JP2012128561A - Abnormal vehicle detection device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an abnormal vehicle detection device that can perform abnormal vehicle detection processing on a road, etc. in consideration of a road condition, store images including the abnormal vehicle to efficiently collect evidence on abnormal traveling, and call attention to a vehicle traveling around the abnormal vehicle to reduce a risk of an accident.SOLUTION: The abnormal vehicle detection device obtains images of a road on which a vehicle travels by a camera, detects the vehicle traveling on the road from the obtained images, determines the road condition based on information of the detected vehicle, detects an abnormal vehicle on the road based on the determined road condition and positional information of the detected vehicle, and stores photographed images taken since the abnormal vehicle is detected until the abnormal vehicle is no longer detected, as a unit of images.

Description

  Embodiments described herein relate generally to an abnormal vehicle detection apparatus that detects an abnormally traveling vehicle on a road.

  On a toll road such as an expressway, an image on the road is captured with a camera installed on the roadside, and a vehicle traveling on the road is detected from the captured image by image processing or the like, and the position of the detected vehicle, A technique for monitoring current road conditions (road traffic jams, stopped vehicles, low-speed vehicles, vehicle escapes, etc.) and suspicious vehicles from information such as speed is known.

  However, the conventional known technique qualitatively and quantitatively determines abnormal driving of a vehicle based on image processing from an image obtained from a camera that captures on the road, thereby automatically detecting obstacles on the road. Although the present invention provides a road traffic abnormality detection method for detection, it does not take into account road conditions in abnormal vehicle detection, and therefore there is a possibility that erroneous detection may occur when the road is congested.

JP-A-6-274786 JP 2010-44502 A

  The problem to be solved by the present invention is that abnormal vehicle detection processing can be performed on a road or the like in consideration of road conditions, and by storing as an image including abnormal vehicles, evidence of abnormal driving can be efficiently obtained. It is possible to provide an abnormal vehicle detection device that can reduce the risk of an accident because it can be collected and can be alerted to a vehicle traveling around the abnormal vehicle.

  In order to solve the above-described problem, an abnormal vehicle detection device according to an embodiment is installed on a road on which a vehicle travels, an imaging unit that captures an image on the road, and an image captured by the imaging unit. Vehicle detection means for detecting a vehicle traveling on the road, road condition determination means for determining the road condition based on vehicle information detected by the vehicle detection means, and road condition determined by the road condition determination means And an abnormal vehicle detection means for detecting an abnormal vehicle on the road based on the position information of the vehicle detected by the vehicle detection means, and when the abnormal vehicle is detected by the abnormal vehicle detection means, the image obtained from the imaging means An abnormal vehicle image acquisition unit that acquires a captured image of the abnormal vehicle, and an abnormal vehicle detected by the abnormal vehicle detection unit are obtained from the image of the imaging unit. Serial detects that no longer detected by the vehicle detection means, and a abnormal vehicle image storage means for storing a captured image of the abnormal vehicle image abnormality vehicle acquired by the acquisition means as the image of one unit.

The schematic diagram which shows schematically the structure of the abnormal vehicle detection apparatus which concerns on embodiment. The block diagram which shows roughly the structure of the road condition grasping | ascertainment apparatus which concerns on embodiment. The figure explaining the process which registers the position of the white line on the road which concerns on embodiment on a white line table. The flowchart explaining the operation | movement which concerns on embodiment. The flowchart explaining the operation | movement which concerns on embodiment. The schematic explanatory drawing of the labeling in the road condition determination part which concerns on embodiment, and a gravity center point coordinate value calculation. The schematic diagram which expands and shows the specific example of the image labeled with respect to the detected vehicle which concerns on embodiment. The figure which illustrates the process which registers the driving | running | working lane etc. of the determined vehicle which concerns on embodiment to a lane table, and there is no lane change. The figure which shows the process which registers the driving lane etc. of the determined vehicle which concerns on embodiment to a lane table, and shows the case where there exists a lane change and a traveling vehicle is a high-speed vehicle. The figure explaining the process which registers the driving | running | working lane etc. of the determined vehicle which concern on embodiment to a lane table, and there exists a lane change and the figure which shows the case where a running vehicle is a low speed vehicle. The schematic diagram for demonstrating the determination result memory which concerns on embodiment. The flowchart explaining operation | movement of the principal part of the roadside processing apparatus which concerns on embodiment. The flowchart explaining operation | movement of the principal part of the display apparatus and roadside radio | wireless apparatus which concern on embodiment. The schematic diagram which shows the example of a display screen in the display apparatus and onboard equipment which concern on embodiment. The schematic diagram for demonstrating the example which preserve | saves the captured image of the acquired abnormal vehicle which concerns on embodiment.

