JP2008181364A - Image processing apparatus, image processing program, image processing method, image processing system with the apparatus, and traffic information provision system with the apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processing apparatus, an image processing program and an image processing method capable of efficiently and accurately recognizing vehicles by processing pickup images obtained by picking up vehicles driving on a road without previously setting an advancing direction of vehicles. <P>SOLUTION: The image processing apparatus 1a recognizes a vehicle on a lane 11b which is a reversible lane target about a pickup image obtained by picking up a predetermined area R on a road 11 on which reversible lane systems 2a, 2b are installed. The driving direction of the vehicle is calculated on the basis of the position of the vehicle recognized on a certain frame and the position of the vehicle recognized in the just preceding frame. By repeating the calculation for a plurality of vehicles, the advancing direction of vehicles on the lane 11b is determined on the basis of the traveling directions of the plurality of vehicles. After the determination, on the lane 11b in the pickup image of the current frame, a predetermined area is set in the advancing direction on the basis of the position of the vehicle recognized in the just preceding frame and the vehicle is recognized in the predetermined area. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an image processing apparatus, a program, a method, an image processing system including the apparatus, and a traffic information providing system that recognize a vehicle by processing a captured image obtained by capturing the vehicle.

  2. Description of the Related Art Conventionally, a captured image obtained by capturing a vehicle traveling on a road is processed to recognize the vehicle and measure traffic volume and speed. For example, in the traffic flow measurement processing apparatus described in Patent Document 1, the luminance at a plurality of sample points is obtained for each frame based on the captured image captured by the camera, and spatial differentiation processing based on the luminance information is performed to obtain a predetermined value. A binarized image binarized by the threshold value is obtained, a predetermined mask is applied to the binarized image, the position of the vehicle head is obtained from the distribution of sample points corresponding to the edges in the mask, and the vehicle is recognized. .

Such traffic and speed measurements are also desired on roads where the Chuo Line Shift System is applied. The central line shift system is a system that alleviates traffic congestion by shifting the central line that separates the up line and down line of the road according to the time zone and increasing the number of lanes with a large amount of traffic. For example, in the morning rush hours, on the roads where there is a lot of up-line traffic, and in the evening rush hours, on the other hand, there is a lot of down-line traffic. By increasing the number of lanes, traffic congestion is reduced.
JP-A-5-307695

  However, once a vehicle is recognized in a captured image of a certain frame, an area where the vehicle is supposed to exist in the next frame can be predicted to some extent based on the traveling direction of the vehicle. If the vehicle is recognized for every region of the captured image every frame, there is a problem that the efficiency may be poor.

  As a method for solving such a problem, a traveling direction of the vehicle is set in advance, a region where the vehicle is supposed to exist in the next frame is predicted based on the traveling direction, and the vehicle is recognized in this region. It is possible to do. However, it takes time and effort to set the traveling direction of the vehicle in advance. Also, when recognizing a vehicle on a road to which the center line shift system is applied, the traveling direction of the vehicle changes depending on the time zone in the lane subject to the center line shift. There is.

  Accordingly, an object of the present invention is to provide an image processing apparatus and program for processing a captured image obtained by capturing a vehicle traveling on a road and recognizing the vehicle efficiently and accurately without setting the traveling direction of the vehicle in advance. And providing a method. Further, the present invention provides an image processing device, a program, and a method for recognizing a vehicle efficiently and accurately by processing a captured image obtained by capturing a vehicle traveling on the road even on a road to which the center line shift system is applied. There is.

  An image processing apparatus according to the present invention includes an acquisition unit that acquires a captured image obtained by capturing a vehicle traveling on a road having a plurality of lanes, and a recognition unit that recognizes a vehicle in a predetermined lane of the road in the acquired captured image. And a calculating means for calculating the traveling direction of the vehicle based on the position of the vehicle recognized in the captured image at a time point before the present and the position of the vehicle recognized in the current captured image; An image processing apparatus comprising: a determining unit that determines a traveling direction of the vehicle in the predetermined lane based on the traveling directions of the vehicles calculated by the calculating unit for the plurality of vehicles, wherein the recognizing unit includes the determining unit After determining the traveling direction of the vehicle, in the predetermined lane in the current captured image, the position of the vehicle recognized in the captured image at a time point before the present is determined. As a quasi, a setting unit that sets a predetermined region in the determined traveling direction of the vehicle and a recognition unit that recognizes the vehicle in the predetermined region of the current captured image are included. ).

  According to the present invention, a captured image obtained by capturing a vehicle traveling on a road having a plurality of lanes is acquired, and the vehicle is recognized in a predetermined lane of the road in the captured image. Based on the position of the vehicle recognized in a certain frame and the position of the vehicle recognized in the previous frame (for example, the previous frame), the traveling direction of the vehicle (the direction in which the vehicle actually traveled) Is calculated. This is repeated for a plurality of vehicles, the traveling directions of the plurality of vehicles are calculated, and the traveling direction of the vehicle in the predetermined lane based on these traveling directions (the direction in which the vehicle must travel in the predetermined lane) To decide. After determining the traveling direction of the vehicle, the traveling direction is determined based on the position of the vehicle recognized in the previous (for example, the previous) frame in the predetermined lane of the captured image of the current frame. A predetermined area is set and the vehicle is recognized in the area. Therefore, since it is not necessary to recognize the vehicle outside the predetermined area of the captured image, the vehicle can be recognized efficiently. Moreover, since the traveling direction of the vehicle in the predetermined lane is determined based on the traveling directions of the plurality of vehicles, it is not necessary to set the traveling direction of the vehicle in advance.

  In the image processing apparatus of the present invention, the determining means may determine the traveling direction of the vehicle in the predetermined lane only when the ratio of the traveling directions of the plurality of vehicles being equal to or greater than a first threshold. The driving direction is determined according to the present invention (claim 2).

  According to this invention, the traveling in which the traveling directions of the vehicles in the predetermined lane coincide with each other only when the traveling direction of the plurality of vehicles is equal to or greater than a first threshold (for example, 95%). Determine the direction. Therefore, since the traveling direction of the vehicle can be determined with high accuracy, the vehicle can be recognized with high accuracy.

  The image processing apparatus of the present invention further includes a detection unit that detects a traffic volume in the predetermined lane, and the determination unit has a traffic volume detected by the detection unit within a first period equal to or greater than a second threshold value. Only in this case, the traveling direction of the vehicle in the predetermined lane is determined (Claim 3).

