JP7565994B2 - OD recording system - Google Patents

Description

The present invention relates to an OD recording system for acquiring OD data for public transportation.

In order to improve convenience for users and make operations more efficient, public transportation such as route buses collect OD data, which is then used to create bus schedules. OD is an abbreviation for origin and destination. OD data is data on the number of users between the departure point (boarding point) and the destination (disembarking point).

IC cards are used for cashless payment of fares on trains and urban buses. OD data can be obtained by using the data collected when using an IC card.

For example, a bus ticket gate device that uses a contactless IC card is known (see Patent Document 1). In the bus ticket gate device, IC card readers/writers are installed at the boarding and disembarking entrances, and data related to boarding and disembarking is acquired. However, such bus ticket gate devices are expensive, and some bus operators cannot introduce them. Also, not all passengers carry IC cards.

A passenger management device has been proposed that identifies boarding and disembarking stations without using IC cards (see Patent Document 2). This passenger management device processes images captured by an onboard camera to identify people, identifies the person's boarding and disembarking stations, and processes the settlement of the fare for paid seats. However, this passenger management device is expensive because it identifies individuals by processing images with high accuracy to prevent errors in the billing destination and billed amount of the fare. Furthermore, because this passenger management device identifies individuals, it is not easy to handle from the perspective of protecting privacy.

Japanese Patent Application Publication No. 9-134460 JP 2020-154917 A

The present invention aims to solve the above problems and provide an OD recording system that is low-cost and easy to protect privacy.

The OD recording system of the present invention is a system for acquiring OD data, which is data on the number of passengers between the boarding and disembarking points in a vehicle, and is equipped with an on-board camera that images people getting on, an off-board camera that images people getting off, a position information acquisition unit that acquires position information of the vehicle at the time of boarding and disembarking, a storage device in which data is recorded, and a control unit connected to the on-board camera, the disembarking camera, the position information acquisition unit, and the storage device, wherein the control unit extracts features of people appearing in images captured by the on-board camera and records the features and the position information at the time of boarding in the storage device as passenger data for that person, and the control unit extracts features of people appearing in images captured by the disembarking camera and compares the features with features in the passenger data for which there is no disembarking information, and writes the disembarking information to the passenger data to which those features are most similar, and the disembarking information includes position information at the time of disembarking.

In this OD recording system, it is preferable that the location information acquired by the location information acquisition unit is bus stop information input to the OD recording system, or latitude and longitude obtained by satellite positioning.

In this OD recording system, it is preferable that the system has a timing function, and the control unit further writes the boarding time and the disembarking time for each passenger data item into the passenger data item.

In this OD recording system, it is preferable that the control unit automatically deletes features in the passenger data in which the disembarking information is written.

In this OD recording system, it is preferable that the images captured by the boarding camera and the disembarking camera are not recorded in the storage device.

According to the OD recording system of the present invention, the person who has disembarked is determined to be the person whose characteristics are most similar to those of the passengers, so there is no need to determine whether the person disembarking is the same as a specific person, which simplifies the control unit that calculates the characteristics, and reduces the cost of the OD recording system. Also, privacy is easily protected by not determining whether a passenger is the same as a specific person.

FIG. 1 is a block diagram showing the configuration of an OD recording system according to an embodiment of the present invention. FIG. 2 is a plan view of the interior of the vehicle, showing the arrangement of the on-board cameras and the off-board cameras in the system. FIG. 3 is a process flow diagram showing data processing in the system. FIG. 4 is a diagram showing the structure of passenger data in the system in tabular form.

An OD recording system according to one embodiment of the present invention will be described with reference to Figs. 1 to 4. The OD recording system is a system for acquiring OD data. OD data is data on the number of passengers between the boarding point and the disembarking point of a vehicle.

As shown in FIG. 1, the OD recording system 1 includes an on-board camera 2, an off-board camera 3, a position information acquisition unit 4, a storage device 5, and a control unit 6.

