Intersecting road detection device, intersecting road detection method, and program

Abstract

To appropriately detect an intersecting road that intersects a road on which a host vehicle is traveling.SOLUTION: An intersecting road detection device 15 includes: an acquisition section 3B that acquires information indicating road pixels corresponding to roads included in a host vehicle front image; a first road width calculation section 3C that calculates the width of road pixels at a first position and a second position in a traveling direction of a host vehicle 1 on the basis of the road pixels indicated by the acquired information; a second road width calculation section 3D that calculates the width of the road pixels at a third position in the traveling direction of the host vehicle 1 on the basis of the road pixels indicated by the acquired information; a road width estimation section 3E that estimates the width of the road pixels at the third position on the basis of the width of the road pixels at the first position and the second position which are calculated by the first road width calculation section 3C by assuming that an intersecting road intersecting a host vehicle traveling road is not included in the image; and a determination section 3F that determines that the intersecting road exists at the third position when calculated width of the road pixels at the third position is greater than estimated width of the road pixels at the third position.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present disclosure relates to an intersection road detection device, an intersection road detection method, and a program.

Background Art

[0002] Patent Document 1 describes that it is determined whether a lateral road (including a horizontal road extending in the left - right direction with respect to the vehicle and an oblique direction) is shown in a captured image.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] By the way, in the technique described in Patent Document 1, in order to determine whether a lateral road is shown in a captured image, the measurement result of an autofocus device is used. Therefore, in the technique described in Patent Document 1, when information indicating the measurement result of the autofocus device is not included in the data of the captured image, the lateral road included in the captured image cannot be detected. While driving, it is necessary to detect intersecting roads (intersections) in order to respond to sudden appearances of cyclists, pedestrians, etc., near intersections, and to reduce collisions between the vehicle and surrounding vehicles traveling on intersecting roads. When GPS (Global Positioning System) and map information are used to detect intersecting roads (intersections), costs may increase, and there is a risk that the location of the intersecting road (intersection) may be incorrectly detected due to GPS positional errors. When intersecting roads (intersections) are detected using lane markings, etc., of the road the vehicle is traveling on, which are included in the image obtained by photographing the area in front of the vehicle, there is a risk that the intersecting road (intersection) may not be properly detected if there are no clear lane markings on the road the vehicle is traveling on. Intersecting roads are places where something may suddenly appear on the road the vehicle is traveling on, so it is necessary to detect them properly.

[0005] In view of the above, the purpose of this disclosure is to provide an intersecting road detection device, an intersecting road detection method, and a program that can appropriately detect intersecting roads that cross the road on which a vehicle is traveling. [Means for solving the problem]

[0006] (1) One aspect of the present disclosure includes: an acquisition unit that acquires information indicating road pixels, which are pixels corresponding to roads included in an image, generated from an image obtained by photographing the front of the vehicle; a first road width calculation unit that calculates, based on the road pixels indicated by the information acquired by the acquisition unit, the width of the road on which the vehicle is traveling, including at least the width of the road pixels at a first position in the direction of travel of the vehicle and the width of the road pixels at a second position different from the first position in the direction of travel of the vehicle; and the road width calculation unit that calculates, based on the road pixels indicated by the information acquired by the acquisition unit, the width of the road at a third position different from the first and second positions in the direction of travel of the vehicle. The intersecting road detection device comprises: a second road width calculation unit that calculates the width of a pixel; a road width estimation unit that estimates the width of the road pixel at a third position based on the width of the road pixel at at least the first position and the width of the road pixel at the second position calculated by the first road width calculation unit, assuming that the intersecting road, which is a road that intersects the road on which the vehicle is traveling, is not included in the image; and a determination unit that determines that the intersecting road exists at the third position if the width of the road pixel at the third position calculated by the second road width calculation unit is greater than the width of the road pixel at the third position estimated by the road width estimation unit.

[0007] (2) In the intersecting road detection device of (1), the determination unit may determine that the intersecting road does not exist at the third position if the width of the road pixel at the third position calculated by the second road width calculation unit is equal to the width of the road pixel at the third position estimated by the road width estimation unit.

[0008] (3) In the intersecting road detection device of (1) or (2), the third position may be further away from the vehicle than the first and second positions.

