Driving assistance device

Abstract

The present application provides a kind of driving assistance device by improving the identification accuracy of boundary structure (such as guardrail, tunnel wall and soundproof wall etc.), to improve the determination accuracy of artifact, reduce the possibility of unnecessary pre-crash control to artifact is executed.The driving assistance device has: first sensor, which detects the object of the outside of vehicle according to the reflection of the transmitted wireless medium;Second sensor, which detects the boundary structure of the boundary between the road area where the vehicle can travel and the area outside the road area;And control unit, which executes pre-crash control, pre-crash control is used to avoid collision with the identified object as the object detected by first sensor.The control unit is configured to: the execution possibility of pre-crash control in the case where back side position condition is established is lower than the execution possibility of pre-crash control in the case where back side position condition is not established, and back side position condition is the condition that the identified object is located on the back side of the boundary structure relative to the vehicle.

Description

Technical Field

[0001] This invention relates to a driving assistance device that performs pre-collision control for avoiding collisions with objects. Background Technology

[0002] Previously, there were known driver assistance devices that performed pre-collision control on objects detected by millimeter-wave radar mounted on vehicles. Millimeter-wave radar transmits radio waves (millimeter waves) with wavelengths of 1mm to 10mm and receives millimeter waves reflected from objects, thereby detecting objects.

[0003] Regarding sensors that detect objects by reflecting wireless media such as radio waves, if the wireless medium reflected once by an object is further reflected by other objects before reaching the sensor, the object may be erroneously detected at a location different from its actual location. This erroneously detected object is called an "artifact." Such further reflections (sub-reflections) caused by other objects are likely to occur, for example, at structures (boundary structures) that separate road areas from areas outside the road area, such as guardrails, tunnel walls, and sound barriers.

[0004] For example, the driving assistance device described in Patent Document 1 (hereinafter referred to as the "existing device") obtains an actual line corresponding to a boundary structure such as a wall based on the detection results of millimeter-wave radar, and obtains a base line based on the actual line. Then, the existing device identifies objects located outside (behind) the base line relative to the vehicle as artifacts.

[0005] Existing technical documents

[0006] Patent documents

[0007] Patent Document 1: Japanese Patent Application Publication No. 2001-116839 Summary of the Invention

[0008] Existing devices use a millimeter-wave radar to detect both objects and actual lines. When the millimeter waves reflected by the object (hereinafter referred to as "object reflected waves") are further reflected by boundary structures, the millimeter-wave radar identifies the points on the boundary structures where the object reflected waves are further reflected (hereinafter referred to as "sub-reflection points") as artifacts.

[0009] Existing devices acquire the "line formed by continuously detected stationary objects" as the actual line (see paragraph 0010). In the event of subreflection, existing devices may fail to detect the subreflection points of boundary structures as the aforementioned "continuously detected stationary objects." As a result, existing devices cannot acquire the actual line, and therefore may be unable to determine artifacts.

[0010] This invention was made to solve the aforementioned technical problems. Specifically, one objective of this invention is to provide a driving assistance device that improves the accuracy of artifact detection by enhancing the recognition accuracy of boundary structures, thereby reducing the likelihood of performing unnecessary pre-collision control on artifacts.

[0011] The driving assistance device of the present invention (hereinafter referred to as "the device of the present invention") comprises:

[0012] The first sensor (22, 24L, 24R) detects objects outside the vehicle based on the reflection of the transmitted wireless medium.

[0013] The second sensor (26, 28L, 28R) detects boundary structures located at the boundary between the road area where the vehicle can travel and the area outside the road area; and

[0014] Control units (20, 30, 40, 50) perform pre-collision control to avoid collisions with identified objects that are detected by the first sensor.

[0015] The control unit is configured to make the probability of executing the pre-collision control when the rear-side position condition is met lower than the probability of executing the pre-collision control when the rear-side position condition is not met, wherein the rear-side position condition is the condition that the identified object is located on the rear side of the boundary structure relative to the vehicle ("Yes" in steps 435 and 615, and step 625).