Hereinafter, embodiments will be described with reference to the drawings.
FIG. 1 schematically illustrates the configuration of an abnormal vehicle detection device according to an embodiment. In FIG. 1, it is assumed that a road 1 is a toll road (hereinafter simply referred to as a road) such as an expressway, and a vehicle 2 travels in the direction indicated by the arrow. The road 1 is composed of two lanes 3a and 3b in this example, and a white line 4 is provided as a dividing line that separates the two lanes 3a and 3b. In addition, the white line 4 shall be located in the center part with respect to the width direction of the road 1 in this example.

On the side of the road 1, a video camera (hereinafter simply referred to as a camera) 5 as an image pickup means is installed. The camera 5 captures images in the detection area set on the road 1 at regular intervals (for example, every 33 ms) and outputs them as frame images. The output frame images are sequentially road condition grasping device 6. Sent to.
In this example, the camera 5 captures an image from diagonally forward of the traveling vehicle 2. However, the camera 5 may capture an image from diagonally behind the vehicle 2 or from above the vehicle 2.

  The road situation grasping device 6 performs predetermined image processing on the image taken by the camera 5 to detect the vehicle 2 traveling on the road 1 and its position, and based on the detection result (vehicle information), the road situation. In addition, the abnormal vehicle is determined and detected, and the determination and detection result are sent to the roadside processing device 7.

  The roadside processing device 7 sends the road condition determined by the road condition grasping device 6 to the display device 8 and the roadside wireless device 9. The display device 8 visually displays various road conditions and the like to the driver of the vehicle 2 to notify (notification means), and is parked on a predetermined position on the road 1 or a service area (toll road such as an expressway). In addition, it is installed as a monitor for displaying information in a section having facilities such as meals and refueling.

The roadside wireless device 9 includes an antenna 10 for performing communication with an external wireless device, and wireless communication using narrow area wireless communication (DSRC) between the vehicle-mounted devices 11 mounted on the vehicle 2. Is used to transmit and receive data, and various road conditions and the like are transmitted to the vehicle-mounted device 11 for notification (notification means), or vehicle-mounted device information transmitted from the vehicle-mounted device 11 is acquired. . The antenna 10 is installed at a predetermined position on the road 1, for example.
An image database 12 is connected to the roadside processing device 7 as abnormal vehicle image storage means for storing a detected image of the abnormal vehicle. As will be described in detail later, the image database 12 stores photographed images from when the detected abnormal vehicle is detected by the vehicle detection means from the image of the camera 5 until no longer detected as one unit of image.

  The roadside processing device 7 and the image database 12 are connected to a host device (host computer) 14 via communication lines 13a and 13b.

  For example, as shown in FIG. 2, the road condition grasping device 6 performs a predetermined image processing on an image captured by the camera 5 to detect a vehicle 2 traveling on the road 1 and a vehicle detection unit 21 that detects the position of the vehicle 2. The road condition determination unit 22 that determines the condition of the road 1 based on the vehicle information detected by the vehicle detection unit 21, the road condition determined by the road condition determination unit 22, and the position of the vehicle 2 detected by the vehicle detection unit 21 It is comprised by the abnormal vehicle determination part 23 which detects the abnormal vehicle in the road 1 based on information.

