, imaging for vehicle color determination It is a substantially rectangular shape above the plate position on the image, and the vertical side of the rectangle forms a predetermined angle with respect to the horizontal axis on the captured image, and is located at the lower center of the substantially rectangular shape. A region having a convex region is specified as a vehicle color determination region.

  More specifically, as shown in FIG. 14, on the picked-up image for vehicle color determination, it is in a direction that forms an angle θ with respect to the horizontal axis from the plate position (xs, ys) and is at a position of distance (s6 + b6). There is a midpoint of the base of the length (a6 + 2 × a7), the parallelogram of the other side length b7 and diagonal α is specified, and the other side is a6 at the lower center of the parallelogram. And a shape having a parallelogram area whose other side is b6 is specified as a vehicle color determination area.

  As shown in FIG. 15, in the case of an ordinary vehicle having a predetermined number of passengers (11 persons) or more, there is no front grille, a rectangular bonnet region is located above the license plate, and two regions are continuously connected to the region. Since there is a rectangular bonnet area with a short height above the headlight, if the detected classification number corresponds to a regular car with a predetermined passenger capacity or higher, it is located above the number sign (number plate). A region having a substantially rectangular shape, in which the vertical side of the rectangular shape forms a predetermined angle with respect to the horizontal axis on the captured image, and has a convex region at the lower center of the substantially rectangular shape. It is specified as a color determination area. In addition, an area in which a rectangular side in the vertical direction forms a predetermined angle with respect to the horizontal axis on the captured image is defined as a vehicle color determination area when a vehicle is imaged obliquely from the front. Is inclined by an angle corresponding to the installation state of the imaging device 200 from the vertical direction on the vehicle color determination captured image. Since the vehicle color determination area is specified based on the classification number, the color of the vehicle can be determined without performing advanced image processing.

  In FIG. 14, the diagonal α (predetermined angle) of the parallelogram is 80 ° to 100 ° with respect to the horizontal axis direction on the vehicle color determination captured image. In the case of a regular vehicle having a predetermined number of passengers or more, the front surface of the vehicle is substantially vertical and has a relatively large area compared to a small vehicle. For this reason, even if the vehicle is imaged from the front rather than in front of the vehicle according to the installation conditions of the imaging device 200, the vehicle color determination area is substantially vertical on the vehicle color determination captured image. By making the vertical side of 80 ° to 100 ° with respect to the horizontal axis, it is possible to prevent unnecessary portions from being included in the vehicle color determination region.

  In FIG. 14, the angle θ can be an angle of 80 ° to 100 ° with respect to the horizontal axis direction on the vehicle color determination captured image from the plate position. The front surface of the vehicle is substantially vertical and has a relatively large area as compared with ordinary cars, small cars, and the like that have a predetermined passenger capacity or more. Therefore, depending on the installation conditions of the imaging device 200, even when the vehicle is imaged from the front rather than in front of the vehicle, the vehicle color determination area is substantially vertical on the vehicle color determination captured image. By specifying the vehicle color determination region at a position along a direction that forms an angle of 80 ° to 100 ° with respect to the horizontal axis direction on the color determination captured image, an unnecessary portion is included in the vehicle color determination region. Can be prevented.

  Note that the angles θ, α, lengths (distances) s6, a6, a7, b6, b7, etc. shown in FIG. 14 are previously empirically used as parameters for specifying the position, shape, and size of the vehicle color determination area. It can be calculated and stored in the storage unit 19 in association with the classification number. Note that the angles θ and α can be appropriately set according to the installation state of the imaging device 200.

  Next, the operation of the vehicle color determination device 100 will be described. FIG. 16 is a flowchart showing a processing procedure of the vehicle color determination apparatus 100 of the present embodiment. The control part 10 acquires the picked-up image for plates (S11), and detects a vehicle (S12). The controller 10 recognizes the plate by extracting the rectangular license plate portion from the vehicle image (S13), detects the classification number from the recognized plate, and plate information other than the classification number (character of the vehicle number) Information) and the size of the plate (number plate) is detected based on the result (S14).

  The control unit 10 calculates the plate position based on the extracted position of the license plate part (S15), and specifies the vehicle color determination region by obtaining the plate position on the color determination captured image (S16). The controller 10 determines the vehicle color (S17). Details of the vehicle color determination process will be described later.

