Obstacle detection method and detection system for vehicles

Abstract

To provide an obstacle detection method and an obstacle detection system for a vehicle, capable of detecting an obstacle in a blind spot using a system with a simple configuration.SOLUTION: In an obstacle detection method for a vehicle, light 7 is emitted from a light source 6 toward a detection target area 4 that is a blind spot from a vehicle 3; a road surface illuminated by the light 7 emitted from the light source 6 and passing through a detection target area 4 is photographed using an imaging unit mounted on the vehicle 3; and a photographed video is image-processed in the vehicle 3, and if there is a shadow 8 extending from an obstacle 5 in the detection target area 4 on the road surface, it is determined that the obstacle 5 is present in the detection target area 4.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present disclosure relates to a method and a system for detecting obstacles for vehicles.

Background Art

[0002] In recent years, technologies related to ADAS (Advanced Driver-Assistance Systems) and autonomous driving installed in vehicles have evolved significantly, and their obstacle detection capabilities have also improved. For vehicles, especially those running in autonomous driving, it is important to determine whether there are obstacles in blind spots for collision avoidance at intersections with poor visibility. The following Patent Document 1 discloses an obstacle detection system at intersections.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the system disclosed in Patent Document 1, it is necessary to install a camera, an arithmetic device, and a communication device at intersections, and it is also necessary to install a communication device in the vehicle in addition to sensors such as radar and cameras. In addition, the camera arithmetic device and communication device at intersections need to be installed for each intersection. If this system attempts to perform the same obstacle detection at locations where blind spots occur, such as building entrances or site entrances that are not intersections, it is also necessary to install devices such as cameras at these entrances. Therefore, it is very costly to put the system into practical use.

[0005] An object of the present disclosure is to provide a method and a system for detecting obstacles for vehicles that can detect obstacles in blind spots using a system with a simple configuration.

Means for Solving the Problems

[0006] In the vehicle obstacle detection method described herein, a light source is used to illuminate a detection target area that is a blind spot from the vehicle. An imaging unit mounted on the vehicle is used to photograph the road surface illuminated by the light that has been emitted from the light source and passed through the detection target area. The captured image is processed in the vehicle, and if there is a shadow extending from within the detection target area onto the road surface, it is determined that there is an obstacle in the detection target area. Furthermore, when the imaging unit captures a code sign installed near the detection target area, the image processing unit compares the image of the road surface captured by the imaging unit with a reference image of the road surface when the obstacle is absent, which is stored in the vehicle's memory beforehand, to determine whether or not the shadow extending from the obstacle within the detection target area is present on the road surface. do.

[0007] In the image processing described above, when the shadow on the road surface that was captured is moving on the road surface, it may be determined that there is an obstacle moving within the detection target area.

[0011] The vehicle obstacle detection system described herein comprises a light source that emits light toward a detection target area that is a blind spot from the vehicle, and an imaging unit mounted on the vehicle. The vehicle is equipped with a storage unit that pre-stores a reference image of the road surface when no obstacles are present in the detection target area, The vehicle is equipped with a detection control unit. The detection control unit is configured to process images of the road surface illuminated by the light emitted from the light source and passing through the detection target area, and to determine that there is an obstacle in the detection target area if there is a shadow extending from within the detection target area on the road surface. When the imaging unit captures a code sign installed near the detection target area, the detection control unit, in the image processing, compares the image of the road surface captured by the imaging unit with the reference image to determine whether or not there is a shadow extending from the obstacle within the detection target area on the road surface. [Effects of the Invention]

[0012] According to the vehicle obstacle detection method and detection system of this disclosure, obstacles in blind spots can be detected using a system with a simple configuration. [Brief explanation of the drawing]

