Bus boarding and alighting counting system and method

Abstract

To provide a bus embarkation / disembarkation counting system capable of relatively simply counting passengers who embark and disembark a bus.SOLUTION: The bus embarkation / disembarkation counting system comprises: a dashboard camera system 4 that at least records image data and audio data of a bus 5; and an embarkation / disembarkation counting unit 1 that, on the basis of information acquired from the dashboard camera system, identifies a bus stop 7 at which a door is opened and closed, and counts the number of passengers at each identified bus stop by analyzing, from among video data, video data at the identified bus stop.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a bus boarding and alighting counting system and method.

Background Art

[0002] Bus operators need to conduct boarding and alighting surveys for optimizing route schedules, applying for subsidies, and reporting transportation performance. However, usually, the boarding and alighting surveys are conducted by investigators, so the survey cost is high, and there are limitations in the surveyable period and routes.

[0003] For this reason, a technique for automatically performing boarding and alighting counting using sensors has been proposed (Patent Document 1). However, since it requires not only a camera for photographing boarding and alighting passengers but also a door sensor for detecting door opening and closing, it is laborious in electrical wiring inside the vehicle and increases the cost.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] The invention described in Patent Document 1 is a technique for improving the accuracy of boarding and alighting counting by camera images, and it is difficult to easily conduct boarding and alighting surveys for all buses on all routes regardless of the period.

[0006] The present invention has been made in view of the above problems, and an object thereof is to provide a bus boarding and alighting counting system and method that can perform boarding and alighting counting relatively easily.

Means for Solving the Problems

[0007] To solve the above problems, a bus boarding and alighting counting system according to one aspect of the present invention comprises a drive recorder system that records at least video data and audio data of the bus, and a boarding and alighting counting unit that identifies bus stops where doors have been opened and closed based on information obtained from the drive recorder system, and counts the number of passengers boarding and alighting at each identified bus stop by analyzing the video data at the identified bus stops. [Effects of the Invention]

[0008] According to the present invention, based on information from a drive recorder system, it is possible to identify bus stops where doors have been opened and closed, and to count the number of passengers getting on and off at each identified bus stop. [Brief explanation of the drawing]

[0009] [Figure 1] This is an overall diagram of the bus boarding and alighting counting system. [Figure 2] This is an explanatory diagram showing the placement of cameras and microphones inside a bus. [Figure 3] This is a hardware configuration diagram of the bus passenger boarding / alighting counting system. [Figure 4] This is an example of data stored on a storage medium that is detachably installed in a drive recorder system. [Figure 5] This is the data format for the stop information storage unit of the passenger boarding / alighting counting server. [Figure 6] This is a flowchart of the process for building a model to recognize voice announcements at bus stops. [Figure 7] This is a flowchart for passenger boarding and alighting counting using voice recognition. [Figure 8] This is a flowchart for the process of analyzing video footage from the front entrance / exit. [Figure 9] This is a flowchart for the process of analyzing video footage from the rear entrance / exit. [Figure 10] This is a flowchart of the process for identifying bus stops using voice recognition. [Figure 11]It is a flowchart of a process for associating boarding / alighting analysis results with voice recognition results of stop names. [Figure 12] It is a flowchart of boarding / alighting count processing using the travel distance according to Example 2. [Figure 13] It is a flowchart of a process for analyzing the video of the front boarding / alighting opening. [Figure 14] It is a flowchart of a process for analyzing the video of the rear boarding / alighting opening. [Figure 15] It is a flowchart of a process for associating boarding / alighting analysis results with recognition results of stop names. [Figure 16] It is a flowchart of boarding / alighting count processing using location information according to Example 3. [Figure 17] It is a flowchart of a process for analyzing the video of the front boarding / alighting opening. [Figure 18] It is a flowchart of a process for analyzing the video of the rear boarding / alighting opening. [Figure 19] It is a flowchart of boarding / alighting count processing using the appearance information of the stop according to Example 4. [Figure 20] It is a flowchart of a process for registering the appearance data of the stop. [Figure 21] It is a flowchart of a process for analyzing the video of the front boarding / alighting opening. [Figure 22] It is a flowchart of a process for analyzing the video of the rear boarding / alighting opening. [Figure 23] It is a flowchart of a process for detecting the opening and closing of the door according to Example 5. [Figure 24] It is an overall configuration diagram of a bus boarding / alighting count system according to Example 6. [Figure 25] It is a configuration diagram of a bus boarding / alighting count system according to Example 7.

Modes for Carrying Out the Invention

[0010] Embodiments of the present invention will be described below with reference to the drawings. In the bus boarding and alighting counting system according to this embodiment, the bus stops where the bus has stopped are identified based on information obtained from the drive recorder system, and the number of passengers who boarded or alighted at the bus stops is counted.

[0011] In this embodiment, as described later, the video data in the video file captured by the drive recorder system 4 is used to count the number of users 6 getting on and off the bus 5 at each stop, and the boarding and alighting summary data D3 is provided to the bus operator.

[0012] Video files (video data and audio data) captured by cameras 41 installed at the front entrance / exit 51F, rear entrance / exit 51R, driver's seat, and rear of the vehicle are stored in a storage medium 3 that is detachably attached to the control device 42. After the end of each day's work, the storage medium 3 is removed from the control device 42 by the crew and connected to the bus operator's depot terminal 2, thereby allowing the data stored during the operation of the bus 5 to be read into the depot terminal 2. The bus operator's depot terminal 2 transmits the data acquired from the storage medium 3 (video file D1, described later in Figure 4) to the passenger count management server 1.

[0013] The passenger boarding / alighting count management server 1 analyzes the video file D1 received from the office terminal 2, compares it with the bus stop information D2 to identify bus stop 7, and counts the number of people who boarded or alighted at the identified bus stop 7 by analyzing the video. It then outputs passenger boarding / alighting summary data D3 showing the number of passengers boarding or alighting at each bus stop.