[0009] (4) One aspect of the present disclosure is an acquisition step in which the intersecting road detection device acquires information indicating road pixels, which are pixels corresponding to roads included in an image, generated from an image obtained by photographing the area in front of the vehicle; a first road width calculation step in which the intersecting road detection device calculates, based on the information acquired in the acquisition step, the width of the road on which the vehicle is traveling, including at least the width of the road pixels at a first position in the direction of travel of the vehicle and the width of the road pixels at a second position different from the first position in the direction of travel of the vehicle; and the intersecting road detection device calculates, based on the information acquired in the acquisition step, the width of the road pixels at a third position different from the first and second positions in the direction of travel of the vehicle. The crossing road detection method comprises: a second road width calculation step of calculating the width of a road pixel; a road width estimation step in which the crossing road detection device estimates the width of the road pixel at a third position based on the width of the road pixel at at least the first position and the width of the road pixel at the second position calculated in the first road width calculation step, assuming that the crossing road that intersects the road on which the vehicle is traveling is not included in the image; and a determination step in which the crossing road detection device determines that the crossing road exists at the third position if the width of the road pixel at the third position calculated in the second road width calculation step is greater than the width of the road pixel at the third position estimated in the road width estimation step.

[0010] (5) One aspect of the present disclosure is a processor that provides information indicating road pixels, which are pixels corresponding to roads included in an image, generated from an image obtained by photographing the front of the vehicle; a first road width calculation step that calculates, based on the road pixels indicated by the information obtained in the acquisition step, the width of the road on which the vehicle is traveling, including at least the width of the road pixels at a first position in the direction of travel of the vehicle and the width of the road pixels at a second position different from the first position in the direction of travel of the vehicle; and the road width calculation step that calculates, based on the road pixels indicated by the information obtained in the acquisition step, the width of the road at a third position different from the first and second positions in the direction of travel of the vehicle. This program is for causing a vehicle to perform the following steps: a second road width calculation step of calculating the width of a pixel; a road width estimation step of estimating the width of a road pixel at a third position based on the width of the road pixel at at least the first position and the width of the road pixel at the second position calculated in the first road width calculation step, assuming that an intersecting road, which is a road that intersects the road the vehicle is traveling on, is not included in the image; and a determination step of determining that an intersecting road exists at the third position if the width of the road pixel at the third position calculated in the second road width calculation step is greater than the width of the road pixel at the third position estimated in the road width estimation step. [Effects of the Invention]

[0011] According to this disclosure, it is possible to appropriately detect intersecting roads that cross the road the vehicle is traveling on. [Brief explanation of the drawing]

[0012] [Figure 1] This figure shows an example of a vehicle 1 to which the intersecting road detection device 15 of the first embodiment is applied. [Figure 2] This figure shows a first example of the processing results of image segmentation performed by the image processing device 3A. [Figure 3] This figure shows a second example of the processing results of image segmentation performed by the image processing device 3A. [Figure 4] This figure shows a third example of the processing results of image segmentation performed by the image processing device 3A. [Figure 5] This is a flowchart illustrating an example of processing performed by the processor 153 of the intersecting road detection device 15 of the first embodiment. [Figure 6] This figure shows an example of a vehicle 1 to which the intersecting road detection device 15 of the second embodiment is applied. [Modes for carrying out the invention]

[0013] Hereinafter, embodiments of the intersecting road detection device, intersecting road detection method, and program of this disclosure will be described with reference to the drawings.

[0014] <First Embodiment> Figure 1 shows an example of a vehicle 1 to which the intersecting road detection device 15 of the first embodiment is applied. In the example shown in Figure 1, the vehicle 1 is equipped with a camera 11, an HMI (Human Machine Interface) 12, a vehicle state sensor 13, a surrounding situation sensor 14, an intersecting road detection device 15, a vehicle control device 16, a steering actuator 16A, a braking actuator 16B, and a drive actuator 16C. The camera 11 takes a picture of the area in front of the vehicle 1 and transmits the image data to the intersecting road detection device 15 and the vehicle control device 16. The HMI 12 has functions such as receiving various operations from the driver of the vehicle 1 and transmits signals indicating the driver's operations to the vehicle control device 16. The vehicle state sensor 13 detects the state of the vehicle 1 and transmits the detection result to the vehicle control device 16. The vehicle state sensor 13 includes, for example, a vehicle speed sensor. The surrounding situation sensor 14 detects, for example, obstacles, surrounding vehicles, pedestrians, etc., that exist around the vehicle 1 and transmits the detection result to the vehicle control device 16. The surrounding environment sensors 14 include, for example, cameras that photograph the sides and rear of the vehicle 1, LiDAR (Light Detection and Ranging), radar, sonar, etc. The intersecting road detection device 15 detects roads that intersect the road the vehicle 1 is traveling on (roads RD1, RD2, RD4 (see Figures 2 to 4)) (intersecting roads RD3, RD5 (see Figures 3 and 4)) and transmits the detection results to the vehicle control device 16. The vehicle control device 16 is composed of a driver assistance ECU (Electronic Control Unit). The vehicle control device 16 controls the steering actuator 16A, braking actuator 16B, and drive actuator 16C based on information (data, signals) transmitted from, for example, the camera 11, HMI 12, vehicle condition sensor 13, surrounding environment sensors 14, and intersecting road detection device 15.

