JP7238615B2 - Vehicle tracking device, vehicle tracking method, and vehicle tracking program - Google Patents

Description

The present invention relates to technology for tracking a vehicle running on a road.

Conventionally, there has been a system that tracks the running of a vehicle by installing cameras at multiple points (hereinafter referred to as vehicle detection points) on the road on which the vehicle travels and identifying the vehicle between the vehicle detection points ( See Patent Document 1, etc.). This system is put into practical use, for example, as a system for measuring the travel time between two vehicle detection points.

Patent document 1 processes a frame image of a vehicle at each vehicle detection point, recognizes a series of designation numbers written on the license plate of the vehicle, and compares the series of designation numbers to detect the vehicle between the vehicle detection points. It is a configuration to identify.

Also, as described in Patent Document 1, there are a plurality of vehicles having the same series designation number. Therefore, if a group of vehicles to be identified between vehicle detection points includes multiple vehicles with the same series designation number, it is impossible to identify the vehicle between vehicle detection points only by matching the series designation number. become inappropriate.

Japanese Patent Laid-Open No. 2004-100003 is a configuration that suppresses improper vehicle identification between vehicle detection points by using vehicle type information in addition to serial designation numbers.

Japanese Patent Publication No. 5-86600

However, on the license plate attached to the vehicle, area names, classification numbers, and hiragana are written in addition to the serial designation numbers described above. Here, the area name, classification number, hiragana, and serial designation number are collectively referred to as a registration symbol. Vehicles can be identified by the registration number printed on the license plate.

The area name, classification number, and hiragana written on the license plate are smaller than the serial number. There are also license plates (regions) where the region name is written in kanji. For these reasons, frame images used for recognizing registered symbols require a higher resolution than when recognizing only serial designation numbers. Generally, a camera capable of capturing frame images with higher resolution is more expensive. Therefore, installing a camera capable of picking up a frame image with a resolution at which the registration mark can be properly recognized at each vehicle detection point increases the cost for constructing the system.

An object of the present invention is to process frame images of a vehicle license plate captured at three or more locations, to suppress improper vehicle identification between locations, and to reduce the cost of system construction. To provide a technique capable of suppressing an increase in

In order to achieve the above objects, the vehicle tracking device of the present invention is constructed as follows.

The recognition result storage unit stores a first recognition result, a second recognition result, and a third recognition result. The first recognition result is the recognition result of the registration symbol written on the license plate of the vehicle that passed the first point on the road. The second recognition result is the recognition result of the registration symbol written on the license plate of the vehicle that passed the second point on the travel route. The third recognition result is the recognition result of the serial designation number written on the license plate of the vehicle that passed the middle point of the travel road.

The first point, the intermediate point, and the second point are arranged in this order from the upstream side of the travel path. In addition, here, the area name, classification number, hiragana, and serial designation number written on the license plate are collectively referred to as a registration symbol. In addition, the first recognition result includes the passing time of the vehicle at the first point. The second recognition result includes the passing time of the vehicle at the second point.

The identification unit identifies the vehicle between points using the first recognition result, the second recognition result, and the third recognition result.

The identification unit identifies a vehicle that has passed through the first point and a vehicle that has passed through the second point by collating the registration symbols. If there is a vehicle that cannot be identified as a vehicle that has passed the first point or the second point by collation of the serial designation number among the vehicles that have passed the intermediate point, the identification unit detects the passage time of the vehicle at the first point. , and the passing time of the vehicle at the second point.

In this way, a vehicle that cannot be identified by collation of a series of designated numbers is identified by using the time of passage through the first point and the time of passage through the second point. can be suppressed.

Also, for intermediate points, a camera with a resolution capable of capturing a frame image capable of recognizing a series of designation numbers can be used. That is, it is not necessary to use a camera with a resolution capable of capturing a frame image capable of recognizing the area name, classification number, and hiragana written on the license plate. Therefore, an increase in the cost required for constructing the system can be suppressed.

If there is a vehicle that cannot be identified as a vehicle that has passed the first point or the second point by collation of the serial designation number among the vehicles that have passed the intermediate point, the identification unit detects the passage time of the vehicle at the first point. , and the passing time of the vehicle at the second point, the distance between the first point and the waypoint and the distance between the second point and the waypoint may be used for identification. By also considering the distance between the first point and the waypoint and the distance between the second point and the waypoint, it is possible to further suppress improper identification of the vehicle between the points.

Further, the configuration may include a travel time calculation unit that calculates the travel time of the vehicle between two points based on the identification result of the vehicle at the corresponding two points by the identification unit.

Furthermore, the character recognition processing is performed on the frame image in which the license plate of the vehicle that has passed the first point is captured, and the registration symbol written on the license plate is obtained as the first recognition result, and the vehicle that has passed the second point is obtained. Character recognition processing is performed on the frame image of the license plate of the vehicle, and the registration symbol written on the license plate is obtained as the second recognition result. A configuration may be provided in which a recognition result acquisition unit acquires, as a third recognition result, a serial designation number written on the license plate by character recognition processing.

