JP5151600B2 - Behavior analysis system - Google Patents

Description

  The present invention relates to a system for analyzing the behavior of a person in a store or an exhibition hall.

  When it is necessary to analyze the behavior of people in stores and exhibition halls, such as when verifying the effectiveness of sales promotion such as POP (Point-Of-Purchase) and grasping the degree of interest in each exhibition at the event hall, A human sensor that detects the location of a person is placed at each location in a store or exhibition hall, and the number of people passing in front of the human sensor or the person in front of the human sensor based on the detection results of each human sensor The residence time of a person is calculated, and the behavior of the person in the store or the exhibition hall is analyzed.

In the background art of Patent Literature 1 filed by the present applicant, several prior inventions of such technology are described. Further, in Patent Literature 1, the number of detection objects (for example, customers) passing through a specific place is described. An optimum correction value calculation device capable of calculating an optimum correction value for determining the above with higher accuracy is disclosed.
Japanese Patent Application No. 2007-253999

  When counting and analyzing the number of people passing by and time spent by time, if a human sensor device is provided for each human sensor device, each human sensor detects the location of the person. Although it is possible to count the number of passing people and the staying time for each time from the time, it is expensive to provide a clock for each human sensor device.

In order to synchronize the time of the clock provided in each human sensor device, the clock provided in the human sensor device may be a radio clock, but in an environment where radio waves of standard time cannot be received, There is a problem that time synchronization becomes impossible.

  Therefore, the present invention provides a behavior analysis system capable of totaling the number of passing people and the staying time for each time without providing a clock for each human sensor device when the number of passing people and the staying time are counted and analyzed. The purpose is to provide.

A first invention for solving the above-described problem is a human sensor device provided with a human sensor for detecting the location of a person, detection data of the human sensor is transmitted from the human sensor device, and the time is A behavior analysis system comprising at least a counting transfer device having a clock to be measured and a behavior analysis server to which the detection data is transferred from the counting transfer device. In the behavior analysis system, the human sensor device And a network used for communication between the total transfer devices. The human sensor device, every time a predetermined detection interval has elapsed, to obtain said detection data from said motion sensor, and transmits the packet data including the detection data in the aggregation transfer device, the motion sensor device The packet data transmitted to the total transfer device includes a device number for identifying the human sensor device, a packet number assigned consecutively every time the packet data is generated, and a packet transmission interval. It includes a mass of the detection data obtained from the human sensor. The aggregate transfer apparatus, by applying a reception time of the packet data, the transferring the packet data received from the human body sensor device to the behavior analysis server, the behavior analysis server, transferred from the aggregate transfer device The detection time of each detection data included in the packet data is calculated using the reception time and the detection interval given to the packet data, and used for behavior analysis.

  In this way, by constructing a communication network that is used only for communication between the human sensor device and the aggregation transfer device, it is possible to avoid congestion of the communication network, and the human sensor device Since the communication time between the aggregation transfer devices is extremely short, it is not necessary to consider the communication time between the human sensor device and the aggregation transfer device.

Therefore, even if the aggregation transfer device is provided with a clock instead of the human sensor device, and the behavior analysis server uses the reception time provided to the packet data by the aggregation transfer device for the behavior analysis, There is no clear difference from human action time.

According to the first invention, the human sensor device transmits the packet data including the device number, the packet number, and a plurality of the detection data, thereby reducing the processing load of the aggregation transfer device. Can do.

Further, a second invention is the behavior analysis system according to the first invention, wherein after the human sensor device is activated, a test signal is transmitted by broadcast in the network, and the response is earliest. A summary transfer device is selected, and the packet data is transmitted to the selected summary transfer device.

Furthermore, the third invention is the behavior analysis system according to the second invention, wherein when the communication with the selected totaling transfer apparatus is not possible for a certain period of time, a test signal is broadcast again in the network. The summary transfer device that transmits and responds earliest is selected.

