Crowding level identification device, control method, and program

Abstract

A congestion degree identification device (2000) acquires a captured image (50) generated by an in-vehicle camera that captures an image of the inside of a target vehicle. The congestion degree identification device (2000) identifies, for each of persons (30) present in the target vehicle, an area (20) where the person (30) is positioned from among a plurality of areas (20) in the target vehicle by using the captured image (50). The congestion degree identification device (2000) identifies the congestion degree of the target vehicle by using the number of persons (30) positioned in each of two or more areas (20).

Description

[Technical field]

[0001] The present disclosure relates to a technology for grasping the degree of congestion in a vehicle. [Background technology]

[0002] Systems have been developed to identify the degree of congestion in train carriages. For example, Patent Document 1 discloses a technology for accurately calculating the occupancy rate of a carriage by using images generated by a surveillance camera capturing images of the interior of the carriage and calculating the occupancy rate in two different ways. The first method for calculating the occupancy rate is to use the total value of the area of ​​the carriage occupied by each person. The second method for calculating the occupancy rate is to calculate the occupancy rate based on the ratio between the total number of people riding in the carriage and the maximum number of passengers that the carriage can accommodate. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2013-025523 A Summary of the Invention [Problem to be solved by the invention]

[0004] In the method of Patent Document 1, unless all people in the vehicle are accurately detected, it is difficult to accurately determine the degree of congestion in the vehicle. The present disclosure has been made in consideration of this problem, and its purpose is to provide a new technology for determining the degree of congestion in a vehicle. [Means for solving the problem]

[0005] The congestion level determination device of the present disclosure includes an acquisition unit that acquires an image generated by a camera that images the interior of a target vehicle, a position determination unit that uses the image to determine, for each person in the target vehicle, an area in which the person is located from among multiple areas within the target vehicle, and a congestion level determination unit that determines the degree of congestion of the target vehicle using the number of people located in each of two or more areas.

[0006] The control method of the present disclosure is executed by a computer and includes an acquisition step of acquiring an image generated by a camera that captures an image of the inside of a target vehicle, a position identification step of identifying, for each person in the target vehicle, an area in which the person is located from among a plurality of areas in the target vehicle using the captured image, and a congestion degree identification step of identifying a congestion degree of the target vehicle using the number of people located in each of two or more areas.

[0007] The non-transitory computer-readable medium of the present disclosure stores a program that causes a computer to execute the control method of the present disclosure. Effect of the Invention

[0008] According to the present disclosure, a new technology for identifying the degree of congestion in a vehicle is provided. [Brief description of the drawings]

[0009] [Figure 1] FIG. 2 is a diagram illustrating an example of an outline of an operation of the congestion degree specifying device of the first embodiment. [Diagram 2] 1 is a block diagram illustrating a functional configuration of a congestion degree specifying device according to a first embodiment. [Diagram 3] FIG. 2 is a block diagram illustrating a hardware configuration of a computer that realizes the congestion degree identification device. [Figure 4] 4 is a flowchart illustrating a flow of a process executed by the congestion degree specifying device of the first embodiment. [Diagram 5]FIG. 11 is a diagram illustrating an example of area information in a table format. [Figure 6] FIG. 13 is a diagram illustrating a case where an area overlapping with a person is identified as the area in which the person is located. [Figure 7] 10 is a flowchart illustrating an example of a process flow for identifying a congestion level of a target vehicle based on the number of people in each area and the total number of people detected from a captured image. [Figure 8] 10 is a flowchart illustrating an example of a process flow for identifying a congestion level of a target vehicle based on the number of people in each area and the total number of people detected from a captured image. [Figure 9] FIG. 11 is a diagram illustrating an example of congestion degree information in a table format. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] Hereinafter, an embodiment of the present disclosure will be described in detail with reference to the drawings. In each drawing, the same or corresponding elements are given the same reference numerals, and duplicated explanations are omitted as necessary for clarity of explanation. Furthermore, unless otherwise specified, predetermined values ​​such as predetermined values ​​and threshold values ​​are stored in advance in a storage device accessible from a device that uses the values. Furthermore, unless otherwise specified, the storage unit is composed of one or more arbitrary number of storage devices.

[0011] Fig. 1 is a diagram illustrating an example of an outline of the operation of the congestion level identification device 2000 of the embodiment 1. Here, Fig. 1 is a diagram for facilitating understanding of the outline of the congestion level identification device 2000, and the operation of the congestion level identification device 2000 is not limited to that shown in Fig. 1.

[0012] The congestion degree identification device 2000 identifies the congestion degree of each of one or more cars of a target train. The target train is any train for which the congestion degree of the car is to be identified. Hereinafter, the car for which the congestion degree is to be identified is referred to as a target car. Hereinafter, all of the cars constituting the target train may be treated as target cars, or only some of the cars constituting the target train may be treated as target cars.

[0013] The degree of congestion of the target vehicle is determined based on the number of people 30 in each of the multiple areas 20. The areas 20 are a portion of the target vehicle, such as an area in front of the entrance / exit, a seating area, or an aisle area. For example, in the example of FIG. 1, the target vehicle has seven areas 20-1 to 20-7. Area 20-1 is the area in front of the entrance / exit. Areas 20-2 to 20-5 are seating areas. Areas 20-6 and 20-7 are aisle areas.

[0014] A person 30 is any person in the target vehicle, for example a passenger. However, a person other than a passenger, such as a crew member, who is in the target vehicle, may also be treated as a person 30.

[0015] The congestion degree identification device 2000 acquires a captured image 50. The captured image 50 is generated by a camera (hereinafter, an in-vehicle camera) that is provided in the target vehicle so as to capture an image of the inside of the target vehicle. For example, the in-vehicle camera is set at a relatively high position, such as the ceiling of the target vehicle, so as to overlook the inside of the target vehicle.

[0016] The congestion degree specifying device 2000 analyzes the captured image 50 to specify in which area 20 each person 30 captured by the in-vehicle camera is located. Furthermore, the congestion degree specifying device 2000 specifies the number of people 30 present in each of two or more areas 20, and specifies the degree of congestion in the target vehicle based on the specified number of people.

[0017] It is possible that one target vehicle is provided with multiple in-vehicle cameras that capture images of different locations. For example, in the example of FIG. 1, the target vehicle has four boarding / alighting doors on each of the left and right sides. In this case, four in-vehicle cameras are provided in one vehicle. In such a case, the congestion level of the target vehicle may be determined using all four in-vehicle cameras, or may be determined using only some of the in-vehicle cameras. This point will be described in detail later. In the following explanation, the left and right boarding / alighting doors are treated as one set.