According to the present invention, improper vehicle identification can be suppressed, and an increase in the cost required for building a system can be suppressed.

1 is a schematic diagram showing a vehicle tracking system; FIG. It is a figure which shows a license plate. It is a block diagram which shows the structure of the principal part of a point processing apparatus. It is a block diagram showing the configuration of the main part of the vehicle tracking device. FIG. 5(A) shows a storage area for storing the recognition result received from the point processing device 2a, and FIG. 5(B) shows a storage area for storing the recognition result received from the point processing device 2b. (C) shows a storage area for storing the recognition result received from the point processing device 2c, and FIG. 5(D) shows a storage area for storing the recognition result received from the point processing device 2d. FIG. 4 shows vehicle tracking data; It is a flowchart which shows operation|movement of a point processing apparatus. 8 is a flowchart showing recognition result storage processing; 4 is a flowchart showing identification processing; It is the figure which showed the recognition result in each point in the passing order of a vehicle. It is the figure which showed the recognition result in each point in the passing order of a vehicle. 12A and 12B are diagrams for explaining the process of estimating an overtaking point. FIGS. 13A and 13B are diagrams showing vehicle identification between points according to estimated passing points. It is a block diagram which shows the structure of the principal part of the vehicle tracking device concerning a modification.

Embodiments of the present invention will be described below.

<1. Application example>
FIG. 1 is a schematic diagram showing an application example of the vehicle tracking system according to this example. FIG. 2 is a diagram showing a license plate attached to a vehicle. The vehicle tracking system according to this example is a system that tracks a vehicle 100 traveling on a road (corresponding to a travel path in the present invention). The vehicle tracking system of this example obtains tracking data for vehicle 100 by detecting vehicle 100 at multiple points and identifying vehicle 100 between the points. From the tracking data of the vehicle 100, it is possible to obtain the congestion status of the vehicle 100 on the road, the travel time between two points, and the like.

In this example, points A, B, C, and D shown in FIG. 1 are passage detection points for detecting passage of the vehicle 100 . They are lined up in order of A point, B point, C point, and D point from the upstream side of the road. In this example, the point A corresponds to the first point referred to in the present invention, and the point D corresponds to the second point referred to in the present invention. Moreover, the B point and the C point correspond to the waypoint referred to in the present invention. That is, this example is an example in which two waypoints are set between the first point and the second point referred to in the present invention.

The vehicle tracking system of this example is provided with corresponding cameras 6 (6a-6d) and point processing devices 2 (2a-2d) for each passage detection point. Each camera 6 is installed at an angle that captures an image of the license plate 101 attached to the vehicle 100 passing through the corresponding detection point. Specifically, the camera 6a is attached at an angle that captures the license plate 101 of the vehicle 100 passing the A point. The camera 6b is attached at an angle that captures the license plate 101 of the vehicle 100 passing through the B point. The camera 6c is attached at an angle that captures the license plate 101 of the vehicle 100 passing the C point. The camera 6d is attached at an angle that captures the license plate 101 of the vehicle 100 passing through the D point.

With reference to FIG. 2, the registration symbols written on license plate 101 will be briefly described. The registered symbols written on the license plate 101 include four items: area name 111 , classification number 112 , hiragana 113 , and series designation number 114 . The area name is a place name indicating the jurisdiction transportation branch office written in kanji, hiragana, or the like. The classification number 112 is a three-digit number indicating the type of the vehicle 100 and its use. Hiragana 113 is one character and indicates private use, business use, or the like. The series designation number 114 is a number of four digits or less.

Also, the series designation number 114 written on the license plate 101 has a larger character size than the area name 111 , the classification number 112 and the hiragana 113 . In other words, the area name 111, classification number 112, and hiragana 113 written on the license plate 101 are smaller in character size than the serial number 114. FIG.

As is well known, the vehicle 100 can be identified by the registration symbols written on the license plate 101 (area name 111, classification number 112, hiragana 113, and serial designation number 114). However, the vehicle 100 cannot be specified only by a part of the registration symbols written on the license plate 101 (for example, a serial designation number).

The cameras 6 (6a-6d) output captured frame images to the associated point processing devices 2 (2a-2d). If the license plate 101 of the vehicle 100 passing through the point A is captured in the frame image captured by the camera 6a, the point processing device 2a displays the registration symbol (area name 111, Recognize classification numbers 112, hiragana 113, and series designation numbers 114). If the license plate 101 of the vehicle 100 passing through the point B is captured in the frame image captured by the camera 6b, the point processing device 2b recognizes the serial designation number 114 written on the license plate 101. (Regional name 111, classification number 112, and hiragana 113 are not recognized). If the license plate 101 of the vehicle 100 passing through the point C is captured in the frame image captured by the camera 6c, the point processing device 2c recognizes the serial designation number 114 written on the license plate 101. (Regional name 111, classification number 112, and hiragana 113 are not recognized). If the license plate 101 of the vehicle 100 passing through the D point is imaged in the frame image captured by the camera 6d, the point processing device 2d displays the registration symbol (area name 111, Recognize classification numbers 112, hiragana 113, and series designation numbers 114).