[0015] The intersecting road detection device 15 is composed of a microcomputer equipped with a communication interface (I / F) 151, a memory 152, and a processor 153. The communication interface 151 has an interface circuit for connecting the intersecting road detection device 15 to the camera 11 and the vehicle control device 16. The memory 152 stores programs and various data used in processing performed by the processor 153. The processor 153 has the functions of an image processing device 3A, an acquisition unit 3B, a first road width calculation unit 3C, a second road width calculation unit 3D, a road width estimation unit 3E, and a determination unit 3F. The image processing device 3A acquires data of the image of the area in front of the vehicle transmitted from the camera 11. The image processing device 3A also performs image segmentation processing (identification of subjects included in the image), such as semantic segmentation, on the image of the area in front of the vehicle.

[0016] Figures 2 to 4 show examples of the processing results of image segmentation performed by the image processing device 3A. Specifically, Figure 2 shows the first example of the processing results of image segmentation performed by the image processing device 3A, Figure 3 shows the second example of the processing results of image segmentation performed by the image processing device 3A, and Figure 4 shows the third example of the processing results of image segmentation performed by the image processing device 3A. In the example shown in Figure 2, the processing result of image segmentation performed by the image processing device 3A on the image of the front of the vehicle captured by the camera 11 includes road pixels RP corresponding to the road on which the vehicle 1 is traveling (road RD1 on which the vehicle is traveling). In the example shown in Figure 3, the processing result of image segmentation performed by the image processing device 3A on the image of the front of the vehicle captured by the camera 11 includes road pixels RP corresponding to the road on which the vehicle 1 is traveling (road RD2 on which the vehicle is traveling) and the road that intersects road RD2 on which the vehicle is traveling (intersecting road RD3). In the example shown in Figure 4, the image segmentation processing result performed by the image processing device 3A on the image of the front of the vehicle captured by the camera 11 includes road pixels RP corresponding to the road on which the vehicle 1 is traveling (vehicle travel road RD4) and roads intersecting the vehicle travel road RD4 (intersecting road RD5), and wall pixels WP corresponding to a wall located adjacent to the vehicle travel road RD4.

[0017] In the example shown in Figure 1, the acquisition unit 3B acquires information indicating road pixels RP from the image segmentation processing results performed by the image processing device 3A. In other words, the acquisition unit 3B acquires information indicating pixels (road pixels RP) corresponding to roads RD1, RD2, RD3, RD4, and RD5 included in the image in front of the vehicle 1, which is generated from the image obtained by photographing the area in front of the vehicle 1 (vehicle front image). More specifically, in the example shown in Figure 2, the acquisition unit 3B acquires information indicating road pixels RP corresponding to road RD1 while the vehicle is traveling. In the example shown in Figure 3, the acquisition unit 3B acquires information indicating road pixels RP corresponding to road RD2 and intersecting road RD3 while the vehicle is traveling. In the example shown in Figure 4, the acquisition unit 3B acquires information indicating road pixels RP corresponding to road RD4 and intersecting road RD5 while the vehicle is traveling. In the example shown in Figure 1, the acquisition unit 3B also acquires information indicating wall pixels WP (see Figure 4) from the image segmentation processing results performed by the image processing device 3A.

[0018] In the example shown in FIG. 1, based on the road pixels RP indicated by the information acquired by the acquisition unit 3B, the first road width calculation unit 3C calculates the widths of the roads RD1, RD2, and RD4 during the travel of the host vehicle as the widths W1, W4, and W7 of the road pixels RP at the first positions P1, P4, and P7 in the traveling direction of the host vehicle 1. Also, based on the road pixels RP indicated by the information acquired by the acquisition unit 3B, the first road width calculation unit 3C calculates the widths of the roads RD1, RD2, and RD4 during the travel of the host vehicle as the widths W2, W5, and W8 of the road pixels RP at the second positions P2, P5, and P8 in the traveling direction of the host vehicle 1. Specifically, in the example shown in FIG. 2, based on the road pixels RP indicated by the information acquired by the acquisition unit 3B, the first road width calculation unit 3C calculates the width W1 of the road pixel RP at the first position P1 in the traveling direction of the host vehicle 1 indicated by the white arrow in FIG. 2 as the width of the road RD1 during the travel of the host vehicle. Also, based on the road pixels RP indicated by the information acquired by the acquisition unit 3B, the first road width calculation unit 3C calculates the width W2 of the road pixel RP at the second position P2 different from the first position P1 in the traveling direction of the host vehicle 1 as the width of the road RD1 during the travel of the host vehicle. In the example shown in FIG. 3, based on the road pixels RP indicated by the information acquired by the acquisition unit 3B, the first road width calculation unit 3C calculates the width W4 of the road pixel RP at the first position P4 in the traveling direction of the host vehicle 1 indicated by the white arrow in FIG. 3 as the width of the road RD2 during the travel of the host vehicle. Also, based on the road pixels RP indicated by the information acquired by the acquisition unit 3B, the first road width calculation unit 3C calculates the width W5 of the road pixel RP at the second position P5 different from the first position P4 in the traveling direction of the host vehicle 1 as the width of the road RD2 during the travel of the host vehicle. In the example shown in FIG. 4, based on the road pixels RP indicated by the information acquired by the acquisition unit 3B, the first road width calculation unit 3C calculates the width W7 of the road pixel RP at the first position P7 in the traveling direction of the host vehicle 1 indicated by the white arrow in FIG. 4 as the width of the road RD4 during the travel of the host vehicle. Also, based on the road pixels RP indicated by the information acquired by the acquisition unit 3B, the first road width calculation unit 3C calculates the width W8 of the road pixel RP at the second position P8 different from the first position P7 in the traveling direction of the host vehicle 1 as the width of the road RD4 during the travel of the host vehicle.

