JP2014085805A - Lane departure prevention apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a lane departure prevention apparatus which performs safe lane departure prevention without annoyance.SOLUTION: A lane departure prevention apparatus includes: a white line recognition processing part 5 which detects a white line 4 of a lane 3; target detection means 12 which detects a plurality of objects on the lane 3 where a vehicle travels and an opposite lane, as targets; a collision prediction determination part 13 which applies ID information and type information, such as displacement, stop, oncoming, or the like, to each of the targets by tracking the targets detected in the target detection means 12, and performs collision prediction on all ID targets; a threshold adjustment part 10 which receives information of the white line recognition processing part 5, estimates lateral displacement 6 of the vehicle 2 to the white line 4, a lane width 7, curvature 8 of the lane 3, and an azimuth 9 of the vehicle 2, sets a threshold to departure accordingly, and adjusts the threshold on the basis of information from the collision prediction determination part 13; and a departure determination part 11 which gives an alarm that departure is likely to occur when the threshold is exceeded.

Description

  The present invention relates to a lane departure prevention apparatus for preventing a vehicle from departing from a lane.

  The lane departure prevention device prevents this when the vehicle departs from or is about to deviate by warning or steering control. However, if the vehicle equipped with the lane departure prevention device is a commercial vehicle, However, there are cases where there is not enough room for the lane width, or appropriate warnings and controls may not be output due to overhang caused by the wheelbase length. In particular, conditions are severe on ordinary roads, and Patent Document 1 has already proposed a plurality of inventions for inhibiting troublesome warnings and controls such as prohibiting lane departure warnings from the relative relationship between lane widths and vehicle widths.

  On the other hand, frontal collision accidents are always listed at the top of the cause of accidents in commercial vehicles. From the viewpoint of occupant protection as well as mitigation of harm, it is desirable to function effectively as a lane departure prevention device when there is an oncoming vehicle. ing.

JP 2005-346269 A JP 2010-92416 A JP 2010-36757 A

  However, there is a trade-off between proper lane departure prevention control in the opposite lane and the above-described troublesome warning and control suppression, and it is often difficult to achieve both. Patent Document 2 and Patent Document 3 propose more appropriate departure prevention control in the lane adjacent to the oncoming lane by detecting the oncoming vehicle, but there are many one-sided roads, and annoying alarms and control are annoying May not be performed sufficiently, and more appropriate response is required for each oncoming vehicle.

  Accordingly, an object of the present invention is to solve the above-mentioned problems, perform collision prediction on an oncoming vehicle, and adjust the control timing of lane departure according to the predicted state, thereby eliminating bothersome and safe lane departure. It is an object of the present invention to provide a lane departure prevention device that performs prevention.

  In order to solve the above problems, the present invention detects a white line recognition processing unit that detects white lines on both sides of a lane in which the host vehicle is traveling, and detects a plurality of objects on the lane in which the host vehicle is traveling or on opposite lanes as targets. A target detection unit and a collision that performs a tracking process of the target detected by the target detection unit to give ID information and type information such as movement / stop / opposite to each target and perform collision prediction for all ID targets Information of the prediction judgment unit and the white line recognition processing unit is input, and the lateral displacement to the white line of the own vehicle, the lane width, the curvature of the traveling lane, and the azimuth angle of the own vehicle are estimated and based on these, the deviation is determined. A threshold is set, and a threshold adjustment unit that adjusts the threshold based on information from the collision prediction determination unit, and a deviation when a threshold set by the threshold adjustment unit is exceeded. Les is that a deviation determination unit for performing an alarm or steering control as it is.

  The threshold adjustment unit adjusts a threshold for avoiding a collision and outputs a warning at a departure control unit when a collision is predicted by the collision prediction determination unit and the time until the collision is a predetermined value or less. The adjusted threshold value is preferably output to the departure control unit so as to avoid a collision.

  Preferably, the threshold adjustment unit controls the vehicle to travel with the adjusted threshold when no collision is predicted by the collision prediction determination unit or when the time until the collision exceeds a predetermined value.

  The target detection means comprises a millimeter wave radar, a laser radar, or a camera, and the collision prediction judgment unit discriminates an oncoming vehicle by comparing relative speed information from the target detection means with speed information of the own vehicle. Is preferred.

  According to the present invention, it is possible to perform safe lane departure prevention control without inconvenience.

It is a block diagram of the lane departure prevention apparatus which concerns on this Embodiment. It is explanatory drawing of the information acquired in a white line recognition process part. It is explanatory drawing of the prediction judgment by a collision prediction judgment part. It is explanatory drawing of the prediction judgment with respect to an oncoming vehicle. It is a flowchart which shows the flow of a process of a lane departure prevention apparatus.

  A lane departure prevention apparatus according to the present invention will be described with reference to the accompanying drawings.

