Vehicle detection device

Abstract

To provide a vehicle detection device for rapidly and reliably detecting a dangerous traveling vehicle even on a road where a safe area is divided by an installation object.SOLUTION: A vehicle detection device 100 comprises: a photographing device 10 for acquiring an image of a monitoring area DA including an installation object 90 which divides a safe area SA; and a determination device 33 for performing first determination of dangerous traveling by a first determination reference based on the safe area SA which is regulated by the installation object 90 in the image, and performing second determination of the dangerous traveling by a second determination reference based on a safe traveling area TA which is set by learning a traveling track RP of a vehicle CA in the image. The determination device 33 performs the first determination before the end of the learning.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to a vehicle detection device that detects dangerously moving vehicles on a road where safety areas are separated by installed objects. [Background technology]

[0002] A known vehicle accident prediction device learns the normal traveling speed, normal traveling direction, and normal traveling area of ​​a vehicle in a monitored area as normal behavior, and detects a vehicle exhibiting behavior that deviates from the learned normal behavior as a dangerous vehicle (Patent Document 1). The device in Patent Document 1 requires behavior data for a certain number of vehicles to learn the normal behavior of vehicles, and has the problem that it cannot be put into operation immediately after installation.

[0003] A known warning device divides a captured image into a safe area and a non-safe area using restrictive members such as cones placed on the road, and detects vehicles moving recklessly from changes in the image in the safe area (Patent Document 2). The device in Patent Document 2 sets the safe area and the non-safe area using lines connecting the restrictive members, but there are cases where the areas cannot be set accurately due to the restrictive members falling over or the distant restrictive members not being detected. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2014-164524 [Patent Document 2] Japanese Patent Application Laid-Open No. 2012-128574 Summary of the Invention

[0005] The present invention has been made in consideration of the above points, and aims to provide a vehicle detection device that can quickly and reliably detect dangerously moving vehicles even on roads where safety areas are separated by installed objects.

[0006] In order to achieve the above-mentioned object, the vehicle detection device of the present invention comprises an imaging device that acquires an image of a monitoring area including installations that demarcate a safe area, and a judgment device that makes a first judgment of dangerous driving based on a first judgment criterion based on the safe area defined by the installations in the image, and makes a second judgment of dangerous driving based on a second judgment criterion based on the safe driving area that is set by learning the vehicle's driving trajectory in the image, and the judgment device makes the first judgment before learning is completed.

[0007] In the vehicle detection device, the first judgment is made based on the first judgment criterion, which can be set relatively quickly, before the judgment device finishes learning about the vehicle's driving trajectory, so that the vehicle detection device can be used immediately after installation without waiting for the learning time. [Brief explanation of the drawings]

[0008] [Figure 1] 1 is a block diagram illustrating a vehicle detection device according to a first embodiment. [Figure 2] FIG. 2 is a bird's-eye view illustrating an installation state of a vehicle detection device. [Figure 3] FIG. 2 is a side view illustrating an installation state of the vehicle detection device. [Figure 4] FIG. 2 is a conceptual diagram illustrating a captured image. [Figure 5] 10(A) and 10(B) are conceptual diagrams illustrating the setting of the first criterion. [Figure 6] 10(A) and 10(B) are conceptual diagrams illustrating the setting of the second criterion. [Figure 7] 4 is a flowchart illustrating the operation of the vehicle detection device of the first embodiment. [Figure 8] 10 is a flowchart illustrating the operation of the vehicle detection device according to the second embodiment. [Figure 9] 10 is a flowchart illustrating the operation of the vehicle detection device according to the third embodiment. [Figure 10] FIG. 10 is a diagram illustrating a change in a safety area. DETAILED DESCRIPTION OF THE INVENTION

[0009] [First embodiment] A first embodiment of a vehicle detection device according to the present invention will be described below with reference to the drawings.

[0010] FIG. 1 is a block diagram showing an example of the configuration of a vehicle detection device 100. FIG. 2 is a conceptual overhead view showing the installation state of the vehicle detection device 100 shown in FIG. 1. FIG. 3 is a conceptual side view showing the installation state of the vehicle detection device 100 shown in FIG. 1. FIG. 4 is a conceptual diagram explaining a captured image IM. As shown in FIG. 2 and other figures, a road RO is a two-lane road, and a maintenance work site MW is surrounded by multiple installations 90 to ensure a safety area SA in the outer lane. The installations 90 separate the safety area SA from a non-safety area NSA (see FIG. 5(B)). The safety area SA is an area into which vehicles CA traveling on the road RO normally do not enter. The non-safety area NSA is an area of ​​the road RO where vehicles CA other than the safety area SA travel. Examples of the installations 90 include cones (pylons), fences, barricades, etc.