According to the second invention, after the human sensor device is activated, it transmits a test signal by broadcast in the network, selects the total transfer device that responded earliest, and selects the total transfer device selected wherein by sending packet data, can be the aggregate transfer device and the reception time to be applied to the packet data, to shorten the difference between the detection time when it is detected that actually the detection data.

Further, as in the third invention, when the human sensor device cannot communicate with the total transfer device for some reason, the packet data can be transmitted to the other total transfer device, The loss of the packet data can be reduced.

Further, a fourth invention is the behavior analysis system according to any one of the first invention to the third invention, wherein the human sensor device sends the packet data to the total transfer device a plurality of times. The aggregation transfer device checks the duplication of the packet data received from the human sensor device using the device number and the bucket number included in the packet data , and transfers the packet data to the behavior analysis server The packet data is not retransmitted.

According to the fourth aspect, the motion sensor device, the packet data by transmitting a plurality of times, it is possible to improve the transmission reliability of the packet data, the aggregate transfer apparatus, the packet data By checking for duplication, the packet data that is duplicated on the behavior analysis server side is not used for behavior analysis.

Further, a fifth invention is the behavior analysis system according to any one of the first invention to the fourth invention, wherein the behavior analysis server is configured to receive the reception time given to the packet data or the Means is provided for summing up the missing period of the packet data for each device number from the packet number included in the packet data .

According to the fifth invention, when the behavior analysis server includes means for counting the packet data loss period for each device number, the packet data loss period is large for a specific device number. Thus, it is possible to take measures to improve communication between the human sensor device specified by the device number and the total transfer device (for example, increase the total transfer device).

  Thus, according to the present invention, when counting and analyzing the number of people passing by and the residence time, the number of people passing and the residence time can be counted for each time without providing a clock for each human sensor. A behavior analysis system can be provided.

  From here, the behavior analysis system 6 according to the present invention will be described in detail with reference to the drawings. FIG. 1 is a diagram showing a configuration of a behavior analysis system 6 to which the present invention is applied.

As shown in FIG. 1, the present embodiment is an embodiment in which the behavior analysis system 6 according to the present invention is implemented in a store, and a human sensor device 1 including a human sensor 10 that detects a visitor, A plurality of total transfer devices 2 that transfer packet data 7 transmitted from the sensor device 1 to the behavior analysis server 3, a behavior analysis server 3 that analyzes the behavior of the store visitor, and behavior analysis results of the behavior analysis server 3 are displayed. The human sensor device 1, the total transfer device 2 and the terminal device 4 are installed in the store, and the behavior analysis server 3 is installed in a place different from the store, and each total transfer is performed. The device 2 and the behavior analysis server 3 are connected via the Internet 5.

  The human sensor device 1 is installed at a place where a person's stay is investigated, for example, in front of a store entrance, in front of a POP advertisement (POP: Point-Of-Purchase), or in front of each display shelf, and performs behavior analysis. In the system 6, a wireless network 60 is constructed between the total transfer device 2 and the human sensor device 1. Note that the network established between the aggregation transfer device 2 and the human sensor device 1 may be a wired network.

  The human sensor device 1 aggregates, as transmission data, packet data 7 including a cluster of detection data detected by the human sensor 10 within the packet transmission interval every time a certain packet transmission interval elapses. Transmit to the transfer device 2.

When the total transfer device 2 receives the packet data 7 from the human sensor device 1, the total transfer device 2 assigns the reception time of the packet data 7 to the packet data 7 and transfers the packet data 7 to the behavior analysis server 3. 7, the detection time at which the detection data included in the packet data 7 is detected is calculated from the reception time given to 7, the behavior of the store visitor is analyzed according to a predetermined procedure, and the analysis result is displayed on the terminal device 4.

