Driving support device

Abstract

To provide a technique capable of accurately detecting a retreat region.SOLUTION: A drive support apparatus 110 mounted on a vehicle 10 having a camera 122 comprises: a recognition unit 111 which recognizes a road end and a lane division line closest to the road end on the basis of a peripheral image captured by the camera by using a low freedom degree model and a high freedom degree model and generates road end information indicating the road end and lane division line information indicating the lane division line; a detection unit 112 which detects a retreat region to which a vehicle existing between the road end and the lane division line can retreat by using the road end information and lane division line information; and a control unit 113 which performs control of the vehicle when the vehicle retreats to the retreat region. The road end information includes low freedom degree road end information recognized by using the low freedom degree model and high freedom degree road end information recognized by using the high freedom degree model. The detection unit uses the high freedom degree road end information when a first condition is satisfied and uses the low freedom degree road end information when the first condition is not satisfied but a second condition is satisfied.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present disclosure relates to a driving support device.

Background Art

[0002] As a driving support device, a technique for retreating a vehicle to the road shoulder is known. Patent Document 1 describes a technique for detecting a retreat area for retreating using road information including the shape of a road recognized by a camera or a radar.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In order to recognize a road with a complex shape such as a parking lot using road information recognized by a camera or a radar, a model with a high degree of freedom in shape expression may be used to recognize the road shape. In this case, the inventors have found a problem that there is a risk of an increase in false detection in the recognition of a road with a simple shape.

Means for Solving the Problems

[0005] The present disclosure has been made to solve the above problems and can be realized in the following forms.

[0006] According to one aspect of the present disclosure, there is provided a driving assistance device (110) mounted on a vehicle (10) having a camera (122). The vehicle providing device uses a low-degree-of-freedom model, which is a model with a low degree of freedom in shape representation, and a high-degree-of-freedom model, which is a model with a higher degree of freedom in shape representation than the low-degree-of-freedom model, to recognize, based on a peripheral image that is an image captured by the camera of the periphery of the vehicle, a road edge that is an edge of the road on which the vehicle is traveling and that is the closest to the vehicle, and a lane line closest to the road edge, and generates peripheral information including road edge information indicating the road edge and lane line information indicating the lane line. A recognition unit (111), a detection unit (112) that uses the road edge information and the lane line information to detect a retreat region that is a region existing between the road edge and the lane line and in which the vehicle can retreat, and a control unit (113) that performs predetermined control of the vehicle according to the detected retreat region when the vehicle retreats to the retreat region. The retreat region is a region in which the length in the extending direction of the road is equal to or greater than the vehicle length of the vehicle, and the width, which is the length in a direction perpendicular to the extending direction, is equal to or greater than the vehicle width of the vehicle. The road edge information includes low-degree-of-freedom road edge information representing the road edge recognized using the low-degree-of-freedom model and high-degree-of-freedom road edge information representing the road edge recognized using the high-degree-of-freedom model. When the peripheral information satisfies a first condition indicating that the accuracy of the low-degree-of-freedom road edge information has decreased or is likely to decrease, the detection unit uses the high-degree-of-freedom road edge information as the road edge information. When the first condition is not satisfied and the peripheral information satisfies a second condition regarding the accuracy of the high-degree-of-freedom road edge information, the detection unit uses the low-degree-of-freedom road edge information as the road edge information.

[0007] When the shape of the road edge is complex, the shape of the road edge can be recognized with higher accuracy using a high-degree-of-freedom model than using a low-degree-of-freedom model. However, when the shape of the road edge is a simple straight line shape, the shape of the road edge can be recognized with higher accuracy using a low-degree-of-freedom model than using a high-degree-of-freedom model. According to the driving assistance device of this aspect, since the low-degree-of-freedom model and the high-degree-of-freedom model can be used appropriately, the retreat region can be detected with high accuracy.

Brief Description of the Drawings

[0008]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Modes for Carrying Out the Invention

[0009] A. First Embodiment: As shown in FIG. 1, the vehicle 10 includes an automatic driving control system 100. In this embodiment, the automatic driving control system 100 executes the automatic driving of the vehicle 10. In this embodiment, the automatic driving control system 100 includes a driving support device 110, a camera 122, an automatic driving control unit 210, a driving force control ECU (Electronic Control Unit) 220, a braking force control ECU 230, and a steering control ECU 240. The driving support device 110, the driving force control ECU 220, the braking force control ECU 230, and the steering control ECU 240 are connected via an in-vehicle network 250. Note that the vehicle 10 may be driven not only by automatic driving but also by manual driving performed by a driver.

[0010] The camera 122 captures the surroundings of the vehicle 10 to acquire an image. The image captured by the camera 122 is output to the driving support device 110.