[0019] In the example shown in Figure 1, the second road width calculation unit 3D calculates the widths W3, W6, and W9 of the road pixel RP at third positions P3, P6, and P9, which are different from the first positions P1, P4, P7 and second positions P2, P5, P8 in the direction of travel of the vehicle 1, based on the road pixel RP indicated by the information acquired by the acquisition unit 3B. More specifically, in the example shown in Figure 2, the second road width calculation unit 3D calculates the width W3 of the road pixel RP at third position P3, which is different from the first position P1 and second position P2 in the direction of travel of the vehicle 1, as indicated by the white arrows in Figure 2, based on the road pixel RP indicated by the information acquired by the acquisition unit 3B. In the example shown in Figure 3, the second road width calculation unit 3D calculates the width W6 of the road pixel RP at third position P6, which is different from the first position P4 and second position P5 in the direction of travel of the vehicle 1, as indicated by the white arrows in Figure 3, based on the road pixel RP indicated by the information acquired by the acquisition unit 3B. In the example shown in Figure 4, the second road width calculation unit 3D calculates the width W9 of the road pixel RP at the third position P9, which is different from the first position P7 and second position P8 in the direction of travel of the vehicle 1, as indicated by the white arrows in Figure 4, based on the road pixel RP indicated by the information acquired by the acquisition unit 3B. In the examples shown in Figures 2 to 4, the third positions P3, P6, and P9 are further away from the vehicle 1 than the first positions P1, P4, P7 and the second positions P2, P5, and P8 (they are set on the upper side in Figures 2 to 4). Therefore, in the examples shown in Figures 2 to 4, the estimation accuracy of the widths W3E, W6E, and W9E of the road pixel RP at the third positions P3, P6, and P9 by the road width estimation unit 3E can be improved compared to when the third positions P3, P6, and P9 are set closer to the vehicle 1 than the first positions P1, P4, P7 and the second positions P2, P5, and P8.

[0020] In the example shown in Figure 1, the road width estimation unit 3E assumes that roads intersecting roads RD1, RD2, and RD4 while the vehicle is traveling (intersecting roads RD3 and RD5) are not included in the forward-facing image of the vehicle captured by the camera 11, and estimates the widths W3E, W6E, and W9E of the road pixel RP at the third position P3, P6, and P9 based on the widths W1, W4, and W7 of the road pixel RP at the first position P1, P4, and P7 calculated by the first road width calculation unit 3C and the widths W2, W5, and W8 of the road pixel RP at the second position P2, P5, and P8. Specifically, in the example shown in Figure 2, the road width estimation unit 3E estimates the width W3E of the road pixel RP at the third position P3 based on the width W1 of the road pixel RP at the first position P1 and the width W2 of the road pixel RP at the second position P2, calculated by the first road width calculation unit 3C, assuming that roads intersecting road RD1 while the vehicle is traveling are not included in the forward image of the vehicle captured by the camera 11. More specifically, in the example shown in Figure 2, the road width estimation unit 3E estimates the coordinates of the left end of the road pixel RP at the third position P3 by extending a line segment connecting the left end of the road pixel RP at the first position P1 and the left end of the road pixel RP at the second position P2 to the third position P3. The road width estimation unit 3E also estimates the coordinates of the right end of the road pixel RP at the third position P3 by extending a line segment connecting the right end of the road pixel RP at the first position P1 and the right end of the road pixel RP at the second position P2 to the third position P3. Furthermore, the road width estimation unit 3E calculates the width W3E of the road pixel RP at the third position P3 as the distance between the coordinates of the left edge of the road pixel RP at the estimated third position P3 and the coordinates of the right edge of the road pixel RP at the estimated third position P3. In the example shown in Figure 3, the road width estimation unit 3E assumes that the road intersecting the road RD2 while the vehicle is traveling (intersecting road RD3) is not included in the forward image of the vehicle captured by the camera 11, and estimates the width W6E of the road pixel RP at the third position P6 based on the width W4 of the road pixel RP at the first position P4 and the width W5 of the road pixel RP at the second position P5 calculated by the first road width calculation unit 3C.Specifically, in the example shown in FIG. 3, the road width estimation unit 3E estimates the coordinates of the left end of the road pixel RP at the third position P6 by extending the line segment connecting the left end of the road pixel RP at the first position P4 and the left end of the road pixel RP at the second position P5 to the third position P6. Further, the road width estimation unit 3E estimates the coordinates of the right end of the road pixel RP at the third position P6 by extending the line segment connecting the right end of the road pixel RP at the first position P4 and the right end of the road pixel RP at the second position P5 to the third position P6. Furthermore, the road width estimation unit 3E calculates the distance between the estimated coordinates of the left end of the road pixel RP at the third position P6 and the estimated coordinates of the right end of the road pixel RP at the third position P6 as the width W6E of the road pixel RP at the third position P6.