  Thus, by constructing the wireless network 60 in which the total transfer device 2 and the human sensor device 1 communicate wirelessly, the communication time between the total transfer device 2 and the human sensor device 1 becomes extremely short, and behavior analysis is performed. The server 3 calculates the time when the human sensor device 1 detects the location of the person from the reception time given to the packet data 7 without considering the communication time between the total transfer device 2 and the human sensor device 1. Will be able to.

  First, the human sensor device 1 will be described in detail. FIG. 2 is a block diagram of the human sensor device 1. As shown in FIG. 2, the human sensor device 1 is an electronic device including a human sensor 10 that detects the presence of a person. In addition to the human sensor 10, a CPU 11 (CPU: Central Processing Unit), a RAM 12 ( RAM: Random Access Memory (ROM), ROM 13 (ROM: Read-Only Memory), and wireless communication interface 15. The ROM 13 of the human sensor device 1 generates packet data 7 including the detection result of the human sensor 10, A computer program 14 (hereinafter referred to as “packet transmission program 14”) for transmitting the generated packet data 7 to the total transfer device 2 using the wireless communication interface 15 is stored.

  The human sensor 10 provided in the human sensor device 1 irradiates light (or ultrasonic waves) and measures the time until the reflected light (or reflected sound) is received, thereby detecting the human sensor 10. It is a sensor that measures the distance from the reference plane to the position where a person is located.

  When the power of the human sensor device 1 is turned on, the packet transmission program 14 stored in the ROM 13 of the human sensor device 1 is automatically started, and the inspection signal of the total transfer device 2 is broadcast in the wireless network 60. Then, the summary transfer device 2 with the earliest response is selected.

When the total transfer device 2 is selected, the packet transmission program 14 measures a certain detection interval (for example, 0.2 seconds) by counting the clock of the CPU 11 or the like, and every time the detection interval passes, The distance data indicating the distance from the reference plane of the sensor 10 to the position where the person is located is acquired from the human sensor 10 as the distance data.

The packet transmission program 14 stores the classification criteria related to the distance, classifies the distance data acquired from the human sensor 10 according to the classification criteria, and temporarily stores the classification value of the distance data as detection data in the RAM 12. .

  FIG. 3 is a diagram for explaining an example of the division criterion related to the distance related to the distance. As illustrated in FIG. 3, the packet transmission program 14 stores a division criterion that associates a distance range with a segment value, and converts the distance data acquired from the human sensor 10 into a segment value.

In the classification criteria of FIG. 3, the distance data of 240 mm or more belongs to the distance classification of “00b” (b indicates binary), and the distance data of 160 mm or more and less than 240 mm has the classification value “01b”. The distance data belonging to the distance category of 80 mm or more and less than 160 mm belongs to the distance category of the category value “02b”, and the distance data of 0 mm or more and less than 240 mm belongs to the distance category of the category value “03b”. It is shown.

  The human sensor device 1 does not need to convert the distance data into segment values every time the distance data is acquired from the human sensor 10, for example, the distance detection value acquired from the human sensor 10 a plurality of times. The average value may be converted into a segment value.

  The packet transmission program 14 stores the segment value obtained by converting the distance data acquired from the human sensor 10 in the RAM 12, and then stores the segment value stored in the RAM 12 within the packet transmission interval every time a certain packet transmission interval elapses. The included packet data 7 is generated, and the generated packet data 7 is transmitted to the selected total transfer device 2 a plurality of times using the wireless communication interface 15, and the segment values stored in the RAM 12 are initialized.

As described above, since the human sensor 10 detects the location of a person at regular detection intervals, the packet data 7 includes a certain number of segment values. For example, when the detection interval is 0.2 seconds / time and the packet transmission interval is 1 minute, one packet data 7 includes 300 segment values.

  FIG. 4 is a diagram for explaining the structure of the packet data 7 generated by the packet transmission program 14. As illustrated in FIG. 4, the packet data 7 generated by the packet transmission program 14 includes a device number 70 that can uniquely identify the human sensor device 1 that transmits the packet data 7 and a serial number in the order in which the packet data 7 is generated. Packet number 71, which is a number, and a data portion 72. In the data portion 72 of the packet data 7, a certain number (for example, 300) of segment values are described in a binary format.