  The control unit 10 associates the recognized plate information (character information of the car number) and the car color and transmits them to the external central device (S18), and determines whether or not there is a processing end instruction (S19). If there is no process end instruction (NO in S19), the control unit 10 continues the process from step S11. If there is a process end instruction (YES in S19), the process ends.

  FIG. 17 is a flowchart showing a processing procedure of vehicle color determination by the vehicle color determination device 100 of the present embodiment. The control unit 10 converts the pixel value (color component of each pixel) in the vehicle color determination area from RGB to HSI (H: hue, S: saturation, I: luminance) (S101), and the saturation S is a predetermined value. It is determined whether it is larger than a threshold (for example, 0.4) (S102). If the saturation S is greater than the predetermined threshold (YES in S102), the control unit 10 calculates the hue H as a chromatic color (S103), and if the saturation S is not greater than the predetermined threshold (NO in S102) ), Luminance I is calculated as an achromatic color (S104).

  The control unit 10 determines the color of the pixel based on the hue H or the luminance I (S105), and determines whether or not the color determination for all the pixels in the vehicle color determination region is completed (S106). If the color determination for all the pixels has not been completed (NO in S106), the control unit 10 continues the process from step S102.

  When the color determination of all the pixels is completed (YES in S106), the control unit 10 adds up the determined number of pixels of each color (S107) and determines whether the daytime mode is set (S108). When it is the daytime mode (YES in S108), the control unit 10 determines the vehicle color based on the chromatic or achromatic color (S109), and when not in the daytime mode (NO in S108), the control unit 10 determines the vehicle color based on the achromatic color ( S110), the process is terminated.

  The vehicle color determination apparatus 100 according to the present embodiment can also be realized using a general-purpose computer that includes a CPU, a RAM, and the like. That is, as shown in FIGS. 16 and 17, a program code defining each processing procedure is loaded into a RAM provided in the computer, and the program code is executed by the CPU, whereby the vehicle color determination device 100 is executed on the computer. Can be realized.

  In the above-described embodiment, the two types of cameras, the plate recognition camera and the vehicle color determination camera, are used. However, the present invention is not limited to this. For example, it is possible to omit the plate recognition camera and use the captured image obtained by imaging with the vehicle color determination camera to recognize the license plate, detect the classification number, and also perform the vehicle color determination.

  In the above-described embodiment, the shape or size of the vehicle color determination region and the position (distance and direction) from the license plate are examples and are not limited. For example, the set value can be changed as appropriate according to the installation conditions of the camera. In addition, the shape or size of the vehicle color determination area can be appropriately changed according to the design and fashion of the automobile. In this embodiment, since the parameters for specifying the vehicle color determination area are stored, the vehicle color determination areas for vehicles of various types or forms can be obtained by replacing only the parameters without changing the processing procedure. Can be easily identified.

  The disclosed embodiments are to be considered in all respects as illustrative and not restrictive. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

DESCRIPTION OF SYMBOLS 10 Control part 11 Image input part for plate recognition 12 Image input part for vehicle color determination 13 Vehicle detection part 14 Plate recognition part (number sign determination means)
15 Plate position calculation unit 16 Classification number detection unit (classification number detection means)
17 vehicle color determination area specifying unit (vehicle color determination area specifying means, special purpose automobile determination means)
18 Car color determination unit (vehicle color determination means)
19 Memory part 20 Day / night judgment part 21 Clock

Claims (14)