[0011] As shown in FIG. 1 and other figures, the vehicle detection device 100 includes an imaging device 10 and an information processing device 20.

[0012] The imaging device 10 is a camera that captures an image of a road RO (see FIG. 2, etc.) on which a vehicle CA is traveling. The imaging device 10 defines a target space including the target road RO, the area above it, and its surroundings as an imaging range PA (see FIG. 2). The imaging range PA corresponds to a monitoring area DA. The imaging device 10 acquires images of the monitoring area DA, including an installed object 90. The imaging device 10 continuously captures an object at a predetermined time interval (for example, about 30 fps). Captured images IM (see FIG. 4, etc.), including road images captured by the imaging device 10, are output to the information processing device 20. The frame rate of continuous imaging by the imaging device 10 may be, for example, 5 fps or 10 fps, or may be about 100 fps or more.

[0013] The information processing device 20 is a computer, and includes a control device 21, a storage device 22, a user interface 23, a notification device 24, and a communication device 25. The information processing device 20 detects a recklessly driving vehicle that may enter a safety area SA using the captured image IM captured by the imaging device 10. As will be described in detail later, the information processing device 20 performs two types of setting processing in advance, and a detection processing that detects vehicle information and determines whether the vehicle is driving recklessly based on this vehicle information.

[0014] The control device 21 operates based on a program stored in the storage device 22. The storage device 22 stores a basic program for operating the information processing device 20 and application software that runs on the basic program. The application software includes software for operating the imaging device 10 to acquire captured images IM, software for executing a pre-setting process, and software for executing a detection process for detecting vehicle information from the captured images IM. The user interface 23 includes a display, a touch panel, a speaker, a microphone, and the like, and receives instructions from an operator and presents the processing results of the information processing device 20 to the operator. The information processing device 20 notifies the worker US (see FIG. 2 ) or the like within the safety area SA of the result of the dangerous driving determination via the notification device 24. The information processing device 20 can also communicate with a mobile terminal 50 carried by the worker US or an external management server (not shown) via a communication network via the communication device 25. The mobile terminal 50 functions in conjunction with or in place of the notification device 24. The notification device 24 notifies the operator of dangerous driving. Examples of the notification device 24 include a display device, a speaker, and a warning light.

[0015] The control device 21 includes an image processing device 31 , a learning device 32 , and a determination device 33 .

[0016] The image processing device 31 performs image processing on the captured image IM acquired by the imaging device 10. The image processing device 31 is also a part for setting the first judgment criterion in the setting processing stage. The first judgment criterion is based on a safety area SA (see FIG. 5(B)) defined by an installed object 90, and is set relatively early in the initial stage of the start of operation of the vehicle detection device 100. In other words, the safety area SA serves as the first judgment criterion for determining whether or not the vehicle CA is driving dangerously. In the first setting processing stage, the image processing device 31 detects the installed object 90 from the captured image IM and determines the safe area SA and the non-safe area NSA (see FIG. 5(B)).

[0017] The learning device 32 is a part that executes setting of the second judgment criterion in the setting processing stage. The second judgment criterion is based on a safe driving area TA (see FIG. 6(B)) that is set by learning the driving trajectory RP (see FIG. 6(A)) of the vehicle CA, and is set after the vehicle detection device 100 starts operating and after a predetermined condition is met or a predetermined time has passed. In other words, the safe driving area TA serves as a second judgment criterion for determining whether the vehicle CA is driving dangerously. In the second setting processing stage, the learning device 32 detects vehicle information from the captured image IM, learns the driving trajectory RP of the vehicle CA, and determines the safe driving area TA.