[0013] [Figure 1] This is a plan view of the area near the entrance of a building where the obstacle detection method according to the first embodiment is implemented. [Figure 2] This is a side view corresponding to the plan view in Figure 1. [Figure 3]This is a plan view of the area near the entrance of a building where the obstacle detection method according to the second embodiment is implemented. [Figure 4] This is a side view corresponding to the plan view in Figure 3. [Figure 5] This is a plan view of the area near the entrance to the site where the obstacle detection method according to the third embodiment is implemented. [Figure 6] This is a plan view of the area near the entrance to the site where the obstacle detection method according to the fourth embodiment is implemented. [Figure 7] This is a plan view of the area near the entrance to the site where the obstacle detection method according to the fifth embodiment is implemented. [Figure 8] This is a block diagram of the vehicle in the obstacle detection system used in the above embodiment. [Figure 9] This is the first part of the flowchart for the obstacle detection method of the above embodiment. [Figure 10] This is the latter half of the flowchart shown above. [Modes for carrying out the invention]

[0014] The following describes embodiments of a vehicle obstacle detection method and detection system with reference to the drawings.

[0015] Figures 1 and 2 show the first embodiment. In the first embodiment, an obstacle 5 is detected in a detection target area 4 inside building 1, which is a blind spot from the vehicle 3 at the entrance 2 of building 1. In this embodiment, when detecting the obstacle 5, a light source 6 is installed on the other side of the detection target area 4 relative to the vehicle 3's own position, i.e., the position where the vehicle 3 detects the obstacle 5 within the detection target area 4. In this embodiment, building 1 is, for example, a workshop or warehouse in a factory, and in this embodiment, assuming that no workers are working inside building 1, the light source 6 is installed at a low position. More specifically, the light source 6 is installed at a low position on the inner wall surface opposite to the entrance 2 of building 1. However, workers are not prohibited from passing inside building 1 or near the entrance 2. Light 7 is emitted from the light source 6 slightly downwards, and the upper boundary of the light 7 is almost horizontal.

[0016] Light 7 is emitted from the light source 6 to extend a shadow 8 from the obstacle 5 within the detection target area 4 and form a shadow 8 on the road surface. The shadow 8 is formed by the light 7 emitted from the light source 6 toward the detection target area 4, extending from the obstacle 5 within the detection target area 4, which is a blind spot, to outside the detection target area 4, that is, from the entrance / exit 2 of the building 1. In this embodiment, the light source 6 is installed at a low position in order to form a long shadow 8. Obstacle detection is performed using the patterns of light 7 and shadow 8 on the road surface, so it is affected by the brightness of the surrounding environment, but obstacle detection can be performed as long as a pattern of light 7 is formed on the road surface and a pattern of shadow 8 from the obstacle 5 is also formed. In this embodiment, the vehicle 3 is, for example, a truck that delivers parts to the building 1 and is driving automatically on the roads within the factory.

[0017] Vehicle 3 is equipped with a camera, which serves as an imaging unit 30 as shown in Figure 8, and a detection control unit 31 that processes the video or images captured by the imaging unit 30. In this embodiment, the imaging unit 30 captures video, but it may also capture still images. If still images are captured continuously at predetermined time intervals, image processing can be performed in the same way as with video. In this case, the imaging unit 30 is a camera that captures the area in front of vehicle 3, which is provided for autonomous driving.

[0018] In this embodiment, the vehicle 3 can detect that it has approached the detection target area 4 for obstacle detection. Two examples of this self-position detection will be described. In the first example, a code label 9 indicating that the detection target area 4 is present is provided at the position photographed by the imaging unit 30 of the vehicle 3 approaching the detection target area 4. Here, the code label 9 is in the form of a traffic sign and is installed in front of the entrance / exit 2. As the code, for example, a QR code (registered trademark) can be used. The code label 9 may be provided on the outer wall surface of the building 1 or on the road surface. Here, the imaging unit 30 always takes pictures, and the video is always processed by the image processing unit 31a of the detection control unit 31. When the code label 9 is detected in the video, the detection necessity determination unit 31b of the detection control unit 31 determines that the vehicle has approached the detection target area 4. That is, the detection necessity determination unit 31b determines that obstacle detection is necessary.