[0014] The bus passenger boarding and alighting counting system can provide passenger boarding and alighting aggregate data D3 to bus operators via a web server (not shown in the diagram). Passenger boarding and alighting aggregate data can also be provided to bus operators using electronic communication methods such as email. The office terminal 2 and the passenger boarding and alighting counting management server 1 may be operated by the bus operator (on-premise), or the bus operator may operate the office terminal 2 and the passenger boarding and alighting counting management server 1 may be operated by a passenger boarding and alighting counting service provider on a so-called cloud. In the latter case, the passenger boarding and alighting counting management server 1 can receive data D1 from each bus 5 from multiple different bus operators, analyze the number of passengers boarding and alighting, and provide passenger boarding and alighting aggregate data D3 to each bus operator.

[0015] In route buses, which are corporate vehicles, the installation of a drive recorder system 4 is mandatory. The drive recorder system 4 stores video data of the interior and exterior of the bus 5 and audio data from inside the bus as video files. The passenger boarding and alighting counting system of this embodiment can identify the bus stops where the bus 5 has stopped and count the users (passengers) who board or alight at the identified bus stops, based solely on the information (video files and sensor data) that can be obtained from the drive recorder system 4.

[0016] Therefore, according to the passenger boarding and alighting counting system of this embodiment, passenger boarding and alighting can be performed easily and at low cost for all buses on all routes, without imposing a burden on the crew, and without limiting the period or time of day, thereby improving convenience for bus operators. In other words, in this embodiment, changes in the number of passengers on a vehicle can be measured for each stopping position of the vehicle, based only on information obtained from specific equipment (e.g., a drive recorder system) installed on the vehicle that is subject to passenger boarding and alighting counting. [Examples]

[0017] The first embodiment will be explained using Figures 1 to 11. Figure 1 is an overall configuration diagram of the bus passenger boarding and alighting counting system. The bus passenger boarding and alighting counting system comprises, for example, a passenger boarding and alighting counting management server 1, a depot terminal 2, a storage medium 3, and a drive recorder system 4, as will be described later. The bus 5, which is the target vehicle, temporarily stops at bus stops 7 that are appropriately set up for each route and where there are passengers boarding or alighting, allowing passengers to get on and off. Bus stops 7 usually have signs, but even if there are no clear signs, if the bus 5 stops at a nearly fixed position and allows passengers to get on and off, it can be considered a bus stop and handled by the bus passenger boarding and alighting counting system.

[0018] The passenger count management server 1 is a server computer located in a so-called cloud, and includes, for example, a data acquisition unit 11, a storage unit 12, an analysis unit 13, a passenger count aggregation unit 14, and a bus stop information storage unit 15. In the diagram, the bus stop information storage unit 15 stores bus stop information D2, which shows information about each bus stop for each route. The functions of each unit will be described later.

[0019] The office terminal 2 is a personal computer owned by the bus operator and includes, for example, a data acquisition unit 21 and a data transmission unit 22. Furthermore, the office terminal 2 may also include a data extraction unit 23 that extracts only predetermined data from the data acquired by the data acquisition unit 21 and sends it to the data transmission unit 22. For example, the data extraction unit 23 can extract only video and audio data for a predetermined time before and after the opening and closing of the front entrance / exit 51F or the rear entrance / exit 51R. In this case, the amount of data transmitted from the office terminal 2 to the management server 1 can be reduced, and the communication time can be shortened.

[0020] The drive recorder system 4 has multiple cameras 41 installed on the bus 5, and stores video data captured by each camera 41 and audio data recorded by a microphone 43 (see Figure 3) in the control device 42. Hereafter, the microphone 43 will be abbreviated as microphone 43. In the figure, it is shown as if there is one microphone 43, but a microphone 43 can be built into each of the multiple cameras 41.

[0021] The drive recorder system 41 may also include, as in the embodiment described later, a GPS (Global Positioning System), a distance meter, a speedometer, an accelerometer, and the like, in addition to the camera 41 and microphone 43.

[0022] Figure 2 shows an example of the arrangement of cameras 41 and microphones 43 in bus 5. Bus 5 has a driver's seat 52 and multiple passenger seats 53. In addition, bus 5 is equipped with multiple cameras 41(1) to 41(4) and a fare box 44.

[0023] Camera 41(1) photographs the front entrance 51F and the driver's seat from inside the vehicle. Camera 41(2) photographs the rear entrance 51R from inside the vehicle. Camera 41(3) photographs the front of bus 5. Camera 41(4) photographs the rear of bus 5. Hereafter, unless otherwise specified, cameras 41(1) to 41(4) will be referred to as camera 41. Unless otherwise specified, the front entrance 51F and the rear entrance 51R will be referred to as entrance 51.

[0024] Figure 3 is a hardware configuration diagram of the boarding / alighting counting system. The boarding / alighting counting management server 1 includes, for example, a microprocessor (CPU in the figure) 101, memory 102, a communication interface unit (communication IF in the figure) 103, and a user interface unit (UI in the figure) 104.

[0025] Here, memory 102 is a device that serves as both main memory and auxiliary memory. Memory 102 stores predetermined computer programs for implementing the data acquisition unit 11, storage unit 12, analysis unit 13, passenger boarding / alighting count aggregation unit 14, and bus stop information D2. The microprocessor 101 reads the predetermined computer programs from memory 102 and executes them to realize each of the functions 11 to 15.

[0026] The data acquisition unit 11 acquires data transmitted from the office terminal 2 (video file D1, which includes video and audio data collected by the drive recorder system 4). The storage unit 12 stores the video file D1 as "data" acquired by the data acquisition unit 11. The analysis unit 13 identifies bus stops 7 by performing video and audio analysis on the video file D1 stored in the storage unit 12, and counts the number of passengers getting on and off at each identified bus stop. The passenger count aggregation unit 14 aggregates the passenger count results for each bus stop on each route by comparing the analysis results of the analysis unit 13 with the bus stop information stored in the bus stop information storage unit 15, and outputs the aggregated data D3.

[0027] The communication interface unit 103 is a circuit for bidirectional communication with the office terminal 2 via the communication network CN. The communication network CN may be a public line such as the internet, or a dedicated line. Furthermore, the communication network CN may include wired and wireless networks.