  Here, the road conditions determined by the road condition determination unit 22 are, for example, four states of “road traffic jam”, “stop vehicle”, “low speed vehicle”, and “normal state”. “Road congestion” means that road 1 is congested with vehicle 2, “stop vehicle” means that vehicle 2 is stopped on road 1, and “low-speed vehicle” means that road 1 is in advance on road 1. It is assumed that the vehicle 2 traveling at a speed equal to or less than a predetermined value is present. In addition, when the road condition is not the above two states, the “normal state” is set.

Hereinafter, the operation in such a configuration will be described.
First, the position of the white line 4 on the road 1 is registered in the white line table (memory) T1 provided in the abnormal vehicle determination unit 23. That is, for example, an image in the detection area set on the road 1 by the camera 5 (preferably an image in which the vehicle 2 does not exist) is captured, and the captured image is displayed on the display 16 as shown in FIG. Thus, the coordinate value indicating the position of the white line 4 on the displayed image, that is, the x coordinate value for the y coordinate value of the white line 4 is registered in the white line table T1 (see FIG. 3B).

  A specific example will be described. An image obtained from the camera 5 is displayed on a monitor (not shown), and a movable cursor in the image is moved with an operable mouse (not shown). If the size of the pixel is 320 (x direction) x 240 (y direction) pixels, the cursor is placed on the white line 4 at the position where the height (coordinate in the y direction) is “0” and the mouse button is clicked. The x coordinate value is acquired (if there is a section where the white line 4 does not exist, optimization is performed by performing interpolation later). By repeating this pixel by pixel, the x coordinate value until the height finally reaches the position of “239” is acquired. The x coordinate value for the y coordinate value thus obtained is registered in the white line table T1 as shown in FIG.

  In the above example, the case where the position of the white line 4 is registered in the white line table T1 by capturing an image in the detection area in advance has been described. However, instead of the position of the white line 4, for example, detection detected by the camera 5 Since the distance in the width direction of the road 1 in the area and the position of the substantially central portion with respect to the distance in the width direction are known in advance, the coordinate value indicating the position of the central portion, that is, x with respect to the y coordinate value of the central portion position. The coordinate value may be registered in the white line table T1. This method of registering 1 of road 1 is effective when, for example, road 1 is not a two-lane road but a single lane, but it is of course applicable to cases where two or more lanes are used.

Next, the operation will be described with reference to the flowcharts shown in FIGS.
The camera 5 always captures images in the detection area set on the road 1 at predetermined intervals, and sequentially sends the captured frame images to the vehicle detection unit 21 of the road condition grasping device 6 (step S1). ). The vehicle detection unit 21 detects the vehicle 2 traveling on the road 1 by performing predetermined image processing for each frame image photographed by the camera 5 (step S2).

  Specifically, for example, first, a background image in which the vehicle 2 does not exist is previously stored in a memory (not shown) as initial data, and the coordinate value of the image is converted into an absolute distance (actual distance) from the camera 5. The conversion table is also stored in the memory.

  In this state, when an image captured by the camera 5 is input, a difference image of brightness is generated by obtaining a difference in brightness value of each pixel between the input image and a background image previously stored in the memory. To do. Next, the vehicle region (vehicle 2) is detected by performing binarization processing on the generated difference image. Next, the detected position of the vehicle 2 (region), the vehicle area, and the traveling speed are obtained by converting the detected coordinate value of the vehicle region into an absolute distance using a conversion table previously stored in a memory. The details of the vehicle detection process are described in, for example, Japanese Patent Application No. 2008-207034, so please refer to it.

  Next, the road condition determination unit 22 determines at least four conditions of the road 1 based on the vehicle information detected by the vehicle detection unit 21: road congestion, stopped vehicle, low-speed vehicle, and normal state (step S3).