[0021] In the example shown in Figure 4, the road width estimation unit 3E assumes that the road intersecting road RD4 while the vehicle is traveling (intersecting road RD5) is not included in the forward-facing image of the vehicle captured by the camera 11, and estimates the width W9E of the road pixel RP at the third position P9 based on the width W7 of the road pixel RP at the first position P7 and the width W8 of the road pixel RP at the second position P8, which are calculated by the first road width calculation unit 3C. More specifically, in the example shown in Figure 4, the road width estimation unit 3E estimates the coordinates of the left end of the road pixel RP at the third position P9 by extending a line segment connecting the left end of the road pixel RP at the first position P7 and the left end of the road pixel RP at the second position P8 to the third position P9. The road width estimation unit 3E also estimates the coordinates of the right end of the road pixel RP at the third position P9 by extending a line segment connecting the right end of the road pixel RP at the first position P7 and the right end of the road pixel RP at the second position P8 to the third position P9. Furthermore, the road width estimation unit 3E calculates the distance between the coordinates of the left edge of the road pixel RP at the estimated third position P9 and the coordinates of the right edge of the road pixel RP at the estimated third position P9 as the width W9E of the road pixel RP at the third position P9. As described above, in the example shown in Figure 1, the first road width calculation unit 3C calculates the widths W1, W4, W7 of the road pixel RP at the first positions P1, P4, P7 in the direction of travel of the vehicle 1, and the widths W2, W5, W8 of the road pixel RP at the second positions P2, P5, P8 in the direction of travel of the vehicle 1, as the widths of the roads RD1, RD2, RD4 while the vehicle is traveling (that is, it calculates the widths of the road pixel RP at two positions in the direction of travel of the vehicle 1). Furthermore, the road width estimation unit 3E estimates the widths W3E, W6E, and W9E of road pixels RP at third positions P3, P6, and P9 based on the widths of road pixels RP at two positions in the direction of travel of the vehicle 1 calculated by the first road width calculation unit 3C. In another example, the first road width calculation unit 3C may calculate the widths of road pixels RP at three or more positions in the direction of travel of the vehicle 1 as the width of the road while the vehicle is traveling, and the road width estimation unit 3E may estimate the widths of road pixels RP at positions different from those three or more positions in the direction of travel of the vehicle 1 based on the widths of road pixels RP at three or more positions in the direction of travel of the vehicle 1 calculated by the first road width calculation unit 3C.

[0022] In the example shown in Figure 1, the determination unit 3F determines that intersecting roads RD3 and RD5 exist at the third positions P6 and P9 if the widths W6 and W9 of the road pixels RP at the third positions P6 and P9, calculated by the second road width calculation unit 3D, are greater than the widths W6E and W9E of the road pixels RP at the third positions P6 and P9, estimated by the road width estimation unit 3E. The determination unit 3F also determines that there are no intersecting roads at the third position P3 if the width W3 of the road pixels RP at the third position P3, calculated by the second road width calculation unit 3D, is equal to the width W3E of the road pixels RP at the third position P3, estimated by the road width estimation unit 3E. Specifically, in the example shown in Figure 2, the width W3 of the road pixel RP at the third position P3 calculated by the second road width calculation unit 3D is equal to the width W3E of the road pixel RP at the third position P3 estimated by the road width estimation unit 3E, so the determination unit 3F determines that there is no intersecting road at the third position P3. In the example shown in Figure 3, the width W6 of the road pixel RP at the third position P6 calculated by the second road width calculation unit 3D is greater than the width W6E of the road pixel RP at the third position P6 estimated by the road width estimation unit 3E, so the determination unit 3F determines that there is an intersecting road RD3 at the third position P6. In the example shown in Figure 4, the width W9 of the road pixel RP at the third position P9 calculated by the second road width calculation unit 3D is greater than the width W9E of the road pixel RP at the third position P9 estimated by the road width estimation unit 3E, so the determination unit 3F determines that there is an intersecting road RD5 at the third position P9.