[0019] In the second example, the vehicle 3 includes a self-position detection unit 31c. As the self-position detection unit 31c, for example, a satellite positioning system such as GPS is common. Here too, the self-position detection unit 31c uses a GPS or the like that detects the self-position of the vehicle 3 provided for automatic driving. The self-position of the vehicle 3 is detected by the self-position detection unit 31c, and the detection necessity determination unit 31b of the detection control unit 31 determines whether the vehicle has approached the detection target area 4. The detection necessity determination unit 31b determines that obstacle detection is necessary when the vehicle has approached the detection target area 4. In this case, after the detection necessity determination unit 31b determines that obstacle detection is necessary, the image processing for obstacle detection may be started by the image processing unit 31a. If it is not necessary to perform the image processing for obstacle detection until it is determined that obstacle detection is necessary, the load on the detection control unit 31 can be reduced.

[0020] Next, two examples of obstacle detection will also be described. In the first example, the obstacle 5 is detected when it is moving within the detection target area 4. Since the light source 6 is fixed, the pattern of the light 7 extending from the entrance / exit 2 onto the road surface does not change. The video captured by the imaging unit 30 is captured from the traveling vehicle 3, so the pattern of the light 7 changes its position within the video frame. However, through image processing, it can be detected that the pattern of the light 7 is not moving on the road surface. At this time, when there is an obstacle 5 moving within the detection target area 4, the pattern of the shadow 8 moving inside the pattern of the light 7 that does not move on the road surface can be detected. Whether the shadow 8 is moving can be determined through image processing based on the pattern of the non-moving light 7. This determination is made by the obstacle detection unit 31d of the detection control unit 31. If a moving shadow 8 is detected within the pattern of the light 7, it can be determined that there is an obstacle 5 moving within the detection target area 4.

[0021] In the second example, the vehicle 3 includes a storage unit 31e in which a reference image of the road surface when there is no obstacle 5 is stored in advance. The reference image is the pattern of the light 7 on the road surface outside the entrance / exit 2 when there is no obstacle 5. That is, by performing image processing on the video captured by the imaging unit 30 to detect the pattern of the light 7 on the road surface and comparing this with the reference image, it is possible to easily determine whether the shadow 8 of the obstacle 5 is within the pattern of the light 7. The reference image is stored for each detection target area 4, and which reference image to use for comparison can be determined from the position of the vehicle 3 detected by the above-described code label 9 and the self-position detection unit 31c. As a result of the comparison with the reference image, if there is a shadow 8 within the pattern of the light 7, the obstacle detection unit 31d determines that there is an obstacle 5 within the detection target area 4. In the second example, the obstacle 5 can be detected even if the obstacle 5 is not moving. Regarding the obstacle detection in this second example, it will be described in detail later by referring to the flowcharts of FIGS. 9 and 10.

[0022] Figures 3 and 4 show a second embodiment. The second embodiment differs from the first embodiment only in the installation position of the light source 6. When workers are performing tasks inside building 1, if the light source 6 is installed in the same position as in the first embodiment, for example, the workers may be dazzled by the light 7 from the light source 6, which could hinder their work. In such cases, as in this embodiment, the light source 6 is installed at a higher position so as not to obstruct the workers' work. In this way, it is possible to achieve compatibility between work inside building 1 and the illumination of light from the light source 6 for obstacle detection.

[0023] Figure 5 shows a third embodiment. In the first and second embodiments described above, the detection target area 4 was inside building 1. In this embodiment, the detection target area 4 is outdoors, more specifically, within the premises near an entrance / exit 2, such as the main gate of a factory. The entrance / exit 2 is located where a part of the factory's outer wall 1X is interrupted. Therefore, the detection target area 4 is a blind spot for vehicles 3 due to the outer wall 1X. The light source 6 is installed above the internal road on the opposite side of the entrance / exit 2 of the detection target area 4. Specifically, a gate is installed on the internal road, and the light source 6 is attached to this gate.