[0028] The user interface unit 104 is an information input / output device used by the administrator of the passenger boarding / alighting counting system. The user interface unit 104 has an information input device for the administrator to input information to the passenger boarding / alighting counting management server 1, and an information output device for providing information from the passenger boarding / alighting counting management server 1 to the administrator. The information input device is, for example, a keyboard, mouse, touch panel, voice recognition device, eye-tracking device, etc. The information output device is, for example, a monitor display, printer, speaker, etc. The information input device and the information output device may be integrated.

[0029] The branch office terminal 2 comprises, for example, a microprocessor 201, memory 202, a communication interface unit 203, a user interface unit 204, and a media interface unit (media IF in the figure) 205. The branch office terminal 2 is, for example, a personal computer such as a desktop, laptop, or tablet. It is not limited to a personal computer; any information processing device that has the function of acquiring information from a storage medium 3 and the function of communicating with the management server 1 is acceptable.

[0030] Memory 202 stores, for example, a predetermined computer program for implementing the data acquisition unit 21 and the data transmission unit 22. The data acquisition unit 21 acquires the video file D1 from the storage medium 3. The data transmission unit 22 transmits the video file D1 acquired by the data acquisition unit 21 to the boarding / alighting count management server 1.

[0031] The data acquisition unit 21 may send all of the video file D1 acquired to the boarding / alighting count management server 1 via the data transmission unit 22, or the data acquisition unit 21 may extract a portion of the video file D1 acquired to the data extraction unit 23, and the extracted data may be sent to the management server 1 via the data transmission unit 22.

[0032] If a data extraction unit 23 is provided in the office terminal 2, the data extraction unit 23 extracts only specific data that matches pre-set extraction criteria, such as video data of the opening and closing of the passenger doors, and audio and video data when voice announcements are played on the bus 5. By sending only the specific data extracted by the data extraction unit 23 from the data transmission unit 22 to the passenger count management server 1, the amount of communication and communication time can be reduced. Furthermore, since video data unrelated to passenger counting is not sent to the passenger count management server 1, the level of protection for personal information is increased. Note that the office terminal 2 may also be equipped with a mosaic processing unit (not shown). The mosaic processing unit can detect the outline of the faces of bus users and reduce the resolution of the user's face image or overlay other images to prevent identification of individuals. In this case, for example, the data extraction unit 23 in the figure can be changed to "mosaic processing unit 23".

[0033] The data extraction unit 23 may also be equipped with a function to edit the video file D1 acquired by the data acquisition unit 21. The bus operator may remove the data showing the time from departure from the depot until arrival at the first stop, and the data showing the time from passing the final stop until returning to the depot, before handing over the data to the data transmission unit 22.

[0034] The communication interface unit 203 is a circuit that communicates bidirectionally with the passenger boarding / alighting count management server 1 via the communication network CN. The user interface unit 204 is a device that exchanges information with the user operating the office terminal 2.

[0035] The media interface unit 205 is a device that inputs and outputs information to and from a storage medium 3, such as an SD (Secure Digital) card. The storage medium 3 is not limited to an SD card; it may also be a memory device, hard disk, optical disk, magneto-optical disk, magnetic tape, etc. The media interface unit 205 can be connected to the storage medium 3 by wire or wireless connection.

[0036] The drive recorder system 4 comprises the aforementioned cameras 41, a drive recorder control device 42, and a microphone 43. Video data captured by each camera 41 and audio data recorded by the microphone 43 are stored in the storage medium 3. The audio data includes voice announcements that read out the names of bus stops, as well as alarm sounds and warning messages to inform passengers of emergencies. The drive recorder system 4 can also incorporate sensors such as a GPS, speedometer, accelerometer, and distance meter (none of which are shown).

[0037] Figure 4 shows an example of the configuration of video file D1 stored on storage medium 3. Video file D1 is an example of "information acquired from the drive recorder". Each video file D1 has a file name that includes a device ID (identifier) ​​that identifies the camera 41 that generated the video file, and the date and time when recording started. Video file D1 may include video data, audio data, speed data, acceleration data, and GPS data.

[0038] If video file D1 exceeds a certain size, the recording start date and time are updated and a new video file D1 is created. If bus 5 travels on multiple routes between leaving and returning to the depot, multiple video files D1 are stored on storage medium 3.

[0039] Figure 5 shows an example of bus stop information D2. Bus stop information D2 is data that is pre-registered in the bus stop information storage unit 15 of the passenger boarding / alighting count management server 1 by the bus operator or the administrator of this system.

[0040] The bus stop information D2 includes, for example, the route ID (identifier) ​​D201, route name D202, bus stop sequence number D203, bus stop name D204, start / end point flag D205, bus stop name in kana D206, bus stop address D207, bus stop location information D208, distance between bus stops D209, destination address D210 for outbound image data, and destination address D211 for return image data.

[0041] Route ID_D201 is information that identifies the route operated by bus 5. Route name D202 is the name of the route. Stop sequence number D203 is the sequence number of a stop on a particular route. Stop name D204 is the name of the stop. Start / end flag D205 is information that indicates whether the stop is the start (origin) or the end (destination). Stop kana name D206 is the kana spelling of the stop name.

[0042] The bus stop address D207 is the address where the bus stop is located. The bus stop location information D208 is the location information (GPS data) of the bus stop. The distance between bus stops D209 indicates the distance to the previous bus stop. If the name of the bus stop is known, the total distance traveled can be calculated by summing the distances between each bus stop D209 up to that bus stop.

[0043] The storage address D210 for the outbound image data is the storage address for the video file D1 captured by each camera 41 during the outbound journey. The storage address D211 for the return image data is the storage address for the video file D1 captured by each camera 41 during the return journey. The outbound and return image data are examples of the appearance data of a bus stop. Since the bus stops are different on the outbound and return journeys, the appearance data will also differ.