[0016] According to the apparatus of the present invention, a second sensor, different from the first sensor that detects the object to be identified as the target of pre-collision control, is used to detect boundary structures. Therefore, even if the first sensor detects artifacts caused by subreflections of the boundary structures, the apparatus of the present invention can detect the boundary structures unaffected by these artifacts. Thus, the accuracy of boundary structure identification can be improved, thereby enabling reliable determination of the validity of the backside position condition. As a result, the possibility of performing unnecessary pre-collision control based on artifacts can be reduced.

[0017] In one embodiment of the device of the present invention, the control unit is configured as follows:

[0018] Based on the detection results of the first sensor, obtain the reliability (RD) representing the probability that the identified object actually exists (steps 415 and 625).

[0019] For identified objects with a reliability exceeding a predetermined threshold, the pre-collision control is performed (steps 435 and 445).

[0020] For identified objects with a reliability less than the threshold, the pre-collision control is not performed (step 435), and a predetermined subtraction value is subtracted from the reliability of identified objects for which the backside position condition is met (step 625).

[0021] Therefore, since the reliability of object recognition when the backside position condition is met is reduced by the subtraction value, the likelihood of executing pre-collision control is lower when the backside position condition is met compared to when it is not met. Thus, the possibility of performing unnecessary pre-collision control on artifacts can be reduced.

[0022] In one embodiment of the device of the present invention, the second sensor (26) is a sensor of a different type than the first sensor (22, 24L, 24R).

[0023] Therefore, even if the first sensor detects artifacts caused by the secondary reflections of the boundary structure, the second sensor, being of a different type than the first sensor, can detect the boundary structure more accurately without being affected by the secondary reflections of the boundary structure.

[0024] In the above method, the second sensor is a camera (26) that acquires images by photographing the external environment.

[0025] The control unit is configured to,

[0026] Based on the image captured at the current moment, the external environment is divided into the road area and the area outside the road area, thereby determining the boundary (steps 505 and 510).

[0027] Based on multiple images captured before the current time, obtain the depth and height of the object captured in the image captured at the current time (step 515).

[0028] The photographed object located within a specified range based on the boundary, having a depth of more than a specified first threshold, and a height of more than a specified second threshold, is identified as the boundary structure (step 525).

[0029] Therefore, objects with a depth greater than the first threshold and a height greater than the second threshold are identified as boundary structures. Thus, the device of this invention can detect boundary structures more accurately.

[0030] In one embodiment of the apparatus of the present invention, the control unit is configured to,

[0031] Obtain an approximate line of the boundary structure in the vehicle coordinate system with the specified reference point of the vehicle as the origin (step 530).

[0032] Based on the relationship between the position coordinates of the identified object in the vehicle coordinate system and the approximate line, determine whether the rear position condition is met (step 615).

[0033] This allows for a more accurate determination of whether an object is located on the back side of a boundary structure relative to a vehicle.

[0034] In one embodiment of the device of the present invention, the second sensor is disposed at a different location than the first sensor.

[0035] If the second sensor is of the same type as the first sensor, and is positioned in the same location as the first sensor, then if the first sensor detects an artifact, the second sensor may also detect the artifact. By positioning the second sensor in a different location than the first sensor, this possibility can be reduced.

[0036] Furthermore, in the foregoing description, to aid in understanding the invention, the names and / or reference numerals used in the embodiments described below are enclosed in parentheses to represent the components of the invention. However, the constituent elements of the invention are not limited to the embodiments specified by the names and / or reference numerals. Other objects, features, and incidental advantages of the invention can be readily understood from the relevant descriptions of the embodiments of the invention described alongside the following drawings. Attached Figure Description

[0037] Figure 1 This is a schematic system configuration diagram of a driving assistance device according to an embodiment of the present invention.

[0038] Figure 2 yes Figure 1 The diagram illustrates the installation locations of each millimeter-wave radar and the front camera.

[0039] Figure 3 This is an explanatory diagram illustrating an example of the operation of a driving assistance device according to an embodiment of the present invention.