  More specifically, for example, it is determined whether or not the road 1 is congested based on the position, vehicle area, and traveling speed of the vehicle 2 acquired by the vehicle detection unit 21 and acquired by the vehicle detection unit 21. The stop vehicle and the low-speed vehicle are determined based on the travel speed. Note that details of the road condition determination processing are described in, for example, Japanese Patent Application No. 2008-207034.

  In step S3, when it is determined that the road condition is “road traffic jam” (step S4), or when it is determined that the vehicle is “stopped vehicle” (step S5), the process returns to step S1 to the next frame image. Repeat the processing operation.

  If it is determined in step S3 that the road condition is not “road traffic jam” and is not “stopped vehicle” (steps S4 and S5), the abnormal vehicle determination unit 23 determines an abnormal vehicle.

  That is, first, the vehicle 2 detected in step S2 is labeled, and the barycentric point coordinate value (Xg, Yg) for each label is obtained (step S6). FIG. 6 shows a schematic explanatory diagram of labeling and centroid point coordinate value calculation. 6A shows the input image, FIG. 6B shows the labeling with respect to the vehicle 2 detected from the input image, and FIG. 6C shows the obtained center-of-gravity point coordinate values (Xg, Yg). FIG. 6D shows lane determination (performed in step S8 described later) based on the obtained barycentric point coordinate value.

Note that the existing calculation method is used for the calculation of the barycentric point coordinate value in step S6. There are various centroid point coordinate value calculation methods, but the following formula 1 is generally used.

The above formula represents the average value of the x-coordinate values and the average value of the y-coordinate values of each label.

FIG. 7 is an enlarged view of a specific example of an image labeled on the detected vehicle 2, and this example shows a case where two vehicles 2 (label 1 and label 2) are detected. When the centroid point coordinate value is calculated with respect to this image using the above equation 1, and the value after the decimal point of the calculation result is rounded off, the centroid point coordinate value of label 1 becomes the following equation 2, and the label The center-of-gravity point coordinate value of 2 is as shown in Equation 3 below.

  Next, based on the y-coordinate value (Yg) of the barycentric point obtained in step S6, the x-coordinate value of the white line 4 in the y-coordinate value (Yg) is referred to by referring to the white line table T1 in FIG. (Xw) is obtained (step S7).

  Next, by comparing the x-coordinate value (Xg) of the barycentric point obtained in step S6 and the x-coordinate value (Xw) of the white line 4 obtained in step S7, the traveling lane on which the vehicle 2 is traveling is determined. (Step S8), the determination result is registered in a lane table (memory) T2 provided in the abnormal vehicle determination unit 23 (Step S9).

  That is, for example, the x coordinate value (Xg) of the barycentric point is compared with the x coordinate value (Xw) of the white line 4, and the x coordinate value (Xg) of the barycentric point is larger than the x coordinate value (Xw) of the white line 4. For example, the travel lane is determined to be the lane 3a (lane 1), and if the travel lane is small, the travel lane is determined to be the lane 3b (lane 2), and the determination result is registered in the lane table T2. At the same time, the road condition at this time (determination result in step S3) is also registered in the lane table T2.

  Here, specific examples of the lane table T2 are shown in FIG. 8, FIG. 9, and FIG. FIG. 8A, FIG. 9A, and FIG. 10A show the movement trajectory of the x-coordinate value (Xg) of the barycentric point obtained for each frame image, and FIG. 8B and FIG. (B) and FIG.10 (b) have shown the lane table T2 corresponding to this. 8 shows the case where there is no lane change, FIG. 9 shows the case where there is a lane change and the traveling vehicle is a high-speed vehicle, and FIG. 10 shows the case where there is a lane change and the traveling vehicle is a low-speed vehicle. .

  Next, in the lane table T2, the travel lane of the current frame image is compared with the travel lane of the image one frame before (step S10), and if the two are equal, it is determined that the vehicle 2 has no lane change (step S10). S11), returning to step S1, the processing operation for the next frame image is repeated, and if they are different, it is determined that the vehicle 2 has changed lanes (step S11), and the number of lane changes in the lane table T2 is incremented. (Step S12).