[0044] Figure 6 is a flowchart of the process (S10) for constructing a bus stop voice model. For each route, the bus stop name and its kana name information are registered (S11), and stored in association with voice data announcing the bus stop name (S12). Using the voice model generated in advance in this process, the voice data extracted from the video file D1 can be recognized and the bus stop name identified.

[0045] Figure 7 is a flowchart of the boarding / alighting counting process using speech recognition. The boarding / alighting counting process (S100) acquires the video file D1 from the storage medium 3 (S101), then performs video data analysis of the front boarding / alighting area (S110), video data analysis of the rear boarding / alighting area (S130), stop identification processing using speech recognition (S150), and linking the boarding / alighting analysis results with the stop name recognition results (S170). The video data analysis of the front boarding / alighting area (S110), the video data analysis of the rear boarding / alighting area (S130), and the stop identification processing (S150) are executed in parallel. In the figure, time information is sometimes referred to as a timestamp.

[0046] Every day, the driver removes the storage medium 3 from the drive recorder system 4, connects it to the depot terminal 2, and saves the video file D1 for the day to the depot terminal 2. The user of the depot terminal 2 associates the video file D1 with the date, route number, and information to determine whether it was the front or rear boarding / alighting entrance, and registers it, then sends it to the boarding / alighting count management server 1. The boarding / alighting count management server 1 retrieves the video file D1 from the depot terminal 2 (S101).

[0047] Figure 8 is a flowchart detailing the process (S110) for analyzing video data of the front entrance / exit.

[0048] Management server 1 identifies and saves the recording start date and time based on the name of the video file D1 of the front entrance / exit (S111), and lowers the frame rate to quickly search for door opening (S112). If management server 1 detects that a door has opened (S113: YES), it saves the time (S114).

[0049] Then, the management server 1 increases the frame rate to perform a highly accurate count of the number of people (S115), and when it detects that the door has closed (S116: YES), it stops the counting process and saves the data along with the time (S117).

[0050] Management Server 1 reduces the frame rate and searches for the next door opening at high speed until the end of video file D1 is reached (S118: NO) (S112). When Management Server 1 detects the end of video file D1 (S118: YES), if there is another target file (S119: NO), it selects that file as the target for analysis (S120) and returns to step S111. Management Server 1 performs the above process for all video files D1 that are to be processed.

[0051] The video data analysis process for the rear entrance / exit (S130) is performed in the same way as the video data analysis process for the front entrance / exit (S110). Steps S131 to S140 in Figure 9 correspond to steps S111 to S120 in Figure 8, simply by replacing "front entrance / exit" with "rear entrance / exit," so no further explanation is provided.

[0052] Figure 10 is a flowchart of the stop identification process using speech recognition (S150). The management server 1 identifies and saves the recording start date and time from the name of the video file (S151), and extracts audio data from the video file (S152). The management server 1 uses speech recognition technology to recognize a specific audio clip from the audio data of the announcement guiding passengers to a stop, which consists of the words "next stop" and the name of the stop (S153), and records the recognized stop name and the time (S154).

[0053] When the management server 1 detects the terminal stop of the route being processed (S155:YES), the speech recognition for one route is completed, so it saves the speech recognition result (S156) and determines whether speech recognition has been performed up to the end of the speech data (S157).

[0054] Management Server 1 repeatedly executes steps S153 to S157 until all audio data has been recognized as speech (S157: NO). Once Management Server 1 has recognized all audio data contained in one video file (S157: YES), it determines whether the speech recognition process has been completed for all video files (S158). If there are any unprocessed video files (S158: NO), Management Server 1 moves on to the next video file (S159) and returns to step S151.

[0055] Figure 11 is a flowchart showing the process of linking the number of passengers getting on and off the bus based on video data analysis with the names of bus stops based on speech recognition from audio data.

[0056] The management server 1 detects the name of the nearest bus stop based on the time of the passenger count (boarding / alighting count) analyzed from the video data, and links the boarding / alighting count, which is the boarding / alighting analysis result, with the bus stop name (S171). This is because the bus stops a short time after the audio data announcing the next bus stop is played inside the bus 5, and passengers board and alight after that time.

[0057] Management Server 1 generates passenger boarding / alighting summary data D3 by linking the passenger boarding / alighting count, which is the result of video data analysis, with the bus stop name, which is the result of voice data recognition (S172). Management Server 1 aggregates the passenger boarding / alighting counts from the starting point to the ending point as one bus trip, generates passenger boarding / alighting summary data D3, and publishes it to bus operators through the web server owned by Management Server 1. Bus operators register route names and bus stop names with Management Server 1 and hold authentication information for viewing passenger boarding / alighting summary data D3. Bus operators can securely exchange information with Management Server 1 via the office terminal 2.

[0058] The passenger boarding and alighting summary data D3 shown at the bottom of Figure 11 is created for each bus route and destination, and for each period of time for which data is collected. The passenger boarding and alighting summary data D3 records the name of each stop and the number of passengers boarding and alighting at each stop for each bus service operated during the day. It is also possible to collect passenger boarding and alighting counts separately for the outbound and return journeys.

[0059] If it can be assumed that the operation, such as front boarding or rear boarding, is predetermined for each route, then by registering this information in advance on the management server 1, the number of people counted at each entrance / exit, which is the result of analyzing the video data, can be determined to represent the number of people boarding or alighting. It can also be determined whether it is the outbound or return journey based on the order of the recognized stops. This is possible because the order of the stop names announced by voice is reversed between the outbound and return journeys.

[0060] While no audio announcement is made at the starting stop, it is possible to infer the route from the names of subsequent stops. When the same vehicle operates on multiple routes, the video file will contain data for multiple routes, but it is possible to identify the route of a series of trips through audio recognition of the stop names. In this case, when uploading the video file to management server 1, the vehicle number should also be entered, for example, and the route information assigned to the vehicle should be registered in advance. This will allow for more accurate recognition of stop names from the audio announcements.