[0040] Figure 4 It is shown Figure 1 The flowchart shown is of the pre-collision control routine executed by the CPU of the driver assistance ECU.

[0041] Figure 5 It is shown Figure 1 The flowchart shown is a flowchart of the boundary structure determination subroutine executed by the CPU of the driving assistance ECU.

[0042] Figure 6 It is shown Figure 1 The flowchart shown is a subroutine for determining the rear position condition executed by the CPU of the driving assistance ECU.

[0043] Figure 7 This is an explanatory diagram showing the arrangement positions of the left-side and right-side millimeter-wave radars in a second variation of an embodiment of the present invention. Detailed Implementation

[0044] A driving assistance device (hereinafter referred to as "this assistance device") 10 according to one embodiment of the present invention is mounted on a vehicle VA. For example... Figure 1 As shown, this auxiliary device 10 includes a driving assistance ECU (hereinafter referred to as "DSECU") 20, an engine ECU 30, a brake ECU 40, and an instrument ECU 50.

[0045] ECU is short for Electronic Control Unit, an electronic control circuit that primarily consists of a microcomputer including a CPU, ROM, RAM, and interfaces (I / F). ECU is sometimes also referred to as a "controller," "controller," or "computer." The CPU performs various functions by executing instructions (routes) stored in the memory (ROM). All or some of the aforementioned ECUs 20, 30, 40, and 50 can be integrated into a single ECU.

[0046] This auxiliary device 10 includes a front millimeter-wave radar 22, a left front millimeter-wave radar 24L, a right front millimeter-wave radar 24R, and a front camera 26. They are connected to the DSECU 20 for exchanging data.

[0047] Hereinafter, without needing to distinguish between the front millimeter-wave radar 22, the left front millimeter-wave radar 24L, and the right front millimeter-wave radar 24R, they will be referred to as "millimeter-wave radars". Furthermore, they are sometimes referred to as "first sensors".

[0048] The millimeter-wave radar 24 detects objects by transmitting millimeter waves and receiving the reflected waves. The millimeter-wave radar 24 determines the distance to the object, the object's lateral position, and its relative velocity, and transmits the object information containing these parameters to the DSECU 20.

[0049] like Figure 2 As shown, the front millimeter-wave radar 22 is disposed near the center of the front end of the vehicle VA in the vehicle width direction. The front millimeter-wave radar 22 receives millimeter waves reflected by objects (three-dimensional objects) located within the detection range DR1 in front of the vehicle VA, and detects the objects. The detection range DR1 is a fan-shaped area extending forward in the direction of the front-rear axle of the vehicle VA with a central angle θ1.

[0050] like Figure 2As shown, the left front millimeter-wave radar 24L is mounted on the left end of the front of the vehicle VA in the vehicle width direction. The left front millimeter-wave radar 24L detects objects located within the detection range DR2L. In addition, the detection range DR2L is a fan-shaped area extending diagonally forward to the left of the vehicle VA with a central angle θ2.

[0051] like Figure 2 As shown, the right front millimeter-wave radar 24R is mounted on the right end of the front of the vehicle VA in the vehicle width direction. The right front millimeter-wave radar 24R detects objects located within its detection range DR2R. Furthermore, the detection range DR2R is a fan-shaped area extending diagonally forward to the right of the vehicle VA with a central angle θ2. The central angle of the detection range DR2R can be the same as or different from the central angle of the detection range DR2R.

[0052] like Figure 2 As shown, the front camera 26 is positioned near the center of the upper part of the front windshield of the vehicle VA in the vehicle width direction. That is, the front camera 26 is positioned differently from the millimeter-wave radar. The front camera 26 acquires images by capturing a shooting range PR in front of the vehicle VA and sends the captured images to the DSECU 20. Furthermore, the shooting range PR is a fan-shaped area extending forward in the front-rear axle direction of the vehicle VA with a central angle θ3. The front camera 26 is sometimes referred to as a "second sensor".