  Note that the packet transmission program 14 of the human sensor device 1 may perform broadcasting again and reselect the aggregation transfer device 2 when communication with the selected aggregation transfer device 2 is interrupted for a certain period.

  Next, the summary transfer device 2 to which the packet data 7 is transmitted from the summary transfer device 2 will be described. FIG. 5 is a block diagram of the summary transfer apparatus 2. As illustrated in FIG. 5, the total transfer device 2 is an electronic device, such as a CPU 20 (CPU: Central Processing Unit), a RAM 21 (RAM: Random Access Memory), a ROM 22 (ROM: Read-Only Memory), and a hard disk. A capacity data storage device 25, a clock 24, a wireless communication interface 26, and a wired communication interface 27 are provided, and the packet data 7 transmitted from the human sensor device 1 is transferred to the behavior analysis server 3 in the ROM 22 of the total transfer device 2. A computer program 23 (hereinafter referred to as “packet transfer program 23”) is stored.

  The packet transfer program 23 of the aggregation transfer device 2 is automatically activated when the aggregation transfer device 2 is turned on, and the wireless communication interface 26 is used to send the inspection signal of the aggregation transfer device 2 from the human sensor device 1. When received, a response signal is transmitted to the wireless network 60, and when the packet data 7 is received from the human sensor device 1, the same packet data 7 is received by using the device number 70 and the packet number 71 of the received packet data 7. Check if has already been received.

  Since the human sensor device 1 has a specification for transmitting the packet data 7 a plurality of times, the aggregation transfer device 2 of the present embodiment has a function of confirming whether the same packet data 7 has already been received.

In order to confirm whether or not the same packet data 7 has been received, the total transfer device 2 stores the device number 70 and the packet number 71 included in the packet data 7 received from the human sensor device 1 in the data storage device 25. It is checked whether the device number 70 and the packet number 71 included in the packet data 7 received from the human sensor device 1 are already stored in the data storage device 25.

  When the packet transfer program 23 of the total transfer device 2 receives the packet data 7 that does not overlap, the packet transfer program 23 acquires the reception time of the packet data 7 from the clock 24 and gives the reception time to the packet data 7 received from the human sensor device 1. Using the wired communication interface 27, the packet data 7a to which the reception time is given (hereinafter referred to as “transfer packet data 7a”) is transferred to the behavior analysis server 3. When receiving duplicate packet data 7, the total transfer device 2 does not transfer the transfer packet data 7 a to the behavior analysis server 3.

  FIG. 6 is a diagram for explaining the structure of the transfer packet data 7 a that the total transfer device 2 transfers to the behavior analysis server 3. As shown in FIG. 6, the transfer packet data 7a has a configuration in which the time data 73 indicating the reception time 73 is added to the packet data 7 described with reference to FIG. 4, and the device number 70a, the packet number 71a, the data portion 72a and time data 73.

The device number 70a of the transfer packet data 7a is the same data as the device number 70 of the packet data 7 received from the human sensor device 1, and the packet number 71a is the bucket number 71 of the packet data 7 received from the human sensor device 1. Is the same data. The data part 72a may be the same data as the data part 72 of the packet data 7. However, the segment value described in the binary format may be converted into text data and used as the data of the data part 72a of the transfer packet data 7a. .

  As described above, the total transfer device 2 gives the packet data 7 the reception time 73 indicating the reception time of the packet data 7, so that the time when each of the segment values included in the data portion 72 of the packet data 7 is generated, That is, it becomes possible to determine the time when the human sensor 10 of the human sensor device 1 detects the location of a person.