  Next, in the lane table T2, it is determined whether or not the number of lane changes exceeds a predetermined value (which can be arbitrarily set) (step S13). Is stored in the determination result memory (storage means) M1 provided in (1), “0” indicating that the current frame abnormal vehicle is not detected (step S14), and the process returns to step S1 to repeat the processing operation for the next frame image.

  In step S13, when the number of lane changes exceeds the specified value, it is determined as an abnormal vehicle (suspicious vehicle) (step S15). That is, when the number of lane changes exceeds a specified value, for example, by referring to the road situation of the frame image determined to be an abnormal vehicle in the lane table T2, whether or not the road situation is “low speed vehicle” is determined. In the case of “low-speed vehicle” (for example, in the case of FIG. 10, “low-speed” is shown in the “road condition” column of the lane table T2 in FIG. 10B), If it is determined that the vehicle is not a “low-speed vehicle” (normal state, for example, in the case of FIG. 9, “normal” is indicated in the “road condition” column of the lane table T2 in FIG. 9B)) Is determined to be a runaway vehicle that repeats in a short time.

  If it is determined in step S15 that the vehicle is abnormal, “1” is stored in the determination result memory M1 provided in the abnormal vehicle determination unit 23 as “current frame abnormal vehicle detected” (step S16). FIG. 11 shows an example of the determination result memory M1. As shown in the figure, “0” is stored as “abnormal vehicle is not detected” and “1” is stored as “abnormal vehicle detected” for each frame image.

  If it is determined in step S15 that the vehicle is abnormal, this is used as a trigger (trigger) to start saving an image of the abnormal vehicle captured by the camera 5 in the image database 12 (step S17). The storage of the captured image of the abnormal vehicle is performed via the roadside processing device 7 and is performed for each frame image obtained from the camera 5 until the abnormal vehicle does not exist on the image, that is, the abnormal vehicle. This is performed until the vehicle is detected from the image of the camera 5 by the vehicle detection unit 21.

  Further, when it is determined that the vehicle is abnormal in step S15, an abnormal vehicle detection signal is sent to the roadside processing device 7 which is a host device of the road condition grasping device 6 (step S18). The roadside processing device 7 performs processing according to the flowchart shown in FIG. That is, when an abnormal vehicle detection signal is received (step S31), an abnormal vehicle detection signal indicating that an abnormal vehicle exists is sent to the display device 8 and the roadside wireless device 9 (step S32), and to the host device 14. Then, the presence of an abnormal vehicle is notified by an alarm (step S33).

  Upon receiving the abnormal vehicle detection signal, the display device 8 and the roadside wireless device 9 process the vehicle according to the flowchart shown in FIG. 13, and the vehicle traveling around the abnormal vehicle (in the example according to the embodiment of FIG. 1, the vehicle 2 is If the vehicle is an abnormal vehicle (for example, a runaway vehicle), a message that warns the abnormal vehicle is notified to the vehicle 2a traveling ahead.

  In other words, when the abnormal vehicle detection signal is received (step S41), the display device 8 displays, for example, an abnormal vehicle for a vehicle 2a traveling in front of the abnormal vehicle (for example, a runaway vehicle) as shown in FIG. A message notifying that there is an abnormal vehicle together with the image 31, for example, a message 32 “Now approaching suspicious vehicle!” And a message 33 “Please pay attention to the rear” are displayed. Similar images 31 and messages 32 and 33 are displayed on the display unit 11a of the vehicle-mounted device 11 mounted on the vehicle 2a traveling in front of the abnormal vehicle 2 (step S42).

  In addition, a message notifying that there is an abnormal vehicle is displayed together with the abnormal vehicle image 31 to the vehicle 2 traveling behind the abnormal vehicle (for example, a suspicious behavior vehicle). A similar image 31 and message are displayed on the display unit 11a of the vehicle-mounted device 11 mounted on the vehicle 2 traveling around the abnormal vehicle 2 (step S42).