[0011] The driving support device 110 is a device that causes the vehicle 10 to retreat to a retreat area. The driving support device 110 includes a recognition unit 111, a detection unit 112, and a control unit 113. The driving support device 110 is composed of a microcomputer including a central processing unit (CPU), a RAM, and a ROM, etc., and the microcomputer executes a pre-installed program to realize the functions of these respective units. However, part or all of the functions of these respective units may be realized by a hardware circuit.

[0012] The recognition unit 111 generates surrounding information based on a surrounding image obtained by imaging the surroundings of the vehicle 10 by the camera 122. The surrounding information is information including a road edge indicating an edge of the road on which the vehicle 10 is traveling and a lane marking on the road edge side. In the present embodiment, the road edge is the edge closest to the vehicle 10 and is the edge on the opposite side of the oncoming lane with respect to the lane in which the vehicle 10 is traveling. The road edge is indicated by, for example, a boundary line between a roadway and a sidewalk or a roadside object.

[0013] The recognition unit 111 recognizes the shape of the road edge using a low-degree-of-freedom model which is a model with a low degree of freedom in shape representation and a high-degree-of-freedom model which is a model with a higher degree of freedom in shape representation than the low-degree-of-freedom model. The higher the degree of freedom in shape representation, the more the shape is represented by a high-dimensional polynomial. The low-degree-of-freedom model recognizes the shape of the road edge by, for example, a quadratic curve or a cubic curve. The high-degree-of-freedom model recognizes the shape of the road edge by, for example, a quartic curve or Occupancy Grid Mapping (OGM).

[0014] In the present embodiment, the recognition unit 111 recognizes a point sequence representing the shape of the road edge from a detection point which is a point indicating the road edge detected in the surrounding image, using the low-degree-of-freedom model or the high-degree-of-freedom model.

[0015] The information indicating the road edge recognized by the recognition unit 111 is referred to as road edge information, and the information indicating the lane line is referred to as lane line information. The road edge information includes low-degree-of-freedom road edge information recognized using a low-degree-of-freedom model and high-degree-of-freedom road edge information recognized using a high-degree-of-freedom model.

[0016] The detection unit 112 detects an evacuation area using the road edge information and the lane line information recognized by the recognition unit 111. The evacuation area is an area that exists between the road edge and the lane line and is an area where the vehicle 10 can evacuate. Also, the evacuation area is an area where the vertical width, which is the length in the extending direction of the road, is equal to or greater than the vehicle length of the vehicle 10, and the horizontal width, which is the length in the direction perpendicular to the extending direction, is equal to or greater than the vehicle width of the vehicle 10. Here, the horizontal width can also be said to be the length in the direction orthogonal to both the extending direction of the road and the gravitational direction.

[0017] Also, when the surrounding information satisfies a predetermined first condition, the detection unit 112 uses the high-degree-of-freedom road edge information as the road edge information, and when the surrounding information does not satisfy the first condition and satisfies a second condition, the detection unit 112 uses the low-degree-of-freedom road edge information as the road edge information. The first condition is information regarding the accuracy of the low-degree-of-freedom road edge information. As the first condition, for example, any one or more of the following conditions can be adopted.

[0018] <Condition 1> The accuracy of the low-degree-of-freedom road edge information has decreased. <Condition 2> There is a high possibility that the accuracy of the low-degree-of-freedom road edge information will decrease.

[0019] The above-mentioned condition 1 is determined to be established, for example, when the distance between the point sequence recognized using the low-degree-of-freedom model and the detection points corresponding to the point sequence is equal to or greater than a predetermined threshold distance. The distance between the point sequence and the detection points can be, for example, the distance between one point in the point sequence and the corresponding detection point in the direction perpendicular to the extending direction of the road on which the vehicle 10 is traveling, or the average value of the distances between a plurality of points in the point sequence and the corresponding detection points.

[0020] Condition 2 described above is determined to be satisfied, for example, when a sign indicating a no-parking zone is detected in the surrounding image. A no-parking zone is, for example, an area established on the shoulder of an expressway for the purpose of parking by vehicles such as breakdown vehicles, emergency vehicles, and road management vehicles.

[0021] Also, the above-mentioned Condition 1, Condition 2, and other conditions can be appropriately combined to form the first condition. In this embodiment, the detection unit 112 adopts the above-mentioned Condition 1 and Condition 2, and when either of these Condition 1 or Condition 2 is satisfied, it is determined that the first condition is satisfied.