[0018] The determination device 33 is a part for executing the detection processing stage. In the detection processing stage, the determination device 33 detects vehicle information from the captured image IM and determines whether or not the vehicle is driving recklessly based on the vehicle information. The determination device 33 determines whether the vehicle is driving recklessly based on two determination criteria. In this embodiment, the determination device 33 performs a first determination based on the first determination criterion until learning is completed, i.e., until the second determination criterion is set, and after learning is completed, the determination criterion is switched to the second determination criterion and a second determination based on the second determination criterion. By switching the determination criterion before and after learning is completed, the determination device 33 can perform highly accurate determinations.

[0019] 2 and 3, the vehicle detection device 100 is mounted on a sign vehicle V1 located adjacent to a maintenance work site MW, where construction work is being carried out or fallen objects are being collected on a road RO, such as a highway. The vehicle equipped with the vehicle detection device 100 is not limited to the sign vehicle V1, but may be other vehicles such as an emergency vehicle or a patrol car. The vehicle detection device 100 includes an imaging device 10 fixed on a sign 3 fixed on a loading platform 2a, an information processing device main body 20a fixed on the loading platform 2a, and a user interface 23 installed in the driver's seat 2b. The information processing device main body 20a includes, for example, a control device 21, a storage device 22, an alarm device 24, and a communication device 25. The maintenance work site MW is located behind the sign vehicle V1. A human-shaped warning sign 4 is located upstream of the sign vehicle V1 in the vehicle's traveling direction, and supports a rotating warning light 4a. The sign vehicle V1, the maintenance work site MW, and the warning sign 4 are surrounded by multiple installed objects 90. The inner area surrounded by the installation 90 is a safety area SA to ensure the safety of workers US and others. Outside the safety area SA, there is a vehicle CA traveling on the road RO, and signs 3 and installations 90 urge the vehicle to slow down and change lanes. The warning sign 4 does not need to be a person-shaped sign, and the sign of the sign vehicle V1 may be used, or it may be replaced with something else. The warning sign 4 may also be omitted. The warning sign 4 may be an alarm device 24. In this case, the warning sign 4 can receive a danger signal notifying the detection of a dangerously traveling vehicle via the communication device 25, and can cause the warning light 4a to light up to warn of the danger.

[0020] The imaging device 10 captures an image of a target area including the road RO behind the sign vehicle V1 and the upstream direction of vehicle travel. The imaging range PA of the imaging device 10 includes the area outside the road RO. The captured image IM (see FIG. 4, etc.) including the road image output from the imaging device 10 is processed by the information processing device main body 20a, and the processing result is output to the user interface 23.

[0021] Returning to FIG. 1 , in the information processing device 20, the image processing device 31, under the supervision of an operator, sets a first judgment criterion as a first setting processing stage. Based on the operator's instructions, the control device 21 of the information processing device 20 adjusts the attitude of the imaging device 10 and sets a shooting range PA including the road RO. The control device 21 operates the imaging device 10 to capture an image of the road RO ahead of the imaging device 10 and the area above it as a preliminary image P0 (see FIG. 4, etc.). The preliminary image P0 is a captured image IM captured initially after the vehicle detection device 100 starts operating. The control device 21 operates the image processing device 31 to extract an image of an installed object 90 from the preliminary image P0. A known pattern recognition technique can be used to extract the installed object 90. As shown in FIG. 5(A), the image processing device 31 connects a line L1 based on, for example, a vertex of the installed object 90, and as shown in FIG. 5(B), sets the area surrounded by the edge of the road RO and the line L1 as a safety area SA. Furthermore, the image processing device 31 sets the area of ​​the road RO excluding the safety area SA as a non-safe area NSA. The safety area SA and the non-safe area NSA are recorded in the storage device 22 as indicators of the first determination criterion.

[0022] As described above, the image processing device 31 automatically determines the first determination criterion for determining reckless driving depending on the position of the vehicle CA relative to the safety area SA.

[0023] In a second setting process or detection process described below, the control device 21 operates the imaging device 10 to capture a captured image IM with the same angle of view as the preliminary image P0. The control device 21 operates the image processing device 31 to extract one or more images of vehicles CA from the captured image IM. Known pattern recognition techniques can be used to extract the vehicles CA. The image processing device 31 incorporates pattern recognition software trained by machine learning, such as deep learning, as a trained model for vehicle extraction, thereby determining the image area of ​​the vehicle CA from the captured image IM and calculating the front area (pixel position) of the vehicle CA or its center coordinates. In this embodiment, the image processing device 31 sets the bottom center of a rectangular frame enclosing the front area of ​​the vehicle CA as the reference point BP for the position of the vehicle CA (see FIG. 6(A) and other figures). Any type of pattern recognition software can be used as long as it can recognize the vehicle CA. However, distinguishing attributes such as the vehicle model and color makes it easier to check the identity of the extracted vehicle images. A specific example of pattern recognition software that can be used is one that uses, for example, HOG (Histograms of Oriented Gradients) to extract features and a support vector machine as a learning model for identifying the features. The method of extracting vehicle images is not limited to the machine learning method described above, and may also be one that extracts artificially set local features using a statistical learning method, such as optical flow or template matching.