[0018] <Example of action and effect> According to the congestion degree identification device 2000 of this embodiment, it is identified in which of a plurality of areas 20 included in the target vehicle a person 30 inside the target vehicle is located. Then, the congestion degree of the target vehicle is identified based on the number of people 30 inside each of two or more areas 20. In this way, according to the congestion degree identification device 2000 of this embodiment, a new technique for identifying the congestion degree of a vehicle is provided.

[0019] However, because there are obstacles such as advertisements inside the vehicle, it is difficult to capture images of all the people inside the vehicle with the in-vehicle camera. In addition, in the method of Patent Document 1, which determines the degree of congestion by focusing only on the total number of people and the total area inside the vehicle, the presence of people who cannot be detected due to obstacles is thought to have a large impact on the accuracy of the degree of congestion in the vehicle.

[0020] In this regard, the congestion degree specifying device 2000 specifies the number of people 30 in each area 20. Therefore, for example, by focusing on the number of people 30 in areas 20 that are less affected by obstacles (areas 20 where most of the images can be captured by the in-vehicle camera), it is possible to reduce the effect of the presence of people 30 who cannot be detected due to obstacles on the accuracy of the congestion degree of the vehicle.

[0021] Furthermore, the congestion degree identification device 2000 identifies the degree of congestion of the target vehicle based on the number of people 30 in each of two or more areas 20. In this regard, a method of identifying the degree of congestion of the target vehicle by focusing on only one specific area 20 is also conceivable. However, when focusing on only one specific area 20, there is a possibility that the entire vehicle is erroneously determined to be congested when only that area 20 happens to be congested.

[0022] For example, suppose that a group of passengers is riding in a carriage with few passengers. In this case, passengers belonging to the same group are usually considered to be in the same area 20. Therefore, it is possible that a situation may arise in which only a specific area 20 is crowded, even though the carriage as a whole is empty.

[0023] In this regard, the congestion degree specifying device 2000 uses the number of people 30 in each of two or more areas 20. Therefore, even if only a specific area 20 happens to be congested, the conditions of the other areas 20 are also taken into consideration, so that the degree of vehicle congestion can be specified more accurately.

[0024] The congestion level identifying device 2000 of this embodiment will be described in more detail below.

[0025] <Example of functional configuration> 2 is a block diagram illustrating a functional configuration of a congestion level identification device 2000 according to the first embodiment. The congestion level identification device 2000 includes an acquisition unit 2020, a position identification unit 2040, and a congestion level identification unit 2060. The acquisition unit 2020 acquires a captured image 50. The position identification unit 2040 uses the captured image 50 to identify the area 20 in which each person 30 is located. The congestion level identification unit 2060 identifies the number of people 30 present in each of two or more areas 20, and identifies the congestion level of the target vehicle based on the number.

[0026] <Example of hardware configuration> Each functional component of the congestion degree identification device 2000 may be realized by hardware that realizes each functional component (e.g., a hardwired electronic circuit, etc.), or may be realized by a combination of hardware and software (e.g., a combination of an electronic circuit and a program that controls it, etc.). Below, a further description will be given of a case where each functional component of the congestion degree identification device 2000 is realized by a combination of hardware and software.

[0027] 3 is a block diagram illustrating an example of a hardware configuration of a computer 500 that realizes the congestion degree identification device 2000. The computer 500 is any computer. For example, the computer 500 is a stationary computer such as a PC (Personal Computer) or a server machine. In addition, for example, the computer 500 is a portable computer such as a smartphone or a tablet terminal. The computer 500 may be a dedicated computer designed to realize the congestion degree identification device 2000, or may be a general-purpose computer.

[0028] For example, by installing a specific application on the computer 500, each function of the congestion level identification device 2000 is realized by the computer 500. The application is configured with a program for realizing each functional component of the congestion level identification device 2000. The method of acquiring the program is arbitrary. For example, the program can be acquired from a storage medium (such as a DVD disk or a USB memory) in which the program is stored. Alternatively, the program can be acquired by downloading the program from a server device that manages the storage device in which the program is stored.

[0029] The computer 500 has a bus 502, a processor 504, a memory 506, a storage device 508, an input / output interface 510, and a network interface 512. The bus 502 is a data transmission path for the processor 504, the memory 506, the storage device 508, the input / output interface 510, and the network interface 512 to transmit and receive data to and from each other. However, the method of connecting the processor 504 and other components to each other is not limited to a bus connection.

[0030] The processor 504 is a processor such as a central processing unit (CPU), a graphics processing unit (GPU), or a field-programmable gate array (FPGA). The memory 506 is a main storage device realized using a random access memory (RAM) or the like. The storage device 508 is an auxiliary storage device realized using a hard disk, a solid state drive (SSD), a memory card, or a read only memory (ROM) or the like.

[0031] The input / output interface 510 is an interface for connecting the computer 500 to an input / output device. For example, the input / output interface 510 is connected to an input device such as a keyboard and an output device such as a display device.

[0032] The network interface 512 is an interface for connecting the computer 500 to a network. This network may be a LAN (Local Area Network) or a WAN (Wide Area Network).

[0033] The storage device 508 stores a program (a program for realizing the above-mentioned application) for realizing each functional component of the congestion degree identification device 2000. The processor 504 reads this program into the memory 506 and executes it to realize each functional component of the congestion degree identification device 2000.

[0034] The congestion level identification device 2000 may be realized by one computer 500 or may be realized by multiple computers 500. In the latter case, the configurations of the computers 500 do not need to be the same, and may be different from each other.

[0035] <<About the in-car camera>> The in-vehicle camera repeatedly captures images to generate a plurality of captured images 50. Here, the in-vehicle camera may be a video camera that generates video data, or a still camera that generates still images. In the former case, the captured images 50 are video frames that constitute the video data.

[0036] A part or all of the functions of the congestion level identification device 2000 may be realized by an in-vehicle camera. As a camera that realizes a part or all of the functions of the congestion level identification device 2000 in this way, for example, a camera called an IP (Internet Protocol) camera, a network camera, an intelligent camera, or the like can be used.