That is, the point processing devices 2a and 2d recognize the registration symbol written on the license plate 101, but the point processing devices 2b and 2c only recognize the series designation number 114 written on the license plate 101. , area name 111, classification number 112, and hiragana 113.

The point processing device 2 transmits the recognition result of the license plate 101 to the vehicle tracking device 1 . This recognition result includes the captured time of the frame image used for recognizing the registration symbol or serial designation number. The imaging time of this frame image is the passage time of the vehicle 100 at the corresponding passage detection point.

The recognition result transmitted from the point processing device 2a to the vehicle tracking device 1 corresponds to the first recognition result referred to in the present invention. The recognition results transmitted from the point processing devices 2b and 2c to the vehicle tracking device 1 correspond to the third recognition results referred to in the present invention. The recognition result transmitted from the point processing device 2d to the vehicle tracking device 1 corresponds to the second recognition result referred to in the present invention.

The vehicle tracking device 1 identifies the vehicle 100 between the passage detection points using the recognition results transmitted from the point processing devices 2a to 2d. Since the vehicle 100 can be specified by the registration symbol written on the license plate 101, the vehicle 100 that has passed the A point and the vehicle 100 that has passed the D point can be accurately identified. On the other hand, the vehicle 100 cannot be identified only by the series designation number 114 (there are a plurality of vehicles 100 to which the license plate 101 having the same series designation number 114 is attached). Therefore, if there are a plurality of vehicles 100 with the same series designation number 114 in the group of vehicles identified between the passing detection points, the license plate 101 with the same series designation number 114 is attached. The vehicle 100 is identified using the passage time of the vehicle 100 at the A point and the passage time of the vehicle 100 at the D point.

<2. Configuration example>
FIG. 3 is a block diagram showing the configuration of the main part of the point processing device. The point processing device 2 (2a to 2d) includes a control unit 21, an image input section 22, and a communication section .

The control unit 21 controls the operation of each part of the point processing device 2 main body. The control unit 21 also has a character recognition section 21a. This character recognition unit 21a will be described later.

The image input unit 22 receives frame images of vehicle detection points (points A to D) captured by the corresponding cameras 6 (6a to 6d). The frame rate of the camera 6 is, for example, several tens of frames/sec (10 to 30 frames/sec). The point processing device 2 and the camera 6 may be wired or wirelessly connected.

The communication unit 23 controls data communication with the vehicle tracking device 1 via the network 7 .

Further, the character recognition unit 21a determines whether or not the license plate 101 of the vehicle 100 is captured in the frame image of the camera 6 input to the image input unit 22 or not. The character recognition section 21a is: If the license plate 101 of the vehicle 100 is imaged, character recognition processing is performed to recognize characters written on the license plate 101 . However, the character recognition unit 21a of the point processing devices 2a and 2d recognizes the registration symbols written on the license plate 101 (area name 111, classification number 112, hiragana 113, and series designation number 114). On the other hand, the character recognition units 21a of the point processing devices 2b and 2c do not recognize the area name 111, the classification number 112, and the hiragana 113 written on the license plate 101, but recognize only the series designation number 114. FIG.

Thus, the point processing devices 2a to 2d have the same hardware configuration. However, while the point processing devices 2a and 2d recognize the registration symbol written on the license plate 101, the point processing devices 2a and 2d recognize only the serial designation number 114 written on the license plate 101. The difference is that Therefore, the point processing devices 2b and 2c have a smaller processing load than the point processing devices 2a and 2d. For this reason, the point processing devices 2b and 2c may have a configuration using a cheaper CPU whose processing speed is slower than that of the point processing devices 2a and 2d.

The control unit 21 of the point processing device 2 is composed of a hardware CPU, memory, and other electronic circuits. When the hardware CPU executes the character recognition program, it operates as the character recognition section 21a. The memory also has an area for expanding the character recognition program and an area for temporarily storing data generated when the character recognition program is executed. The control unit 21 may be an LSI that integrates a hardware CPU, memory, and the like.

FIG. 4 is a block diagram showing the configuration of the main parts of the vehicle tracking device. The vehicle tracking device 1 includes a control unit 11 , a communication section 12 , a recognition result storage section 13 , a tracking data storage section 14 and an output section 15 .

The control unit 11 controls the operation of each part of the vehicle tracking device 1 main body. The control unit 11 also has an identification section 11a and a travel time calculation section 11b. Details of the identification unit 11a and the travel time calculation unit 11b will be described later.