[0024] The image processing device 31 receives captured images IM as successive frames from the imaging device 10 and sequentially extracts vehicle images from each captured image IM. The image processing device 31 determines the pixel position or screen coordinates of the vehicle image in each of the successive captured images IM over time. By tracking a specific vehicle image, the image processing device 31 can determine the movement vector (vehicle vector) of the vehicle CA as the positional deviation of the vehicle image between the captured images IM, and can obtain a traveling trajectory RP (see FIG. 6(A)) that is the movement trajectory of the vehicle CA. At night, the images from the imaging device 10 mainly include headlights. Even in such a case, the image processing device 31 can identify and track the vehicle CA using pattern recognition technology.

[0025] The image processing device 31 can set a rectangular speed measurement area along the road RO in the captured image IM, monitor the vehicle CA passing through this speed measurement area, and determine the speed of the vehicle CA from the time it passes through. The size of the speed measurement area is known. The speed measurement area is made possible by setting four reference points on the road RO in advance. Information on the world coordinates (latitude and longitude) of the reference points is acquired, and they can be made possible by identifying an object in the captured image IM, placing an actual indicator, or marking the captured image IM. Note that instead of the speed measurement area, a speed measurement start line and a speed measurement end line set to cross the road RO can also be used. Two or more speed measurement areas can be set in the captured image IM.

[0026] The learning device 32 sets the second judgment criterion through learning as a second setting processing stage. The learning device 32 accumulates information while learning from the captured images IM captured by the imaging device 10 and processed by the image processing device 31. The number of vehicles CA to be captured is, for example, 10 or more, but the number of vehicles captured in advance can be increased or decreased depending on the situation. At this time, the speed of the vehicle CA can be measured, and vehicles CA that are exceeding the speed limit can be excluded from the learning target.

[0027] The learning device 32 learns the driving trajectories RP obtained for multiple vehicles CA in the image processing device 31, and determines a safe driving area TA based on the learning results. The driving trajectories RP may be straight lines, curves, or approximate curves. The safe driving area TA is a strip-shaped region including multiple driving trajectories RP. The safe driving area TA is an area in which the vehicle CA maintains normal driving without entering the safety area SA. The safe driving area TA is recorded in the storage device 22 as an index of the second judgment criterion. In other words, the learning results in the learning device 32 are recorded as a learned model corresponding to the second judgment criterion.

[0028] The operation of the image processing device 31 in the second setting processing stage will be described with reference to FIG. 6(A). A road RO including two lanes is captured in the captured image IM, and the image processing device 31 detects a large number of driving trajectories RP acquired for a large number of vehicles CA. These driving trajectories RP are virtual, not images of real objects, and can be superimposed on the captured image IM on the display of the user interface 23. Note that the image used in determining the safe driving area TA (see FIG. 6(B)) may be the captured, uncompressed captured image IM, but is not limited to this. The determination of the safe driving area TA itself does not need to be highly accurate; using compressed images can make the pre-processing stage efficient and rapid. When compressed images are used in determining the safe driving area TA, they can be compressed as part of the image processing, or compressed images can be prepared separately.

[0029] The operation of the learning device 32 in the second setting processing stage will be described with reference to FIG. 6(B). The learning device 32 determines a safe driving area TA including the driving trajectory RP in the captured image IM, i.e., the shooting range PA (see FIG. 2). Specifically, the learning device 32 accumulates the driving trajectories RP of the number of vehicles CA previously captured through learning as information (learned data), and determines a band- or frame-shaped safe driving area TA that encompasses these. The safe driving area TA is virtual, but can be displayed superimposed on the captured image IM on the display of the user interface 23. The learning device 32 may further determine a no-driving area FA that excludes the safe driving area TA from the road RO. The safe driving area TA can also be expanded horizontally within the shooting range PA.