[0037] When some of the functions of the congestion degree identification device 2000 are realized by an in-vehicle camera, for example, the acquisition unit 2020 and the position identification unit 2040 are realized by the in-vehicle camera. In this case, the in-vehicle camera analyzes a captured image 50 generated by itself to detect people 30 from the captured image 50 and determine in which area 20 the people 30 are located. Then, information indicating in which area 20 each person 30 is located is provided to a device realizing the congestion degree identification unit 2060. Then, the congestion degree of the target vehicle is identified by this device.

[0038] <Processing flow> 4 is a flowchart illustrating a process executed by the congestion level identification device 2000 of the first embodiment. The acquisition unit 2020 acquires a captured image 50 (S102). The position identification unit 2040 detects people 30 from the captured image 50 (S104). For each person 30 detected from the captured image 50, the position identification unit 2040 identifies the area 20 in which the person 30 is located (S106). The congestion level identification unit 2060 identifies the number of people 30 present in each of two or more areas 20, and identifies the degree of congestion of the target vehicle based on the number (S108).

[0039] <Opportunity for identifying the degree of congestion in the target vehicle> The congestion degree of the target vehicle may change over time, so it is preferable for the congestion degree specifying device 2000 to repeatedly specify the congestion degree of the target vehicle (the series of processes exemplified in FIG. 4).

[0040] There are various triggers for identifying the congestion degree of the target vehicle. For example, the congestion degree identifying device 2000 periodically acquires a captured image 50 and identifies the congestion degree of the target vehicle using the captured image 50. In another example, the congestion degree identifying device 2000 acquires a captured image 50 when a specific event occurs and identifies the congestion degree of the target vehicle using the captured image 50.

[0041] An event that triggers the determination of the congestion degree is, for example, an event that "a target train departs from a station." The congestion degree of the target car can change significantly as passengers get on and off the target train when it stops at a station. On the other hand, it is considered that the congestion degree of the target car does not change much while the target train is running. Therefore, by acquiring a captured image 50 in response to the target train leaving a station and using the captured image 50 to grasp the congestion degree of the target car, it is possible to grasp the congestion degree of the target car at an appropriate time.

[0042] It should be noted that immediately after the target train departs from the station, people who boarded the target train just before the doors closed may move around inside the target car. Therefore, the congestion level identification device 2000 may identify the congestion level of the target car after a predetermined time (e.g., 30 seconds) has elapsed since the target train departs, instead of at the time the target train departs. In this way, the congestion level of the target car is identified after the movement of people 30 in the target car slows down. Therefore, the congestion level of the target car can be identified with higher accuracy.

[0043] <Acquisition of captured image 50: S102> The acquiring unit 2020 acquires the captured image 50 (S102). There are various methods for acquiring the captured image 50. For example, the in-vehicle camera is configured to store the generated captured image 50 in a storage unit that can also be accessed by the congestion level identification device 2000. In this case, the acquiring unit 2020 acquires the captured image 50 by accessing the storage unit. In another example, the in-vehicle camera is configured to transmit the captured image 50 to the congestion level identification device 2000. In this case, the acquiring unit 2020 acquires the captured image 50 by receiving the captured image 50 transmitted by the in-vehicle camera. Furthermore, when the position identification unit 2040 is realized by an in-vehicle camera, the in-vehicle camera acquires the captured image 50 generated by itself.

[0044] <Detection of Person 30: S104> The position identifying section 2040 detects the person 30 from the captured image 50 (S104). More specifically, the position identifying section 2040 detects, from the captured image 50, an image area representing the person 30 (hereinafter, a person area).

[0045] Here, various existing methods can be used to detect a person region from an image. For example, a feature amount representing the characteristics of a person in an image is determined in advance and stored in an arbitrary storage unit in a manner that can be acquired by the congestion level identification device 2000. The position identification unit 2040 detects an image region having a feature amount matching the determined feature amount from the captured image 50, and treats each detected image region as a person region. Note that the feature amount of a person may be a feature amount of the entire body, or a feature amount of a characteristic part (for example, a face).

[0046] Alternatively, for example, the position identification unit 2040 may detect a human region from the captured image 50 by using a trained model (hereinafter, a human detection model). The human detection model is trained in advance so as to output a human region included in an input image in response to the input image. Any type of model such as a neural network can be used as the human detection model.

[0047] <Identifying the location of person 30: S106> The position identifying unit 2040 identifies, for each person 30 detected from the captured image 50, in which area 20 the person 30 is located (S106). For example, the position identifying unit 20 identifies an image area on the captured image 50 that represents each area 20. Then, for each person 30 detected from the captured image 50, the position identifying unit 20 identifies the area 20 in which the person 30 is located, based on the person area of ​​the person 30 and the image area that represents each area 20.

[0048] In order to identify the image region on the captured image 50 that represents each area 20, for example, the position identifying unit 2040 acquires information (hereinafter, area information) that represents the positional relationship between the area 20 and the image region on the captured image 50. The area information is generated in advance by an administrator of the congestion degree identifying device 2000 or the like, and is stored in an arbitrary storage unit in a form that can be acquired from the congestion degree identifying device 2000.

[0049] FIG. 5 is a diagram illustrating an example of area information in a table format. The area information 100 in FIG. 5 has three columns: area identification information 102, area name 104, and image area 106. The area identification information 102 indicates the identification information of the area. The area name 104 indicates the name of the area. The image area 106 indicates which part of the captured image 50 the corresponding area 20 corresponds to. Note that in this example, it is assumed that the shape of the area 20 is rectangular. Thus, the image area 106 indicates the coordinates of the upper left and lower right of the image area corresponding to the area 20.

[0050] Here, the size of the captured image 50 may differ depending on the resolution of the in-car camera. Therefore, for example, so that the area information 100 does not depend on the resolution of the in-car camera, the image region 106 may express the position of the corresponding area 20 in relative coordinates on the image. For example, the relative coordinates are expressed based on the vertical or horizontal length of the image. As a specific example, the vertical length of the image is taken as the reference length 1, and the image region 106 indicates "top left: (x1, y1), bottom right: (x2, y2)". In this case, if the vertical length of the captured image 50 is h, the image region representing the corresponding area 20 in the captured image 50 is "top left (h*x1, h*y1), bottom right (h*x2, h *y 2 )".

[0051] The layout of each area 20 on the captured image 50 may differ depending on the type of target vehicle. Therefore, for example, the area information 100 is prepared for each type of vehicle. In this case, the position specifying unit 2040 acquires the captured image 50 corresponding to the type of target vehicle.