[0061] According to this embodiment, boarding and alighting counts for each bus stop can be performed based on information obtained from a specific device among the devices required to be installed on bus 5. Therefore, according to this embodiment, there is no need to combine multiple devices or install special sensors, and boarding and alighting counts for each bus stop can be achieved at low cost regardless of the period.

[0062] In this embodiment, since the drive recorder system 4 is adopted as a specific device, video data taken near the passenger doors (entrance / exit) 51 and audio data including voice announcements played inside the bus before arriving at a bus stop can be easily utilized, eliminating the need to use any special sensors or devices other than the drive recorder system 4. Therefore, the bus operator only needs to upload the data (video files) from the drive recorder system 4 to the management server 1, requiring no additional costs, and passenger boarding / alighting counts can be easily implemented for all routes, all vehicles, and any desired period. [Examples]

[0063] Example 2 will be explained using Figures 12 to 15. In this example, the number of passengers boarding and alighting at each bus stop is calculated based on the distance traveled by the drive recorder system 4. The following explanation will focus on the differences from Example 1.

[0064] Figure 12 is a flowchart of the boarding and alighting counting process based on distance traveled. The boarding and alighting counting process (S200) acquires the video file D1 from the storage medium 3 (S201), then performs video data analysis of the front boarding / alighting area (S210), video data analysis of the rear boarding / alighting area (S230), and a process of linking the boarding / alighting analysis results with the detection results of the bus stop name (S270). The video data analysis of the front boarding / alighting area (S210) and the video data analysis of the rear boarding / alighting area (S230) are performed in parallel.

[0065] As described in Figure 7, the user of the office terminal 2 associates the date, route number, and information to determine whether it is the front or rear boarding / alighting entrance with the video file D1 obtained from the storage medium 3, registers it, and sends it to the boarding / alighting count management server 1. The boarding / alighting count management server 1 obtains the video file D1 from the office terminal 2 (S201).

[0066] Figure 13 is a flowchart detailing the process (S210) for analyzing video data of the front entrance / exit.

[0067] The management server 1 identifies and saves the recording start date and time based on the name of the video file D1 of the front entrance / exit (S211), and then reduces the frame rate to quickly search for door openings (S212). If the management server 1 detects that a door has opened (S213: YES), it saves the time and distance traveled (S214). The name of the first stop is also saved in step S214. As shown in Figure 5, the stop information D2, which is pre-registered in the management server 1, contains the distance D209 between stops. Therefore, the distance between stops where the bus has stopped can be calculated based on the stop information D2.

[0068] Management server 1 increases the frame rate to perform a high-precision count of the number of people (S215), and when it detects that the doors have closed (S216: YES), it stops the counting process and saves the data along with the time (S217). If the bus is at its starting or ending point, in step S217, the distance traveled, which had been calculated in management server 1, is reset.

[0069] Management Server 1 reduces the frame rate and searches for the next door opening at high speed until it reaches the end of video file D1 (S218: NO) (S212). When Management Server 1 detects the end of video file D1 (S218: YES), if there is another target file (S219: NO), it selects that file as the target for analysis (S220) and returns to step S211. Management Server 1 performs the above process for all video files D1 that are to be processed.

[0070] The video data analysis process for the rear entrance / exit (S230) shown in Figure 14 is performed in the same way as the video data analysis process for the front entrance / exit (S210). Steps S231 to S240 in Figure 14 correspond to steps S211 to S220 in Figure 13, simply by replacing "front entrance / exit" with "rear entrance / exit," so no further explanation is provided.

[0071] Figure 15 is a flowchart showing the process of linking bus stop names, which are identified based on the number of passengers getting on and off the bus and the distance traveled, as determined by the analysis of video data.

[0072] The management server 1 detects the names of bus stops that differ little from the distance traveled at the time of the passenger count (boarding and alighting count) analyzed from the video data, and links the boarding and alighting count, which is the result of the boarding and alighting analysis, with the names of the bus stops (S271).

[0073] As described above, the video file D1 of the drive recorder system 4 shown in Figure 4 contains speed data and acceleration data, from which the distance traveled can be calculated. The distance traveled between bus stops is stored in the bus stop information D2. Therefore, by accumulating the distance traveled from the starting point, the name of the bus stop can be identified based on the distance traveled.

[0074] Management Server 1 generates passenger boarding / alighting summary data D3 by linking passenger boarding / alighting counts with bus stop names (S272). Management Server 1 publishes the passenger boarding / alighting summary data D2 to bus operators via a web server.

[0075] This embodiment, configured in this way, achieves the same effects as Embodiment 1. Furthermore, in this embodiment, based on stop information D2, which includes the distance traveled between stops, and a drive recorder system 4 that outputs measurement data for calculating the distance traveled by the bus 5, it is possible to detect the name of the stop where the bus 5 stopped and the number of passengers who boarded or alighted at that stop. In this embodiment, since speech recognition is not performed, there is no need to create dictionary data for speech recognition of stop names, and the configuration can be simplified compared to Embodiment 1. [Examples]

[0076] Embodiment 3 will be explained using Figures 16 to 18. Figure 16 is a flowchart of the boarding / alighting counting process based on location information. The boarding / alighting counting process (S300) acquires the video file D1 from the storage medium 3 (S301), then performs video data analysis of the front boarding / alighting area (S310), video data analysis of the rear boarding / alighting area (S330), and linking the boarding / alighting analysis results with the bus stop name detection results (S370). The video data analysis of the front boarding / alighting area (S310) and the video data analysis of the rear boarding / alighting area (S330) are performed in parallel.

[0077] Similarly, the user of the office terminal 2 associates the date, route number, and information to determine whether it is the front or rear boarding / alighting entrance with the video file D1 obtained from the storage medium 3, registers it, and sends it to the boarding / alighting count management server 1. The boarding / alighting count management server 1 obtains the video file D1 from the office terminal 2 (S301).

[0078] Figure 17 is a flowchart detailing the process (S310) for analyzing video data of the front entrance / exit.

[0079] Management server 1 identifies and saves the recording start date and time based on the name of the video file D1 of the front entrance / exit (S311), and lowers the frame rate to quickly search for door openings (S312). When management server 1 detects a door opening (S313: YES), it saves the time and location information (GPS positioning data) (S314).