  This is because the human sensor device 1 and the total transfer device 2 directly communicate with each other, so that there is no congestion of lines as occurs in the Internet 5 and no time loss is considered between the human sensor device 1 and the total transfer device 2. This is because communication can be performed and the reception time when the total transfer device 2 receives the packet data 7 and the time when the human sensor device 1 generates the packet data 7 can be regarded as the same.

As described above, since the detection interval at which the human sensor device 1 detects the location of the person and the number of segment values included in the data portion of the packet data 7 are determined, the reception time 73 given to the packet data 7 is determined. , It is possible to determine the detection time when each segment value included in the data portion 72 of the transfer packet data 7a is detected.

  For example, if the detection interval of the human sensor 10 is 0.2 seconds and the packet transmission interval of the human sensor device 1 is 1 minute, the data portion 72a of the transfer packet data 7a includes 300 segment values. The first segment value of the data portion 72a of the transfer packet data 7a can be determined as the segment value when the location of a person is detected one minute before the reception time 73.

  Next, the behavior analysis server 3 to which the transfer packet data 7a is transferred from the total transfer device 2 will be described. FIG. 7 is a block diagram of the behavior analysis server 3. As illustrated in FIG. 7, the behavior analysis server 3 is a device realized by a general-purpose server. The behavior analysis server 3 includes a CPU 30 (CPU: Central Processing Unit), a RAM 31 (RAM: Random Access Memory), and a ROM 32. (ROM: Read-Only Memory), a large-capacity data storage device 33 such as a hard disk, and a wired communication interface 35. The data storage device 33 is a computer that analyzes the transfer packet data 7a transmitted from the total transfer device 2. A program 34 (hereinafter referred to as “behavior analysis program 34”) is stored.

  When the behavior analysis program 34 analyzes the transfer packet data 7a transmitted from the total transfer device 2, first, the behavior analysis program 34 analyzes the segment value included in the data portion 72a of the transfer packet data 7a, and The passage start time and end time are calculated.

  The behavior analysis program 34 analyzes the segment value included in the data portion 72a of the transfer packet data 7a, and calculates the segment value included in the data portion 72a of the transfer packet data 7a when calculating the passage start time and end time of the person. Reference is made and a character string not including “0” is determined as one pass. When it is desired to analyze the passage of a person based on the distance, a character string having the same number may be determined as a single passage of the distance.

  FIG. 8 is a diagram for explaining the content of calculating the passage start time and end time of a person. FIG. 8 shows only the data part 72a and the reception time 73 of the transfer packet data 7a. The data part 72a of the transfer packet data 7a shown in FIG. Includes a character string “00120033110”.

  The character string “00120033110” includes two character strings of a character string A “12” and a character string B “3311” as character strings that do not include “0”. The character strings B are determined as different passages, and the passage start time and end time are calculated.

When the behavior analysis program 34 calculates the passage start time and end time, as described above, the behavior analysis program 34 performs a reverse calculation using the detection interval from the reception time 73 of the transfer packet data 7a. Then, the passage start time and end time of the passage are calculated.

For example, when the detection interval is 0.2 seconds, the time interval between the characters is 0.2 seconds, so the time of the first character “0” in the character string “00120033110” is “12:00:00”. If there is, the passage start time corresponding to the character string A “12” is the time “12:00:04” of the character “1”, and the end time is the time “12:00:06” of the character “2”. "become. Further, the passage start time corresponding to the character string “3311” is the time “12:00:12” of the character “3”, and the end time is the time “12:00:18” of the character “1”. Become.

  When analyzing the passage of a person based on the distance, the behavior analysis program 34 determines that the character string of the same number is one pass of the distance, and the behavior analysis program 34 starts from the character string “00120033110”. , “1”, “2”, “33”, and “11” are determined to be passed four times, and the passage start time and end time of each passage are calculated.

  The behavior analysis program 34 analyzes the data portion 72a of the transfer packet data 7a to determine passage, calculates the passage start time and end time, and associates them with the device numbers included in the transfer packet data 7a. Data is stored in the data storage device 33.