[0052] Furthermore, even if the captured image 50 is of the same vehicle, the arrangement of the area 20 in the captured image 50 may differ depending on the position of the in-vehicle camera. For example, the arrangement of the area 20 may differ between a captured image 50 generated by an in-vehicle camera installed on the ceiling near the first entrance and a captured image 50 generated by an in-vehicle camera installed on the ceiling near the second entrance from the front. Therefore, for example, the area information 100 may be prepared for each pair of the type of vehicle and the position of the in-vehicle camera. In this case, the position identification unit 2040 acquires the area information 100 corresponding to "the type of target vehicle and the position of the in-vehicle camera that generated the captured image 50".

[0053] Here, any method may be used to identify the type of the target vehicle and the position of the in-vehicle camera that generated the captured image 50. For example, information in which the identification information of the in-vehicle camera is associated with the type of vehicle in which the in-vehicle camera is installed and information indicating the position within the vehicle in which the in-vehicle camera is installed is stored in advance in any storage unit in a manner that can be acquired from the congestion degree identification device 2000. The position identification unit 2040 can identify the type of the target vehicle and the position of the in-vehicle camera that generated the captured image 50 by acquiring, from this storage unit, information associated with the identification information of the in-vehicle camera that generated the captured image 50.

[0054] After identifying the image region representing each area 20, the position identifying unit 2040 identifies, for each person 30, the area 20 in which the person 30 is located, based on the person region of the person 30 and the image region representing each area 20. For example, the position identifying unit 2040 calculates coordinates representing the position of the person 30 on the captured image 50, based on the person region of the person 30. Then, the position identifying unit 2040 identifies, among the areas 20, the area 20 represented by an image region including the coordinates, as the area 20 in which the person 30 is located. For example, the coordinates representing the position of the person 30 are represented by a predetermined position (e.g., a central position) in the person region of the person 30.

[0055] For example, assume that the predetermined position is the center position. In this case, the position identifying unit 2040 identifies, from among the areas 20, an image region representing the area 20 that includes the center position of the person 30. Then, the position identifying unit 2040 identifies that the person 30 is located in the identified area 20.

[0056] As another example, the position identification unit 2040 identifies, among the areas 20, an area 20 in which an image area representing the area 20 overlaps with an image area representing the person 30, as the area 20 in which the person 30 is located. For the sake of simplicity, hereinafter, the overlap between the image area representing the person 30 and the image area representing the area 20 will also be expressed as "the person 30 and the area 20 overlap."

[0057] Fig. 6 is a diagram illustrating a case where an area 20 overlapping with a person 30 is identified as the area 20 in which the person 30 is located. In Fig. 6, a person area 32 representing the person 30 overlaps with an image area representing an area 20-1. Therefore, the position identifying unit 2040 identifies the person 30 as being located in the area 20-1.

[0058] Here, it is possible that the person 30 overlaps with each of the multiple areas 20. In this case, the position identification unit 2040 identifies the area 20 in which the person 30 is located from among the areas 20 that overlap with the person 30, based on a predetermined rule.

[0059] For example, the predetermined rule may be that "among the areas 20 overlapping with the person 30, the area 20 with the highest priority is identified as the area 20 in which the person 30 is located." In this case, a priority is assigned to each area 20 in advance. For example, the priority is assigned in the order of "area in front of the boarding / alighting entrance, area with seats in the back, area with seats in the front, and area in the aisle." In this case, for example, assume that the person area 32 overlaps with both the area 20 in front of the boarding / alighting entrance and the area 20 with seats in the back. In this case, the position identifying unit 2040 identifies that the person 30 is located in the area 20 in front of the boarding / alighting entrance, which has a higher priority.

[0060] Another predetermined rule is that "the area 20 having the largest area overlapping with the person 30 is identified as the area 20 in which the person 30 is located." In this case, the position identification unit 2040 calculates the area of ​​the overlapping portion between the image area representing each area 20 that overlaps with the person 30 and the person area 32. Then, the position identification unit 2040 identifies the area 20 having the largest calculated area as the area 20 in which the person 30 is located.

[0061] A trained model (hereinafter, area classification model) may be used to identify the area 20 in which the person 30 is located. The area classification model is trained in advance to output identification information of the area 20 in which the person 30 is located in response to input of the captured image 50 and information identifying the person area 32 of the person 30 (for example, the coordinates of the upper left and upper right). In this case, the position identification unit 2040 identifies the area 20 in which each person 30 detected from the captured image 50 is located, using the area classification model. Here, when the area classification model is used, it is not necessary to identify the image area representing each area 20 using the area information 100.

[0062] The area classification model is trained in advance using multiple training data. The training data has, for example, a pair of "an image captured by an in-vehicle camera and information for identifying a person area" as input data, and also has a correct label (identification information of the area 20) as correct output data. Any type of model such as a neural network can be used for the area classification model.

[0063] As described above, the location of the area 20 may differ depending on the type of vehicle and the position of the in-vehicle camera (such as the front entrance). Therefore, for example, an area classification model is prepared for each pair of "vehicle type, in-vehicle camera position." The position identification unit 2040 inputs the captured image 50 and the position of the person region of the person 30 to an area classification model corresponding to the type of target vehicle and the position of the in-vehicle camera that generated the captured image 50 (such as "the front entrance"). The position identification unit 2040 then acquires the identification information of the area 20 output from the area classification model, and identifies that the person 30 is located in the area 20 identified by the identification information.

[0064] <Identifying the degree of congestion: S108> The congestion level identification unit 2060 identifies the degree of congestion of the target vehicle based on the number of people in each of the two or more areas 20 (S108). Below, several specific examples of methods for identifying the degree of congestion of the target vehicle will be described.

[0065] For example, the order of the areas 20 to be used when determining the congestion degree is determined in advance. Then, the congestion degree identification unit 2060 compares the number of people 30 in each of two or more areas 20 with a threshold value according to the predetermined order, and identifies the congestion degree of the target vehicle based on the comparison result.

[0066] Here, the congestion level identification unit 2060 may further use the total number of people 30 detected from the captured image 50 to identify the degree of congestion in the target vehicle. For example, the congestion level identification unit 2060 calculates the total value of the numbers of people 30 in all areas 20, and treats this total value as the total number of people 30. However, the congestion level identification unit 2060 may treat the total value of the numbers of people 30 in each of any number of areas 20, two or more, as the total number of people 30, instead of all areas 20.