The communication unit 12 controls data communication with the point processing devices 2 (2a to 2d) via the network .

The recognition result storage unit 13 accumulatively stores the recognition results received by the communication unit 12 for each point processing device 2 . FIG. 5 is a diagram for explaining recognition results stored in a recognition result storage unit. The recognition result storage unit 13 is provided with a storage area for storing the recognition result for each point processing device 2 . FIG. 5(A) shows a storage area for storing the recognition result received from the point processing device 2a, and FIG. 5(B) shows a storage area for storing the recognition result received from the point processing device 2b. (C) shows a storage area for storing the recognition result received from the point processing device 2c, and FIG. 5(D) shows a storage area for storing the recognition result received from the point processing device 2d.

As is clear from the above description, the recognition result received from the point processing device 2a is the recognition result of the vehicle 100 passing through the point A (see FIG. 5A). The recognition result received from the point processing device 2b is the recognition result of the vehicle 100 passing through the point B (see FIG. 5B). Further, the recognition result received from the point processing device 2c is the recognition result of the vehicle 100 passing through the point C (see FIG. 5(C)). Further, the recognition result received from the point processing device 2d is the recognition result of the vehicle 100 that has passed the point D (see FIG. 5(D)).

As shown in FIGS. 5A and 5D, the recognition results of the vehicle 100 passing through the points A and D are the area name 111, the classification number 112, It is data in which hiragana 113, series designation number 114, and passage time are associated with each other. Further, as shown in FIGS. 5B and 5C, the recognition result of the vehicle 100 passing through the B point and the C point is the serial designation number 114 written on the license plate 101 of the vehicle 100, and This is data in which passage times are associated with each other.

The tracking data storage unit 14 stores, for each vehicle 100, tracking data in which the passage time of each passage detection point is associated. FIG. 6 is a diagram showing vehicle tracking data. As shown in FIG. 6, this tracking data includes the area name 111 written on the license plate 101, the classification number 112, the hiragana 113, the serial designation number 114, the passage time of each passage detection point (passage time of point A, It is data in which the passage time of point B, the passage time of point C, and the passage time of point D) are associated with each other. This tracking data is obtained by the identification unit 11a performing identification processing for identifying the vehicle 100 that has passed each passage detection point (point A, point B, point C, and point D) between points. Details of the identification process will be described later.

The recognition result storage unit 13 and the tracking data storage unit 14 may be nonvolatile memories (EPROM, EEPROM, etc.), or may be auxiliary storage devices such as hard disk drives and solid state drives. Further, the recognition result storage unit 13 and the tracking data storage unit 14 may be configured with a single storage medium, or may be configured with different storage media.

The output unit 15 outputs the tracking data of the vehicle 100 stored in the tracking data storage unit 14 or the travel time between passage detection points (travel time between two points) to a host device (not shown) as needed. Output.

Next, the identification section 11a and the travel time calculation section 11b included in the control unit 11 will be described. Based on the recognition result stored in the recognition result storage unit 13, the identification unit 11a identifies the vehicle 100 whose passage between the passage detection points is detected. The details of the identification process for identifying the vehicle 100 whose passage is detected between the passage detection points will be described later. Further, the travel time calculation unit 11b calculates the travel time between two specified passage detection points based on the tracking data stored in the tracking data storage unit 14. FIG. These two passing detection points may be adjacent points (for example, points A and B) or may be points that are not adjacent (for example, points A and C).

A control unit 11 of the vehicle tracking device 1 is composed of a hardware CPU, memory, and other electronic circuits. When the hardware CPU executes the vehicle tracking program according to the present invention, it operates as an identification section 11a and a travel time calculation section 11b. The memory also has an area for developing the vehicle tracking program according to the present invention and an area for temporarily storing data generated when the vehicle tracking program is executed. The control unit 11 may be an LSI that integrates a hardware CPU, memory, and the like. Also, the hardware CPU is a computer that executes the vehicle tracking method according to the present invention.

<3. Operation example>
The operations of the vehicle tracking device 1 and the point processing device 2 will be described below. FIG. 7 is a flow chart showing the operation of the point processing device. A frame image captured by the corresponding camera 6 is input to the image input unit 22 of the point processing device 2 . The point processing device 2 has an image memory (not shown) that temporarily stores the frame image input to the image input unit 22 .

The point processing device 2 selects a frame image to be processed (s1). The frame image to be processed selected here is a frame image captured after a predetermined number of frames from the previously selected frame image to be processed. This predetermined number of frames may be one frame, or may be two or more frames.

The point processing device 2 determines whether or not the license plate 101 of the vehicle 100 is imaged in the processing target frame image selected in s1 (s2). In s2, it is determined whether or not the entire license plate 101 is imaged.