As shown in FIG. 2, the boarding camera 2 is a camera that captures an image of a person P1 getting on to a vehicle 7. The disembarking camera 3 is a camera that captures an image of a person P2 getting off the vehicle.

The position information acquisition unit 4 acquires the position information of the vehicle 7 when boarding and disembarking. The storage device 5 records data (see FIG. 1 ). The control unit 6 is connected to the on-board camera 2, the off-board camera 3, the position information acquisition unit 4, and the storage device 5.

The configuration of the OD recording system 1 will now be described in further detail. In this embodiment, the OD recording system 1 is mounted on a bus. For example, one OD recording system 1 is shared among multiple buses, and is temporarily installed on the bus being surveyed during an OD survey. The OD recording system 1 may also be permanently installed on a bus. The OD recording system 1 may also be mounted on a railway vehicle.

The boarding camera 2 and the disembarking camera 3 are, for example, network cameras (IP cameras) having wireless LAN functions. When the network cameras are registered in the control unit 6, the camera installed at the boarding entrance of the vehicle 7 is set as the boarding camera 2, and the camera installed at the disembarking entrance is set as the disembarking camera 3 (see FIG. 2). The boarding entrance and the disembarking entrance may be the same. Even if the boarding entrance and the disembarking entrance are the same, the boarding camera 2 and the disembarking camera 3 have different lens optical axis directions and imaging ranges because the boarding and disembarking people move in different directions. In addition, when the boarding and disembarking entrances are the same, one camera may serve as both the boarding camera 2 and the disembarking camera 3. For example, one 360-degree camera (a camera that captures images in all 360-degree directions) may capture images of people getting on and off, and the control unit 6 may distinguish between people getting on and people disembarking based on the person's direction of movement. In addition, multiple boarding cameras 2 and disembarking cameras 3 may be installed. Images captured by the on-board camera 2 and the off-board camera 3 are input to the control unit 6 (see FIG. 1). The on-board camera 2 and the off-board camera 3 may be connected to the control unit 6 by wires.

In this embodiment, the location information acquired by the location information acquisition unit 4 is the bus stop information input to the OD recording system 1. If the vehicle 7 is equipped with a device that has bus stop information indicating at which bus stop the vehicle is stopped, the bus stop information held by the device is input to the OD recording system 1. In this case, the location information acquisition unit 4 is an input interface. If the driver or investigator of the vehicle 7 manually inputs the location information into the OD recording system 1, the location information acquisition unit 4 is an input device that receives the input operation.

The location information acquired by the location information acquisition unit 4 may be latitude and longitude by satellite positioning. The satellite positioning is, for example, GPS (Global Positioning System) or Quasi-Zenith Satellite System. In this case, the location information acquisition unit 4 is a GPS receiver or a Quasi-Zenith Satellite System receiver. From the latitude and longitude of the bus stop and the latitude and longitude acquired by the location information acquisition unit 4, the bus stop at the time of boarding (boarding location) and the bus stop at the time of disembarking (disembarking location) can be determined. As an additional function, the control unit 6 may generate bus stop information from the latitude and longitude acquired by the location information acquisition unit 4 and output it to a voice synthesizer or fare display installed on the bus.

The control unit 6 is a computer that has a CPU, memory, etc., and functions by executing a program. The storage device 5 is an SSD (solid state drive) or a hard disk built into the computer. The storage device 5 may also be an SSD or a hard disk external to the computer.

The processing in the OD recording system 1 configured as above will now be described. As shown in FIG. 3, first, the control unit 6 extracts the feature amounts of people captured in the image captured by the on-board camera 2. At this time, the control unit 6 detects people's faces from the image and extracts the differences between individuals' faces as a numerical sequence. This numerical sequence is the feature amount. Information other than the face, such as body size, may also be included in the feature amount. Note that feature amount is a term described in the Japanese Industrial Standards.