[0053] The engine ECU 30 is connected to the engine actuator 32. The engine actuator 32 includes a throttle actuator that changes the opening of the throttle valve of the engine 32a. The engine ECU 30 can change the torque generated by the engine 32a by driving the engine actuator 32. The torque generated by the engine 32a is transmitted to the drive wheels via a transmission (not shown). Alternatively, the vehicle VA may replace the engine 32a or, in addition to the engine 32a, also have an electric motor as the vehicle drive source.

[0054] The brake ECU 40 is connected to the brake actuator 42. The brake actuator 42 includes a hydraulic circuit. The hydraulic circuit includes a master cylinder, a flow path for brake fluid, multiple valves, a pump, and a motor that drives the pump. In response to an instruction from the brake ECU 40, the brake actuator 42 adjusts the hydraulic pressure supplied to the wheel cylinders built into the brake mechanism 42a. Using this hydraulic pressure, the wheel cylinders generate frictional braking force against the wheels.

[0055] The instrument cluster ECU 50 is connected to a display 52 and a speaker 54. The display 52 is positioned opposite the driver seated in the driver's seat. For example, the display 52 is a multi-information display. The speaker 54 is located inside the vehicle's VA cabin and emits a buzzing sound.

[0056] (A summary of the action)

[0057] The auxiliary device 10 obtains the collision time, which represents the time elapsed until the object identified based on the object information from the millimeter-wave radar (hereinafter referred to as the "identified object") collidees with or comes closest to the target. Hereinafter, the collision time is referred to as "TTC (an abbreviation for Time To Collision)".

[0058] When the TTC is less than or equal to a threshold time Tth, the auxiliary device 10 performs pre-collision control to avoid a collision with the object. The pre-collision control includes at least one of "deceleration control to slow down the vehicle VA" and "warning control to warn the driver of the possibility of a collision".

[0059] As mentioned above, millimeter-wave radar can detect artifacts when the millimeter waves reflected by an object are further reflected by other objects (such reflections are called "subreflections"). In particular, objects that are prone to producing subreflections include guardrails, tunnel walls, and sound barriers. Guardrails and the like are structures located at the boundary between a road and the area outside the road, hereinafter referred to as "boundary structures S".

[0060] The auxiliary device 10 determines whether a boundary structure S exists based on the image captured by the front camera 26. If a boundary structure S exists, the auxiliary device 10 determines whether the condition that the identified object is located on the rear side of the boundary structure S relative to the vehicle VA is met.

[0061] Typically, millimeter-wave radar will not detect objects located behind a boundary structure S relative to the vehicle VA. This is because the millimeter waves emitted by the radar are reflected by the boundary structure S, and therefore do not reach objects located behind it. Thus, objects appearing to be behind the vehicle are likely artifacts falsely detected by the millimeter-wave radar due to secondary reflections from the boundary structure S.

[0062] Therefore, this auxiliary device 10 makes it less likely to perform pre-collision control on an object whose rear-side position conditions are met than on an object whose rear-side position conditions are not met. Thus, this auxiliary device 10 can reduce the likelihood of performing pre-collision control on artifacts, thereby reducing the likelihood of the driver feeling uncoordinated with pre-collision control.

[0063] Furthermore, this auxiliary device 10 uses a second sensor (front-facing camera) different from the first sensor (millimeter-wave radar) that detects the object being controlled before collision to determine whether a boundary structure S exists. Therefore, even if the first sensor detects an artifact due to the subreflection of the boundary structure S, this auxiliary device 10 determines the existence of the boundary structure S based on the detection result of the second sensor, thus enabling accurate detection of the boundary structure S regardless of the detection of artifacts.

[0064] (Example of action)

[0065] Reference Figure 3 Explain the operation of this auxiliary device 10.

[0066] The auxiliary device 10 detects the identified object N at a position (Xn, Yn) in the vehicle coordinate system (X, Y) with the origin O of the vehicle VA as a specified reference point. The vehicle coordinate system (X, Y) has an X-axis in the width direction of the vehicle VA and a Y-axis in the front and rear axle directions of the vehicle VA.