When the point processing device 2 determines in s2 that the license plate 101 is not imaged in the frame image to be processed, the process returns to s1.

When the point processing device 2 determines in s2 that the license plate 101 is imaged in the frame image to be processed, it performs recognition processing (s3). Point processing devices 2 a and 2 d recognize area name 111 , classification number 112 , hiragana 113 , and series designation number 114 written on license plate 101 . The point processing devices 2b and 2d recognize the serial designation number 114 written on the license plate 101. FIG.

In this way, the recognition target of s3 in the point processing devices 2a and 2d and the recognition processing of s3 in the point processing devices 2b and 2c are different.

The processing related to s1 to s3 is executed by the character recognition section 21a.

The point processing device 2 transmits the recognition result of the recognition processing related to s3 to the vehicle tracking device 1 (s4), and returns to s1. The point processing devices 2a and 2d send the recognition results including the region name 111, the classification number 112, the hiragana 113, and the sequence designation number 114 as the recognition results, and the imaging time of the frame image to be processed this time to the vehicle tracking device. Send to 1. The point processing devices 2b and 2c transmit to the vehicle tracking device 1 the recognition result of the series designation number 114 and the recognition result including the imaging time of the frame image to be processed this time. The point processing device 2 communicates with the vehicle tracking device 1 in the communication unit 23 .

Here, the point processing device 2 transmits the recognition result to the vehicle tracking device 1 each time it executes the recognition processing of s3. A configuration may be adopted in which, when executed, these recognition results of the vehicle 100 are collectively transmitted to the vehicle tracking device 1 . Further, the point processing device 2 may be configured to collectively transmit to the vehicle tracking device 1 the recognition results of the vehicle 100 that has executed the recognition process of s3 during the time interval set in advance.

Next, the operation of the vehicle tracking device 1 will be described. First, recognition result storage processing executed by the vehicle tracking device 1 will be described. FIG. 8 is a flowchart showing recognition result storage processing. The vehicle tracking device 1 waits for the communication unit 12 to receive the recognition result from one of the point processing devices 2 connected via the network 7 (s11). Upon receiving the recognition result, the vehicle tracking device 1 determines the passage detection point of the recognition result received this time (s12). The process of s12 is a process of determining the spot processing device 2 that is the transmission source of the recognition result received this time. As is well known, in data communication via the network 7, transmission data includes data indicating a transmission source. The vehicle tracking device 1 stores the recognition result received this time in the recognition result storage unit 13 (s13), and returns to s11. In s13, the recognition result of the passage detection point determined in s12 is stored in the recognition result storage unit 13. FIG.

The vehicle tracking device 1 can store the recognition result shown in FIG. 4 in the recognition result storage unit 13 by executing the process shown in FIG.

Next, identification processing of the vehicle tracking device 1 will be described. FIG. 9 is a flowchart showing identification processing. The vehicle tracking device 1 executes this identification process at a predetermined timing. For example, the vehicle tracking device 1 starts this identification process when a predetermined time (5 minutes, 10 minutes, etc.) has passed since the previous identification process.

The identification unit 11a executes this identification processing. The identification unit 11a determines the current target vehicle group (s21). In s21, the vehicles 100 that passed point D after the start of the previous identification process and before the start of the current identification process are determined as the target vehicle group for the current identification process.

The identification unit 11a identifies the vehicle 100 belonging to the target vehicle group determined in s21 between the point A and the point D (s22). In s22, the registration symbols written on the license plate 101 (area name 111, classification number 112, hiragana 113, and series designation number 114) are used. Therefore, it is possible to accurately identify the vehicle 100 that has passed the A point and the vehicle 100 that has passed the D point.

The identification unit 11a determines whether or not there is a vehicle 100 having the same series designation number 114 among the vehicles 100 belonging to the target vehicle group determined in s21 (s23). In s23, it is determined whether or not there is a vehicle 100 having a duplicate series designation number among the vehicles 100 belonging to the target vehicle group determined in s21. If there is no vehicle 100 with the same series designation number 114, the identifying unit 11a identifies the vehicle 100 between the passage detection points using the series designation number 114 (s24). That is, in s24, if there is no vehicle 100 with a duplicate series designation number among the vehicles 100 belonging to the target vehicle group determined in s21 (in other words, if there is no vehicle 100 belonging to the target vehicle group determined in s21, If the series designation number 114 is different, identification of the vehicle 100 between each passage detection point is performed using the series designation number 114 .

If there are vehicles 100 with the same series designation number 114, the identification unit 11a determines that the passing order of the vehicles 100 with the same series designation number 114 at the point A and the series designation number 114 at the point D are the same. It is determined whether or not the passing order of the vehicles 100 is the same (s25). When the identifying unit 11a determines in s25 that the passing order is the same, the passing order of the vehicles 100 having the same series designation number 114 is the same at the points B and C as at the points A and D. Then, the vehicle 100 is identified between each passing detection point (s26). At this time, the identification unit 11a uses the series designation number 114 to identify other vehicles 100 that do not have the same series designation number 114 between the passing detection points.