[0030] As described above, the learning device 32 automatically determines the second criterion for determining dangerous driving based on the position of the vehicle CA relative to the safe driving area TA. Note that the learning device 32 may automatically determine an upper speed limit for determining whether the speed of the vehicle CA of interest is abnormal, or may accept a specification from the operator, as an additional criterion to the second criterion.

[0031] Returning to FIG. 1, the determination device 33 determines whether or not the vehicle is driving dangerously based on the vehicle information obtained by the image processing device 31, specifically, the driving path RP and speed of the vehicle CA.

[0032] The determination device 33 makes a first determination of dangerous driving based on the first determination criterion after the vehicle detection device 100 starts operating and before learning is completed. In addition, the determination device 33 makes a second determination of dangerous driving based on the second determination criterion after learning is completed. This allows for more accurate determination without being affected by inaccuracies in the safety area SA.

[0033] In the first determination, the determination device 33 determines that the vehicle CA is driving dangerously when it enters the safety area SA. In other words, the determination of dangerous driving is made based on whether the vehicle CA is within the safety area SA. Note that the determination device 33 may also determine that the vehicle CA of interest is driving dangerously when it approaches the safety area SA by a predetermined distance or less. In this case, the "predetermined distance" is set automatically or manually based on pixel units.

[0034] In the second determination, the determination device 33 determines that the vehicle CA is driving unsafely if it is driving outside the safe driving area TA. That is, the determination of unsafe driving is made based on whether the vehicle CA is within the safe driving area TA. Specifically, the determination device 33 determines that the vehicle is driving unsafely if the center of the lower part (reference point BP) of the vehicle CA of interest goes beyond the safe driving area TA and goes outside it. Note that the determination device 33 may also determine that the vehicle is driving unsafely if the driving trajectory RP of the vehicle CA of interest goes outside the safe driving area TA. Furthermore, the determination device 33 may also determine that the vehicle is driving unsafely if the vehicle CA of interest is driving at a distance of more than a predetermined distance outside the safe driving area TA. The above-mentioned "distance of more than a predetermined distance" is based on pixel units. Furthermore, the "distance of more than a predetermined distance" is automatically determined from the width of the safe driving area TA when determining the safe driving area TA, but is not limited to this and may also be set in advance by an operator.

[0035] In addition, the judgment device 33 may also determine that the vehicle CA in question is driving dangerously if (1) the vehicle CA in question is driving in a large, zigzag manner even within the safe driving area TA, or (2) the vehicle CA in question exceeds a predetermined upper speed limit.

[0036] FIG. 7 is a flowchart illustrating the operation of the vehicle detection device 100 shown in FIG. 1 and other figures.

[0037] First, the control device 21 operates the imaging device 10 to capture an image of the monitoring area DA and obtain a captured image IM (step S11).

[0038] Next, the control device 21 checks whether the first judgment criterion has been set (step S12). If the first judgment criterion has not been set (N in step S12), for example, if the vehicle detection device 100 has just started operating, the control device 21 operates the image processing device 31 to set the first judgment criterion (step S13). Specifically, as shown in FIGS. 5(A) and 5(B), the image processing device 31 extracts installed objects 90 from the preliminary image P0, connects them with a line L1, and determines a safety area SA, which is an index of the first judgment criterion. After step S13, or if the first judgment criterion has been set (Y in step S12), the control device 21 operates the image processing device 31 to perform vehicle detection (step S14).

[0039] After step S14, the control device 21 operates the image processing device 31 to measure the vehicle vector (step S15). The vehicle vector indicates the direction and speed of movement of the vehicle CA, and can be used as an auxiliary parameter for determining dangerous driving.

[0040] Next, the control device 21 checks whether learning has been completed, i.e., whether the second judgment criterion has been set (step S16). If learning has not been completed (N in step S16), the control device 21 operates the learning device 32 to perform learning (step S17). This sets the second judgment criterion. Specifically, as shown in FIG. 6(B) and other figures, the learning device 32 determines a safe driving area TA, which is an index of the second judgment criterion, from the multiple learned driving trajectories RP.