[0067] The auxiliary device 10 detects a boundary structure S based on an captured image, for example, by using a least squares method to obtain an approximate line SL of the boundary structure S. Then, the auxiliary device 10 determines whether the line segment L between the identified object N and the origin O intersects the approximate line SL. If the line segment L does not intersect the approximate line SL, the auxiliary device 10 determines that the back-side position condition is not met. If the line segment L intersects the approximate line SL, the auxiliary device 10 determines that the back-side position condition is met. Figure 3 In the example shown, line segment L intersects with approximate line SL, therefore the auxiliary device 10 determines that the object N is in a back position condition.

[0068] The auxiliary device 10 subtracts a predetermined first phase subtraction value SV1 from the reliability RD of the identified object N when the back-side position condition is met. The reliability RD represents the probability that the identified object N actually exists; the higher the reliability RD, the higher the probability that the identified object N actually exists.

[0069] When the following condition A1 is met, the auxiliary device 10 adds a predetermined sum value AV to the reliability RD of the identified object N, and when the following condition A2 is met, subtracts a predetermined second subtraction value SV2 from the reliability RD of the identified object N.

[0070] Condition A1: The same object N is detected more than a specified number of times.

[0071] Condition A2: The condition of suddenly detecting an previously undetected object N.

[0072] This auxiliary device 10 obtains the time-to-market (TTC) of the identified object N with a reliability RD of RDth or higher. Specifically, this auxiliary device 10 obtains the TTC by dividing the distance to the identified object N by the relative velocity of the identified object N.

[0073] Then, the auxiliary device 10 performs pre-collision control when the minimum TTC is below the threshold time Tth.

[0074] When performing deceleration control as pre-collision control, this auxiliary device 10 controls the engine actuator 32 in such a way that the engine 32a does not generate torque, and controls the brake actuator 42 in such a way that the wheels generate a specified frictional braking force.

[0075] In detail, DSECU 20 sends a deceleration command containing a specified negative target acceleration to engine ECU 30 and brake ECU 40. Upon receiving the deceleration command, engine ECU 30 closes the throttle valve by controlling engine actuator 32, preventing engine 32a from generating torque. Upon receiving the deceleration command, brake ECU 40 controls brake actuator 42 to generate frictional braking force at the wheels to match the aforementioned target acceleration of vehicle VA.

[0076] When the auxiliary device 10 performs warning control as a pre-collision control, it causes the instrument ECU 50 to display a warning image on the display 52 to inform the driver of the possibility of a collision, and causes the speaker 54 to emit a beeping sound.

[0077] In detail, DSECU 20 sends a warning command to instrument cluster ECU 50. Upon receiving the warning command, instrument cluster ECU 50 displays a warning image on display 52 and emits a beeping sound from speaker 54.

[0078] This auxiliary device 10 reduces the reliability RD of the identified object N when the backside position condition is met, thereby lowering the probability of executing pre-collision control for the identified object N when the backside position condition is met compared to the probability of executing it for the identified object N when the backside position condition is not met. Therefore, this auxiliary device 10 can reduce the likelihood of performing pre-collision control on artifacts.

[0079] (Specific actions)

[0080] <Pre-collision control routine>

[0081] The CPU of DSECU 20 (hereinafter, unless otherwise specified, refers to the CPU of DSECU 20) executes at specified intervals. Figure 4 The flowchart shows the pre-collision control routine.

[0082] Therefore, when the specified time is reached, the CPU starts from... Figure 4 The process begins at step 400 and proceeds sequentially from step 405 to step 430.

[0083] Step 405: The CPU obtains object information from the millimeter-wave radar.

[0084] Step 410: The CPU determines the object N to be identified based on the object information.

[0085] In detail, the CPU determines the position coordinates (Xn, Yn) representing the location of the identified object N in the vehicle coordinate system (X, Y) based on the object information. Further, in step 410, the CPU sets the reliability RD of identifying object N to the initial value IV.