  According to this invention, in the predetermined lane, the number of recognized vehicles is counted to detect the traffic volume within the first period (for example, 5 minutes), and this traffic volume is the second threshold (for example, 100 cars). Only in the above case, the traveling direction of the vehicle in the predetermined lane is determined. This is because it is considered that the accuracy is high if the traveling direction of the vehicle is determined based on the traveling directions of the plurality of vehicles equal to or greater than the second threshold. Therefore, since the traveling direction of the vehicle can be determined with high accuracy, the vehicle can be recognized with high accuracy.

  Further, the road is a road to which a center line shift system is applied, and the predetermined lane is a lane that is a target of the center line shift.

  According to this invention, since the traveling direction of the vehicle in the lane subject to the center line shift is determined based on the traveling directions of the plurality of vehicles, even if the traveling direction of the vehicle in the lane changes depending on the time zone. The traveling direction of the vehicle can be determined with high accuracy, and the vehicle can be recognized with high accuracy.

  The image processing apparatus according to the present invention further includes a detection unit that detects a traffic volume in the predetermined lane, and the determination unit is configured such that the traffic volume detected by the detection unit within a second period is less than a third threshold value. Only, the traveling direction of the vehicle in the predetermined lane is determined again (Claim 5).

  In the Chuo Line Transition System, the vehicle is prohibited from traveling in the lane that is the subject of the Chuo Line Transition for a predetermined period from the start of the Chuo Line Transition. A vehicle traveling in the traveling direction of the vehicle before the line shift is completely driven out of this lane. And after the said predetermined period progress, driving | running | working of a vehicle is permitted in the advancing direction of the vehicle after a center line change. Therefore, the traffic on this lane will be very low during the predetermined period (for example, less than 5 cars in 5 minutes), so if it can be detected that the traffic is very low, the center line will change before and after this. Thus, it can be determined that the traveling direction of the vehicle has been changed in the lane subject to the change of the center line.

  According to this invention, in the lane subject to the change of the center line, the number of recognized vehicles is counted to detect the traffic volume in the second period (for example, 5 minutes), and this traffic volume is (for example, 5 cars). Only when it is less than the third threshold, it is determined that the traveling direction of the vehicle has changed in the lane, and the traveling direction of the vehicle is re-determined, so that the traveling direction of the vehicle in the lane is surely determined after the change of the center line. Can do.

  In the image processing apparatus of the present invention, the second period is a period required for the vehicle to travel from one end to another end of the lane targeted for the center line shift. (Claim 6).

  During the above-mentioned predetermined period of prohibiting the vehicle from traveling in the lane subject to the center line transition, at least until the vehicle that has entered the lane subject to the center line transition exits this lane immediately before the start of the center line transition The period required for the vehicle, that is, the period required for the vehicle to travel from one end of the lane to the other end is equal to or longer. Therefore, if the traffic during the period required for the vehicle to travel from one end of the lane to the other end can be detected and it can be detected that the traffic is very low, it is possible to accurately detect that the center line has changed around this time. Can be recognized.

  According to the present invention, the traffic in the period required for the vehicle to travel from one end to the other end of the lane subject to the change of the center line is compared with the third threshold, so the center line has changed in this lane. This can be recognized with high accuracy, and the traveling direction of the vehicle in this lane can be reliably determined after the transition to the center line.

  The image processing apparatus according to the present invention recognizes a vehicle in a predetermined lane of the road in the acquired captured image and an acquisition unit that acquires a captured image of a vehicle traveling on a road having a plurality of lanes. An image processing apparatus including a recognizing unit, wherein the recognizing unit is configured to use the predetermined lane in the predetermined lane in the current captured image with reference to the position of the vehicle recognized in the captured image before the current time. Setting a first predetermined area in one direction along the predetermined direction, and setting a second predetermined area in another direction along the predetermined lane, and the first and And a recognition unit for recognizing the vehicle in a second predetermined area.

  According to the present invention, two predetermined areas are set in both directions along the lane based on the position of the vehicle recognized in the previous frame without considering the traveling direction of the vehicle in the predetermined lane. The vehicle is recognized within. Therefore, even if the traveling direction of the vehicle in the lane is unknown, the vehicle can be reliably recognized within the two predetermined areas, and the vehicle need not be recognized outside the predetermined area. It can be recognized efficiently.

  In addition, the road is a road to which a center line shift system is applied, and the predetermined lane is a lane that is a target of the center line shift.

  The image processing apparatus of the present invention also includes first acquisition means for acquiring a captured image obtained by imaging a vehicle traveling on a road having a plurality of lanes, and information on the traveling direction of the vehicle in the predetermined lane on the road. An image processing apparatus comprising: a second acquisition unit that acquires from a recognition unit that recognizes a vehicle in the predetermined lane in the acquired captured image, wherein the recognition unit includes the predetermined lane in a current captured image. , The setting unit for setting a predetermined region in the traveling direction of the vehicle with reference to the position of the vehicle recognized in the captured image at a time point before the current time, and recognizing the vehicle in the predetermined region of the current captured image And a recognizing unit (claim 9).

  According to this invention, the traveling direction of the vehicle in the predetermined lane is acquired from the outside instead of being determined based on the traveling direction of the vehicle recognized based on the captured image. A predetermined area is set in the traveling direction of the vehicle with reference to the position of the vehicle recognized in the previous frame, and the vehicle is recognized in the area. Therefore, the vehicle can be reliably recognized within the predetermined area, and it is not necessary to recognize the vehicle outside the predetermined area, so that the vehicle can be efficiently recognized.

  In addition, the road is a road to which a center line shift system is applied, the predetermined lane is a lane that is a target of the center line shift, and the second acquisition means is a travel direction of the vehicle from the center line shift system. Is obtained (claim 10).

  The image processing apparatus of the present invention includes an acquisition unit that acquires a captured image obtained by capturing a vehicle traveling on a road to which the central line shift system is applied, and an imaging time of the captured image, and a centerline shift of the road. An image processing apparatus comprising: a storage unit that stores correspondence information that associates a traveling direction of a vehicle with a time in a target lane; and a recognition unit that recognizes the vehicle in the lane in the acquired captured image. The recognizing means is based on the position of the vehicle recognized in the captured image before the current time in the lane in the current captured image, and the acquisition time of the captured image at the time prior to the current A setting unit that sets a predetermined area in the traveling direction of the vehicle determined based on the correspondence information, and the vehicle is recognized in the predetermined area of the current captured image Characterized in that it comprises a 識部 (claim 11).