[0024] Alternatively, the light source 6 can be mounted on a pole that extends cantilevered from the side of the internal road, similar to a traffic light on a public road. Since factories have traffic lights for their internal roads, these traffic light poles can also be used. Factories may also have signs indicating destinations for suppliers, so the gates and poles for these signs can also be used. Furthermore, piping racks for routing various pipes between buildings may be stretched across the internal roads within the factory, so the light source 6 can also be mounted on such racks. In Figure 5, the shadow 8 extending from the obstacle 5 reaches the public road in front of the factory entrance 2. The other configurations are the same as those of the first and second embodiments described above.

[0025] Figure 6 shows a fourth embodiment. In this embodiment, the configuration of the detection system itself is the same as in the third embodiment described above. Furthermore, the detection of obstacles 5 moving within the detection target area 4 is the same as in the first and second examples of obstacle detection described in the first embodiment. This embodiment shows that it is also possible to detect how the obstacle 5 is moving. In this embodiment, the obstacle 5 moves within the detection target area 4 while keeping its distance from the light source 6 almost unchanged. In other words, the obstacle 5 moves almost perpendicular to the direction of irradiation of light 7 from the light source 6.

[0026] In this case, a pattern of stationary light 7 is formed on the road surface outside the entrance / exit 2, and the shadow 8 changes position within it, but its size remains almost unchanged. As a result of image processing by the image processing unit 31a, the obstacle detection unit 31d also detects how the obstacle 5 within the detection target area 4 is moving. In this embodiment, since the obstacle 5 is unlikely to be located in front of the vehicle 3 in the direction of travel, it can be determined that the risk of a head-on collision between the vehicle 3 and the obstacle 5 is low.

[0027] Figure 7 shows the fifth embodiment. In this embodiment, the configuration of the detection system itself is the same as in the third and fourth embodiments described above. In this embodiment, the obstacle 5 is moving toward the entrance / exit 2. In other words, the obstacle 5 moves away from the light source 6 along the direction of illumination of the light 7 from the light source 6. In this case, the shadow 8 formed on the road surface outside the entrance / exit 2 does not change its position much, but its size becomes smaller. As a result of image processing by the image processing unit 31a, the obstacle detection unit 31d detects that the obstacle 5 is approaching the entrance / exit 2, even with this movement of the shadow 8. In this embodiment, it can be determined that there is a high risk of a head-on collision between the vehicle 3 and the obstacle 5. On the other hand, if the position of the shadow 8 does not change much, but its size becomes larger, it can be determined that the obstacle 5 is moving away, and there is a low risk of a head-on collision between the vehicle 3 and the obstacle 5.

[0028] Next, the system configuration on the vehicle 3 side will be described with reference to the block diagram in Figure 8. Some of these components were mentioned in the first to fifth embodiments described above, but they will be explained again here. The vehicle is equipped with a camera as an imaging unit 30. As described above, obstacle detection is performed based on the video captured by the imaging unit 30. The vehicle 3 is also equipped with a detection control unit 31 that performs obstacle detection based on the video captured by the imaging unit 30. The detection control unit 31 is an electronic device consisting of a CPU, ROM, RAM, storage such as an SSD or HDD, and I / O devices.

[0029] The detection control unit 31 includes an image processing unit 31a that processes the video captured by the imaging unit 30. Therefore, the imaging unit 30 is connected to the image processing unit 31a, and the video captured by the imaging unit 30 is sent to the image processing unit 31a. The detection control unit 31 also includes a detection necessity determination unit 31b connected to the image processing unit 31a. The detection necessity determination unit 31b determines whether obstacle detection is necessary. For example, if the code mark 9 described above is detected as a result of image processing by the image processing unit 31a, the detection necessity determination unit 31b determines that obstacle detection is necessary.

[0030] The detection control unit 31 also includes a self-position detection unit 31c connected to the detection necessity determination unit 31b. For example, as described above, if the satellite positioning system acting as the self-position detection unit 31c detects that the unit is approaching the detection target area 4, the detection necessity determination unit 31b determines that obstacle detection is necessary. However, if the detection necessity determination unit 31b makes the determination based solely on the code markings 9 described above, the self-position detection unit 31c does not need to be provided.