[0023] In the example shown in Figure 1, when the determination unit 3F determines that intersecting roads RD3 and RD5 are located ahead of the vehicle 1 in the direction of travel (third positions P6 and P9), the vehicle control device 16 (driving support ECU) outputs a warning to the HMI 12, for example by display or sound, indicating that intersecting roads RD3 and RD5 are located ahead of the vehicle 1 in the direction of travel, based on the detection result of the intersecting road detection device 15 (determination result of the determination unit 3F). In another example, when the determination unit 3F determines that intersecting roads RD3 and RD5 are located ahead of the vehicle 1 in the direction of travel (third positions P6 and P9), the vehicle control device 16 (driving support ECU) may perform braking and / or steering assistance to avoid a collision between the vehicle 1 and other vehicles, pedestrians, etc. that suddenly emerge from intersecting roads RD3 and RD5 onto roads RD2 and RD4 while the vehicle is traveling.

[0024] Figure 5 is a flowchart illustrating an example of processing performed by the processor 153 of the intersecting road detection device 15 of the first embodiment. In the example shown in Figure 5, in step S10, the acquisition unit 3B acquires information indicating road pixels RP. In step S11, the first road width calculation unit 3C calculates the widths of road pixels RP at first positions P1, P4, P7 in the direction of travel of the vehicle 1 and the widths of road pixels RP at second positions P2, P5, P8 in the direction of travel of the vehicle 1, as the widths of roads RD1, RD2, RD4 while the vehicle is traveling, based on the road pixels RP indicated by the information acquired in step S10. In step S12, the second road width calculation unit 3D calculates the widths of road pixels RP at third positions P3, P6, P9, which are different from the first positions P1, P4, P7 and second positions P2, P5, P8 in the direction of travel of the vehicle 1, based on the road pixels RP indicated by the information acquired in step S10. In step S13, the road width estimation unit 3E assumes that the intersecting roads RD3 and RD5 are not included in the image in front of the vehicle, and estimates the widths W3E, W6E, and W9E of the road pixel RP at the third positions P3, P6, and P9 based on the widths W1, W4, and W7 of the road pixel RP at the first positions P1, P4, and P7 calculated in step S11, and the widths W2, W5, and W8 of the road pixel RP at the second positions P2, P5, and P8.

[0025] In step S14, the determination unit 3F determines whether the widths W3, W6, and W9 of the road pixels RP at the third positions P3, P6, and P9 calculated in step S12 are greater than the widths W3E, W6E, and W9E of the road pixels RP at the third positions P3, P6, and P9 estimated in step S13. If the answer is YES, the process proceeds to step S15; otherwise, the process proceeds to step S16. In step S15, the determination unit 3F determines that intersecting roads RD3 and RD5 exist at the third positions P6 and P9. In step S16, the determination unit 3F determines that there are no intersecting roads at the third position P3.

[0026] If the width of roads RD1, RD2, and RD4 remains constant from the foreground to the background while the vehicle is traveling, then the apparent width of roads RD1, RD2, and RD4 in the image in front of the vehicle should decrease as the vehicle moves further into the distance. On the other hand, if intersecting roads RD3 and RD5 exist, at the depth position where intersecting roads RD3 and RD5 exist (third position P6, P9), the road width (width of road pixel RP W5, W8) becomes larger compared to the depth position before the third position P6, P9 (second position P5, P8) because the parts of intersecting roads RD3 and RD5 that extend to the left and right are added to the image in front of the vehicle. Based on this, the intersecting road detection device 15 of the first embodiment measures (calculates) the road width (widths W1 to W9 of road pixels RP) on the image in front of the vehicle at each depth position (first position P1, P4, P7, second position P2, P5, P8, and third position P3, P6, P9) in the direction from near to far on the image in front of the vehicle, and detects the parts (third position P6, P9) where the road width (widths W6, W9 of road pixels RP) on the image in front of the vehicle in the depth direction is not narrower than the road width (widths W6E, W9E of road pixels RP) that can be assumed (estimated) from the road width (widths W4, W5, W7, W8) of the earlier depth position (first position P4, P7, and second position P5, P8), as places where there is traffic intersecting with roads RD2, RD4 while the vehicle is traveling (intersecting roads RD3, RD5). In the intersecting road detection device 15 of the first embodiment, in order to detect road pixels RP included in the image in front of the vehicle (that is, to detect the attributes of the pixels), for example, "pixel-specific labels" calculated (output) in semantic segmentation processing (processing in the image processing device 3A) which assigns attributes to each of the multiple pixels included in the image are used.