[0067] 7 and 8 are flowcharts illustrating a process for identifying the degree of congestion in a target vehicle based on the number of people 30 in each area 20 and the total number of people 30 detected from a captured image 50. In the examples of FIGS. 7 and 8, the degree of congestion is classified into levels 1 to 5, with the higher the level number, the higher the degree of congestion. That is, level 5 is the most congested state, and level 1 is the least congested state. The area captured by the in-vehicle camera includes seven areas 20, as illustrated in FIG. 1.

[0068] First, Fig. 7 will be described. The congestion level identification unit 2060 determines whether or not the condition that "the total number of people 30 detected from the captured image 50 is equal to or less than a threshold Th1, and the number of people 30 detected from the boarding / alighting area is equal to or less than a threshold Th2" is satisfied (S202). If this condition is not satisfied (S202: NO), S212 shown in Fig. 8 is executed. Fig. 8 will be described later.

[0069] If the condition of S202 is satisfied (S202: YES), the congestion level identification unit 2060 determines whether or not the condition that "the total number of people is equal to or less than a threshold value Th3, and the number of people in the rear seating area is equal to or less than Th4" is satisfied (S204). The threshold value Th3 is set to satisfy, for example, Th1>Th3. If the condition of S204 is satisfied (S204: YES), the congestion level identification unit 2060 determines that the congestion level of the target vehicle is level 1. On the other hand, if the condition of S204 is not satisfied (S204: NO), S208 is executed.

[0070] Here, in the captured image 50 of FIG. 1, there are two seating areas at the back, area 20-2 and area 20-3. For the comparison with the threshold value Th4, only one of these two areas 20 may be used, or both may be used. In the former case, it is preferable to use the area 20 (area 20-3 in FIG. 1) of the two areas 20 which has fewer obstacles such as advertisements and allows the number of people 30 therein to be more accurately determined. Since the congestion degree identification device 2000 employs a method of dividing the target vehicle into a plurality of areas 20 and identifying the number of people in each area 20, it is possible to focus on the area 20 in which the number of people 30 can be accurately determined in this manner.

[0071] Here, the position of the advertisement is determined, for example, by the type of vehicle and the position of the in-vehicle camera. For example, for a pair of the type of vehicle and the position of the in-vehicle camera, which of the area 20-2 and the area 20-3 is to be used for comparison with the threshold value Th4 is determined in advance. The congestion degree identification unit 2060 identifies which of the area 20-2 and the area 20-3 is to be used as the back seat area based on the type of the target vehicle and the position of the in-vehicle camera that generated the captured image 50.

[0072] Alternatively, for example, the congestion level determination unit 2060 may compare the number of people 30 in the area 20-2 with the number of people 30 in the area 20-3, and use the greater number for comparison with the threshold value Th4. This is because it is considered that the more accurately the number of people 30 can be detected in an area 20, the greater the number of people 30 that will be detected.

[0073] It is preferable to set the threshold value Th4 to be different when using one area 20 and when using two areas 20. However, this does not apply when a statistical value (average value, maximum value, etc.) of the number of people in the two areas 20 is used instead of the sum of the number of people in the two areas 20.

[0074] If the condition of S204 is not satisfied (S204: NO), the congestion level identification unit 2060 determines whether or not the condition that "the total number of people is equal to or less than a threshold Th3, and the number of people in the front seat area is equal to or less than a threshold Th5" is satisfied (S208). If the condition of S208 is satisfied (S208: YES), the congestion level identification unit 2060 determines that the congestion level of the target vehicle is level 1 (S206). On the other hand, if the condition of S208 is not satisfied (S208: NO), the congestion level identification unit 2060 determines that the congestion level of the target vehicle is level 2 (S210). Here, Th4 and Th5 may be set to the same value or different values.

[0075] 1, there are two front seating areas, similar to the back seating areas. Thus, similar to the front seating areas, one or both of these two areas 20 may be used for comparison with the threshold value Th5. In the former case, the method of determining the area to be used for comparison with the threshold value Th5 from the two front seating areas is the same as the method of determining the seating area to be used for comparison with the threshold value Th4 from the two back seating areas.

[0076] Here, in the process flow of Fig. 7, the determination focusing on the number of people in the back seating area (S204) is performed before the determination focusing on the number of people in the front seating area (S208). This is because the face of a person 30 located in the back seating area is captured more clearly by the in-car camera than that of a person 30 located in the front seating area, and therefore can be detected with higher accuracy. In this way, by performing the determination focusing on the area where the detection accuracy of the person 30 is high first, the amount of processing performed by the congestion degree identification device 2000 can be reduced. The time required for the congestion degree identification device 2000 to perform processing can be shortened.

[0077] However, it is not essential to perform the determination focusing on the number of people in the back seating area first. Therefore, the order of the determination focusing on the number of people in the front seating area and the determination focusing on the number of people in the back seating area may be reversed from that shown in FIG.

[0078] Next, Fig. 8 will be described. If the condition of S202 in Fig. 7 is not satisfied (S202: NO), the congestion level identification unit 2060 determines whether or not the condition that "the total number of people is equal to or less than a threshold Th6, and the number of people 30 detected from the boarding / alighting area is equal to or less than a threshold Th7" is satisfied (S212). Here, the threshold Th6 is set so as to satisfy, for example, Th6>Th1. Moreover, the threshold Th7 is set so as to satisfy, for example, Th7>Th2.

[0079] If the condition of S212 is satisfied (S212: YES), the congestion level identification unit 2060 identifies the congestion level as level 3. On the other hand, if the condition of S212 is not satisfied (S212: NO), the congestion level identification unit 2060 determines whether or not the condition "the total number of people is equal to or less than a threshold value Th8" is satisfied (S216). Here, the threshold value Th8 is set so as to satisfy, for example, Th8>Th6.

[0080] If the condition of S216 is satisfied (S216: YES), the congestion degree identification unit 2060 determines that the congestion degree of the target vehicle is level 4. On the other hand, if the condition of S216 is not satisfied (S216: NO), the congestion degree identification unit 2060 determines that the congestion degree of the target vehicle is level 5 (S220).

[0081] The magnitude relationship between the threshold values ​​described above is merely an example and is not essential.