As an example, in the vehicle group determined in s21, as vehicles 100 having the same series designation number 114, a vehicle 100 having a registration code of "Kyoto 550 Ta 77" and a vehicle 100 having a registration code of "Naniwa 381 Mi 77" are selected. and exist. Also, in this example, the passing order of the vehicles 100 having the same series designation number 114 at the point A and the passing order of the vehicles 100 having the same series designation number 114 at the point D are the same. In this case, the identification unit 11a determines that the passing order of these vehicles 100 at points B and C is the same as that at points A and D. For example, as shown in FIG. Vehicle 100 is identified between the detection points. FIG. 10 is a diagram showing recognition results at each point in the order in which vehicles pass. Also, "→" shown in FIG. 10 indicates the vehicle 100 identified between two adjacent points.

In the example shown in FIG. 10, the vehicle 100 whose registration symbol is "Kyoto 550 ta 77" passes through the points A and D before the vehicle 100 whose registration symbol is "Naniwa 381 mi 77". Therefore, the identification unit 11a determines that the vehicle 100 whose registration symbol is "Kyoto 550 ta 77" passes through the points B and C before the vehicle 100 whose registration symbol is "Naniwa 381 mi 77". determine and identify the vehicle 100 between points.

Further, when the identification unit 11a determines in s25 that the passing order is not the same, the identification unit 11a identifies the vehicles 100 having the same serial designation number 114 by identifying the passing time of the point A and the passing time of the point D, (s27).

As an example, in the vehicle group determined in s21, as vehicles 100 having the same series designation number 114, a vehicle 100 having a registration code of "Kyoto 550 Ta 77" and a vehicle 100 having a registration code of "Naniwa 381 Mi 77" are selected. and exist. Also, in this example, as shown in FIG. 11, the passing order of the vehicles 100 having the same series designation number 114 at the point A and the passing order of the vehicles 100 having the same series designation number 114 at the point D are the same. not. Specifically, at point A, the vehicle 100 whose registration symbol is "Kyoto 550 ta 77" passes before the vehicle 100 whose registration symbol is "Naniwa 381 mi 77", but at point D, The vehicle 100 whose registration code is "Naniwa 381 Mi 77" has passed before the vehicle 100 whose registration code is "Kyoto 550 Ta 77". In this case, the vehicle 100 whose registration code is "Naniwa 381 Mi 77" overtakes the vehicle 100 whose registration code is "Kyoto 550 Ta 77" between the A point and the D point. Therefore, unless the vehicle 100 whose registration symbol is "Naniwa 381 Mi 77" has overtaken the vehicle 100 whose registration symbol is "Kyoto 550 Ta 77" (overtaking point), as shown in FIG. Vehicles 100 having the same designation number 114 cannot be identified between points.

In s27, the identification unit 11a assumes that the vehicles 100 having the same series designation number 114 traveled from point A to point D at a constant speed, and estimates the overtaking point. Specifically, as shown in FIG. 12, for each vehicle 100 having the same series designation number 114, a linear function between the time and the position is obtained, and the point of intersection thereof is estimated to be the overtaking point. FIG. 12(A) is an example in which the overtaking point is located between points C and D, and FIG. 12(B) is an example in which the overtaking point is located between points B and C. For example.

As described above, at points A and D, the vehicle 100 can be specified by authenticating the registration symbol, and the transit time of each vehicle 100 can be obtained. Therefore, the above linear function is obtained.

◯ and ◇ shown in FIG. 12 are passage times included in the recognition result of each point. Whether the linear function of the time and position obtained for each vehicle 100 having the same series designation number 114 is appropriate, and the passage time of the vehicle 100 having the same series designation number 114 at the B point and the C point are determined. You may make it judge using.

The identification unit 11a identifies the vehicle 100 between points according to the passing point estimated in the above process. FIG. 13(A) shows the identification result of the vehicle between points when the overtaking point is between points C and D. FIG. FIG. 13(B) shows the identification result of the vehicle between the points when the overtaking point is between the B point and the C point. 13A and 13B are diagrams showing the recognition results at each point in the order in which the vehicles pass, similarly to FIG. Also, "→" shown in FIGS. 13A and 13B indicates the vehicle 100 identified between two adjacent points.

The identification unit 11a generates the tracking data shown in FIG. 5 based on the identification result of the vehicle 100 in s24, s26, or s27, stores it in the tracking data storage unit 14 (s28), and ends this process.

Also, the vehicle tracking device 1 according to this example can calculate the travel time of the vehicle 100 between two vehicle passage detection points based on the tracking data stored in the tracking data storage unit 14 . The travel time is calculated by the travel time calculator 11b.