  According to the present invention, instead of determining the traveling direction of the vehicle in the lane subject to the center line shift based on the traveling direction of the vehicle recognized based on the captured image, the vehicle in the lane stored in advance is stored. By comparing the correspondence information in which the traveling direction and the time are associated with the imaging time of the captured image of the previous frame, the traveling direction of the vehicle in the captured image is determined. A predetermined area is set in the determined traveling direction of the vehicle based on the position of the vehicle recognized in the captured image, and the vehicle is recognized in the area. Therefore, the vehicle can be reliably recognized within the predetermined area, and it is not necessary to recognize the vehicle outside the predetermined area, so that the vehicle can be efficiently recognized.

  Moreover, the image processing system of this invention is provided with the imaging means which images the vehicle which drive | works on a road, and said image processing apparatus (Claim 12).

  The traffic information providing system of the present invention is a traffic information providing system comprising the above-described image processing device and a traffic information calculating device that calculates and provides traffic information, wherein the image processing device includes the recognition means. And a means for outputting the recognition result of the vehicle in the predetermined lane recognized by the vehicle and the traveling direction of the vehicle in the predetermined lane determined by the determining means, and the traffic information calculation device includes the recognition result and the vehicle Means for acquiring a traveling direction; and means for calculating traffic information in the predetermined lane based on the acquired recognition result and the traveling direction of the vehicle. .

  According to this invention, since traffic information in the lane is calculated based on the recognition result of the vehicle in the predetermined lane acquired from the image processing device and the traveling direction of the vehicle, traffic information such as traffic volume in the up line Traffic information such as traffic on the down line can be calculated with high accuracy.

  An image processing program and an image processing method according to the present invention are the same program and method according to the same invention as the image processing apparatus described above (claims 14 and 15).

  As described above, according to the image processing device, the program, and the method of the present invention, it is possible to efficiently process a captured image obtained by capturing a vehicle traveling on a road without setting the traveling direction of the vehicle in advance. The vehicle can be recognized with high accuracy. In addition, even on a road to which the center line shift system is applied, a captured image obtained by capturing a vehicle traveling on the road can be processed to recognize the vehicle efficiently and accurately.

Embodiment 1:
Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. FIG. 1 is a schematic diagram showing an outline of a traffic information providing system including an image processing apparatus and an image processing system according to the present invention. The traffic information providing system includes an image processing system 1, center line shift systems 2 a and 2 b, vehicle detectors 3 a and 3 b, roadside devices 4 such as optical beacons, traffic signals 5 a and 5 b, a traffic control center (not shown), and the like. Including. The traffic signal 5a is connected to the image processing system 1, vehicle detectors 3a and 3b, and the roadside device 4 through a communication line (not shown). The traffic signal 5b is connected to an image through a communication line (not shown). A processing system 1 and vehicle detectors 3a and 3b are connected. The traffic control center is connected to the center line transition systems 2a, 2b and traffic signals 5a, 5b through a communication line (not shown).

  The image processing system 1 includes an image processing device 1a, an imaging device 1b, an installation pole 1c, and the like. The imaging device 1b is a camera for imaging a vehicle on the road 11, and is attached to the upper portion of the installation pole 1c on the side of the road 11 with a predetermined dip so that the field of view is set to a predetermined region R on the road 11. It has been. The image processing device 1a is a device that processes a captured image obtained by imaging with the imaging device 1b to recognize a vehicle and is attached to a side portion of the installation pole 1c. In addition, you may divert a traffic light, a footbridge, etc. instead of the installation pole 1c. The image processing apparatus 1a transmits information such as a vehicle recognition result to the traffic signals 5a and 5b through a communication line. The predetermined region R is provided so as to extend over all lanes of the first lane 11a, the second lane 11b, and the third lane 11c of the road 11.

  The center line shift systems 2a and 2b are intended for the center line shift of the second lane 11b. The center line transition system 2a includes a center line transition device 2aa, an arm 2ab, and the like. The center line shifting device 2aa is provided with a sign for indicating that it is a center line with respect to a vehicle entering the road 11 from the intersection 12a, moves along the arm 2ab according to the time zone, and moves to the first lane. The center line is shifted to the white line 13a that separates the 11a and the second lane 11b or the white line 13b that separates the second and third lanes 11b and 11c. The lane in which a vehicle entering the road 11 from the intersection 12a can travel is only the first lane 11a when the sign is on the white line 13a, and the first lane 11a when the sign is on the white line 13b. The second lane 11b becomes two lanes.

  Similarly, the center line transition system 2b includes a center line transition device 2ba, an arm 2bb, and the like. The center line shifting device 2ba is provided with a sign for indicating that it is a center line for vehicles entering the road 11 from the intersection 12b, and moves along the arm 2bb according to the time zone. The center line is shifted to the white line 13b. The lane in which a vehicle entering the road 11 from the intersection 12b can travel is the second lane 11b and the third lane 11c when the sign is on the white line 13a, and the sign is on the white line 13b. Only the third lane 11c. Both the central line transition systems 2a and 2b can acquire a command from the traffic control center through the communication line, and the time zone for shifting the central line can be changed according to this command.

  Here, the procedure in which the center line transition systems 2a and 2b shift the center line will be described. FIG. 2 is a diagram for explaining a procedure in which the center line transition systems 2a and 2b shift the center line, and illustrates a case where the center line is shifted from the white line 13a to the white line 13b. FIG. 2A shows a state before the transition of the center line, and the center line transition devices 2aa and 2ba are both on the white line 13a. First, as shown in FIG. 2B, only the center line shifting device 2ba is moved on the white line 13b so that the vehicle does not enter the second lane 11b from the intersection 12b. Since the vehicle cannot enter the second lane 11b from the intersection 12a, the center line shifting devices 2aa and 2ba maintain the same position until a predetermined time (hereinafter, this predetermined time is referred to as a vehicle travel prohibition period). The vehicle waits for the vehicle that has entered the second lane 11b to leave the lane immediately before the line shift device 2ba moves on the white line 13b. When the vehicle travel prohibition period elapses, the center line transition device 2aa is moved onto the white line 13b, and the transition of the center line from the white line 13a to the white line 13b is completed. Conversely, when the center line is changed from the white line 13b to the white line 13a, the center line change device 2aa is first moved onto the white line 13a, and after the vehicle travel prohibition period has elapsed, the center line change device 2ba is moved to the white line 13b. Move up.