[0080] As shown in Figure 4, the video file D1 stored on the storage medium 3 contains GPS data corresponding to the bus's movement. And, as shown in Figure 5, the bus stop information D2, which is pre-registered on the management server 1, includes the bus stop location information D208.

[0081] Management server 1 increases the frame rate to perform a highly accurate count of people (S315), and when it detects that a door has closed (S316: YES), it stops the people counting process and saves the data along with the time (S317).

[0082] Management Server 1 reduces the frame rate and searches for the next door opening at high speed until it reaches the end of video file D1 (S318: NO) (S312). When Management Server 1 detects the end of video file D1 (S318: YES), if there is another target file (S319: NO), it selects that file as the target for analysis (S320) and returns to step S311. Management Server 1 performs the above process for all video files D1 that are to be processed.

[0083] The video data analysis process for the rear entrance / exit (S330) shown in Figure 18 is performed in the same way as the video data analysis process for the front entrance / exit (S310). Steps S331 to S340 in Figure 18 correspond to steps S311 to S320 in Figure 17, simply by replacing "front entrance / exit" with "rear entrance / exit," so no further explanation is provided.

[0084] The linking process S370 shown in Figure 16, similar to the process S270 shown in Figure 15, includes the steps of linking passenger counts with bus stop names and creating passenger alighting summary data D3 (neither of which is shown), and provides this data to the bus operator. In Figure 15, S271, "Linking passenger alighting analysis results and bus stop names based on distance traveled," can be renamed to "Linking bus stop names identified by bus stop location information with passenger alighting analysis results."

[0085] This embodiment, configured in this way, also achieves the same effects as Embodiment 1. Furthermore, in this embodiment, stop information D2, which includes the location information of each stop, is registered in advance with the management server 1, and the stop where the bus 5 has stopped is identified based on GPS data measured by the drive recorder system 4, making it easier and more accurate to identify the stop. [Examples]

[0086] Embodiment 4 will be explained using Figures 19 to 22. In this embodiment, the appearance data of the bus stop is prepared in advance and included in the bus stop information D2. Then, in this embodiment, the appearance of the bus stop is extracted from the video file D1 captured by the drive recorder system 4, and the bus stop where the bus 5 stopped is identified by comparing the pre-registered appearance data of the bus stop with the appearance data of the bus stop captured during the bus's operation. The appearance data of the bus stop only needs to be image data that can identify which bus stop it is on the route, and it is not necessarily required that the bus stop sign is visible.

[0087] Figure 19 is a flowchart of the passenger boarding / alighting counting process based on the appearance data (appearance information) of bus stops. First, the appearance data for each bus stop for each route is acquired and registered with the management server 1 (S4000). Details of this process will be described later in Figure 20.

[0088] After the visual data of each bus stop is registered in the management server 1 as part of the bus stop information D2, the boarding and alighting count processing is performed based on the visual data of the bus stops. In Figure 19, the process of registering the visual data of the bus stops (S4000) is shown as part of the boarding and alighting count processing (S400) for convenience, but in reality, they are separate processes.

[0089] The boarding / alighting counting process (S400) retrieves the video file D1 from the storage medium 3 (S401), then executes video data analysis processing for the front boarding / alighting area (S410), video data analysis processing for the rear boarding / alighting area (S430), and linking the boarding / alighting analysis results with the bus stop name detection results (S470). The video data analysis processing for the front boarding / alighting area (S410) and the video data analysis processing for the rear boarding / alighting area (S430) are executed in parallel.

[0090] Using Figure 20, we will explain the process of collecting bus stop appearance data and registering it as part of bus stop information D2 to the management server 1. The following process can be performed automatically by the management server 1, manually by the bus operator or system administrator, or collaboratively by the bus operator or system administrator and the management server 1. The following explanation will describe the case where the management server 1 performs the process.

[0091] The management server 1 analyzes the image data of the bus stop's exterior from the video file D1 acquired from the storage medium 3, automatically extracts the bus stop's exterior data, and registers it as part of the bus stop information D2.

[0092] As shown in Figure 5, the bus stop information D2 stores the storage address D210 for image data from the outbound journey (video data captured by the outward-facing camera 41) and the storage address D211 for image data from the return journey. Even if the bus stop name is the same, the location of installation differs between the outbound and return journeys, so the appearance data also differs.

[0093] Management server 1 initializes the classification number of the visual data on a per-route basis (S4001). Management server 1 saves image data (video data) captured by camera 41 that photographs the direction of the bus stop from storage medium 3 and tags it with route information (S4002).

[0094] Management server 1 extracts and saves the starting date and time information from the name of the video file D1, which is a recording of the front entrance / exit (S4003). Management server 1 initializes the sequence number SN (S4004) and searches for door opening and closing in the video file D1 (S4005).

[0095] When the management server 1 detects the opening and closing of a door (S4006: YES), it increments the sequence number SN by one (S4007) and saves the image data captured by the outward-facing camera 41 with the same timestamp (S4008). In other words, the management server 1 saves the appearance data of the bus stop when the bus stop is detected.

[0096] Management Server 1 determines whether there is similar appearance data to already registered stop appearance data in the same system (S4009). If no similar appearance data is registered (S4009: NO), Management Server 1 assigns a classification number, associates it with a sequence number SN, and registers it in association with the image data saved in step S4008 (S4010). Conversely, if there is similar already registered appearance data (S4009: YES), Management Server 1 associates the classification number of that already registered data with the sequence number SN and saves it (S4011).

[0097] Management server 1 determines whether classification numbers have been assigned to all stops on the outbound and return routes (S4012), and if classification numbers for classifying appearance data have been set for all stops (S4012:YES), it determines whether the order of all classification numbers can be determined from the sequence number SN (S4013).

[0098] If the management server 1 cannot determine the order of all classification numbers from the sequence number (S4013:NO), it determines whether it has searched the video files of the front entrance / exit to the end (S4014). If it has not finished the search (S4014:NO), it returns to step S4005.