[0027] <Second Embodiment> The vehicle 1 to which the intersecting road detection device 15 of the second embodiment is applied is configured in the same way as the vehicle 1 to which the intersecting road detection device 15 of the first embodiment is applied, except for the points described later.

[0028] Figure 6 shows an example of a vehicle 1 to which the intersection road detection device 15 of the second embodiment is applied. As described above, in the example of a vehicle 1 to which the intersection road detection device 15 of the first embodiment is applied (the example shown in Figure 1), the processor 153 of the intersection road detection device 15 functions as an image processing device 3A. On the other hand, in the example shown in Figure 6, the processor 153 of the intersection road detection device 15 does not function as an image processing device 3A, and an image processing device 17 having the same function as the image processing device 3A is provided outside the intersection road detection device 15. In other words, in the example shown in Figure 6, the vehicle 1 is equipped with an image processing device 17 separately from the intersection road detection device 15.

[0029] In the example shown in Figure 6, camera 11 transmits data of the image in front of the vehicle to vehicle control device 16 and image processing device 17. Image processing device 17 acquires the data of the image in front of the vehicle transmitted from camera 11. Image processing device 17 also performs image segmentation processing (identification of subjects included in the image), such as semantic segmentation, on the image in front of the vehicle. Acquisition unit 3B acquires information indicating road pixels RP from the image segmentation processing results performed by image processing device 17.

[0030] <Third Embodiment> The vehicle 1 to which the intersecting road detection device 15 of the third embodiment is applied is configured in the same way as the vehicle 1 to which the intersecting road detection device 15 of the first embodiment is applied, except for the points described later.

[0031] As described above, in an example of the vehicle 1 to which the intersection road detection device 15 of the first embodiment is applied (the example shown in Figure 1), the vehicle control device 16 is configured by a driver assistance ECU. On the other hand, in an example of the vehicle 1 to which the intersection road detection device 15 of the third embodiment is applied, the vehicle control device 16 is configured by an automatic driving ECU.

[0032] As described above, in an example of the vehicle 1 to which the intersecting road detection device 15 of the first embodiment is applied (the example shown in Figure 1), when the determination unit 3F determines that intersecting roads RD3 and RD5 are located ahead of the vehicle 1 in the direction of travel (third positions P6 and P9), the vehicle control device 16 (driving support ECU) outputs a warning to the HMI 12, for example by display, voice, etc., indicating that intersecting roads RD3 and RD5 are located ahead of the vehicle 1 in the direction of travel, based on the detection result of the intersecting road detection device 15 (determination result of the determination unit 3F). On the other hand, in an example of the vehicle 1 to which the intersecting road detection device 15 of the third embodiment is applied, when the determination unit 3F determines that intersecting roads RD3 and RD5 are located ahead of the vehicle 1 in the direction of travel (third positions P6 and P9), the vehicle control device 16 (autonomous driving ECU) generates a driving plan for the vehicle 1 to avoid collisions with other vehicles, pedestrians, etc. that suddenly appear from the intersecting roads RD3 and RD5 onto the roads RD2 and RD4 that the vehicle is traveling on, and controls the steering actuator 16A, braking actuator 16B, and drive actuator 16C based on that driving plan.

[0033] <Fourth Embodiment> The vehicle 1 to which the intersection road detection device 15 of the fourth embodiment is applied is configured in the same way as the vehicle 1 to which the intersection road detection device 15 of the second embodiment is applied, except for the points described later.

[0034] As described above, in an example of the vehicle 1 to which the intersection road detection device 15 of the second embodiment is applied (the example shown in Figure 6), the vehicle control device 16 is configured by a driver assistance ECU. On the other hand, in an example of the vehicle 1 to which the intersection road detection device 15 of the fourth embodiment is applied, the vehicle control device 16 is configured by an automatic driving ECU. In the example of the vehicle 1 to which the intersection road detection device 15 of the fourth embodiment is applied, similar to the example of the vehicle 1 to which the intersection road detection device 15 of the third embodiment is applied described above, when the determination unit 3F determines that intersection roads RD3 and RD5 are located in front of the vehicle 1 in the direction of travel (third positions P6 and P9), the vehicle control device 16 (automatic driving ECU) generates a driving plan for the vehicle 1 to avoid collisions with other vehicles, pedestrians, etc. that suddenly appear from intersection roads RD3 and RD5 onto roads RD2 and RD4 on which the vehicle is traveling, and controls the steering actuator 16A, braking actuator 16B, and drive actuator 16C based on that driving plan.