[0031] The detection control unit 31 also includes an obstacle detection unit 31d connected to the detection necessity determination unit 31b. When the detection necessity determination unit 31b determines that obstacle detection is necessary, the obstacle detection unit 31d uses the image processing results from the image processing unit 31a to detect obstacles 5 within the detection target area 4. More specifically, as described above, the obstacle detection unit 31d performs obstacle detection based on the patterns of light 7 and shadow 8 on the road surface. As in the fourth and fifth embodiments described above, the obstacle detection unit 31d also detects the state of the obstacle 5. Since the obstacle detection unit 31d uses the image processing results from the image processing unit 31a, it is also connected to the image processing unit 31a.

[0032] In the second example of obstacle detection described in the first embodiment, the vehicle 3 is required to have a storage unit 31e that stores a reference image of the road surface when no obstacle 5 is present. For this reason, the detection control unit 31 also includes a storage unit 31e connected to the obstacle detection unit 31d that uses the reference image. However, if obstacle detection is performed only in the first example of obstacle detection described in the first embodiment, the storage unit 31e does not need to be provided. As described above, the storage unit 31e stores multiple reference images and selects which reference image to use based on the vehicle 3's own position. For this reason, the storage unit 31e is also connected to the self-position detection unit 31c.

[0033] Although not described in the embodiments described above, if an obstacle 5 is detected, some kind of action will need to be taken. This could simply involve notifying the occupants of the autonomously driven vehicle 3, or it could involve avoiding a collision with the obstacle 5 through autonomous driving. For this reason, the detection control unit 31 also includes an action determination unit 31f connected to the obstacle detection unit 31d. A display unit 31g, such as a display monitor, is connected to the action determination unit 31f for notification. The display unit 31g may be a touchscreen or the like and used as an input device to the detection control unit 31.

[0034] As described above, since the vehicle 3 in the above embodiment is driven by automatic driving, the vehicle 3 is also equipped with an automatic driving control unit 32. The automatic driving control unit 32 is connected to the response determination unit 31f of the detection control unit 31 in order to operate the vehicle 3 by automatic driving when an obstacle 5 is detected by the response determination unit 31f in order to avoid a collision with the obstacle 5. In automatic driving, the vehicle 3 can be controlled, including acceleration and deceleration, as well as turning, and in order to perform these controls, the automatic driving control unit 32 is also connected to the drive unit 33 of the vehicle 3.

[0035] Next, we will explain obstacle detection control in more detail, referring to the flowcharts in Figures 9 and 10. Here, we will explain a second example in which vehicle 3, which is driving autonomously on the factory's internal roads, detects its own position using the self-position detection unit 31c, and obstacle detection utilizes the reference image described above. Since vehicle 3 is driving autonomously, the camera, acting as the imaging unit 30, is constantly taking pictures.

[0036] In step S1, the vehicle 3 detects its own position using the self-position detection unit 31c. Following step S1, in step S2, the detection necessity determination unit 31b determines whether the vehicle 3 is located within the detection processing target area near the detection target area 4. In the specific example of the above embodiment, the detection processing target area is the area approaching the entrance / exit 2 where obstacle detection control should be performed. The processing of steps S1 and S2 is looped until step S2 is affirmed, that is, until the detection necessity determination unit 31b determines that obstacle detection is necessary.

[0037] If step S2 is affirmed, in step S3, the image processing unit 31a starts image processing of the video captured by the imaging unit 30 for obstacle detection. Specifically, as described above, this image processing involves analyzing the patterns of light 7 and shadow 8 on the road surface. As a result of step S3, in step S4, the obstacle detection unit 31d first determines whether or not a pattern of light 7, i.e., an exposure pattern, can be detected on the road surface.