Then, the control unit 6 records the extracted feature amount and the position information at the time of boarding (the position information acquired by the position information acquisition unit 4) in the storage device 5 as passenger data for that person. As shown in FIG. 4, the passenger data is recorded in the storage device 5 as a table in a tabular format. The records corresponding to the rows of the table are passenger data for each passenger. In this embodiment, the passenger data has columns of ID, feature amount, position information at the time of boarding, boarding time, disembarking flag, position information at the time of disembarking, and disembarking time. The ID is a code for distinguishing passengers, for example, a serial number. The control unit 6 assigns a new ID to a new passenger. Therefore, the passenger data for each passenger who boards has a unique ID. The multiple passenger data recorded in the storage device 5 constitute a database. The database is the source of the OD data. Note that in FIG. 4, specific values of the passenger data are omitted.

The control unit 6 extracts the features of people appearing in the image captured by the disembarking camera 3 (see FIG. 3). At this time, the control unit 6 detects people's faces from the image and extracts the differences between individual faces as a numerical sequence. This numerical sequence is the feature.

The control unit 6 then compares the extracted features with features in passenger data that does not have alighting information. Alighting information is information that indicates that a passenger has already alighted. In this embodiment, the alighting information is an alighting flag (see FIG. 4). In general data processing, a flag is represented by a one-bit true/false value (Boolean value) or a binary number. The alighting flag is, for example, "true" or "1" if the passenger has alighted, and "false" or "0" if the passenger has not alighted. In this way, the control unit 6 compares the features of people alighting with the features of passengers who have not alighted (see FIG. 3).

The control unit 6 then writes the disembarking information into the passenger data whose features are most similar. In other words, the control unit 6 determines that the passenger who most resembles the person disembarking has disembarked. The approximation of the features is quantitatively evaluated, for example, based on the distance between the features. Since the features are numeric strings, if there are n numeric strings (n is a positive integer), the distance between the features is a distance in a multidimensional space (n-dimensional space) in mathematics. In other words, the features being most similar means that the mathematical distance between the numeric strings of the features is the smallest.

The disembarking information includes location information at the time of disembarking. That is, in this embodiment, the disembarking information is a disembarking flag and location information at the time of disembarking (see FIG. 4). Note that if the passenger data for a passenger includes location information at the time of disembarking, it is possible to know that the passenger has disembarked, and therefore the disembarking flag in the passenger data can be omitted.

The OD recording system 1 has a timekeeping function. The control unit 6 is a computer, and the computer has a clock function (timekeeping function). The control unit 6 further writes the boarding time and the disembarking time to each passenger data. The disembarking time indicates that the passenger has already disembarked, and is therefore included in the disembarking information. The boarding time and the disembarking time written to the passenger data are called time stamps.

In general facial recognition, the face of the person is compared with a facial image registered in advance in a database to determine whether they are the same person. This makes it possible to charge the person based on the facial recognition. In contrast, the OD recording system 1 of this embodiment determines that the passenger who most resembles the person getting off has got off, but does not determine that the person getting off is the same person as a specific person. Therefore, the processing in the control unit 6 is simpler than general facial recognition, the hardware and software of the control unit 6 can be simplified, and the OD recording system 1 can be made low cost. Because the OD recording system 1 is not used for charging, no monetary damage is incurred by passengers or businesses even if there is an error in the determination.

The OD recording system 1 does not have information that can identify individuals, such as names, i.e., personal information. This is because personal information is not necessary for calculating OD data (data on the number of passengers between the boarding and disembarking points). For this reason, the OD recording system 1 does not compare or link features extracted from facial images with personal information and does not identify individuals, so passenger privacy is protected.

In this embodiment, the control unit 6 has a function of automatically deleting the feature values in the passenger data in which the disembarking information is written. The feature values of individual passengers are not necessary for calculating OD data, so deleting them does not interfere with the investigation of OD data. According to the Japanese Industrial Standard JIS X9250:2017 "Information technology - Security technology - Privacy framework (privacy protection framework and principles)," "Biometric data" refers to the physical or behavioral characteristics and attributes of a person that are used in biometric authentication and identification processing, and biometric data is considered to be personally identifiable information (PII) (see 4.4.2 of the same JIS). For this reason, in the OD recording system 1, the feature values extracted from the face image correspond to biometric data and are considered to be personally identifiable information (PII). By deleting the feature values, the OD recording system 1 does not have personally identifiable information (PII), and passenger privacy is further protected.