  Next, the abnormal vehicle determination unit 23 refers to the determination result memory M1 so that the previous frame image immediately before the current (current) frame image is abnormal vehicle detected (1) and the current frame image. Determines whether there is no abnormal vehicle detected (0) (step S19), otherwise, that is, the previous frame image has an abnormal vehicle detected (1), and the current frame image also has an abnormal vehicle detected (1) If so, it is determined that the abnormal vehicle is still detected by the vehicle detection means from the image of the camera 5, and the process returns to step S1 to repeat the process for the next frame image.

  In step S19, if the previous frame image has an abnormal vehicle detection (1) and the current frame image has no abnormal vehicle detection (0), the abnormal vehicle is detected from the image of the camera 5 by the vehicle detection means. It is determined that the frame has disappeared, and the frame image of the abnormal vehicle stored in the image database 12 is converted into a single moving image (image) and stored in the image database 12 (step S20).

  That is, for example, as shown in FIG. 15, a plurality of frame images F <b> 1 to Fn from when the abnormal vehicle stored in the image database 12 is detected by the vehicle detection unit 21 from the image of the camera 5 are detected. For example, it is converted into a moving image 34 of one scene and stored in the image database 12. In that case, date information or the like is used as the file name 35 for easy access later.

  When the process of step S20 ends, the process returns to step S1 to repeat the process for the next frame image.

  Further, in the host device 14 that has received an alarm from the roadside processing device 7, for example, as shown in FIG. 15, a worker or an attendant accesses the image database 12 by using the file name 35 as a key (a clue to search). The stored abnormal vehicle moving image 34 is displayed and confirmed visually. After confirmation, contact the police if necessary.

  As described above, according to the above-described embodiment, the abnormal vehicle detection process is performed in consideration of the road conditions, so that it is possible to distinguish and detect the behavior suspicious vehicle and the runaway vehicle. In addition, since the abnormal vehicle detection process is not performed when the road condition is traffic jam, the meandering operation for avoiding traffic jam is not erroneously detected.

Furthermore, since it can preserve | save as an image containing an abnormal vehicle, the evidence of abnormal driving | running | working can be collected efficiently.
Moreover, the risk of an accident can be reduced by urging attention to a vehicle traveling around an abnormal vehicle.

  Although several embodiments of the present invention have been described, these embodiments are presented by way of example and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the scope of the invention. These embodiments and their modifications are included in the scope and gist of the invention, and are also included in the invention described in the claims and the equivalents thereof.

  DESCRIPTION OF SYMBOLS 1 ... Road, 2, 2a ... Vehicle, 3a, 3b ... Lane, 4 ... White line (separation line), 5 ... Video camera (imaging means), 6 ... Road condition grasping device, 7 ... Roadside processing device, 8 ... Display device , 9: Roadside wireless device, 10: Antenna, 11: On-board unit, 11a ... Display unit, 12 ... Image database, 13a, 13b ... Communication line, 14 ... Host device (host computer), 16 ... Display, 21 ... Vehicle detection , 22 ... Road condition determination unit, 23 ... Abnormal vehicle determination unit, 31 ... Abnormal vehicle image, 32 ... Message, 33 ... Message, 34 ... Movie, 35 ... File name, F1, Fn ... Frame image, T1 ... White line Table, T2 ... Lane table, M1 ... Determination result memory.

Claims (4)