  When the behavior analysis program 34 of the behavior analysis server 3 analyzes the passage based on the data stored in the data storage device 33, by counting the number of passages for each device number, the human identification specified by the device number By counting the number of people who have passed in front of the sensor device 1, the flow of visitors can be analyzed.

In addition, when displaying the number of people passing by, divided by time zone and day of the week, displaying the number of people passing by a line graph, or memorizing the position information where the human sensor device 1 is installed for each device number, It is good to display the number of passing people on the map.

  Furthermore, by calculating the residence time in which a person has stayed in front of the human sensor device 1 from the passage start time and end time of the passage, it is possible to identify a POP or the like that the visitor has shown interest in. Become.

  Note that the behavior analysis server 3 checks the continuity of the packet numbers included in the transfer packet data 7a received from the total transfer device 2 to total the defect time of the transfer packet data 7, and the continuity of the packet numbers. For the human sensor device 1 that is remarkably inferior, communication between the human sensor device 1 and the total transfer device 2 may be improved by newly installing a total transfer device 2 or the like.

The figure which showed the structure of the behavior analysis system to which this invention was applied. The block diagram of a human sensitive sensor device. The figure explaining an example of the division | segmentation reference | standard regarding the distance regarding a distance. The figure explaining the structure of packet data. The block diagram of an aggregation transfer apparatus. The figure explaining the structure of the transfer packet data. The block diagram of a behavior analysis server. The figure explaining the content which calculates a person's passage start time and end time.

Explanation of symbols

1 Human sensor device 10 Human sensor 14 Packet transmission program 2 Total transfer device 23 Packet transfer program 3 Behavior analysis server 34 Behavior analysis program 4 Terminal device 5 Internet 6 Behavior analysis system 7 Packet data 7a Transfer packet data

Claims (5)

A human sensor device provided with a human sensor for detecting a person's location, detection data of the human sensor is transmitted from the human sensor device, and a tally transfer device including a clock for measuring time; and A behavior analysis system comprising at least a behavior analysis server to which the detection data is transferred from a total transfer device, and constructed a network used for communication between the human sensor device and the total transfer device;
The human sensor device, every time a predetermined detection interval has elapsed, to obtain said detection data from said motion sensor, and transmits the packet data including the detection data in the aggregation transfer device,
The packet data transmitted by the human sensor device to the aggregation transfer device includes a device number for identifying the human sensor device, a packet number assigned in a serial number every time the packet data is generated, Including a mass of the detection data obtained from the human sensor within a packet transmission interval;
The aggregate transfer apparatus, by applying a reception time of the packet data, and transfers the packet data received from the human body sensor device to the behavior analysis server,
The behavior analysis server calculates a detection time of each of the detection data included in the packet data by using the reception time and the detection interval given to the packet data transferred from the total transfer device. A behavior analysis system characterized by being used for analysis. 2. The behavior analysis system according to claim 1 , wherein after the human sensor device is activated, an inspection signal is transmitted by broadcast in the network, and the aggregation transfer device that has responded earliest is selected and selected. A behavior analysis system, wherein the packet data is transmitted to the aggregation transfer device. 3. The behavior analysis system according to claim 2 , wherein when the communication with the selected aggregation transfer device is not possible for a certain period of time, the inspection signal is broadcast again in the network, and the response is earliest. A behavior analysis system characterized by selecting a transfer device. A behavior analysis system according to any one of claims 1 to 3, wherein the human sensor device transmits a plurality of times the packet data in the aggregate transfer device, the aggregate transfer device, the the included in the packet data by using the device number and the bucket number, check the duplication of the packet data received from the motion sensor device, not to retransmit the packet data transferred to the behavior analysis server Characteristic behavior analysis system. A behavior analysis system according to any one of claims 1 to 4, from the behavior analysis server, the packet number included in the packet data and the reception time or the packet data is assigned to, A behavior analysis system comprising means for counting the packet data loss period for each device number.