In a vehicle color determination device that determines the color of a vehicle based on a captured image obtained by imaging a vehicle traveling on a road with an imaging unit,
Classification number detecting means for detecting a classification number on a number sign of a vehicle based on a captured image;
Vehicle color determination area specifying means for specifying the shape of the vehicle color determination area on the captured image and the position with respect to the number mark based on the classification number detected by the classification number detection means;
A vehicle color determination device comprising: vehicle color determination means for determining a color of a vehicle based on a color component of each pixel of the vehicle color determination area specified by the vehicle color determination area specifying means. The vehicle color determination area specifying means includes
When the classification number detected by the classification number detection means corresponds to a small vehicle or a regular vehicle having a predetermined passenger capacity or less, a rectangular area located above the number mark on the captured image, 2. The vehicle color according to claim 1, wherein a region where a vertical side of the shape forms a predetermined angle with respect to a horizontal axis on the captured image is specified as a vehicle color determination region. Judgment device. The vehicle color determination area specifying means includes
When the classification number detected by the classification number detection means is a small automobile for transporting people, a regular automobile with a predetermined passenger capacity or less, or a small automobile for special purposes, a vertically long rectangular area is specified as a vehicle color determination area The vehicle color determination device according to claim 2, wherein the vehicle color determination device is configured as described above. The vehicle color determination area specifying means includes
3. The structure according to claim 2, wherein when the classification number detected by the classification number detection means is a small vehicle for transporting cargo, a horizontally long rectangular area is specified as a vehicle color determination area. The vehicle color determination device described in 1.   5. The vehicle color determination device according to claim 2, wherein the predetermined angle is an angle formed by a horizontal axis on the captured image and a road extending direction. The vehicle color determination area specifying means includes
When the classification number detected by the classification number detection means corresponds to a small vehicle or a normal vehicle having a predetermined passenger capacity or less, a vehicle color determination region is provided at a position along the road extending direction from the number sign on the captured image. The vehicle color determination device according to any one of claims 2 to 5, wherein the vehicle color determination device is configured to be specified. The vehicle color determination area specifying means includes
When the classification number detected by the classification number detection means corresponds to a normal automobile for transporting cargo, the classification number is a substantially rectangular shape above the number mark on the captured image, and the vertical direction of the rectangular shape A side is formed at a predetermined angle with respect to the horizontal axis on the captured image, and a region obtained by cutting out the substantially rectangular lower center in a concave shape is specified as a vehicle color determination region. The vehicle color determination device according to claim 1. The vehicle color determination area specifying means includes
When the classification number detected by the classification number detection means corresponds to a normal vehicle having a predetermined number of passengers or more, it is substantially rectangular in the upper position of the number mark on the captured image, and the rectangular vertical direction The side of the image has a predetermined angle with respect to the horizontal axis on the captured image, and a region having a convex region at the lower center of the substantially rectangular shape is specified as the vehicle color determination region. The vehicle color determination device according to claim 1. The vehicle color determination area specifying means includes
When the classification number detected by the classification number detection means corresponds to a special-purpose ordinary vehicle or a large special vehicle, it is two rectangular shapes separated by an appropriate length with the upper position of the number sign in between on the captured image. 2. The apparatus according to claim 1, wherein a region in which the rectangular vertical side forms a predetermined angle with respect to a horizontal axis on the captured image is specified as a vehicle color determination region. The vehicle color determination device described.   The vehicle color determination device according to claim 7, wherein the predetermined angle is 80 ° to 100 ° with respect to a horizontal axis direction on the captured image. The vehicle color determination area specifying means includes
When the classification number detected by the classification number detection means corresponds to a regular vehicle for transporting cargo, a regular vehicle having a predetermined passenger capacity or more, a regular vehicle for special purposes, or a large special vehicle, the number sign on the captured image The vehicle color determination area is specified at a position along a direction that forms an angle of 80 ° to 100 ° with respect to the horizontal axis direction of the vehicle. The vehicle color determination device according to item 1. Number sign judging means for judging the size of the number sign of the vehicle;
Special-purpose vehicle determination means for determining either a special-purpose ordinary vehicle or a special-purpose small vehicle based on the size of the number-sign determined by the number-mark determination means and the classification number detected by the classification number detection means The vehicle color determination device according to any one of claims 2 to 11, further comprising: In a computer program for causing a computer to execute a step of determining a color of a vehicle based on a captured image obtained by imaging a vehicle traveling on a road by an imaging unit,
On the computer,
Detecting a classification number on a vehicle number sign based on the captured image;
Based on the detected classification number, specifying the shape of the vehicle color determination region on the captured image and the position with respect to the number mark;
And a step of determining a color of the vehicle based on a color component of each pixel of the specified vehicle color determination area. In the vehicle color determination method by the vehicle color determination device that determines the color of the vehicle based on the captured image obtained by imaging the vehicle traveling on the road in the imaging unit,
Detecting a classification number on a vehicle number sign based on the captured image;
Identifying the shape of the vehicle color determination area on the captured image and the position with respect to the number mark based on the detected classification number;
Determining the color of the vehicle based on the color component of each pixel in the specified vehicle color determination area.

Images