[0022] The second condition is a condition different from the first condition. The second condition is information regarding the accuracy of high-degree-of-freedom road edge information. The second condition is determined to be satisfied when there is a possibility that the accuracy of high-degree-of-freedom road edge information is lower than that of low-degree-of-freedom road edge information, for example, when the shape of the road edge indicated by the low-degree-of-freedom road edge information is a straight line. Note that the second condition may be a condition that is always satisfied during the operation of the automatic driving control system 100. That is, when the surrounding conditions do not satisfy the first condition, the detection unit 112 may always use the low-degree-of-freedom road edge information as the road edge information.

[0023] When the vehicle 10 retreats to the evacuation area, the control unit 113 performs predetermined control of the vehicle 10 according to the evacuation area detected by the detection unit 112. When the vehicle 10 evacuates to the evacuation area, for example, it is when the driver of the vehicle 10 loses consciousness, when the sensor of the vehicle 10 fails, or when the driver requests evacuation. The case where the sensor of the vehicle 10 fails is, for example, when an abnormality of some sensors used for automatic driving is detected during automatic driving such as lane keeping, and it becomes difficult to continue automatic driving. In this embodiment, the control unit 113 evacuates the vehicle 10 to the evacuation area as the predetermined control. Note that the predetermined control does not always mean the same control, but means control determined according to the position of the evacuation area.

[0024] The automatic driving control unit 210 consists of a microcomputer composed of a central processing unit (CPU), a RAM, a ROM, etc. The automatic driving function is realized by the microcomputer executing a pre-installed program. The automatic driving control unit 210 realizes the automatic driving function by controlling the driving force control ECU 220, the braking force control ECU 230, and the steering control ECU 240. For example, the automatic driving control unit 210 controls the driving force control ECU 220 and the braking force control ECU 230, and automatically changes lanes using the steering control ECU 240.

[0025] The driving force control ECU 220 is an electronic control unit that controls an actuator that generates the driving force of the vehicle 10 such as an engine. When the driver manually drives, the driving force control ECU 220 controls a power source such as an engine or an electric motor according to the operation amount of the accelerator pedal. On the other hand, when performing automatic driving, the driving force control ECU 220 controls the power source according to the required driving force calculated by the control unit 113.

[0026] The braking force control ECU 230 is an electronic control unit that controls a brake actuator that generates the braking force of the vehicle 10. When the driver manually drives, the braking force control ECU 230 controls the brake actuator according to the operation amount of the brake pedal. On the other hand, when performing automatic driving, the braking force control ECU 230 controls the brake actuator according to the required braking force calculated by the control unit 113.

[0027] The steering control ECU 240 is an electronic control unit that controls a motor that generates the steering torque of the vehicle 10. When the driver manually drives, the steering control ECU 240 controls the motor according to the operation of the steering wheel to generate an assist torque for the steering operation. Thereby, the driver can operate the steering with a small amount of force, and the steering of the vehicle 10 is realized. On the other hand, when performing automatic driving, the steering control ECU 240 performs steering by controlling the motor according to the required steering angle calculated by the control unit 113.

[0028] The evacuation process shown in FIG. 2 is a process in which the driving support device 110 evacuates the vehicle 10 to an evacuation area. This process is a process performed when it is necessary to evacuate the vehicle 10 in the present embodiment.

[0029] In step S100, the recognition unit 111 recognizes the road edge and the lane line. More specifically, it generates surrounding information including road edge information and lane line information.

[0030] As shown in FIG. 3, the vehicle 10 is traveling in the lane Ln1. The recognition unit 111 recognizes the road edge LnE that is the closest road edge to the vehicle 10 and the lane line LnL that is the closest lane line to the road edge LnE. More specifically, the recognition unit 111 recognizes the shape of the road edge LnE indicated by the low-degree-of-freedom road edge information (hereinafter also referred to as "low-degree-of-freedom road edge information ELl"), the shape of the road edge LnE indicated by the high-degree-of-freedom road edge information (hereinafter also referred to as "high-degree-of-freedom road edge information ELh"), and the shape of the lane line LnL indicated by the lane line information (hereinafter also referred to as "lane line information LL").

[0031] In step S110 (see FIG. 2), the detection unit 112 detects an evacuation area using the surrounding information generated by the recognition unit 111 in step S100.

[0032] The road edge information determination process shown in FIG. 4 is a process in which the detection unit 112 determines the road edge information used when detecting an evacuation area in step S110 (see FIG. 2). First, in step S200, the detection unit 112 determines whether the surrounding information satisfies the first condition. If the surrounding information satisfies the first condition, the detection unit 112 proceeds to the process of step S210 and determines to use the high-degree-of-freedom road edge information as the road edge information. On the other hand, if the surrounding information does not satisfy the first condition, the detection unit 112 proceeds to the process of step S205.