[0041] If the learning has not yet been completed in step S17, the control device 21 operates the determination device 33 to perform a first determination based on the first determination criterion (step S18). On the other hand, if the learning has been completed (Y in step S16), the control device 21 operates the determination device 33 to perform a second determination based on the second determination criterion (step S19).

[0042] If it is determined that the vehicle is driving unsafely after the determination in step S18 or step S19 (Y in step S21), the control device 21 activates the alarm device 24 to notify the driver that the vehicle is driving unsafely (step S22), and returns to step S11. At this time, the control device 21 may transmit a danger signal or the like to the mobile terminal 50 carried by the worker US or the like via the communication device 25. On the other hand, if it is determined that the vehicle is driving safely (N in step S21), the control device 21 checks whether or not to terminate the operation of the vehicle detection device 100 (step S23).

[0043] If the operation of the vehicle detection device 100 is not to be ended (N in step S23), the control device 21 returns to step S11. On the other hand, if the operation of the vehicle detection device 100 is to be ended (Y in step S23), the control device 21 ends the operation of the device.

[0044] In the embodiment of the vehicle detection device 100 described above, the first judgment is made based on the first judgment criterion, which can be set relatively quickly, before the judgment device 33 finishes learning about the driving trajectory RP of the vehicle CA, so that after installing the vehicle detection device 100, the device can be used immediately without waiting for the learning time.

[0045] Second Embodiment An example of a vehicle detection device according to the second embodiment will be described below. In the second embodiment, the same matters as in the first embodiment will not be described.

[0046] In this embodiment, the determination device 33 makes a first determination of dangerous driving based on the first determination criterion after the vehicle detection device 100 starts operating and before learning is completed. Furthermore, the determination device 33 makes a first determination and a second determination of dangerous driving based on the first determination criterion and the second determination criterion after learning is completed. In other words, the determination criteria for the determination device 33 after learning is completed are both the first determination criterion and the second determination criterion. This allows for more accurate determination by taking into account both the safe area SA and the safe driving area TA.

[0047] FIG. 8 is a flowchart illustrating the operation of the vehicle detection device 100 according to the second embodiment.

[0048] In this embodiment, when the control device 21 has completed learning, i.e., when the second judgment criterion has been set (Y in step S16), it operates the judgment device 33 and performs the first judgment and the second judgment based on the first judgment criterion and the second judgment criterion (step S219).

[0049] Third Embodiment An example of a vehicle detection device according to the third embodiment will be described below. In the third embodiment, the same matters as in the first embodiment will not be described.

[0050] In this embodiment, the vehicle detection device 100 periodically monitors the safety area SA. As a result, the first judgment criterion changes in accordance with changes in the safety area SA. The judgment device 33 makes a first judgment of dangerous driving based on the first judgment criterion after the vehicle detection device 100 starts operating and before learning is completed. Furthermore, the judgment device 33 makes a second judgment if there is a change in the first judgment criterion after learning is completed, and makes the first judgment if there is no change in the first judgment criterion. This makes it possible to switch between the first judgment and the second judgment while periodically updating the safety area SA.

[0051] FIG. 9 is a flowchart illustrating the operation of the vehicle detection device 100 according to the third embodiment.

[0052] First, the control device 21 operates the imaging device 10 to capture an image of the monitoring area DA and obtain a captured image IM (step S11).

[0053] Next, the control device 21 operates the image processing device 31 to set a first determination criterion (step S13). Thereafter, the control device 21 operates the image processing device 31 to perform vehicle detection (step S14) and measure the vehicle vector (step S15).

[0054] Next, the control device 21 checks whether learning has been completed, i.e., whether the second judgment criterion has been set (step S16). If learning has not been completed (N in step S16), the control device 21 operates the learning device 32 to perform learning (step S17). If learning has not yet been completed in step S17, the control device 21 operates the determination device 33 to perform a first judgment based on the first judgment criterion (step S18). On the other hand, if learning has been completed (Y in step S16), the control device 21 checks whether there has been a change in the first judgment criterion (step S324). If the first judgment criterion has been set for the second time or later in step S13, the safety area SA and the first judgment criterion may change due to, for example, movement of the installed object 90, as shown in FIG. 10.

[0055] If there is a change in the first judgment criterion (Y in step S324), the control device 21 operates the judgment device 33 to make a first judgment based on the updated first judgment criterion (step S18). On the other hand, if there is no change in the first judgment criterion (N in step S324), the control device 21 operates the judgment device 33 to make a second judgment based on the second judgment criterion (step S19).