[0099] When management server 1 finishes the search (S4014:YES), it selects the next video file (S4015) and returns to step S4002.

[0100] If the management server 1 can determine the order of all classification numbers based on the sequence number (S4013:YES), it associates the starting point bus stop, such as the starting or ending point, with the appearance data, and stores the other appearance data with the bus stop name based on the order of the classification numbers (S4016).

[0101] In Figure 20, the user of the office terminal 2 associates the date, route number, and information to determine whether it is the front or rear entrance / exit with the video file D1 acquired from the storage medium 3, and registers it, then sends it to the passenger count management server 1. The passenger count management server 1 acquires the video file D1 from the office terminal 2 (S401). As a result, the video data analysis process for the front entrance / exit S410 and the video data analysis process for the rear entrance / exit S430 are processed in parallel.

[0102] Figure 21 is a flowchart detailing the process (S410) for analyzing video data of the front entrance / exit.

[0103] Management server 1 identifies and saves the recording start date and time based on the name of the video file D1 of the front entrance / exit (S411), and lowers the frame rate to quickly search for door openings (S412). When management server 1 detects a door opening (S413: YES), it saves the time (S414).

[0104] Management server 1 increases the frame rate to perform a highly accurate count of people (S415), and when it detects that a door has closed (S416: YES), it stops the people counting process and saves the data along with the time (S417).

[0105] Management Server 1 reduces the frame rate and searches for the next door opening at high speed until it reaches the end of video file D1 (S418: NO) (S412). When Management Server 1 detects the end of video file D1 (S418: YES), if there is another target file (S419: NO), it selects that file as the target for analysis (S420) and returns to step S411. Management Server 1 performs the above process for all video files D1 that are to be processed.

[0106] The video data analysis process for the rear entrance / exit (S430) shown in Figure 22 is performed in the same way as the video data analysis process for the front entrance / exit (S410). Steps S431 to S440 in Figure 22 correspond to steps S411 to S420 in Figure 21, simply by replacing "front entrance / exit" with "rear entrance / exit," so no further explanation is provided.

[0107] The linking process S470 shown in Figure 19 includes, similarly to the above, the steps of linking the boarding / alighting count with the bus stop name and creating the boarding / alighting summary data D3 (neither of which is shown), and is provided to the bus operator.

[0108] This embodiment, configured in this way, also produces the same effects as Embodiment 1. In this embodiment, the appearance data of bus stop 7 is registered in advance on the management server 1, and the name of the bus stop can be identified by performing pattern matching between the registered appearance data of the bus stop and the appearance data of the bus stop detected from the video file D1. Therefore, according to this embodiment, the name of the bus stop can be identified even if the exact location or address of the bus stop is unknown. [Examples]

[0109] Using Figure 23, we will explain how to detect the opening and closing of the entrance door 51 based on the video file D1 captured by the drive recorder system 4.

[0110] The management server 1 reads the video file D1 from the storage medium 3 and obtains and saves the starting date and time information based on the name of the video file D1 for the front entrance / exit or the name of the video file D1 for the rear entrance / exit (S1001).

[0111] The management server 1 extracts audio data from the video file D1 (S1002) and searches for the sound produced when the doors of the passenger entrance / exit are opened and closed (S1003). For example, if the doors are opened and closed by an opening / closing mechanism such as a pneumatic cylinder, the sound of the opening / closing mechanism is detected by the microphone 43 and stored in the video file D1.

[0112] When the management server 1 detects the opening or closing of a door from the sound of the door moving (S1004: YES), it saves the timestamp of the detection (S1005). The management server 1 continues to execute step S1003 until it detects the opening or closing of a door.

[0113] Management Server 1 determines whether the search for the target video file D1 is complete (S1006). If the search is not complete (S1006: NO), it returns to step S1001. If Management Server 1 has completed the search for the target video file D1 (S1006: YES), it determines whether the search for all video files D1 to be processed is complete (S1007). If there are any unprocessed video files D1 (S1007: NO), it selects the next video file D1 (S1008) and returns to step S1001.

[0114] When management server 1 detects the opening and closing of a door for all video files D1 (S1007: YES), it terminates this process.

[0115] This embodiment, configured in this way, also provides the same effects as Embodiment 1. In this embodiment, since the opening and closing of the doors is detected based on the sound of the doors opening and closing, the opening and closing of the doors can be detected even when the view of the camera 41 that photographs the entrance and exit is obstructed by passengers or luggage. In this embodiment, the accuracy of door opening and closing detection can be improved, so the bus stop where the bus has stopped can be identified more accurately, and the accuracy of passenger boarding and alighting counts can be improved. This embodiment can be combined with any of Embodiments 1 to 4 described above. [Examples]

[0116] Based on Figure 24, Example 6 will be described. In this example, the office terminal 2A is equipped with an analysis unit that analyzes the video data of video file D1 to measure the number of passengers getting on and off the bus.

[0117] In the bus passenger boarding and alighting counting system of this embodiment, the depot terminal 2A is equipped with a storage unit 24 and an analysis unit 25. The video file D1 read from the storage medium 3 is stored in the storage unit 24 and analyzed by the analysis unit 25. The data transmission unit 22 transmits the analysis results of the analysis unit 25 to the management server 1A. For example, the data transmission unit 22 can also transmit at least a portion of the video file D1 to the management server 1A to the extent necessary for the management server 1A to create the passenger boarding and alighting summary data D3.

[0118] The management server 1A acquires the analysis results and video file D1 from the office terminal 2A using the data acquisition unit 11 and stores them in the storage unit 12. Based on the analysis results (number of passengers getting on and off, timestamps, etc.) and the bus stop information D2, the management server 1A aggregates the number of passengers getting on and off at each bus stop and creates boarding / alighting aggregation data D3. Door opening and closing detection based on audio data may also be performed by the analysis unit 25 in the office terminal 2A.