Specifically, the travel time calculation unit 11b calculates the travel time between the two points for each vehicle 100 that passed through the downstream point during the most recent predetermined time (for example, 5 minutes, 10 minutes), The time required for the vehicle 100 to travel between these two points is calculated. Then, the average travel time of these vehicles 100 is calculated as the travel time between the two specified points at that time.

In addition, the vehicle tracking device 1 detects the vehicle 100 that has flowed out of the branch road existing between the A point and the D point, and It is also possible to specify the vehicle 100 that has flowed into the branch road existing between .

Thus, in the vehicle tracking system according to this example, the cameras 6b and 6c installed at points B and C that recognize only the serial designation number 114 are installed at points A and D that authenticate the registration symbols. Since cameras with a resolution lower than that of the cameras 6a and 6d can be used, the cost required for constructing the system can be suppressed. In addition, the above-described identification processing also suppresses deterioration in the accuracy of identification of the vehicle 100 between passing detection points.

In the above example, between point A, which authenticates the registration symbol, and point D, there are two points, point B and point C, where only the serial designation number 114 is authenticated. , and point D, and the number of points where only the serial designation number 114 is authenticated may be one, or three or more.

<4. Variation>
FIG. 14 is a block diagram showing the configuration of main parts of a vehicle tracking device according to another example. The vehicle tracking device 1A according to this example has substantially the same configuration as the above example, but differs in that the control unit 11 has a character recognition section 11c. The character recognition section 11c has the same configuration as the character recognition section 21a of the point processing device 2 in the above example. In FIG. 14, the same reference numerals are assigned to the same configurations as those shown in FIG.

This vehicle tracking device 1A of another example is connected via a network 7 to cameras 6 installed at respective passage detection points (points A to D). That is, the point processing device 2 shown in the above example is not required.

Also, the communication unit 12 receives frame images transmitted from the camera 6 . The character recognition unit 11c recognizes the registration symbol written on the license plate 101 of the captured vehicle 100 in the frame images received from the cameras 6a and 6d. Further, the character recognition unit 11c recognizes the serial designation number written on the license plate 101 of the vehicle 100 being imaged for the frame images received from the cameras 6b and 6c. The character recognition unit 11 c stores the recognition result of the license plate 101 in the recognition result storage unit 13 .

Therefore, with the vehicle tracking device 1A according to this other example, the point processing device 2 does not have to be provided at the passage detection point, so the cost for constructing the system can be further reduced.

Further, in the above description, it is assumed that only the series designation number 114 written on the license plate 101 is recognized for the B point and the C point, but the registration symbol ( It may be configured to recognize area name 111, classification number 112, hiragana 113, and serial designation number 114). In this case, at a certain point, one or more of the area name 111, the classification number 112, or the hiragana 113 cannot be recognized due to the influence of a disturbance such as a shadow. If the vehicle 100 cannot be identified between the points, the identification process shown in FIG. 9 may be executed to identify the vehicle 100 that has passed through each point.

It should be noted that the present invention is not limited to the above-described embodiments as they are, and can be embodied by modifying constituent elements without departing from the scope of the present invention at the implementation stage. Also, various inventions can be formed by appropriate combinations of the plurality of constituent elements disclosed in the above embodiments. For example, some components may be omitted from all components shown in the embodiments. Furthermore, constituent elements of different embodiments may be combined as appropriate. Also, the order of the steps in the flowcharts used in the above description may be changed as appropriate.

Furthermore, the correspondence relationship between the configuration according to the present invention and the configuration according to the above-described embodiment can be described as the following additional remarks.
<Appendix>
A first recognition result of recognizing a registration symbol written on a license plate (101) attached to a vehicle (100) that has passed a first point on a travel path, downstream of the first point on the travel path. A second recognition result of recognizing the registration symbol written on the bumper plate (101) attached to the vehicle (100) that passed through the second point located in the travel path, and the first point and the A serial designation number (114), which is a part of the registration symbol written on the license plate (101) attached to the vehicle (100) that passed the intermediate point between the second point and the second point, was recognized. a recognition result storage unit (13) for storing a third recognition result;
an identification unit (11a) that identifies a vehicle between points using the first recognition result, the second recognition result, and the third recognition result;
The first recognition result includes the passing time of the vehicle (100) at the first point,
The second recognition result includes the passing time of the vehicle (100) at the second point,
The identification unit (11a)
identifying a vehicle (100) that has passed through the first point and a vehicle (100) that has passed through the second point by matching the registration symbols;
Among the vehicles (100) passing through the intermediate point, if there is a vehicle (100) that cannot be identified as a vehicle passing through the first point or the second point by collation of the serial designation number (114), the above A vehicle tracking device (1) that is identified using the passing time of a vehicle (100) at a first point and the passing time of the vehicle (100) at the second point.