  The vehicle sensor 3a includes sensing units 3aa1, 3aa2, 3aa3, an arm 3ab, and the like. The sensing units 3aa1, 3aa2, 3aa3 sense vehicles traveling in the first lane 11a, the second lane 11b, and the third lane 11c, respectively. The vehicle detector 3a transmits information on the detection results of these sensors to the traffic signals 5a and 5b through a communication line.

  Similarly, the vehicle sensor 3b includes sensing units 3ba1, 3ba2, 3ba3, an arm 3bb, and the like. The sensing units 3ba1, 3ba2, and 3ba3 sense vehicles traveling in the first lane 11a, the second lane 11b, and the third lane 11c, respectively. The vehicle detector 3b transmits information on the detection results of these sensors to the traffic signals 5a and 5b through a communication line.

  The roadside device 4 includes communication units 4a1, 4a2, 4a3, an arm 4b, and the like. The communication units 4a1, 4a2, 4a3 perform wireless two-way communication with vehicles traveling in the first lane 11a, the second lane 11b, and the third lane 11c, respectively. When the vehicle passes through the communication area of these communication units, uplink information including an identification code of the in-vehicle device is transmitted from the in-vehicle device mounted on the vehicle to the communication unit, and the communication unit Downlink information including traffic information is transmitted to the in-vehicle device. The downlink information is acquired by the roadside device 4 from the traffic signal 5a through a communication line. Further, the received uplink information is transmitted from the roadside device 4 to the traffic signal 5a through the communication line.

  The traffic signal 5a includes a traffic signal controller 5aa, a signal lamp 5ab, a signal pole 5ac, and the like. The traffic signal controller 5aa controls lighting and extinction of each signal lamp of the signal lamp 5ab attached to the upper part of the signal pole 5ac. Further, the traffic signal controller 5aa acquires information such as vehicle recognition results from the image processing system 1, acquires information on vehicle detection results from the vehicle detectors 3a and 3b, and receives uplink information from the roadside device 4. Acquire and send the acquired information to the traffic control center. The traffic signal controller 5aa acquires downlink information from the traffic control center, and transmits this information to the roadside device 4.

  Similarly, traffic signal 5b includes traffic signal controller 5ba, signal lamp 5bb, signal pole 5bc, and the like. The traffic signal controller 5ba controls lighting and extinction of each signal lamp of the signal lamp 5bb attached to the upper part of the signal pole 5bc. Further, the traffic signal controller 5ba acquires information such as vehicle recognition results from the image processing system 1, acquires vehicle detection result information from the vehicle detectors 3a and 3b, and uses the acquired information as traffic control. Send to center.

  FIG. 3 is a block diagram showing a configuration of the image processing system 1. The imaging device 1b outputs a captured image obtained by imaging to the interface (I / F) unit 1a1 as a video signal (analog signal). The interface unit 1a1 outputs the acquired video signal to the A / D conversion unit 1a2, and the A / D conversion unit 1a2 converts the input video signal into a digital signal, which is converted under the control of the processing unit 1a7. The obtained digital signal is stored in the image memory 1a3 as image data.

  The processing unit 1a7 performs the following processing by executing the computer program PG stored in the RAM 1a4. That is, the processing unit 1a7 uses the captured image input from the imaging device 1b via the interface unit 1a1 as image data and synchronizes with the frame rate of the imaging device 1b (interval at the time of imaging, for example, 30 frames per second). And stored in the image memory 1a3 in units of one frame (for example, 480 × 640 pixels). In addition, the processing unit 1a7 performs a vehicle recognition process, which will be described later, and calculates a traffic volume, an average traveling speed, and the like in each lane based on the vehicle recognition result.

  The storage unit 1a5 stores information on the recognition result of the vehicle obtained by executing the computer program PG in the processing unit 1a7 corresponding to each frame. The storage unit 1a5 stores the traveling direction of the vehicle in each lane (the direction in which the vehicle must travel). The storage unit 1a5 stores information such as traffic volume and average travel speed in each lane.

  The communication unit 1a6 is connected to a communication line (not shown) and stores information on vehicle recognition results, traffic volume, average travel speed, etc., and information on the traveling direction of the vehicle in each lane stored in the storage unit 1a5. And transmitted to the traffic signals 5a and 5b via the communication line.

  Next, the procedure of the vehicle recognition process of the image processing apparatus 1a will be described. FIG. 4 is a flowchart showing the procedure of the vehicle recognition process of the image processing apparatus 1a. The vehicle recognition process is performed every predetermined time (for example, 33 msec, which is the same as the frame rate of the imaging device 1b). When the image processing apparatus 1a is started, the direction from the intersection 12a to the intersection 12b is set as the traveling direction of the vehicle in the first lane 11a, and the direction from the intersection 12a to the intersection 12b is set as the traveling direction of the vehicle in the second lane 11b. Both directions of the direction from the intersection 12b to the intersection 12a are stored as the traveling direction of the vehicle in the third lane 11c, and the direction from the intersection 12b to the intersection 12a is stored in the storage unit 1a5. The reason why the traveling direction of the vehicle in the second lane 11b is set to both directions is that the second lane 11b is the target of the center line shift and it is unknown which direction the traveling direction of the vehicle is.

  First, the image processing device 1a reads a captured image from the image memory 1a3, and sequentially selects a lane to be processed (S1). FIG. 5 is a diagram illustrating an example of the area of each lane in the captured image and the traveling direction of the vehicle in each lane. The areas of the lanes in the captured image are the area PRc of the third lane 11c, the area PRb of the second lane 11b, and the area PRa of the first lane 11a in order from the left, and the traveling direction of the vehicle in each lane is the left From top to bottom, they are upward, both upward and downward, and downward.