An imaging means that is installed on a road on which the vehicle travels and captures an image on the road;
Vehicle detection means for detecting a vehicle traveling on the road from an image captured by the imaging means;
Road condition determination means for determining the condition of the road based on the vehicle information detected by the vehicle detection means;
Abnormal vehicle detection means for detecting an abnormal vehicle on the road based on the road status determined by the road status determination means and the position information of the vehicle detected by the vehicle detection means;
When an abnormal vehicle is detected by the abnormal vehicle detection unit, an abnormal vehicle image acquisition unit that acquires a captured image of the abnormal vehicle obtained from the imaging unit;
It is detected that the abnormal vehicle detected by the abnormal vehicle detection unit is no longer detected by the vehicle detection unit from the image of the imaging unit, and one unit is a captured image of the abnormal vehicle acquired by the abnormal vehicle image acquisition unit. Abnormal vehicle image storage means for storing as an image of,
An abnormal vehicle detection device comprising: An imaging means that is installed on a road on which the vehicle travels and captures an image on the road;
Vehicle detection means for detecting a vehicle traveling on the road from an image captured by the imaging means;
Based on the vehicle information detected by the vehicle detection means, at least the road congestion where the road is congested with the vehicle and the stopped vehicle where there is a vehicle stopped on the road are determined as the road situation. Road condition judging means to
When the road condition determining means determines that the road condition is not a road jam or a stopped vehicle, the vehicle detected by the vehicle detecting means is labeled, and the barycentric point coordinate value for each labeled vehicle A calculation means for obtaining
Based on the center-of-gravity point coordinate value obtained by this calculation means, the center-of-gravity point coordinate is obtained by referring to the table data in which the road center part coordinate value indicating the position of the substantially center part in the width direction of the road is registered in advance. Road center position detecting means for obtaining a road center coordinate value in the value;
Abnormality for detecting whether or not the vehicle is an abnormal vehicle that repeats meandering operation from the relationship between the road center portion coordinate value obtained by the road center portion position detection means and the barycentric point coordinate value obtained by the calculation means. Vehicle detection means;
When an abnormal vehicle is detected by the abnormal vehicle detection unit, an abnormal vehicle image acquisition unit that acquires a captured image of the abnormal vehicle obtained from the imaging unit;
It is detected that the abnormal vehicle detected by the abnormal vehicle detection unit is no longer detected by the vehicle detection unit from the image of the imaging unit, and one unit is a captured image of the abnormal vehicle acquired by the abnormal vehicle image acquisition unit. Abnormal vehicle image storage means for storing as an image of,
An abnormal vehicle detection device comprising: An imaging unit that is installed on a road composed of at least two lanes on which the vehicle travels, and that captures an image on the road;
Vehicle detection means for detecting a vehicle traveling on the road from an image captured by the imaging means;
Based on the vehicle information detected by the vehicle detection means, at least the road congestion where the road is congested with the vehicle and the stopped vehicle where there is a vehicle stopped on the road are determined as the road situation. Road condition judging means to
When the road condition determining means determines that the road condition is not a road jam or a stopped vehicle, the vehicle detected by the vehicle detecting means is labeled, and the barycentric point coordinate value for each labeled vehicle A calculation means for obtaining
Based on the center-of-gravity point coordinate value obtained by the calculation means, the center of gravity is obtained by referring to the table data in which the delimiter line position coordinate value indicating the position of the demarcation line that delimits the two lanes of the road is registered in advance. Break line position detection means for obtaining a break line position coordinate value in a point coordinate value;
From the relationship between the position coordinate value of the dividing line obtained by the dividing line position detecting means and the barycentric point coordinate value obtained by the calculating means, it is detected whether or not the vehicle is an abnormal vehicle that repeatedly changes lanes. Abnormal vehicle detection means;
When an abnormal vehicle is detected by the abnormal vehicle detection unit, an abnormal vehicle image acquisition unit that acquires a captured image of the abnormal vehicle obtained from the imaging unit;
It is detected that the abnormal vehicle detected by the abnormal vehicle detection unit is no longer detected by the vehicle detection unit from the image of the imaging unit, and one unit is a captured image of the abnormal vehicle acquired by the abnormal vehicle image acquisition unit. Abnormal vehicle image storage means for storing as an image of,
An abnormal vehicle detection device comprising:   2. The apparatus according to claim 1, further comprising: an informing unit configured to notify the vehicle traveling forward and rearward of the abnormal vehicle when a abnormal vehicle is detected by the abnormal vehicle detection unit. The abnormal vehicle detection apparatus in any one of -3.

Images