[0038] If step S4 is rejected, step S10 determines whether or not obstacle detection is unnecessary. For example, if the image processing results in a shutter being closed on entrance 2 of building 1, it is determined that obstacle detection is unnecessary because the shutter needs to be opened. This is because the obstacle 5 will be dealt with when the shutter is opened. Note that whether or not the shutter is closed on entrance 2 is determined from the image processing results. It is also possible to improve the judgment based on image processing by attaching a code mark to the surface of the shutter. Alternatively, there may be situations where it is possible to determine that obstacle detection is unnecessary simply from the date and time.

[0039] If step S10 is affirmed, that is, if it is determined that obstacle detection is unnecessary, the response determination unit 31f determines in step S11 that detection is unnecessary and displays information indicating that detection is unnecessary on the display unit 31g mounted on the vehicle 3 in step S12. If there are no passengers in the automatically driven vehicle 3, step S12 may be omitted.

[0040] On the other hand, if step S10 is rejected, it means that for some reason the pattern of light 7, i.e., the exposure pattern, cannot be detected on the road surface, so the response determination unit 31f determines in step S13 that detection has failed. Then, in step S14, the response determination unit 31f displays on the display unit 31g that detection has failed and sends a signal to the automatic driving control unit 32 to stop the vehicle 3. The automatic driving control unit 32 controls the drive unit 33 based on the received stop signal to stop the vehicle 3. As described above, obstacle detection is affected by the brightness of the surrounding environment, so obstacle detection may not be possible due to the brightness of the surrounding environment. In the case of rainy weather when the road surface is wet, it is possible that the pattern on the road surface cannot be accurately detected due to road surface reflection. It is also possible that when it is snowing, the contrast due to the snow accumulation on the road surface may be more dominant than the contrast of light 7 and shadow 8.

[0041] If step S4 is affirmed, in step S5, the obstacle detection unit 31d reads the above-mentioned reference image, i.e., the exposure pattern when there is no obstacle 5, from the storage unit 31e based on the vehicle's position detected in step S1. The obstacle detection unit 31d then compares the exposure pattern read from the storage unit 31e with the exposure patterns detected in steps S3 and S4. In step S6, it is determined whether the exposure patterns match based on the results of the above comparison.

[0042] If step S6 is affirmed, that is, if the compared exposure patterns match, then there is no shadow 8 in the pattern of light 7 on the road surface, and therefore it can be determined that there is no obstacle 5 within the detection target area 4. Accordingly, the response determination unit 31f determines in step S15 that there is no obstacle 5 and displays information on the display unit 31g in step S16 that there is no obstacle 5. Note that if there are no passengers in the automatically driven vehicle 3, step S16 may be omitted.

[0043] On the other hand, if step S6 is denied, it can be determined that there is a shadow 8 in the pattern of light 7 on the road surface and that there is an obstacle 5 within the detection target area 4. In this case, in step S7, the obstacle detection unit 31d compares the current exposure pattern with the exposure pattern from one to several frames ago to analyze the movement of the shadow 8. That is, in step S8, the obstacle detection unit 31d determines whether or not the shadow 8 in the exposure pattern has changed. If step S8 is denied, that is, if the shadow 8 has not changed, it can be determined that the obstacle 5 has not moved. Therefore, in step S17, the response determination unit 31f determines that the obstacle 5 is unlikely to collide, and in step S18, it displays warning information on the display unit 31g and sends a signal to the automatic driving control unit 32 to decelerate the vehicle 3. Based on the received deceleration signal, the automatic driving control unit 32 controls the drive unit 33 to decelerate the vehicle 3.

[0044] If step S8 is affirmed, in step S9, the obstacle detection unit 31d determines whether the change in shadow 8 indicates that shadow 8 is approaching vehicle 3 or that the width of shadow 8 is decreasing. This is the obstacle detection described above with reference to Figure 7. Here, "shadow 8 is approaching vehicle 3" refers to the case where the outline of shadow 8 is closed on the road surface, as shown in Figure 3. In this case, it is easy to detect the approach of obstacle 5 by the movement of shadow 8 itself. On the other hand, "the width of shadow 8 is decreasing" refers to the case where the outline of shadow 8 is not closed on the road surface, as shown in Figure 1. In this case, it is difficult to detect the approach of obstacle 5 by the movement of shadow 8 itself, and the movement of obstacle 5 is determined based on the change in the width of shadow 8.