[0035] As described above, embodiments of the crossroads detection device, crossroads detection method, and program of this disclosure have been explained with reference to the drawings. However, the crossroads detection device, crossroads detection method, and program of this disclosure are not limited to the embodiments described above, and can be modified as appropriate without departing from the spirit of this disclosure. The configurations of each example of the embodiments described above may be combined as appropriate. In each example of the embodiments described above, the processing performed by the crossroads detection device 15 was described as software processing performed by executing a program. However, the processing performed by the crossroads detection device 15 may be hardware processing. Alternatively, the processing performed by the crossroads detection device 15 may be a combination of both software and hardware processing. Furthermore, the program stored in the memory 152 of the crossroads detection device 15 (the program that realizes the functions of the processor 153 of the crossroads detection device 15) may be recorded on a computer-readable storage medium such as a semiconductor memory, magnetic recording medium, or optical recording medium and provided and distributed. [Explanation of symbols]

[0036] 1...Vehicle, 11...Camera, 12...HMI, 13...Vehicle status sensor, 14...Surroundings sensor, 15...Intersecting road detection device, 151...Communication interface, 152...Memory, 153...Processor, 3A...Image processing device, 3B...Acquisition unit, 3C...First road width calculation unit, 3D...Second road width calculation unit, 3E...Road width estimation unit, 3F...Determination unit, 16...Vehicle control device, 16A...Steering actuator, 16B...Brake actuator, 16C...Drive actuator, 17...Image processing device

Claims

1. An acquisition unit that acquires information indicating road pixels, which are pixels corresponding to roads included in an image, generated from an image obtained by photographing the front of the vehicle, A first road width calculation unit calculates, based on the road pixels indicated by the information acquired by the acquisition unit, the width of the road on which the vehicle is traveling, including at least the width of the road pixels at a first position in the direction of travel of the vehicle and the width of the road pixels at a second position different from the first position in the direction of travel of the vehicle. A second road width calculation unit calculates the width of the road pixels at a third position different from the first and second positions in the direction of travel of the vehicle, based on the road pixels indicated by the information acquired by the acquisition unit, Assuming that the intersecting roads that cross the road on which the vehicle is traveling are not included in the image, a road width estimation unit estimates the width of the road pixel at the third position based on the width of the road pixel at at least the first position and the width of the road pixel at the second position calculated by the first road width calculation unit, A crossing road detection device comprising: a determination unit that determines that the crossing road exists at the third position if the width of the road pixel at the third position calculated by the second road width calculation unit is greater than the width of the road pixel at the third position estimated by the road width estimation unit.

2. The intersection road detection device according to claim 1, wherein the determination unit determines that the intersection road does not exist at the third position if the width of the road pixel at the third position calculated by the second road width calculation unit is equal to the width of the road pixel at the third position estimated by the road width estimation unit.

3. The crossing road detection device according to claim 1, wherein the third position is further away from the vehicle than the first and second positions.

4. The intersecting road detection device acquires information indicating road pixels, which are pixels corresponding to roads included in an image, generated from an image obtained by photographing the area in front of the vehicle. The intersecting road detection device performs a first road width calculation step, which calculates, based on the road pixels indicated by the information acquired in the acquisition step, the width of the road on which the vehicle is traveling, including at least the width of the road pixels at a first position in the direction of travel of the vehicle and the width of the road pixels at a second position different from the first position in the direction of travel of the vehicle. The intersecting road detection device performs a second road width calculation step, which calculates the width of the road pixels at a third position different from the first and second positions in the direction of travel of the vehicle, based on the road pixels indicated by the information acquired in the acquisition step. The intersecting road detection device performs a road width estimation step in which it estimates the width of the road pixel at a third position based on the width of the road pixel at at least a first position and the width of the road pixel at a second position, assuming that the intersecting road that intersects the road on which the vehicle is traveling is not included in the image, A method for detecting an intersecting road, comprising: a determination step in which the intersecting road detection device determines that the intersecting road exists at the third position if the width of the road pixel at the third position calculated in the second road width calculation step is greater than the width of the road pixel at the third position estimated in the road width estimation step.

5. In the processor, An acquisition step of acquiring information indicating road pixels, which are pixels corresponding to roads included in an image, generated from an image obtained by photographing the front of the vehicle, A first road width calculation step, based on the road pixels indicated by the information acquired in the acquisition step, calculates the width of the road on which the vehicle is traveling, including at least the width of the road pixels at a first position in the direction of travel of the vehicle and the width of the road pixels at a second position different from the first position in the direction of travel of the vehicle. A second road width calculation step, based on the road pixels indicated by the information acquired in the acquisition step, calculates the width of the road pixels at a third position different from the first and second positions in the direction of travel of the vehicle, Assuming that the image does not include any intersecting roads that cross the road on which the vehicle is traveling, the road width estimation step estimates the width of the road pixel at the third position based on the width of the road pixel at at least the first position and the width of the road pixel at the second position calculated in the first road width calculation step, A program for executing a determination step which determines that the intersecting road exists at the third location if the width of the road pixel at the third location calculated in the second road width calculation step is greater than the width of the road pixel at the third location estimated in the road width estimation step.

Images