[0086] Step 415: The CPU sets the reliability RD of the identified object N.

[0087] Specifically, the CPU determines whether conditions A1 and A2 are true. Then, if condition A1 is true, the CPU adds an additional value AV to the reliability RD of the identified object N, and if condition A2 is true, it subtracts a second subtraction value SV2 from the reliability RD of the identified object N.

[0088] Step 420: The CPU acquires the captured image from the front camera 26.

[0089] Step 425: The CPU executes the boundary structure determination subroutine for determining the boundary structure S based on the captured image. Further details regarding the boundary structure determination subroutine will be provided later. Figure 5 Please provide an explanation.

[0090] Step 430: The CPU determines whether the boundary structure S exists.

[0091] If the boundary structure S does not exist, the CPU determines "no" in step 430 and executes steps 435 to 445 in sequence.

[0092] Step 435: The CPU obtains the TTC of the identified object N with a reliability RD above the threshold RDth and a possibility of collision.

[0093] For example, the threshold RDth, the initial value IV, the sum value AV, the first subtraction value SV1, and the second subtraction value SV2 can be set such that, under the condition that the backside position condition is met, the reliability RD is less than the threshold RDth.

[0094] As an example, the above values ​​are set as follows.

[0095] Threshold RDth: 35

[0096] Initial value IV: 50

[0097] Added value AV: 10

[0098] First phase subtraction SV1: 30

[0099] Second phase subtraction SV2: 10

[0100] Furthermore, the threshold RDth, the initial value IV, the sum value AV, the first subtraction value SV1, and the second subtraction value SV2 can also be set such that when only the back-side position condition is met, the reliability RD will not be less than the threshold RDth, while when the back-side position condition is met and the above condition A2 is met, the reliability RD is less than the threshold RDth.

[0101] The CPU will determine an object N that is likely to collide with the vehicle VA or approach the vehicle VA within a specified distance from the vehicle VA, provided that the vehicle VA and the identified object N maintain their current speed and direction of movement.

[0102] The CPU determines the direction of movement of the vehicle VA based on the vehicle speed (vehicle speed) and the steering angle of the vehicle VA. Additionally, the vehicle speed is measured by a speed sensor (not shown), and the steering angle is measured by a steering angle sensor (not shown).

[0103] The CPU determines the direction of movement of the object N based on the history of the detected positions (Xn, Yn) of the object N.

[0104] Step 440: The CPU determines the object N with the smallest TTC.

[0105] Step 445: The CPU determines whether the minimum TTC is below the threshold time Tth.

[0106] If the minimum TTC is greater than the threshold time Tth, the CPU determines "No" in step 445, proceeds to step 495, and temporarily terminates this routine.

[0107] On the other hand, if the minimum TTC is below the threshold time Tth, the CPU determines "yes" in step 445 and proceeds to step 450. In step 450, the CPU performs pre-collision control and proceeds to step 495, temporarily ending this routine.

[0108] When the CPU enters step 430, if a boundary structure S exists, the CPU determines "yes" in step 430 and proceeds to step 455. In step 455, the CPU executes a backside position condition determination subroutine to determine whether the backside position condition is valid, and then executes the processing after step 435. The backside position condition determination subroutine will be discussed later. Figure 6 Please provide an explanation.

[0109] In step 435 of this routine, the CPU does not acquire the TTC of the identified object N with a reliability RD less than the threshold RDth. Therefore, pre-collision control will not be performed on the identified object N with a reliability RD less than the threshold RDth.

[0110] <Boundary Structure Determination Subroutine>

[0111] CPU enters Figure 4 In step 425, execute Figure 5 The boundary structures shown in the flowchart determine the subroutine. That is, the CPU enters... Figure 4 After step 425, from Figure 5 The process begins from step 500 and proceeds sequentially from step 505 to step 530.

[0112] Step 505: The CPU obtains the area where the vehicle (VA) can travel, i.e., the road area, based on the captured image. For example, the CPU can use Semantic Segmentation (SS) to divide the captured image into the road area and other areas.