[0033] In step S205, the detection unit 112 determines whether the surrounding information satisfies the second condition. When the surrounding information satisfies the second condition, the detection unit 112 proceeds to the process of step S215 and determines to use the low-degree-of-freedom road edge information as the road edge information. On the other hand, when the surrounding information does not satisfy the second condition, the detection unit 112 proceeds to the process of step S210.

[0034] In the present embodiment, since the sign S1 indicating the no-parking zone is detected in the surrounding image, that is, since the surrounding information satisfies the first condition, the detection unit 112 uses the high-degree-of-freedom road edge information as the road edge information.

[0035] As shown in FIG. 5, the detection unit 112 detects the evacuation area A1 using the lane line information LL and the high-degree-of-freedom road edge information ELh. In the present embodiment, the detection unit 112 detects the evacuation area A1 such that the distance Lm between the peripheral edge on the road edge LnE side of the evacuation area A1 and the high-degree-of-freedom road edge information ELh is equal to or greater than a predetermined surplus distance. The distance Lm is, for example, the smallest distance among the distances between the peripheral edge on the road edge LnE side of the evacuation area A1 and the high-degree-of-freedom road edge information ELh in the direction perpendicular to the extending direction of the lane Ln1. The distance Lm is not limited to this, and may also be the average value of the distances between the peripheral edge on the road edge LnE side of the evacuation area A1 and the high-degree-of-freedom road edge information ELh in the direction perpendicular to the extending direction of the lane Ln1. Further, the detection unit 112 detects the evacuation area A1 such that the peripheral edge on the lane line LnL side of the evacuation area A1 overlaps with the lane line information LL. That is, the detection unit 112 detects the evacuation area A1 such that the distance between the peripheral edge on the lane line LnL side of the evacuation area A1 and the lane line information LL becomes 0 m.

[0036] In the present embodiment, the detection unit 112 detects a retreat region A1 from a target region TA1 which is a region where the distance Lw between the road edge LnE and the lane line LnL is wider than the vehicle width and the length Lh in the extending direction of the road is longer than the vehicle length. For the sake of illustration, the target region TA1 is described in an extended manner. Actually, the peripheral edge in the extending direction in the target region TA1 and the peripheral edge in the direction perpendicular to the extending direction in the target region TA1 and close to the vehicle 10 overlap with the peripheral edge of the retreat region A1. The detection unit 112 detects, as the retreat region A1, a region in the target region TA1 that is closest to the vehicle 10, has a length in the extending direction of distance h1, and has a lateral width, which is the length in the direction perpendicular to the extending direction, of distance w1. The distance h1 is a value larger than the vehicle length, and the distance w1 is a value larger than the vehicle width.

[0037] In step S120 (see FIG. 2), the control unit 113 performs predetermined control of the vehicle 10 according to the retreat region A1 detected in step S110.

[0038] When the shape of the road edge LnE is complex, it is possible to recognize the shape of the road edge LnE with higher accuracy by using a high-degree-of-freedom model rather than using a low-degree-of-freedom model. However, when the shape of the road edge LnE is a simple straight line shape, it is possible to recognize the shape of the road edge LnE with higher accuracy by using a low-degree-of-freedom model rather than using a high-degree-of-freedom model. According to the driving support device 110 of the present embodiment described above, since the detection unit 112 can selectively use a low-degree-of-freedom model and a high-degree-of-freedom model according to whether the surrounding information satisfies the first condition or the second condition, the retreat region A1 can be detected accurately. Further, the detection unit 112 detects the retreat region A1 only using the surrounding information generated based on the surrounding image captured by the camera 122. Therefore, the automatic driving control system 100 can detect the retreat region A1 without using sensors other than the camera 122.

[0039] In addition, the first condition indicating that the accuracy of the shape of the road edge LnE indicated by the low-degree-of-freedom road edge information has decreased includes that the distance between the point sequence recognized using the low-degree-of-freedom model and the detection points corresponding to the point sequence is equal to or greater than a predetermined threshold distance. Therefore, even for a road edge LnE whose shape cannot be accurately represented by the low-degree-of-freedom model, the evacuation area A1 can be accurately detected.

[0040] In addition, the first condition includes information indicating that the surrounding information includes information indicating that a sign S1 indicating a no-parking zone has been detected in the surrounding image for setting the point sequence representing the shape of the road. Therefore, even for a road edge LnE whose shape cannot be represented by the low-degree-of-freedom model such as a no-parking zone, the evacuation area A1 can be accurately detected.

[0041] B. Second Embodiment: The automatic driving control system 100B of the second embodiment shown in FIG. 6 is different from the first embodiment in that it includes a host vehicle situation sensor 124, and the other configurations are the same. The evacuation process of the second embodiment shown in FIG. 7 is different from the evacuation process of the first embodiment shown in FIG. 2 in that the evacuation area A1 detected in the past is updated according to whether the evacuation has been completed. In the present embodiment, the evacuation area detected in the past is the evacuation area detected immediately before.