[0082] The method of identifying the degree of congestion of the target vehicle is not limited to a method of comparing the number of people 30 in the area 20 with a threshold value. For example, a score (hereinafter, a congestion score) representing the degree of congestion of the target vehicle may be calculated from the number of people 30 in each area 20, and the congestion degree of the target vehicle may be identified based on the congestion score. In this case, for example, a regression model for calculating the congestion score from the number of people 30 in each area 20 is determined in advance. The congestion degree identification unit 2060 can calculate the congestion degree of the target vehicle by inputting the number of people 30 in each area 20 into the regression model. For example, this regression model is expressed by the following formula (1).

number

[0083] The regression model is trained in advance using training data, such as the number of people in each area 20 and the correct congestion score. Through this training, the weight α[i] assigned to each area 20 is determined.

[0084] Here, the weighting assigned to each area 20 is considered to reflect whether or not the number of people 30 in that area 20 can be accurately grasped. For an area 20 in which it is difficult to accurately grasp the number of people 30 due to obstacles such as advertisements, the correlation between the degree of vehicle congestion and the number of people 30 detected therein is considered to be relatively small. Therefore, in the regression model obtained as a result of learning, the weighting assigned to such an area 20 is considered to be relatively small. On the other hand, for an area 20 in which the number of people 30 can be accurately grasped, the correlation between the degree of vehicle congestion and the number of people 30 therein is considered to be relatively large. Therefore, in the regression model obtained as a result of learning, the weighting assigned to such an area 20 is considered to be relatively large.

[0085] In this way, the congestion degree specifying device 2000 employs a method of dividing the target vehicle into a plurality of areas 20 and specifying the number of people in each area 20, thereby making it possible to grasp the effect on the congestion degree for each area 20. Therefore, it is possible to specify the congestion degree of the target vehicle more accurately, compared to a case where the congestion degree of the target vehicle is specified by focusing on the number of people in the entire target vehicle.

[0086] Although equation (1) is a linear regression model, the model for calculating the congestion degree score is not limited to a linear regression model.

[0087] Here, the congestion degree identification unit 2060 may convert the congestion degree score into the above-mentioned congestion degree level. For example, the numerical range of the congestion degree score is divided into a plurality of partial ranges that do not overlap with each other, and a level is assigned to each partial range in advance. In this case, after calculating the congestion degree score, the congestion degree identification unit 2060 identifies the partial range that includes the congestion degree score, and identifies the congestion degree level corresponding to the identified partial range as the congestion degree level of the target vehicle.

[0088] In the above description, the congestion level of the target vehicle is identified based on the number of people 30 detected from one captured image 50. However, as described above, the target vehicle may be equipped with multiple in-vehicle cameras. Therefore, multiple captured images 50 may be generated for the target vehicle at a specific time point. Therefore, the congestion level identification device 2000 may identify the congestion level of the target vehicle by using the captured images 50 obtained from one or more of the multiple in-vehicle cameras of the target vehicle.

[0089] For example, the congestion level identification device 2000 uses only one of the multiple in-vehicle cameras installed in the target vehicle to identify the congestion level of the target vehicle. In this case, the congestion level of the target vehicle is identified from one captured image 50 by the various methods described above.

[0090] Alternatively, for example, the congestion level identification device 2000 uses the captured images 50 obtained from each of the multiple in-vehicle cameras to perform a process of identifying the congestion level of the target vehicle described above for each captured image 50, and identifies the overall congestion level based on the results. Specifically, the congestion level identification device 2000 uses a statistical value (such as an average value, a mode value, a median value, a maximum value, or a minimum value) of the congestion level identified for each captured image 50 as the overall congestion level of the target vehicle. Hereinafter, the congestion level identified for each captured image 50 is also referred to as a partial congestion level. In addition, the overall congestion level of the target vehicle identified using the partial congestion levels identified by each in-vehicle camera in the target vehicle is also referred to as an overall congestion level.

[0091] For example, assume that there are four sets of entrances in a target vehicle. Assume that the in-vehicle cameras are installed at four locations, namely, near the first entrance, near the second entrance from the front, near the third entrance from the front, and near the last entrance. In this case, the congestion degree identification device 2000 identifies the partial congestion degree for each of these four locations, and then identifies the overall congestion degree using these statistics. Note that, when part of the function of the congestion degree identification device 2000 is realized by the in-vehicle cameras, each in-vehicle camera may identify the partial congestion degree, and the results may be collected to identify the overall congestion degree. The device that identifies the overall congestion degree may be any of the in-vehicle cameras, or may be another device (such as a server device connected to each in-vehicle camera so as to be able to communicate with them).

[0092] Alternatively, for example, the congestion degree identification device 2000 may treat a set of partial congestion degrees identified for the target vehicle as information representing the congestion degree of the target vehicle. In the above example, the congestion degree identification device 2000 identifies a partial congestion degree for each of the four in-vehicle cameras, and treats the set of the identified four partial congestion degrees as the congestion degree of the target vehicle.

[0093] <Result output> The congestion degree identifying device 2000 generates and outputs information representing the results of the various processes described above. For example, the congestion degree identifying device 2000 identifies the congestion degree of each car of the target train by treating each car of the target train as a target car. The congestion degree identifying device 2000 then generates and outputs information indicating the congestion degree of each car of the target train (hereinafter, congestion degree information). However, the congestion degree information may be generated for only specific cars, rather than for all cars of the target train.

[0094] FIG. 9 is a diagram illustrating congestion degree information in a table format. The congestion degree information 200 in FIG. 9 has four columns: vehicle identification information 202, entrance / exit identification information 204, partial congestion degree 206, and overall congestion degree 208. The vehicle identification information 202 indicates the identification information of the vehicle. The entrance / exit identification information 204 indicates a number assigned to the entrance / exit. Note that, in the example of FIG. 9, it is assumed that one vehicle is provided with four sets of entrances, and a partial congestion degree is identified for each set. The partial congestion degree 206 indicates the partial congestion degree identified for the corresponding set of entrances. The overall congestion degree 208 indicates the overall congestion degree identified for the corresponding vehicle.

[0095] For example, the congestion degree identifying device 2000 generates congestion degree information 200 for each of a plurality of trains. The congestion degree identifying device 2000 also generates congestion degree information 200 for each different time for one train. For example, the congestion degree information 200 is generated periodically or at the time of departure from each station. The congestion degree information 200 is output in association with the train identification information and the time of generation. For example, the congestion degree information 200 generated for train R1 at time T1 is output in association with the pair of "train identification information=R1, time=T1".