  The image processing device 1a ends the recognition process when all the lanes, that is, all three lanes have been processed (YES in S2). If all the lanes have not been processed (NO in S2), the image processing device 1a determines whether or not a vehicle is present in the lane selected in S1. Specifically, the vehicle is detected by edge processing, background difference processing, time difference processing, or the like in the lane region of the captured image. When there is a vehicle (YES in S3), the image processing device 1a reads the traveling direction of the vehicle in the lane stored in the storage unit 1a5, determines whether the vehicle is in both directions, and both directions If YES (YES in S4), the vehicle is tracked in both directions (S5). The process of S5 is performed for each vehicle. FIG. 6 is a diagram for explaining the tracking process of the vehicle in the region PRb of the lane when the traveling direction of the vehicle in the second lane 11b is both directions. FIG. 6A shows the vehicle head position Hb recognized one frame before. First, as shown in FIG. 6B, two tracking areas T1 and T2 are set in both directions with the vehicle head position Hb as a reference. In the region PRb, when the x-axis is directed to the right and the y-axis is directed downward, for example, the tracking region T1 starts from the position y0 in the y direction at the center of the vehicle head position Hb during the period (33 msec) of this processing. A position y1 in the y direction that is separated in the -y direction by the length l obtained by converting the distance traveled at the assumed speed (for example, 60 km / h) on the captured image is obtained, and the α, + y directions are obtained from this y1 in the -y direction. Is set to the shaded area in β. Further, for example, the tracking region T2 obtains a position y2 in the y direction that is separated from the y0 by a length l in the + y direction, and is in a hatched area in the α from the y2 to the + y direction and β in the −y direction. Set. Next, as shown in FIG. 6C, the vehicle is recognized in the tracking areas T1 and T2. If the vehicle can be recognized, it is determined that this vehicle is the same as the vehicle one frame before. Accordingly, it is not necessary to recognize the vehicle in all of the regions PRb, so that the vehicle can be recognized efficiently.

  Based on the number of vehicles that have been tracked in S5 within a predetermined time (for example, 5 minutes; hereinafter, this predetermined time will be referred to as the first period) in the past, the image processing device 1a is able to track the cumulative number of tracks in both directions. Is calculated (S6). Here, the cumulative tracked number refers to a cumulative value of the number of vehicles that can be tracked in S5 in each of both directions. The vehicle is captured several times before one vehicle passes through the predetermined region R on the road 11, so that the vehicle is recognized several times from the captured image, but all are the same in S5. Therefore, the number of vehicles that can be tracked is one.

  The image processing apparatus 1a determines whether or not the total value of the cumulative number of tracks in both directions is equal to or greater than a predetermined number (for example, 100) (YES in S7), the tracking direction. It is determined whether or not there is a bias. That is, the ratio of the upward direction and the ratio of the downward direction to the total value of the total number of tracked vehicles in both directions are respectively obtained, and if any of the ratios is equal to or greater than a predetermined ratio (for example, 95%), the bias is Judge that there is. As a result, if there is a bias (YES in S8), it is determined that the traveling direction of most vehicles (the direction in which the vehicle has actually traveled) matches, and the traveling direction of the vehicle in the lane 11b is predetermined. Is set in a direction that is equal to or greater than the ratio, and is stored in the storage unit 1a5 (S9), and the set traveling direction of the vehicle is transmitted to the traffic signals 5a and 5b (S10). As described above, the image processing apparatus 1a is configured so that the vehicle is only in the case where the total value of the cumulative number of tracked vehicles in both directions is equal to or greater than the predetermined number and the ratio of the cumulative tracked number in the upward or downward direction is equal to or greater than the predetermined rate. Therefore, the traveling direction of the vehicle can be determined with high accuracy.

  The image processing apparatus 1a continues the processing after S1 when the total value of the cumulative number of tracks in both directions is less than the predetermined number (NO in S7), or when there is no bias in the tracking direction (NO in S8).

  When the traveling direction of the vehicle in the lane selected in S1 is not bi-directional (NO in S4), the image processing device 1a reads the traveling direction of the vehicle in the lane stored in the storage unit 1a5 and moves the vehicle in that direction. Is tracked (S11). The process of S11 is also performed for each vehicle. FIG. 7 is a diagram for explaining the vehicle tracking process in the region PRb of the lane when the traveling direction of the vehicle in the second lane 11b is upward. The difference from FIG. 6 is that only the tracking region T1 is set. FIG. 7A shows the vehicle head position Hb recognized one frame before. First, as shown in FIG. 7 (b), the tracking area T1 is obtained as a position y1 in the y direction that is separated from the position y0 in the y direction at the center of the vehicle head position Hb by a length l in the -y direction. It is set to the shaded area in α in the −y direction and β in the + y direction. Next, as shown in FIG. 7C, the vehicle is recognized in the tracking area T1. If the vehicle can be recognized, it is determined that this vehicle is the same as the vehicle one frame before. Therefore, it is not necessary to recognize the vehicle in all the regions PRb, and the vehicle is recognized only in the tracking region T1 in the traveling direction of the vehicle, so that the vehicle can be recognized efficiently. Since the traveling direction of the vehicle in the first lane 11a is upward, the vehicle tracking process in the region PRa of the lane is the same.

  On the other hand, FIG. 8 is a diagram for explaining the vehicle tracking process in the region PRb of the lane when the traveling direction of the vehicle in the second lane 11b is downward. The difference from FIG. 6 is that only the tracking region T2 is set. FIG. 8A shows the vehicle head position Hb recognized one frame before. First, as shown in FIG. 8 (b), the tracking region T2 is obtained as a position y2 in the y direction that is separated from the position y0 in the y direction at the center of the vehicle head position Hb by the length l in the + y direction. It is set to the shaded area in α in the direction and β in the −y direction. Next, as shown in FIG. 8C, the vehicle is recognized in the tracking area T2. If the vehicle can be recognized, it is determined that this vehicle is the same as the vehicle one frame before. Therefore, it is not necessary to recognize the vehicle in all the regions PRb, and the vehicle is recognized only in the tracking region T2 in the traveling direction of the vehicle, so that the vehicle can be recognized efficiently. Since the traveling direction of the vehicle in the third lane 11c is downward, the vehicle tracking process in the region PRc of the lane is the same.

  If there is no vehicle in the lane selected in S1 (YES in S3), the image processing device 1a performs the predetermined time (for example, 5 minutes; hereinafter, this predetermined time is referred to as the second period). It is determined whether the traffic amount of the lane (the number of vehicles that can be tracked in S4 or S11 in the lane) is less than a predetermined number (for example, five). As a result, if it is less than the predetermined number (YES in S12), it is determined that the traveling direction of the vehicle has been changed in the lane, and the traveling direction of the vehicle is set to both the upward direction and the downward direction (S13). .