[0045] If step S9 is not met, it can be determined that the obstacle 5 is moving but the likelihood of collision is low. Therefore, the response determination unit 31f determines in step S17 that the obstacle 5 is unlikely to collide, and in step S18 displays warning information on the display unit 31g and sends a signal to the automatic driving control unit 32 to decelerate the vehicle 3. The automatic driving control unit 32 controls the drive unit 33 based on the received deceleration signal to decelerate the vehicle 3.

[0046] On the other hand, if step S9 is affirmed, it can be determined that the obstacle 5 is moving forward in the direction of travel of the vehicle 3 and that there is a high probability of collision. Therefore, the response determination unit 31f determines in step S19 that the obstacle 5 is highly likely to collide, and in step S20 displays warning information on the display unit 31g and sends a signal to the automatic driving control unit 32 to stop the vehicle 3. Based on the received stop signal, the automatic driving control unit 32 controls the drive unit 33 to stop the vehicle 3.

[0047] In the vehicle obstacle detection method of the above embodiment, light 7 is emitted from a light source 6 towards a detection target area 4, which is a blind spot from the vehicle 3. An imaging unit 30 mounted on the vehicle 3 is used to capture images of the road surface illuminated by the light 7 emitted from the light source 6 and passing through the detection target area 4. The captured images are processed in the vehicle 3, and if there is a shadow 8 extending from within the detection target area 4 on the road surface, it is determined that there is an obstacle 5 in the detection target area 4. Therefore, the detection determination process itself is completed only on the vehicle 3 side. In addition, only the light source 6 needs to be installed in the detection target area 4, and no communication devices are required. Thus, an obstacle 5 in a blind spot can be detected using a system with a simple configuration. Furthermore, even if there are many detection target areas 4, only the light source 6 needs to be installed in each detection target area 4, and everything else is completed on the vehicle 3 side, making it easy to construct the system.

[0048] In this case, during image processing, if the shadow 8 extending from the obstacle 5 on the road surface is moving across the road surface, it may be determined that the obstacle 5 is moving within the detection target area 4. In this case, the obstacle 5 can be detected without the need to save the reference image described later on the vehicle 3 side. However, detection is not possible if the obstacle 5 is not moving.

[0049] Alternatively, in image processing, the presence or absence of a shadow 8 extending from the obstacle 5 within the detection target area 4 is determined by comparing it with a reference image of the road surface when no obstacle 5 is present, which is previously stored in the storage unit 31e of the vehicle 3. In this case, although it is necessary to store the reference image on the vehicle 3 side, it is possible to detect not only moving obstacles 5 but also stationary obstacles 5.

[0050] Furthermore, if the self-position detection unit 31c mounted on the vehicle 3 detects that it is approaching the detection target area 4, obstacle detection may be performed by comparing it with a reference image. In this case, the self-position detection unit 31c is required, but obstacle detection does not need to be performed until the vehicle approaches the detection target area 4 where obstacle detection is to be performed, thus reducing the control load. In addition, an appropriate reference image can be retrieved from the storage unit 31e based on the detection result of the self-position detection unit 31c.

[0051] Alternatively, when the imaging unit 30 captures a code marker 9 installed near the detection target area 4, obstacle detection may be performed by comparing it with a reference image. In this case, it becomes necessary to install a code marker 9 in addition to the light source 6 near the detection target area 4, but the self-position detection unit 31c is not necessary for obstacle detection by using the imaging unit 30 which detects light 7 and shadow 8. In this case as well, obstacle detection does not need to be performed until approaching the detection target area 4 where obstacle detection is to be performed, thus reducing the control load. Furthermore, since the vehicle 3's own position can be determined based on the code marker 9, an appropriate reference image can be retrieved from the storage unit 31e.