[0042] The host vehicle situation sensor 124 (see FIG. 6) includes a vehicle sensor and a yaw rate sensor. The host vehicle situation sensor 124 detects the speed and yaw rate of the vehicle 10 as the situation of the vehicle 10.

[0043] The process of step S120 (see FIG. 7) in the second embodiment includes steps S121, S122, and S123. In step S121, the control unit 113 starts evacuation control. In the present embodiment, the control unit 113 controls the automatic driving control unit 210 to evacuate the vehicle 10 to the evacuation area A1.

[0044] In step S122, the driving support device 110 determines whether the evacuation of the vehicle 10 is completed. For example, when the driving support device 110 can determine that the vehicle 10 has evacuated to the evacuation area A1, it determines that the evacuation of the vehicle 10 is completed. When the evacuation of the vehicle 10 is completed, the evacuation process ends. On the other hand, when the evacuation of the vehicle 10 is not completed, the detection unit 112 updates the evacuation area A1 in step S123 and returns to the process of step S122. That is, the driving support device 110 continues the evacuation control and repeats the processes of steps S122 and S123 until the evacuation of the vehicle 10 is completed.

[0045] In the present embodiment, in step S123, the detection unit 112 updates the evacuation area A1 using the own vehicle situation detected by the own vehicle situation sensor 124 and the lane line information. More specifically, the detection unit 112 fixes the periphery on the lane Ln1 side of the evacuation area A1 so as to overlap the lane line information LL, and updates the evacuation area A1. That is, the detection unit 112 fixes the lateral width and the longitudinal width in the evacuation area A1 detected in step S110, determines the position of the evacuation area A1 in the direction perpendicular to the extending direction of the lane Ln1, and updates the position of the evacuation area A1 in the extending direction of the lane Ln1 using the own vehicle situation detected by the own vehicle situation sensor 124. The lane line information may be the lane line information recognized in step S100 or the lane line information newly recognized by the recognition unit 111.

[0046] The lane line information has a smaller error and is more stable than the own vehicle situation, which is an ego motion such as the speed and yaw rate of the vehicle 10. According to the detection unit 112 of the second embodiment described above, in addition to the own vehicle situation, the evacuation area A1 is updated using highly stable lane line information. Therefore, the evacuation area A1 can be updated with higher accuracy than when updating using only the own vehicle situation.

[0047] C. Third Embodiment: The third embodiment is different from the second embodiment in that the detection unit 112 further updates the evacuation area A1 using road edge information. Since the configuration of the automatic driving control system according to the third embodiment is the same as that of the automatic driving control system according to the first embodiment, the description of the configuration of the automatic driving control system will be omitted.

[0048] In step S123, the detection unit 112 updates the evacuation area A1 using the road edge information in addition to the vehicle status detected by the vehicle status sensor 124 and the lane line information. More specifically, the detection unit 112 updates the lateral width of the lane Ln1 in the evacuation area A1 detected in step S110. The road edge information may be the road edge information recognized in step S100 or the road edge information newly recognized by the recognition unit 111.

[0049] According to the detection unit 112 of the third embodiment described above, the evacuation area A1 is updated using the road edge information. The recognition accuracy of the road edge LnE increases as the vehicle 10 approaches. Therefore, the evacuation area can be updated with higher accuracy.

[0050] D. Other Embodiments: (D1) In the above-described embodiments, the driving support device 110 performs an evacuation process when the vehicle 10 needs to evacuate. However, the present invention is not limited to this. For example, the driving support device 110 may detect the evacuation area A1 during the running of the vehicle 10 regardless of whether the vehicle 10 needs to evacuate. More specifically, the driving support device 110 may repeatedly execute the processes of steps S100 and S110 (see FIG. 2) every 500 ms, for example.

[0051] (D2) In the above-described embodiments, the recognition unit 111 recognizes the surrounding information based on the surrounding image captured by the camera 122. In addition to this, the recognition unit 111 may further recognize the surrounding information based on the distances to the surrounding objects of the vehicle 10 obtained using a laser radar, a millimeter-wave radar, a Lidar (Light Detection and Ranging), an ultrasonic sensor, or the like.

[0052] (D3) In the above-described embodiment, the recognition unit 111 recognizes a point sequence representing the shape of the road edge LnE from the detection points, which are points indicating the road edge LnE detected in the surrounding image, using a low-degree-of-freedom model or a high-degree-of-freedom model. Not limited to this, for example, the recognition unit 111 may recognize the shape of the road edge LnE from the region indicating the road edge LnE detected in the surrounding image, using a low-degree-of-freedom model or a high-degree-of-freedom model.