  Here, when the traffic volume of the lane in the second period is less than the predetermined number, it is determined that the traveling direction of the vehicle is changed in the lane for the following reason. As described above, since the vehicle cannot enter the second lane 11b during the vehicle travel prohibition period, the traffic volume of the second lane 11b in the second period becomes very small. Therefore, if it can be detected that the traffic volume of the lane in the second period is very small, it can be determined that the center line has changed before and after this, and as a result, the traveling direction of the vehicle has been changed in the lane. Judgment can be made.

  The second period may be a fixed value as described above, or may be a time required for the vehicle to travel from one end of the second lane 11b to the other end. The vehicle travel prohibition period is at least the time required for a vehicle that has entered the second lane 11b to exit the lane immediately before the start of the center line transition, that is, from one end of the second lane 11b to the other end. This is because it is set to be longer than the time required for the vehicle to travel to the section. In this case, the predetermined time can be obtained, for example, by dividing the length of the road 11 by an assumed speed (for example, 60 km / h).

  In order to re-determine the traveling direction of the vehicle in the lane, the image processing device 1a clears the cumulative number of tracks in both directions (S14), and continues the processing after S1.

  When the traffic volume of the lane in the predetermined time is larger than the predetermined number (NO in S12), the image processing device 1a continues the processing after S1.

  Next, the procedure of the measurement data collection process of the traffic signals 5a and 5b will be described. The traffic signals 5a and 5b refer to the information on the traveling direction of the vehicle in the second lane 11b acquired from the image processing device 1a. When the traveling direction is the direction from the intersection 12a to the intersection 12b, the traffic signal 5a Acquires information such as traffic volume and average travel speed in the first lane 11a and the second lane 11b from the image processing device 1a, and the detection results in the first lane 11a and the second lane 11b from the vehicle detectors 3a and 3b. The traffic signal 5b obtains information such as traffic volume and average travel speed in the third lane 11c from the image processing device 1a, and obtains information on the sensing result in the third lane 11c from the vehicle detectors 3a and 3b. Acquire and send each to the traffic control center.

  On the other hand, when the traveling direction is the direction from the intersection 12b to the intersection 12a, the traffic light 5a acquires information such as the traffic volume and the average traveling speed in the first lane 11a from the image processing device 1a, and the vehicle detector 3a, 3b to obtain information on the sensing result in the first lane 11a, the traffic signal 5b obtains information on the traffic volume, average travel speed, etc. in the second lane 11b and the third lane 11c from the image processing device 1a, Information on sensing results in the second lane 11b and the third lane 11c is acquired from the vehicle detectors 3a and 3b, and transmitted to the traffic control center.

  Next, a procedure for calculating and providing traffic information at the traffic control center will be described. When the traffic control center acquires measurement data in the direction from the traffic signal 5a, 5b to the intersection 12b in the second lane 11b, in the first lane 11a, the second lane 11b, the intersection 12a to the intersection 12b. Traffic information such as traffic jam information in the direction to is calculated, and traffic information such as traffic jam information in the direction from the intersection 12b to the intersection 12a is calculated in the third lane 11c. Then, the traffic control center provides downlink information including traffic information of the first lane 11a and the second lane 11b to the vehicle from the communication units 4a1 and 4a2 of the roadside device 4 via the traffic signal 5a. Downlink information including traffic information of the three lanes 11c is provided to the vehicle from the communication unit 4a3 of the roadside device 4 via the traffic signal 5a.

  On the other hand, when the traffic control center acquires measurement data in the direction from the traffic signal 5a, 5b to the intersection 12a in the second lane 11b, the direction from the intersection 12a to the intersection 12b in the first lane 11a. Traffic information such as traffic jam information is calculated, and traffic information such as traffic information in the direction from the intersection 12b to the intersection 12a is calculated in the second lane 11b and the third lane 11c. The traffic control center then provides the downlink information including the traffic information of the first lane 11a to the vehicle from the communication unit 4a1 of the roadside device 4 via the traffic signal 5a, and the second lane 11b, the third lane. Downlink information including traffic information 11c is provided to the vehicle from the communication units 4a2 and 4a3 of the roadside device 4 via the traffic signal 5a.

  As described above, the traffic control center calculates the traffic information in the lane in consideration of the traveling direction of the vehicle in the second lane 11b that is the target of the center line shift, so the traffic from the intersection 12a to the intersection 12b is calculated. The traffic information of the direction and the traffic information of the direction from the intersections 12b to 12a can be calculated with high accuracy.

  Further, the traffic control center provides traffic information in the direction from the intersection 12a to the intersection 12b only to vehicles traveling in the direction, and the traffic information in the direction from the intersection 12b to the intersection 12a is provided in the vehicle traveling in the direction. Therefore, only the traffic information necessary for the vehicle driver can be efficiently provided.

  In the above-described embodiment, the image processing device 1a is configured to determine the traveling direction of the vehicle in all lanes, but is not limited to this, and the lane that is the target of the center line transition. The structure which determines the advancing direction of a vehicle only by may be sufficient.

  In the above-described embodiment, the configuration recognizes a vehicle traveling on a road to which the central line shift system is applied. However, the present invention is not limited to this, and the road to which the central line shift system is not applied. The structure which recognizes the vehicle which drive | works above may be sufficient.

Embodiment 2:
In Embodiment 1 described above, the configuration is such that the traveling direction of the vehicle is determined in all lanes or lanes that are subject to centerline transition. However, the present invention is not limited to this. The structure which acquires the advancing direction of the vehicle in the lane used as the object of center line transition from a transition system may be sufficient.

  In the second embodiment, the image processing system 1 is connected to the center line transition systems 2a and 2b through communication lines, and the second lane 11b that is the target of the center line transition from the center line transition systems 2a and 2b. Information on the traveling direction of the vehicle at is always acquired.

  The image processing device 1a stores the direction from the intersection 12a to the intersection 12b as the traveling direction of the vehicle in the first lane 11a, and the direction from the intersection 12b to the intersection 12a as the traveling direction of the vehicle in the third lane 11c. It is stored in 1a5. Moreover, if the information of the advancing direction of the vehicle in the 2nd lane 11b is acquired from the center line transition system 2a, 2b, it will memorize | store in the memory | storage part 1a5.