[0052] The vehicle obstacle detection system of the above embodiment comprises a light source 6 that emits light 7 from the vehicle 3 toward a detection target area 4 which is a blind spot, an imaging unit 30 mounted on the vehicle 3, and a detection control unit 31 mounted on the vehicle 3. The detection control unit 31 processes the image captured by the imaging unit of the road surface illuminated by the light 7 emitted from the light source 6 and passing through the detection target area 4. The detection control unit 31 is configured to determine that there is an obstacle 5 in the detection target area 4 if, as a result of the image processing, there is a shadow 8 extending from within the detection target area 4 on the road surface. Therefore, the detection determination process itself is completed only by the imaging unit 30 and detection control unit 31 on the vehicle 3 side. In addition, only the light source 6 needs to be installed in the detection target area 4, and no communication devices are required. Thus, obstacles 5 in blind spots can be detected using a system with a simple configuration. Furthermore, even if there are many detection target areas 4, only the light source 6 needs to be installed in each detection target area 4, and everything else is completed on the vehicle 3 side, making it easy to construct the system.

[0053] This disclosure is not limited to the embodiments described above. For example, in the embodiments described above, the camera, which is the imaging unit 30, photographed the area in front of the vehicle 3. However, a vehicle 3, such as a truck, may also enter the building entrance 2 from the rear cargo bed. In this case, obstacle detection may be performed based on the image from a rear camera that photographs the area behind the vehicle 3. Furthermore, although the above embodiments were described on the premise that the vehicle 3 is driving autonomously, the above-described obstacle detection may of course also be performed on a vehicle being driven by a human driver.

[0054] Furthermore, an advantage of this disclosure is that the obstacle detection process can be completed solely on the vehicle 3 side without the need for communication devices or other means to communicate with the outside. However, this does not preclude the use of communication devices for obstacle detection. For example, when determining whether or not obstacle detection is unnecessary in step S10 described above, information obtained from the outside via a communication device may be used. [Explanation of symbols]

[0055] 3 vehicles 30 Imaging Unit 31 Detection Control Unit 31e Storage section 4. Detection Target Area 5 Obstacles 6 light source 7 light 8 Shadows 9 Code Sign

Claims

1. A method for detecting obstacles in vehicles, The light source illuminates the detection target area, which is a blind spot from the vehicle, Using the imaging unit mounted on the vehicle, the road surface illuminated by the light emitted from the light source and passing through the detection target area is photographed. The captured video is processed in the vehicle, and if a shadow is seen on the road surface extending from within the detection target area, it is determined that there is an obstacle in the detection target area. A vehicle obstacle detection method, wherein when the imaging unit captures a code sign installed near the area to be detected, the image processing compares the image of the road surface captured by the imaging unit with a reference image of the road surface when no obstacle is present, which is stored in the vehicle's memory beforehand, to determine whether or not there is a shadow extending from the obstacle within the area to be detected on the road surface.

2. The vehicle obstacle detection method according to claim 1, wherein, in the image processing described above, when the shadow on the road surface that has been photographed is moving on the road surface, it is determined that there is an obstacle moving within the detection target area.

3. Vehicle obstacle detection system, A light source that illuminates the detection target area, which is a blind spot from the vehicle, The vehicle includes an imaging unit, The vehicle is equipped with a storage unit that pre-stores a reference image of the road surface when no obstacles are present in the detection target area, The vehicle is equipped with a detection control unit, The detection control unit is configured to process images of the road surface illuminated by the light emitted from the light source and passing through the detection target area, and to determine that there is an obstacle in the detection target area if there is a shadow extending from within the detection target area on the road surface. Vehicle obstacle detection system, wherein when the imaging unit captures a code sign installed near the area to be detected, the detection control unit, in the image processing, compares the image of the road surface captured by the imaging unit with the reference image to determine whether or not there is a shadow extending from the obstacle within the area to be detected on the road surface.

Images