Every time the vehicle 7 stops and a passenger gets on or off, the OD recording system 1 performs the above-mentioned process (see FIG. 3). As a result, passenger data of passengers who get on and off the vehicle 7 while it is in operation is stored in the storage device 5 (see FIG. 4).

OD data (data on the number of passengers between the boarding point and the disembarking point) is calculated from the position information at the time of boarding and the position information at the time of disembarking in the passenger data recorded in the storage device 5. In this embodiment, the control unit 6 calculates such OD data. The OD data may also be calculated by a computer external to the OD recording system 1.

The number of passengers per time period is calculated from the boarding time and disembarking time in the passenger data recorded in the storage device 5. In this embodiment, the control unit 6 calculates the number of passengers per time period. The control unit 6 also calculates the number of passengers in real time. The number of passengers per time period may be calculated by a computer external to the OD recording system 1 after the vehicle 7 has finished operating.

The sojourn time (boarding time) of each passenger is calculated from the boarding time and disembarking time in the passenger data recorded in the storage device 5. Then, from the sojourn time of each passenger, the average sojourn time by time period and the overall average sojourn time from the first departure to the final destination are calculated. In this embodiment, the control unit 6 calculates such sojourn times. Sojourn times may also be calculated by a computer external to the OD recording system 1.

The OD recording system 1 does not record images captured by the on-board camera 2 and the off-board camera 3 in the storage device 5.

As described above, according to the OD recording system 1 of this embodiment, the person (passenger) who has the most similar characteristic values to those of the person riding in the vehicle is determined to have disembarked, so there is no need to determine whether the person disembarking is the same as a specific person, the control unit 6 that calculates the characteristic values can be simplified, and the OD recording system 1 can be made low cost. Furthermore, privacy can be easily protected by not determining whether the passenger is the same as a specific person.

The control unit 6 automatically deletes features from passenger data containing disembarking information, so personal features are not collected, making it easy to protect privacy.

By not recording the images captured by the boarding camera 2 and the disembarking camera 3 in the storage device 5, privacy is easily protected and the management of the storage device 5 is easier, resulting in low operating costs.

The present invention is not limited to the configuration of the above embodiment, and various modifications are possible without departing from the spirit of the invention. For example, the OD recording system 1 may obtain information that is not personally identifiable, such as gender, from the feature before deleting the feature.

1 OD recording system 2 Boarding camera 3 Exiting camera 4 Position information acquisition unit 5 Storage device 6 Control unit

Claims (5)

An OD recording system for acquiring OD data, which is data on the number of passengers between boarding and alighting points in a vehicle, comprising:
A passenger camera that captures images of passengers;
A disembarking camera that captures images of people getting off the train;
A position information acquisition unit that acquires position information of the vehicle when boarding and disembarking;
A storage device in which data is recorded;
a control unit connected to the boarding camera, the disembarking camera, the position information acquisition unit, and the storage device;
The control unit extracts features of a person captured by the on-board camera, and records the features and position information at the time of boarding in the storage device as passenger data of the person.
the control unit extracts features of people captured by the image captured by the alighting camera, compares the features with features in the passenger data that does not have alighting information, and writes the alighting information into the passenger data that has the closest features;
The OD recording system, wherein the alighting information includes location information at the time of alighting. The OD recording system according to claim 1, characterized in that the location information acquired by the location information acquisition unit is bus stop information input to the OD recording system, or latitude and longitude obtained by satellite positioning. It has a timing function,
2. The OD recording system according to claim 1, wherein the control unit further writes a boarding time and an alighting time to the passenger data for each passenger. The OD recording system of claim 1, characterized in that the control unit automatically deletes features in the passenger data in which the disembarking information is written. The OD recording system according to any one of claims 1 to 4, characterized in that the images captured by the on-board camera and the off-board camera are not recorded in the storage device.

Images