  The image processing apparatus 1a performs vehicle recognition processing according to the following procedure. First, the image processing device 1a reads a captured image from the image memory 1a3, and sequentially selects a lane to be processed. If all the lanes, that is, all three lanes have been processed, the image processing device 1a ends the recognition process. If all the lanes have not been processed, the image processing device 1a determines whether a vehicle is present in the selected lane. When the vehicle exists, the traveling direction of the vehicle in the lane is read from the storage unit 1a5, and the vehicle is tracked in this direction. This vehicle tracking is the same as in the first embodiment. Therefore, it is not necessary to recognize the vehicle in the entire area of the second lane 11b of the captured image, and the vehicle is recognized only in the tracking area in the traveling direction of the vehicle, so that the vehicle can be recognized efficiently.

  In Embodiment 2 described above, the configuration recognizes a vehicle traveling on a road to which the center line shift system is applied. However, the present invention is not limited to this, and the center line shift system is not applied. The structure which recognizes the vehicle which drive | works on a road may be sufficient. In this case, the image processing system 1 may always acquire information on the traveling direction of each lane from the outside (for example, a traffic control center or a traffic signal).

Embodiment 3:
In the first embodiment described above, the traveling direction of the vehicle is determined in all lanes or lanes that are subject to center line transition. However, the present invention is not limited to this. The correspondence information in which the traveling direction of the vehicle in the lane is associated in advance may be stored in advance, and the traveling direction of the vehicle in the lane may be determined from the imaging time of the captured image and the correspondence information.

  In the third embodiment, the image processing device 1a uses the direction from the intersection 12a to the intersection 12b as the traveling direction of the vehicle in the first lane 11a, and the direction from the intersection 12b to the intersection 12a as the traveling direction of the vehicle in the third lane 11c. Is stored in the storage unit 1a5. Further, correspondence information in which the time and the traveling direction of the vehicle in the second lane 11b are associated with each other is stored in the storage unit 1a5.

  FIG. 9 is a diagram for explaining an example of correspondence information. Correspondence information is composed of four items: date type, start time, end time, and vehicle traveling direction. In the item numbers, integers are stored in order from 1. As the day type, four types of item number, weekday, Saturday, holiday, and special day are stored. As the start time and end time, a time in 24-hour notation is stored. In the traveling direction of the vehicle, the top (direction from the intersection 12a to the intersection 12b) and the bottom (direction from the intersection 12b to the intersection 12a) are stored. For example, the data of item number 1 indicates that the traveling direction of the vehicle in the second lane 11b is upward from 0:00 to 12:00 on weekdays. The data of item number 2 indicates that the traveling direction of the vehicle in the second lane 11b is downward from 12:00 to 24:00 on weekdays. The data of item number 9 indicates that the traveling direction of the vehicle in the second lane 11b is downward from 10:00 to 24:00 on a singular day, that is, throughout the day.

  The image processing apparatus 1a performs vehicle recognition processing according to the following procedure. First, the image processing device 1a reads a captured image from the image memory 1a3, and sequentially selects a lane to be processed. If all the lanes, that is, all three lanes have been processed, the image processing device 1a ends the recognition process. If all the lanes have not been processed, the image processing device 1a determines whether a vehicle is present in the selected lane. When the vehicle exists, if the lane is the first lane 11a or the third lane 11c, the traveling direction of the vehicle is read from the storage unit 1a5, and the vehicle is tracked in this direction. If the lane is the second lane 11b, the corresponding information is read from the storage unit 1a5, and the imaging time and date type of the captured image, that is, the current time and date type (vehicle recognition processing is performed in synchronization with the frame rate). In this case, the traveling direction of the vehicle is determined with reference to data corresponding to the captured time of the captured image (which substantially coincides with the current time), and the vehicle is tracked in this direction. For example, if the current time and day type is 12:34 on Saturday, the data of item number 4 in the corresponding information is applicable, so the traveling direction of the vehicle is determined upward (direction from the intersection 12a to the intersection 12b). To do. Therefore, it is not necessary to recognize the vehicle in the entire area of the second lane 11b of the captured image, and the vehicle is recognized only in the tracking area in the traveling direction of the vehicle, so that the vehicle can be recognized efficiently.

  In the above-described embodiment, the correspondence information is configured to set the traveling direction of the vehicle with respect to the combination of the day type and the start time / end time. However, the present invention is not limited to this. Alternatively, the vehicle traveling direction may be set for the combination of the start time and the end time.

  The embodiment disclosed this time should be considered as illustrative in all points and not restrictive. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

It is a schematic diagram which shows the outline | summary of the traffic information provision system containing the image processing apparatus and image processing system which concern on this invention. It is a figure for demonstrating the procedure in which a center line transition system changes a center line. It is a block diagram which shows the structure of an image processing system. It is a flowchart which shows the procedure of the recognition process of the vehicle of an image processing apparatus. It is a figure which shows the example of the area | region of the lane in a captured image, and the advancing direction of the vehicle in each lane. It is a figure for demonstrating the tracking process of the vehicle in the area | region of the said lane in case the advancing direction of the vehicle in a 2nd lane is both directions. It is a figure for demonstrating the tracking process of the vehicle in the area | region of the said lane when the advancing direction of the vehicle in a 2nd lane is an upward direction. It is a figure for demonstrating the tracking process of the vehicle in the area | region of the said lane when the advancing direction of the vehicle in a 2nd lane is a downward direction. It is a figure for demonstrating the example of correspondence information.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Image processing system 1a Image processing apparatus 1a1 Interface part 1a2 A / D conversion part 1a3 Image memory 1a4 RAM
1a5 Storage unit 1a6 Communication unit 1a7 Processing unit 1b Imaging device 1c Installation pole 2a, 2b Center line shift system 2aa, 2ba Center line shift device 2ab, 2bb Arm 3a, 3b Vehicle sensor 3aa1, 3aa2, 3aa3, 3ba1, 3ba2, 3ba3 Sensor 3ab, 3bb Arm 4 Roadside device 4a1, 4a2, 4a3 Communication unit 4b Arm 5a, 5b Traffic signal 5aa, 5ba Traffic signal controller 5ab, 5bb Signal lamp 5ac, 5bc Signal pillar 11 Road 11a Road first lane 11b Road Second lane 11c Third lane 12a, 12b of road Intersection 13a White line 13b separating first lane and second lane White line R separating second lane and third lane R Predetermined region PRa on road First image Lane PRb Second lane region PRc in the captured image Region of the third lane in Hb headway position T1, T2 tracking region

Claims (15)