1, 1A Vehicle tracking device 2 (2a to 2d) Point processing device 6 (6a to 6d) Camera 7 Network 11 Control unit 11a Identification unit 11b Travel time calculation unit 11c Character recognition unit 12 Communication Part 13... Recognition result storage part 14... Tracking data storage part 15... Output part 21... Control unit 21a... Character recognition part 22... Image input part 23... Communication part 100... Vehicle 101... License plate 111... Region name 112... Classification number 113... Hiragana 114... Serial number

Claims (6)

A first recognition result of recognizing a registration symbol written on a license plate attached to a vehicle that has passed a first point on a travel path, and a second point located downstream of the first point on the travel path. A second recognition result of recognizing the registration symbol written on the hump plate attached to the vehicle that passed through, and an intermediate point located between the first point and the second point on the travel path a recognition result storage unit that stores a third recognition result of recognizing a serial designation number that is a part of the registration symbol written on the license plate attached to the passing vehicle;
an identification unit that identifies a vehicle between points using the first recognition result, the second recognition result, and the third recognition result;
The first recognition result includes the passing time of the vehicle at the first point,
The second recognition result includes the passing time of the vehicle at the second point,
The identification unit
identifying a vehicle that has passed through the first point and a vehicle that has passed through the second point by matching the registration symbols;
If, among the vehicles that have passed through the intermediate point, there is a vehicle that cannot be identified as a vehicle that has passed through the first point or the second point by collation of the serial designation number, the passing time of the vehicle at the first point, and a vehicle tracking device that identifies using the passing time of the vehicle at the second point. The identification unit
If, among the vehicles that have passed through the intermediate point, there is a vehicle that cannot be identified as a vehicle that has passed through the first point or the second point by collation of the serial designation number, the passing time of the vehicle at the first point, Identify using the passing time of the vehicle at the second point, the distance between the first point and the waypoint, and the distance between the second point and the waypoint,
Vehicle tracking device according to claim 1. 3. The vehicle tracking device according to claim 1, further comprising a travel time calculation unit that calculates the travel time of the vehicle between two points based on the identification result of the vehicle at the corresponding two points by the identification unit. . The registration symbol written on the license plate is obtained as the first recognition result by performing character recognition processing on a frame image in which the license plate of the vehicle that has passed the first point is captured, and the vehicle has passed the second point. The registration symbol written on the license plate is acquired as the second recognition result by character recognition processing on the frame image in which the license plate of the vehicle is captured, and the license plate of the vehicle passing through the intermediate point is captured. 4. The vehicle tracking device according to any one of claims 1 to 3, further comprising a recognition result acquisition unit that acquires, as the third recognition result, a serial designation number written on the license plate by character recognition processing on the frame image. . A first recognition result of recognizing a registration symbol written on a license plate attached to a vehicle that has passed a first point on a travel path, and a second point located downstream of the first point on the travel path. A second recognition result of recognizing the registration symbol written on the hump plate attached to the vehicle that passed through, and an intermediate point located between the first point and the second point on the travel path a first step of storing, in a recognition result storage unit, a third recognition result obtained by recognizing a serial designation number that is a part of a registration symbol written on the license plate attached to the passing vehicle;
an identification step of identifying a vehicle between points using the first recognition result, the second recognition result, and the third recognition result;
The first recognition result includes the passing time of the vehicle at the first point,
The second recognition result includes the passing time of the vehicle at the second point,
The identification step includes:
identifying a vehicle that has passed through the first point and a vehicle that has passed through the second point by matching the registration symbols;
If, among the vehicles that have passed through the intermediate point, there is a vehicle that cannot be identified as a vehicle that has passed through the first point or the second point by collation of the serial designation number, the passing time of the vehicle at the first point, and a step of identifying using the passage time of the vehicle at the second point. A first recognition result of recognizing a registration symbol written on a license plate attached to a vehicle that has passed a first point on a travel path, and a second point located downstream of the first point on the travel path. A second recognition result of recognizing the registration symbol written on the hump plate attached to the vehicle that passed through, and an intermediate point located between the first point and the second point on the travel path a first step of storing, in a recognition result storage unit, a third recognition result obtained by recognizing a serial designation number that is a part of a registration symbol written on the license plate attached to the passing vehicle;
causing a computer to perform an identification step of identifying a vehicle between points using the first recognition result, the second recognition result, and the third recognition result;
The first recognition result includes the passing time of the vehicle at the first point,
The second recognition result includes the passing time of the vehicle at the second point,
The identification step includes:
identifying a vehicle that has passed through the first point and a vehicle that has passed through the second point by matching the registration symbols;
If, among the vehicles that have passed through the intermediate point, there is a vehicle that cannot be identified as a vehicle that has passed through the first point or the second point by collation of the serial designation number, the passing time of the vehicle at the first point, and identifying using the passing time of the vehicle at the second point.

Images