[0096] The congestion degree information 200 may be output in various ways. For example, the congestion degree information 200 may be stored in an arbitrary storage unit. Alternatively, for example, the congestion degree information 200 may be displayed on an arbitrary display device. Alternatively, for example, the congestion degree information 200 may be transmitted to an arbitrary terminal. For example, the terminal may be a customer terminal, a driver terminal, or a terminal installed in a facility that manages train operations. The congestion degree information 200 received by the terminal may be displayed on a display device of the terminal.

[0097] For example, a customer can specify a train using a web page or a specific application on his / her own terminal to know the congestion level of the train. The customer's terminal transmits a request indicating the identification information of the specified train to the congestion level identifying device 2000. The congestion level identifying device 2000, which receives this request, generates congestion level information 200 for the specified train and transmits it to the customer's terminal. The customer can view the received congestion level information 200 to know the congestion level of the train he / she wants to use.

[0098] Instead of generating the congestion degree information 200 in response to a request from a customer, the congestion degree identifying device 2000 may generate the congestion degree information 200 at various times as described above and store it in the storage unit. In this case, the congestion degree identifying device 2000 reads out the congestion degree information 200 that matches the request from the customer from the storage unit and provides the read congestion degree information 200 to the customer.

[0099] When viewed by a customer or the like, the congestion degree information 200 is preferably converted into a format that makes the information easy to understand by using a picture, a graphic, etc. This conversion may be performed by the congestion degree identification device 2000 or by each terminal that receives the congestion degree information 200.

[0100] The congestion degree information 200 does not necessarily need to be provided in real time. For example, the congestion degree identification device 2000 generates the congestion degree information 200 at each of a plurality of timings once a day for each car of each train operated on that day. The results can be used, for example, by a railway company for the purpose of business management. For example, by grasping the degree of congestion of each train for each day of the week and each time period, fares can be appropriately set according to the day of the week and each time period.

[0101] Although the present invention has been described above with reference to the embodiment, the present invention is not limited to the above embodiment. Various modifications that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.

[0102] In the above example, the program includes a set of instructions (or software code) for making the computer perform one or more functions described in the embodiment when the program is loaded into the computer. The program may be stored in a non-transitory computer-readable medium or a tangible storage medium. By way of example and not limitation, the computer-readable medium or tangible storage medium includes random-access memory (RAM), read-only memory (ROM), flash memory, solid-state drive (SSD) or other memory technology, CD-ROM, digital versatile disc (DVD), Blu-ray (registered trademark) disk or other optical disk storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage device. The program may be transmitted on a transitory computer-readable medium or communication medium. By way of example and not limitation, the transitory computer-readable medium or communication medium includes electrical, optical, acoustic, or other forms of propagating signals.

[0103] A part or all of the above-described embodiments can be described as, but is not limited to, the following supplementary notes. (Appendix 1) An acquisition unit that acquires an image generated by a camera that captures an image of the inside of a target vehicle; a position identification unit that identifies, for each person in the target vehicle, an area in which the person is located from among a plurality of areas in the target vehicle, using the captured image; and a congestion degree identification unit that identifies a degree of congestion of the target vehicle by using the number of people located in each of the two or more areas. (Appendix 2) Area information in which the position of an image region representing each of the areas in the captured image is associated with the area is stored in a storage unit; The position identification unit is Using the area information, an image region representing each of the areas is identified in the captured image; Detecting a person area representing each of the people from the captured image; 2. The congestion degree identification device according to claim 1, wherein for each person, the area in which the corresponding image area includes the person's person area is identified as the area in which the person is located. (Appendix 3) A priority order is set for the plurality of areas; The position identification unit is 3. The congestion degree identification device according to claim 2, wherein, when there are multiple areas in which the person's area is included in the corresponding image area, the area with the highest priority among the multiple areas is identified as the area in which the person is located. (Appendix 4) The storage unit stores the area information for each type of vehicle, The congestion degree identification device according to claim 2 or 3, wherein the position identification unit uses the area information corresponding to the type of the target vehicle. (Appendix 5) The congestion level determination unit compares the total number of people detected from the captured image and the number of people located in the area representing the boarding / exiting door of the target vehicle with a threshold value, and determines the degree of congestion of the target vehicle based on the result of the comparison. (Appendix 6) The congestion degree identification unit, when the total number is equal to or less than a first threshold and the number of people located in the area representing the boarding / alighting entrance is equal to or less than a second threshold, compares the total number with a third threshold and compares the number of people located in the area representing the seats with a fourth threshold, and identifies the degree of congestion of the target vehicle based on the result of the comparison. (Appendix 7) The congestion degree identification device according to any one of appendixes 1 to 6, wherein, when there are multiple areas of the same type, the congestion degree identification unit uses the number of people in an area among the multiple areas in which the people can be detected most accurately to identify the degree of congestion of the target vehicle. (Appendix 8) 1. A computer-implemented control method comprising: An acquisition step of acquiring an image generated by a camera that captures an image of the inside of a target vehicle; a position identification step of identifying, for each person in the target vehicle, an area in which the person is located from among a plurality of areas in the target vehicle, using the captured image; and a congestion level determination step of determining a congestion level of the target vehicle using the number of people located in each of the two or more areas. (Appendix 9) Area information in which the position of an image region representing each of the areas in the captured image is associated with the area is stored in a storage unit; In the position specifying step, Using the area information, an image region representing each of the areas is identified in the captured image; Detecting a person area representing each of the people from the captured image; The control method of claim 8, wherein for each person, the area in which the corresponding image area includes the person's person area is identified as the area in which the person is located. (Appendix 10) A priority order is set for the plurality of areas; In the position specifying step, 10. The control method of claim 9, wherein, when there are multiple areas in which the person's area is included in the corresponding image area, the area with the highest priority among the multiple areas is identified as the area in which the person is located. (Appendix 11) The storage unit stores the area information for each type of vehicle, The control method according to claim 9 or 10, wherein the location identification step utilizes area information corresponding to the type of the target vehicle. (Appendix 12) A control method described in any one of Appendix 8 to 11, in which, in the congestion degree determination step, the total number of people detected from the captured image and the number of people located in the area representing the boarding / exiting door of the target vehicle are each compared with a threshold value, and the degree of congestion of the target vehicle is determined based on the result of the comparison. (Appendix 13) The control method described in Appendix 12, in which, in the congestion degree determination step, if the total number is equal to or less than a first threshold and the number of people located in the area representing the boarding / alighting entrance is equal to or less than a second threshold, the total number is compared with a third threshold, and the number of people located in the area representing the seats is compared with a fourth threshold, and the degree of congestion of the target vehicle is determined based on the result of the comparison. (Appendix 14) The control method according to any one of appendixes 8 to 13, wherein, in the congestion level determination step, if there are multiple areas of the same type, the number of people in an area among the multiple areas in which the people can be detected most accurately is used to determine the congestion level of the target vehicle. (Appendix 15) On the computer, An acquisition step of acquiring an image generated by a camera that captures an image of the inside of a target vehicle; a position identification step of identifying, for each person in the target vehicle, an area in which the person is located from among a plurality of areas in the target vehicle, using the captured image; A non-transitory computer-readable medium having stored thereon a program for executing a congestion level determination step of determining the congestion level of the target vehicle using the number of people located in each of two or more of the areas. (Appendix 16) Area information in which the position of an image region representing each of the areas in the captured image is associated with the area is stored in a storage unit; In the position specifying step, Using the area information, an image region representing each of the areas is identified in the captured image; Detecting a person area representing each of the people from the captured image; 16. The computer-readable medium of claim 15, wherein for each person, the area in which the corresponding image region contains the person's person region is identified as the area in which the person is located. (Appendix 17) A priority order is set for the plurality of areas; In the position specifying step, 17. The computer-readable medium of claim 16, wherein if there are multiple areas in which the person's person area is included in the corresponding image area, the area with the highest priority among the multiple areas is identified as the area in which the person is located. (Appendix 18) The storage unit stores the area information for each type of vehicle, 18. The computer-readable medium of claim 16 or 17, wherein the location step utilizes area information corresponding to a type of vehicle of the target. (Appendix 19) A computer-readable medium as claimed in any one of Appendix 15 to 18, wherein in the congestion level determination step, the total number of people detected from the captured image and the number of people located in the area representing the boarding and alighting entrance of the target vehicle are each compared with a threshold value, and the degree of congestion of the target vehicle is determined based on the result of the comparison. (Appendix 20) The computer-readable medium of claim 19, wherein in the congestion level determination step, if the total number is less than or equal to a first threshold and the number of people located in the area representing the boarding / alighting entrance is less than or equal to a second threshold, the total number is compared with a third threshold and the number of people located in the area representing the seats is compared with a fourth threshold, and the congestion level of the target vehicle is determined based on the result of the comparison. (Appendix 21) 21. The computer-readable medium of claim 15, wherein in the congestion level determination step, if there are multiple areas of the same type, the number of people in the area among the multiple areas in which the people can be detected most accurately is used to determine the congestion level of the target vehicle. [Explanation of symbols]