[0053] (D4) In the above-described embodiment, the detection unit 112 detects the evacuation area A1 using the surrounding information. In addition to this, the detection unit 112 may further detect the evacuation area A1 using map information including information such as the position and shape of the previously stored lane line LnL and the position and shape of the road edge LnE. The detection unit 112 detects the evacuation area A1, for example, using map information including the position where the vehicle 10 is traveling estimated using a global navigation satellite system (GNSS) such as GPS (Global Positioning System) and the surrounding information.

[0054] (D5) In the above-described embodiment, as predetermined control, the control unit 113 may control the notification device included in the automatic driving control system 100 to notify the passengers of the vehicle 10 of the position of the evacuation area A1 detected by the detection unit 112.

[0055] (D6) In the above-described embodiment, the surplus distance may be 0 m. That is, the detection unit 112 may have the periphery on the road edge LnE side of the evacuation area A1 overlapping the road edge LnE. Also, the surplus distance does not have to be constant. For example, the surplus distance may be larger when using high-degree-of-freedom road edge information as the road edge information than when using low-degree-of-freedom road edge information as the road edge information.

[0056] (D7) In the above-described embodiment, the detection unit 112 detects the evacuation area A1 such that the distance between the peripheral edge on the section line LnL side of the evacuation area A1 and the section line information LL is 0 m. Not limited to this, the evacuation area A1 may be detected such that the distance between the peripheral edge on the section line LnL side of the evacuation area A1 and the section line information LL is equal to or greater than a predetermined distance. The predetermined distance can be arbitrarily determined and may be the same length as the surplus distance.

[0057] (D8) In the above-described embodiment, the evacuation area A1 has a rectangular shape. Not limited to this, the evacuation area A1 may have any shape.

[0058] (D9) In the above-described embodiment, the detection unit 112 detects, as the evacuation area A1, an area in the target area TA1 that is closest to the vehicle 10, has a length in the extending direction of distance h1, and a width, which is the length in the direction perpendicular to the extending direction, of distance w1. Not limited to this, for example, as shown in FIG. 8, an area in the target area TA2 that has a length in the extending direction equal to or greater than the vehicle length and the largest width of distance w2 may be detected as the evacuation area A2.

[0059] (D10) In the above-described second or third embodiment, when the detection unit 112 changes the lateral width of the evacuation area A1 in the update of the evacuation area A1 in step S123 (see FIG. 7), if the lateral width expands from the evacuation area A1 before the update, the degree of deformation of the evacuation area A1 in the update may be made smaller than when the lateral width narrows from the evacuation area A1 before the update. The case of changing the lateral width of the evacuation area A1 is, for example, when the distance Lm between the road edge LnE and the division line LnL increases or decreases by a predetermined threshold distance or more. The degree of deformation is, for example, the length of the lateral width that increases or decreases per unit time. More specifically, when the lateral width expands from the evacuation area A1 before the update, it is widened by 0.1 m per unit time, and when the lateral width narrows from the evacuation area A1 before the update, it is narrowed by 0.3 m per unit time. The length of the lateral width that increases or decreases per unit time may be fixed or may be determined according to the size of the evacuation area. Also, when the lateral width expands from the evacuation area A1 before the update, it is widened by a distance of 30% of the change amount of the distance Lm between the road edge LnE and the division line LnL per unit time, and when the lateral width narrows from the evacuation area A1 before the update, it is narrowed by a distance of 80% of the change amount of the distance Lm between the road edge LnE and the division line LnL per unit time.

[0060] (D11) In the above-described third embodiment, when the detection unit 112 uses high-degree-of-freedom road edge information as the road edge information, the degree of deformation of the evacuation area A1 in the update is made smaller than when using low-degree-of-freedom road edge information as the road edge information.

[0061] E. Other Forms: The present disclosure is not limited to the above-described embodiments, and can be realized in various configurations without departing from the gist thereof. For example, the technical features in the embodiments corresponding to the technical features in each form described in the summary of the invention can be appropriately replaced or combined in order to solve the above-described problems or to achieve some or all of the above-described effects. Also, if the technical feature is not described as essential in this specification, it can be appropriately deleted.

[0062] The control unit and its method described in the present disclosure may be implemented by a dedicated computer provided by configuring a processor and a memory programmed to execute one or more functions embodied by a computer program. Alternatively, the control unit and its method described in the present disclosure may be implemented by a dedicated computer provided by configuring a processor with one or more dedicated hardware logic circuits. Or, the restriction unit and its method described in the present disclosure may be implemented by one or more dedicated computers constituted by a combination of a processor programmed to execute one or more functions and a memory and a processor constituted by one or more hardware logic circuits. Further, the computer program may be stored in a computer-readable non-transitory tangible recording medium as instructions to be executed by a computer.