[0056] After the determination in step S18 or step S19, the control device 21 performs the processes in step S21 and thereafter.

[0057] In this embodiment, the vehicle detection device 100 can also be moved.

[0058] 〔others〕 The present invention is not limited to the above-described embodiment, and can be embodied in various forms without departing from the spirit and scope of the present invention.

[0059] In the above embodiment, the image capture device 10 captures an image of the vehicle CA from the front, and the reference point BP of the travel trajectory RP is set to the bottom center of a rectangular frame surrounding the front of the vehicle CA, but another part may be used as the reference point BP. For example, if the image capture device 10 captures an image of the vehicle CA from the rear, the reference point BP of the travel trajectory RP may be set to the bottom center of a rectangular frame surrounding the rear of the vehicle CA. Furthermore, the reference point BP of the travel trajectory RP may be the center of gravity of the frame surrounding the vehicle CA, for example.

[0060] In the above embodiment, an example has been described in which the vehicle detection device 100 is installed on a road RO where traffic drives on the left, but it may also be installed on a road RO where traffic drives on the right.

[0061] In the above embodiment, the safety area SA is set by connecting the vertices of the installation 90 with the line L1, but the location where the line L1 is connected on the installation 90 can be changed as appropriate. In addition, the shape of the end of the safety area SA can also be changed as appropriate.

[0062] In the above embodiment, the image processing device 31 performs image processing on the travel trajectory RP of the vehicle CA and the like in the setting processing stage, but the learning device 32 may perform similar image processing.

[0063] In the above embodiment, the driving trajectory RP of the vehicle CA of interest may be predicted from the position and vehicle vector of the vehicle CA, and the prediction result may be used to determine whether the vehicle is driving recklessly.

[0064] In the above embodiment, the setting of the non-safe area NSA and the no-travel area FA may be omitted.

[0065] The vehicle detection device 100 and the imaging device 10 do not have to be mounted on the sign vehicle V1. For example, one or more imaging devices 10 can be installed near or at a remote location from the sign vehicle V1 to detect dangerously driving vehicles remotely or simultaneously in parallel. Note that if the imaging device 10 is installed at a location remote from the information processing device 20, a communication circuit will be required to remotely operate the imaging device 10 and its associated drive circuit, and a support stand will also be required to set the imaging device 10 at a relatively high viewpoint position. [Explanation of symbols]

[0066] 3...sign, 4...warning sign, 10...imaging device, 20...information processing device, 21...control device, 22...storage device, 23...user interface, 24...alarm device, 25...communication device, 31...image processing device, 32...learning device, 33...judgment device, 50...mobile terminal, 90...installed object, 100...vehicle detection device, BP...reference point, CA...vehicle, DA...monitoring area, FA...no driving area, IM...captured image, MW...maintenance work site, NSA...non-safe area, P0...preliminary image, PA...capture range, RO...road, RP...driving trajectory, SA...safe area, TA...safe driving area, V1...signposted vehicle

Claims

1. an imaging device that captures an image of a monitoring area including installations that separate the safety area; a determination device that makes a first determination of dangerous driving using a first determination criterion based on the safe area defined by the installed object in the image, and makes a second determination of dangerous driving using a second determination criterion based on a safe driving area that is set by learning the driving trajectory of the vehicle in the image; Equipped with The determination device performs the first determination before learning is completed. Vehicle detection device.

2. The determination device switches the determination criteria before and after the end of the learning. The vehicle detection device according to claim 1 .

3. The determination device determines in the first determination that the vehicle is driving unsafely when the vehicle enters the safety area. The vehicle detection device according to claim 1 .

4. In the second determination, the determination device determines that the vehicle is traveling unsafely when the vehicle is traveling outside the safe traveling area. The vehicle detection device according to claim 1 .

5. the determination device performs the second determination after the learning is completed. The vehicle detection device according to claim 1 .

6. the determination device performs the first determination and the second determination after the learning is completed. The vehicle detection device according to claim 1 .

7. the first determination criterion changes in accordance with a change in the safety area; the determination device performs the second determination if there is a change in the first determination criterion after the learning is completed, and performs the first determination if there is no change in the first determination criterion; The vehicle detection device according to claim 1 .

Images