[0113] Step 510: The CPU determines the boundary between the road area and other areas in the captured image.

[0114] Step 515: The CPU, based on... Figure 4 The image captured in 420 (the image captured at the current moment) is one of several previously captured images. The depth D and height H of the object captured in the current image are obtained. Specifically, since these multiple images are taken from different viewpoints, the three-dimensional information (depth D, height H, and width) of the object can be determined. For example, the CPU can obtain the depth D and height H of the object using SfM (Structure from Motion).

[0115] Step 520: The CPU converts the boundaries and positions of the captured objects in the captured image into positions in the vehicle coordinate system (X,Y).

[0116] Step 525: The CPU determines the photographed object that satisfies all of the following construction conditions B1 to B3 as the boundary construction S.

[0117] Condition B1: The object being photographed is located within a boundary area of ​​a specified width based on the boundary.

[0118] Condition B2: The depth D of the object being photographed is greater than or equal to the depth threshold Dth.

[0119] Condition B3: The height H of the object being photographed is above the height threshold Hth.

[0120] Step 530: The CPU obtains the approximate line SL of the boundary structure S.

[0121] Furthermore, if the boundary structure S is not determined in step 525 because there is no photographed object that satisfies all the structure conditions B1 to B3, the CPU does not obtain the approximate line SL in step 530.

[0122] After this, the CPU proceeds to step 595, temporarily terminating the current routine and entering... Figure 4 Step 430 is shown.

[0123] <Subroutine for determining dorsal position condition>

[0124] CPU enters Figure 4 In step 455, execute Figure 6 The flowchart shows the back-side position condition determination subroutine. That is, the CPU enters... Figure 4 After step 455, from Figure 6 The process begins from step 600 and proceeds sequentially from step 605 to step 615.

[0125] Step 605: The CPU selects the object N to be processed. Hereinafter, the object N selected in step 605 will be referred to as "processing object N".

[0126] Step 610: The CPU obtains the line segment L that connects the object N and the origin O of the vehicle coordinate system (X,Y).

[0127] Step 615: The CPU determines whether the approximate line SL of the boundary structure S intersects with the line segment L.

[0128] If the approximate line SL does not intersect with line segment L (i.e., the back-side position condition is not met), the CPU determines "No" in step 615 and proceeds to step 620. In step 620, the CPU determines whether all identified objects N are selected as processing objects N.

[0129] If not all identified objects N are selected as processing objects N, the CPU determines "no" in step 620 and returns to step 605. The CPU then selects a processing object N from the identified objects N that have not yet been selected as processing objects N and proceeds to step 610.

[0130] If the approximate line SL intersects with line segment L when the CPU enters step 615 (i.e., the back-side position condition is met), the CPU determines "yes" in step 615 and proceeds to step 625. In step 625, the CPU subtracts the first phase subtraction value SV1 from the reliability RD of the processed object N and proceeds to step 620.

[0131] If all identified objects N are selected as processing objects when the CPU proceeds to step 620, the CPU determines "yes" in step 620, proceeds to step 695, temporarily terminates this routine, and then proceeds to... Figure 4 Step 435 is shown.

[0132] In this embodiment, a second sensor (front-facing camera 26), different from the first sensor (millimeter-wave radar) used to detect and identify object N, is used to detect the boundary structure S. Therefore, even if the first sensor detects artifacts caused by subreflections from the boundary structure S, the boundary structure S can be reliably determined without being affected. Thus, situations where the backside position condition cannot be determined due to the inability to identify the boundary structure S can be prevented, thereby enabling more accurate artifact identification and reducing the likelihood of performing pre-collision control based on artifacts.

[0133] The present invention is not limited to the foregoing embodiments, and various modifications of the present invention may be used.

[0134] (First variation)

[0135] The first sensor is not limited to millimeter-wave radar. Any sensor that detects objects by reflecting the light through a transmitted wireless medium can be considered a first sensor.