[0063] <Aspect 1>The driving assistance device (110) mounted on the vehicle (10) having the camera (122) uses a low-degree-of-freedom model which is a model with a low degree of freedom in shape expression and a high-degree-of-freedom model which is a model with a higher degree of freedom in shape expression than the low-degree-of-freedom model. Based on the peripheral image which is an image captured by the camera of the periphery of the vehicle, it recognizes a road edge which is an edge of the road on which the vehicle is traveling and is the closest edge to the vehicle on the road, and a lane line closest to the road edge, and generates peripheral information including road edge information indicating the road edge and lane line information indicating the lane line. A recognition unit (111), a detection unit (112) that uses the road edge information and the lane line information to detect a retreat region which is a region existing between the road edge and the lane line and in which the vehicle can retreat, and a control unit (113) that performs predetermined control of the vehicle according to the detected retreat region when the vehicle retreats to the retreat region. The retreat region is a region where the length in the extending direction of the road is equal to or greater than the vehicle length of the vehicle, and the length in the direction perpendicular to the extending direction is equal to or greater than the vehicle width of the vehicle. The road edge information includes low-degree-of-freedom road edge information representing the road edge recognized using the low-degree-of-freedom model and high-degree-of-freedom road edge information representing the road edge recognized using the high-degree-of-freedom model. When the peripheral information satisfies a first condition indicating that the accuracy of the low-degree-of-freedom road edge information has decreased or the accuracy of the low-degree-of-freedom road edge information has decreased, the detection unit uses the high-degree-of-freedom road edge information as the road edge information. When the first condition is not satisfied and the peripheral information satisfies a second condition regarding the accuracy of the high-degree-of-freedom road edge information, the detection unit uses the low-degree-of-freedom road edge information as the road edge information.

[0064] <Aspect 2>In the driving assistance device according to the aspect of Aspect 1, the vehicle has a host vehicle situation sensor (124) that detects a host vehicle situation including the speed and yaw rate of the vehicle. The detection unit updates the retreat region detected in the past using the host vehicle situation and the lane line information, and the control unit may perform the control according to the updated retreat region.

[0065] <Mode 3>In the driving assistance device according to the first or second mode, the recognition unit can recognize a point sequence representing the shape of the road edge from a detection point, which is a point indicating the road edge detected in the peripheral image, using the low-degree-of-freedom model and the high-degree-of-freedom model. The first condition indicating that the accuracy of the shape of the road edge indicated by the low-degree-of-freedom road edge information has decreased may include that the distance between the point sequence recognized using the low-degree-of-freedom model and the detection point corresponding to the point sequence is equal to or greater than a predetermined threshold distance.

[0066] <Mode 4>In the driving assistance device according to any one of the first to third modes, the first condition may include information indicating that the peripheral information includes information indicating that a sign indicating a no-parking zone is detected in the peripheral image.

[0067] <Mode 5>In the driving assistance device according to any one of the second to fourth modes, the detection unit may further update the evacuation area detected in the past using the road edge information.

[0068] <Mode 6>In the driving assistance device according to any one of the second to fifth modes, the evacuation area has a rectangular shape. When the detection unit changes the lateral width of the evacuation area detected in the past in the update, when the lateral width expands from the evacuation area detected in the past before the update, the degree of deformation of the evacuation area in the update may be made smaller than when the lateral width narrows from the evacuation area detected in the past before the update.

[0069] <Mode 7>In the driving assistance device according to the fifth or sixth mode, when the detection unit uses the high-degree-of-freedom road edge information as the road edge information, the degree of deformation of the evacuation area in the update may be made smaller than when using the low-degree-of-freedom road edge information as the road edge information.

[0070] <Embodiment 8>In the driving support device according to any one of Embodiments 1 to 7 above, the detection unit sets the evacuation area so that the distance between the peripheral edge on the road edge side of the evacuation area and the road edge is equal to or greater than a predetermined surplus distance, and the surplus distance may be larger when using the high-degree-of-freedom road edge information as the road edge information than when using the low-degree-of-freedom road edge information as the road edge information.

[0071] <Embodiment 9>In the driving support device according to any one of Embodiments 1 to 8 above, the evacuation area is rectangular, and when the detection unit detects the evacuation area from a target area that is an area where the distance between the road edge and the lane line is wider than the vehicle width and the length in the extending direction is longer than the vehicle length, the detection unit may detect, as the evacuation area, the area in the target area where the length in the extending direction is equal to or greater than the vehicle length and the lateral width is the largest.