Acquisition means for acquiring a captured image obtained by imaging a vehicle traveling on a road having a plurality of lanes;
Recognition means for recognizing a vehicle in a predetermined lane of the road in the acquired captured image;
Calculation means for calculating the traveling direction of the vehicle based on the position of the vehicle recognized in the captured image at a time point before the present and the position of the vehicle recognized in the current captured image;
An image processing apparatus comprising: a determining unit configured to determine a traveling direction of the vehicle in the predetermined lane based on a traveling direction of the vehicles calculated by the calculating unit for a plurality of vehicles;
The recognition means is
After the determining means determines the traveling direction of the vehicle, the determined vehicle traveling direction is determined based on the position of the vehicle recognized in the captured image at a time earlier than the present in the predetermined lane in the current captured image. A setting unit for setting a predetermined area;
An image processing apparatus comprising: a recognition unit that recognizes the vehicle within the predetermined area of the current captured image.   The determining means determines that the traveling direction of the vehicle in the predetermined lane is the coincident traveling direction only when the traveling direction of the plurality of vehicles is equal to or greater than a first threshold. The image processing apparatus according to claim 1, wherein: It further comprises detection means for detecting traffic in the predetermined lane,
The determination means determines the traveling direction of the vehicle in the predetermined lane only when the traffic detected by the detection means within a first period is equal to or greater than a second threshold value. An image processing apparatus according to 1. The road is a road to which a central line shift system is applied,
The image processing apparatus according to claim 1, wherein the predetermined lane is a lane to be subjected to a center line shift. Comprising detection means for detecting traffic in the predetermined lane,
5. The determination unit according to claim 4, wherein the determination unit re-determines a traveling direction of the vehicle in the predetermined lane only when the traffic volume detected by the detection unit is less than a third threshold value in the second period. The image processing apparatus described.   6. The image processing according to claim 5, wherein the second period is a period required for the vehicle to travel from one end of the lane to be subjected to the center line shift to another end. apparatus. Acquisition means for acquiring a captured image obtained by imaging a vehicle traveling on a road having a plurality of lanes;
Recognizing means for recognizing a vehicle in a predetermined lane of the road in the acquired captured image,
The recognition means is
In the predetermined lane in the current captured image, a first predetermined area is set in one direction along the predetermined lane with reference to the position of the vehicle recognized in the captured image at a time point before the current time, A setting unit for setting a second predetermined area in another direction along the predetermined lane;
An image processing apparatus comprising: a recognition unit that recognizes the vehicle in the first and second predetermined areas of the current captured image. The road is a road to which a central line shift system is applied,
The image processing apparatus according to claim 7, wherein the predetermined lane is a lane to be subjected to a center line shift. First acquisition means for acquiring a captured image of a vehicle traveling on a road having a plurality of lanes;
Second acquisition means for acquiring information on the traveling direction of the vehicle in a predetermined lane of the road from the outside;
Recognizing means for recognizing a vehicle in the predetermined lane in the acquired captured image,
The recognition means is
In the predetermined lane in the current captured image, a setting unit that sets a predetermined region in the traveling direction of the vehicle with reference to the position of the vehicle recognized in the captured image at a time point prior to the present time;
An image processing apparatus comprising: a recognition unit that recognizes the vehicle within the predetermined area of the current captured image. The road is a road to which a central line shift system is applied,
The predetermined lane is a lane that is a target of a center line shift,
The image processing apparatus according to claim 9, wherein the second acquisition unit acquires information on a traveling direction of the vehicle from the center line shift system. An acquisition means for acquiring a captured image obtained by capturing a vehicle traveling on a road to which the center line shift system is applied, and an imaging time of the captured image;
Storage means for storing correspondence information associating a traveling direction of a vehicle with a time in a lane to be subjected to a change in the center line of the road;
Recognizing means for recognizing a vehicle in the lane in the acquired captured image,
The recognition means is
In the lane in the current captured image, the acquisition means is based on the captured time of the captured image before the current time and the correspondence information with reference to the position of the vehicle recognized in the captured image at the previous time point. A setting unit for setting a predetermined area in the traveling direction of the vehicle determined by
An image processing apparatus comprising: a recognition unit that recognizes the vehicle within the predetermined area of the current captured image. Imaging means for imaging a vehicle traveling on a road;
An image processing system comprising the image processing apparatus according to claim 1. An image processing apparatus according to any one of claims 1 to 6,
A traffic information providing system comprising a traffic information calculating device that calculates and provides traffic information,
The image processing apparatus includes means for outputting a vehicle recognition result in the predetermined lane recognized by the recognition means and a traveling direction of the vehicle in the predetermined lane determined by the determination means,
The traffic information calculation device
Means for acquiring the recognition result and the traveling direction of the vehicle;
A traffic information providing system comprising: means for calculating traffic information in the predetermined lane based on the obtained recognition result and the traveling direction of the vehicle. An acquisition procedure for acquiring a captured image of a vehicle traveling on a road having a plurality of lanes;
A recognition procedure for recognizing a vehicle in a predetermined lane of the road in the acquired captured image;
A calculation procedure for calculating the traveling direction of the vehicle based on the position of the vehicle recognized in the captured image at a time point before the current in the recognition procedure and the position of the vehicle recognized in the current captured image;
An image processing program comprising: a determination procedure for determining a traveling direction of the vehicle in the predetermined lane based on the traveling directions of the vehicles calculated in the calculation procedure;
The recognition procedure is:
After the traveling direction of the vehicle is determined in the determination procedure, in the predetermined lane in the current captured image, the position of the determined vehicle is determined with reference to the position of the vehicle recognized in the captured image at a time earlier than the current time. A procedure for setting a predetermined area in the traveling direction;
And a procedure for recognizing the vehicle in the predetermined area of the current captured image. An acquisition step of acquiring a captured image obtained by imaging a vehicle traveling on a road having a plurality of lanes;
A recognition step of recognizing a vehicle in a predetermined lane of the road in the acquired captured image;
A calculation step of calculating a traveling direction of the vehicle based on the position of the vehicle recognized in the captured image at a time point before the current in the recognition step and the position of the vehicle recognized in the current captured image;
A determination step of determining a traveling direction of the vehicle in the predetermined lane based on the traveling directions of the vehicles calculated in the calculating step,
The recognition step includes
After the direction of travel of the vehicle is determined in the determination step, in the predetermined lane in the current captured image, the position of the determined vehicle is determined based on the position of the vehicle recognized in the captured image at a time point prior to the current time. Setting a predetermined area in the direction of travel;
Recognizing the vehicle in the predetermined area of the current captured image.

Images