[0119] This embodiment, configured in this way, also provides the same effects as Embodiment 1. Furthermore, in this embodiment, since the analysis unit 25 is provided within the branch office terminal 2A, the amount of data transmitted from the branch office terminal 2A to the management server 1A can be reduced, thereby reducing the processing load on the management server 1A.

[0120] In this embodiment, analysis results are mainly transmitted from the branch office terminal 2A to the management server 1A, and the entire video file D1 is not transmitted as is. Therefore, the possibility of passengers' faces and other information captured in the video file D1 being leaked can be reduced, and the reliability of the bus boarding and alighting counting system can be improved. This embodiment can be combined with any of embodiments 1 to 5. [Examples]

[0121] Embodiment 7 will be explained using Figure 25. In this embodiment, the analysis unit function is provided within the drive recorder system 4B, and data is transmitted directly from the drive recorder system 4B to the management server 1B. In this embodiment, the office terminal 2, which serves as a relay device in the bus boarding / alighting counting system, is not provided.

[0122] The drive recorder control device 42B of the drive recorder system 4B includes a data acquisition unit 421, a storage unit 422, an analysis unit 423, and a data transmission unit 424.

[0123] The data acquisition unit 421 acquires video data captured by the camera 41 and audio data detected by the microphone 43, and stores them in the storage unit 422 as video file D1. The analysis unit 423 measures the number of people who boarded or alighted the bus when it stopped. The data transmission unit 424 transmits the analysis results from the analysis unit 423 to the management server 1B via the communication network CN.

[0124] Based on the analysis results received from the drive recorder system 4B and the bus stop information D2, the management server 1B calculates the number of passengers getting on and off at each bus stop and creates the passenger boarding and alighting summary data D3.

[0125] This embodiment, configured in this way, achieves the same effects as Embodiment 1. Furthermore, according to this embodiment, the drive recorder system 4B analyzes the video footage inside the bus and directly transmits the analysis results to the management server 1B, eliminating the need for the office terminal 2 and enabling automatic daily passenger boarding and alighting counts. Moreover, since mainly the analysis results are transmitted from the drive recorder system 4B to the management server 1B, the amount of communication can be reduced, suppressing the increase in operating costs of the bus passenger boarding and alighting counting system. This embodiment can be combined with any of Embodiments 1 to 5.

[0126] It should be noted that the present invention is not limited to the embodiments described above. Those skilled in the art can make various additions and modifications within the scope of the present invention. The embodiments described above are not limited to the configuration examples shown in the accompanying drawings. The configuration and processing methods of the embodiments can be appropriately modified within the scope of achieving the objectives of the present invention.

[0127] Furthermore, by combining the above-described embodiments 1 to 4, the bus operator may be able to select one or more of the following identification processes: identification of bus stop names by voice recognition, identification of bus stop names based on distance traveled, identification of bus stop names based on location information, and identification of bus stop names based on appearance data of the bus stop. The bus operator can select multiple identification processes. If the results of each selected identification process match, the bus stop name can be adopted. If the results of each selected identification process do not match, the system administrator or the bus operator may verify the information from video data or the like.

[0128] Furthermore, each component of the present invention can be arbitrarily selected, and an invention comprising the selected components is also included in the present invention. Moreover, the components described in the claims can be combined in combinations other than those explicitly stated in the claims. [Explanation of Symbols]

[0129] 1,1A,1B: Passenger / Alighting Count Management Server, 2,2A: Office Terminal, 3: Storage Medium, 4,4B: Drive Recorder System, 5: Bus, 51F,51R: Door, 6: Passenger, 7: Bus Stop, 11: Data Acquisition Unit, 12: Storage Unit, 13: Analysis Unit, 14: Passenger / Alighting Count Aggregation Unit, 15: Bus Stop Information Storage Unit, 21: Data Acquisition Unit, 22: Data Transmission Unit, 23: Data Extraction Unit, 41: Camera, 42,42B: Drive Recorder Control Device, 43: Microphone, 421: Data Acquisition Unit, 422: Storage Unit, 423: Analysis Unit, 424: Data Transmission Unit

Claims

1. A boarding and alighting counting system for counting boarding and alighting of buses, a drive recorder system that records at least video data and audio data of the bus; a boarding and alighting counting unit that identifies bus stops at which doors are opened and closed based on information acquired from the drive recorder system, and counts the number of passengers getting on and off at each of the identified bus stops by analyzing video data of the identified bus stops among the video data; A bus boarding and alighting counting system.

2. The boarding / alighting counting unit identifies the bus stop where the door was opened or closed by analyzing the voice announcement in the voice data acquired from the drive recorder system. The bus boarding and alighting counting system according to claim 1.

3. The boarding / alighting counting unit identifies a stop at which the door is opened or closed based on a travel distance acquired from the drive recorder system. The bus boarding and alighting counting system according to claim 1.

4. The boarding / alighting counting unit identifies a stop at which the door is opened or closed based on the position information acquired from the drive recorder system. The bus boarding and alighting counting system according to claim 1.

5. The boarding / alighting counting unit identifies a stop at which the door is opened or closed based on video data from an outward-facing camera among video data acquired from the drive recorder system. The bus boarding and alighting counting system according to claim 1.

6. The entry / exit counting unit detects the opening and closing of the door by analyzing operation sounds when the door is opened and closed in the audio data acquired from the drive recorder system. The bus boarding and alighting counting system according to claim 1.

7. The entry / exit counting unit acquires the video data and the audio data from a storage medium detachably attached to the drive recorder system. The bus boarding and alighting counting system according to claim 1.

8. a relay terminal that acquires the video data and audio data from the storage medium and transmits the acquired video data and audio data to the boarding / alighting counting unit; The bus boarding and alighting counting system according to claim 7.

9. A method for counting boarding and alighting of buses using a bus boarding and alighting counting system, comprising: The bus boarding and alighting counting system includes: acquiring information from a drive recorder system that records at least video data and audio data of the bus; Identifying the bus stop where the door was opened or closed based on the acquired information; The video data at the specified bus stops is analyzed to count the number of passengers getting on and off at each specified bus stop. How to count bus boarding and alighting.

Images