Explanation of Reference Numerals

[0072] 10…Vehicle, 100, 100B…Autonomous Driving Control System, 110…Driving Support Device, 111…Recognition Unit, 112…Detection Unit, 113…Control Unit, 122…Camera, 124…Vehicle Status Sensor, 210…Autonomous Driving Control Unit, 220…Driving Force Control ECU, 230…Braking Force Control ECU, 240…Steering Control ECU, 250…In-Vehicle Network

Claims

1. A driving support device (110) mounted on a vehicle (10) having a camera (122), using a low-degree-of-freedom model which is a model with a low degree of freedom in shape expression and a high-degree-of-freedom model which is a model with a higher degree of freedom in shape expression than the low-degree-of-freedom model, based on a peripheral image which is an image obtained by the camera imaging the periphery of the vehicle, recognizing a road edge which is an edge of the road on which the vehicle is traveling and which is the closest edge to the vehicle on the road, and a lane line closest to the road edge, and generating peripheral information including road edge information indicating the road edge and lane line information indicating the lane line, a recognition unit (111); a detection unit (112) which uses the road edge information and the lane line information to detect a retreat region which is a region existing between the road edge and the lane line and in which the vehicle can retreat; a control unit (113) which, when the vehicle retreats to the retreat region, performs predetermined control of the vehicle according to the detected retreat region, and comprising, the retreat region is a region in which the length in the extending direction of the road is equal to or greater than the vehicle length of the vehicle, and the width which is the length in a direction perpendicular to the extending direction is equal to or greater than the vehicle width of the vehicle, the road edge information includes low-degree-of-freedom road edge information representing the road edge recognized using the low-degree-of-freedom model and high-degree-of-freedom road edge information representing the road edge recognized using the high-degree-of-freedom model, the detection unit, when the peripheral information satisfies a first condition indicating that the accuracy of the low-degree-of-freedom road edge information has decreased or there is a high possibility that the accuracy of the low-degree-of-freedom road edge information will decrease, uses the high-degree-of-freedom road edge information as the road edge information, a driving support device which, when the first condition is not satisfied and the peripheral information satisfies a second condition regarding the accuracy of the high-degree-of-freedom road edge information, uses the low-degree-of-freedom road edge information as the road edge information.

2. The driving support device according to claim 1, wherein the vehicle has a host vehicle situation sensor (124) which detects a host vehicle situation including the speed and yaw rate of the vehicle, the detection unit updates the retreat region detected in the past using the host vehicle situation and the lane line information, the control unit performs the control according to the updated retreat region. A driving support device.

3. The driving support device according to claim 1 or claim 2, The recognition unit can recognize a point sequence representing the shape of the road edge from a detection point that is a point indicating the road edge detected in the peripheral image, using the low-degree-of-freedom model and the high-degree-of-freedom model. The first condition indicating that the accuracy of the shape of the road edge indicated by the low-degree-of-freedom road edge information has decreased includes that the distance between the point sequence recognized using the low-degree-of-freedom model and the detection point corresponding to the point sequence is equal to or greater than a predetermined threshold distance. The driving support device.

4. The driving support device according to claim 1 or claim 2, The first condition is that the peripheral information includes information indicating that a sign indicating a no-parking zone has been detected in the peripheral image. The driving support device.

5. The driving support device according to claim 2, The detection unit further updates the evacuation area detected in the past using the road edge information. The driving support device.

6. The driving support device according to claim 2 or claim 5, The evacuation area is rectangular, When the detection unit changes the width of the evacuation area detected in the past in the update, when the width expands from the evacuation area detected in the past before the update, compared with the case where the width narrows from the evacuation area detected in the past before the update, the driving support device reduces the degree of deformation of the evacuation area in the update.

7. The driving support device according to claim 2 or claim 5, When the detection unit uses the high-degree-of-freedom road edge information as the road edge information, the driving support device reduces the degree of deformation of the evacuation area in the update compared with the case where the low-degree-of-freedom road edge information is used as the road edge information.

8. The driving support device according to claim 1 or claim 2, The detection unit sets the evacuation area so that the distance between the peripheral edge on the road edge side of the evacuation area and the road edge is equal to or greater than a predetermined surplus distance. The surplus distance is larger when the high-degree-of-freedom road edge information is used as the road edge information than when the low-degree-of-freedom road edge information is used as the road edge information. The driving support device.

9. The driving support device according to claim 1 or claim 2, The evacuation area is rectangular, When the detection unit detects the retreat region from a target region that is an area where the distance between the road edge and the lane line is wider than the vehicle width and the length in the extending direction is longer than the vehicle length, the driving support device detects, as the retreat region, a region in the target region where the length in the extending direction is equal to or greater than the vehicle length and the lateral width is the largest.

Images