[0104] 20 Area 30 people 32 People area 50 Captured images 100 Area Information 102 Area Identification Information 104 Area Name 106 Image Area 200 Congestion Information 202 Vehicle Identification Information 204 Boarding / Alighting Gate Identification Information 206 Partial congestion degree 208 Overall congestion level 500 Computers 502 Bus 504 Processor 506 Memory 508 Storage Devices 510 Input / Output Interface 512 Network Interface 2000 Congestion level identification device 2020 Acquisition Department 2040 Location identification part 2060 Congestion Degree Identification Section

Claims

1. An acquisition unit that acquires a captured image showing the interior of a target vehicle; a position identification unit that identifies an area in which each person in the target vehicle is located from among a plurality of areas in the target vehicle based on the captured image; and a congestion degree identification unit that identifies the degree of congestion of the target vehicle using the number of people located in an area where the people can be detected most accurately among multiple areas of the same type.

2. a storage unit stores area information in which each of the areas is associated with a position of an image region representing the area in the captured image; The position identification unit is Using the area information, an image region representing each of the areas is identified in the captured image; Detecting a person area representing each of the people from the captured image; The crowdedness level specifying device according to claim 1 , wherein for each of the persons, the area in which the corresponding image area includes the person area of ​​the person is specified as the area in which the person is located.

3. A priority order is set for the plurality of areas; The position identification unit is The congestion degree identification device according to claim 2, wherein when there are multiple areas in which the person's area is included in the corresponding image area, the area with the highest priority among the multiple areas is identified as the area in which the person is located.

4. The storage unit stores the area information for each type of vehicle, The congestion level specifying device according to claim 2 , wherein the location specifying unit uses the area information corresponding to a type of the target vehicle.

5. 5. The congestion degree determination device according to claim 1, wherein the congestion degree determination unit compares the total number of people detected from the captured image and the number of people located in the area representing the boarding and alighting entrance of the target vehicle with a threshold value, and determines the degree of congestion of the target vehicle based on the result of the comparison.

6. 6. The congestion degree identification device according to claim 5, wherein, when the total number is equal to or less than a first threshold and the number of the people located in the area representing the boarding / alighting entrance is equal to or less than a second threshold, the congestion degree identification unit compares the total number with a third threshold and compares the number of the people located in the area representing the seats with a fourth threshold, and identifies the degree of congestion of the target vehicle based on a result of the comparison.

7. 1. A computer-implemented control method comprising: An acquisition step of acquiring a captured image showing the interior of a target vehicle; a position specifying step of specifying an area in which each person in the target vehicle is located from among a plurality of areas in the target vehicle based on the captured image; and a congestion degree determination step of determining the degree of congestion of the target vehicle using the number of people located in an area where the people can be detected most accurately among multiple areas of the same type.

8. a storage unit stores area information in which each of the areas is associated with a position of an image region representing the area in the captured image; In the position specifying step, Using the area information, an image region representing each of the areas is identified in the captured image; Detecting a person area representing each of the people from the captured image; The control method according to claim 7 , further comprising identifying, for each of the persons, the area in which the corresponding image area includes the person area of ​​the person as the area in which the person is located.

9. On the computer, An acquisition step of acquiring a captured image showing the interior of a target vehicle; a position specifying step of specifying an area in which each person in the target vehicle is located from among a plurality of areas in the target vehicle based on the captured image; A program that executes a congestion degree determination step of determining the degree of congestion of the target vehicle using the number of people located in an area where the people can be detected most accurately among multiple areas of the same type.

Images