[0136] (Second variation)

[0137] The second sensor is not limited to a camera. For example, the second sensor can also be a sensor of the same type as the first sensor (i.e., it can also be a millimeter-wave radar). In the case where the second sensor is a millimeter-wave radar, the CPU identifies objects with a length greater than or equal to a length threshold Lth as boundary structures S.

[0138] like Figure 7 As shown, this auxiliary device 10 has a left-side millimeter-wave radar 28L and a right-side millimeter-wave radar 28R as second sensors to replace the front camera 26.

[0139] likeFigure 7 As shown, the left-side millimeter-wave radar 28L is mounted on the left side of the vehicle VA, detecting objects within the detection range DR3L. The detection range DR3L is a fan-shaped area extending to the left in the width direction of the vehicle VA.

[0140] like Figure 7 As shown, the right-side millimeter-wave radar 28R is mounted on the right side of the vehicle VA, detecting objects within the detection range DR3R. The detection range DR3R is a fan-shaped area extending to the right in the width direction of the vehicle VA.

[0141] If the left-side millimeter-wave radar 28L and the left-front-side millimeter-wave radar 24L are positioned in the same location, then if the left-front-side millimeter-wave radar 24L detects an artifact, the left-side millimeter-wave radar 28L may also detect the artifact. Since the left-side millimeter-wave radar 28L is positioned differently from the left-front-side millimeter-wave radar 24L, this possibility is reduced.

[0142] For the same reason, the right-side millimeter-wave radar 28R is located in a different position than the right-front millimeter-wave radar 24R.

[0143] (Third variation)

[0144] This auxiliary device 10 can be installed in vehicles such as engine vehicles, hybrid electric vehicles (HEV), plug-in hybrid electric vehicles (PHEV), fuel cell electric vehicles (FCEV), and electric vehicles (BEV).

[0145] Label Explanation

[0146] 10: Driving assistance device, 20: Driving assistance ECU, 22: Front millimeter-wave radar, 24L: Left front side millimeter-wave radar, 24R: Right front side millimeter-wave radar, 26: Front camera.

Claims

1. A driving assistance device, comprising: The first sensor detects objects outside the vehicle based on the reflection of the transmitted wireless medium. A second sensor detects boundary structures located at the boundary between the road area where the vehicle can travel and the area outside the road area; and A control unit that performs pre-collision control to avoid collisions with an identified object that is detected by the first sensor. The control unit is configured such that the probability of executing the pre-collision control when the rear-side position condition is met is lower than the probability of executing the pre-collision control when the rear-side position condition is not met, wherein the rear-side position condition is the condition that the identified object is located on the rear side of the boundary structure relative to the vehicle. The control unit is also configured to: Based on the detection results of the first sensor, a reliability level representing the probability that the identified object actually exists is obtained. The pre-collision control is performed on identified objects whose reliability is above a predetermined threshold. For identified objects with a reliability less than the threshold, the pre-collision control is not performed. Subtract a predetermined subtraction value from the reliability of the identified object when the back-side position condition is met.

2. The driving assistance device according to claim 1, wherein, The second sensor is a different type of sensor than the first sensor.

3. The driving assistance device according to claim 2, wherein, The second sensor is a camera that acquires images by photographing the external environment. The control unit is configured to, Based on the currently captured image, the external environment is divided into the road area and the area outside the road area, thereby determining the boundary. Based on multiple images captured before the current time, obtain the depth and height of the object captured in the image captured at the current time. An object located within a specified range based on the boundary, having a depth greater than or equal to a specified first threshold, and a height greater than or equal to a specified second threshold, is defined as the boundary structure.

4. The driving assistance device according to claim 1, wherein, The control unit is configured to, Obtain an approximate line of the boundary structure in the vehicle coordinate system with the vehicle's designated reference point as the origin. Based on the relationship between the position coordinates of the identified object in the vehicle coordinate system and the approximate line, it is determined whether the rear-side position condition is met.

5. The driving assistance device according to claim 1, wherein, The second sensor is located at a different position than the first sensor.

Images