  As shown in FIGS. 1 and 2, the lane departure prevention device 1 includes a white line recognition processing unit 5 that detects white lines 4 on both sides of a lane 3 on which the host vehicle 2 travels, a lane 3 on which the host vehicle 2 travels, and the opposite side. Target detection means 12 for detecting a plurality of objects on a lane (not shown) as targets T1, T2, and T3, and tracking processing of the targets T1, T2, and T3 detected by the target detection means 12 to perform each target A collision prediction determination unit 13 that assigns ID information and type information such as movement, stop, and facing to T1, T2, and T3, and performs collision prediction for all ID targets T1, T2, and T3, and a white line recognition processing unit 5 Information is input, the lateral displacement 6 of the own vehicle 2 to the white line 4, the lane width 7, the curvature 8 of the traveling lane 3, the azimuth angle 9 of the own vehicle 2 are estimated, and a threshold until departure is set based on these. And A threshold adjustment unit 10 that adjusts the threshold based on information from the collision prediction determination unit 13, and a deviation that performs warning or steering control on the assumption that there is a possibility of departure when the threshold set by the threshold adjustment unit 10 is exceeded And a determination unit 11.

  The white line recognition processing unit 5 is connected to a camera that captures an image in front of the host vehicle 2. The white line recognition processing unit 5 performs noise removal, edge extraction, and binarization from the front image 14 acquired from the camera, and extracts a straight line and a curve component by Hough transform or the like based on the result, and the vehicle 2 travels. The white line 4 is detected by narrowing down the white line 4 corresponding to the lane 3 to be performed.

  The target detection means 12 comprises a millimeter wave radar, a laser radar, or a camera, and detects the relative speed of the targets T1, T2, T3 ahead in the traveling direction, the distance and position with respect to the host vehicle 2, and the like. If the camera is a stereo camera, the distance, position, etc. can be easily detected.

  As shown in FIG. 3, the collision prediction determination unit 13 compares the relative speed information of the detection targets T1, T2, and T3 acquired from the target detection unit 12 with the vehicle speed information of the host vehicle 2, thereby detecting the detection targets T1, T2, It is determined whether T3 is an oncoming vehicle, and whether the detection targets T1, T2, and T3 are moving bodies that approach, follow or separate from the own vehicle, or are stationary objects such as parked vehicles. . Specifically, when there are a plurality of detection targets T1, T2, and T3, the collision prediction determination unit 13 performs tracking processing based on a plurality of target information obtained from the target detection unit 12, and performs ID processing and movement. Attribute such as classification information such as body, stop, oncoming vehicle and other information is given to each detection target T1, T2, T3, and collision prediction according to attribute information for all detection targets T1, T2, T3 The detection target with the highest possibility of collision is extracted.

  Collision prediction for an oncoming vehicle is based on information from the target detection means 12 and estimates the future travel trajectory of all the detection targets T1, T2, and T3, and at the same time, the vehicle speed 18 and the yaw rate (tilt angle) from the threshold adjustment unit 10 described later. 19) Based on the steering angle 20, the future travel locus of the host vehicle 2 on which the host vehicle 2 travels with the lateral displacement 6 with respect to the estimated white line 4, the azimuth angle 9 of the host vehicle 2, and the traveling curvature 8 is estimated, and the detection target It is determined whether or not the vehicle collides based on the relative relationship between the future travel locus of T1, T2, and T3 and the future travel locus of the host vehicle 2. Further, when the collision prediction determination unit 13 determines that there is a collision, the collision prediction determination unit 13 calculates a time until the collision.

  The threshold adjustment unit 10 estimates the lane width 7 (the interval between the right white line 4 and the left white line 4) using the narrowed candidate points of the white line 4, and at the same time, from the information on the vehicle speed 18, the yaw rate 19, and the steering angle 20, The lateral displacement 6 with respect to the white line 4 detected by the white line recognition processing unit 5, the azimuth angle 9 of the host vehicle 2, and the traveling curvature 8 are estimated, and these estimated values are compared with preset threshold values. This threshold value is automatically adjusted according to the vehicle speed 18, the traveling curvature 8, and the vehicle width so as to be adapted to vehicle specifications such as the vehicle width and wheelbase of the host vehicle 2. When the vehicle speed is wide and the vehicle speed 18 is high, the threshold value is adjusted so that the control can be performed more safely, the lane width 7 is narrow such as a general road, and the curvature 8 of the traveling lane 3 is less than a certain value. In this case, the threshold is adjusted so that frequent and troublesome control is not performed by narrowing down the conditions. Moreover, the threshold value adjustment part 10 adjusts a threshold value according to the collision prediction, when the collision with an oncoming vehicle is estimated.

  Threshold adjustment when the detection target with the highest possibility of collision is an oncoming vehicle will be described with reference to FIG.

  As shown in FIG. 4, the collision prediction determination unit 13 is based on the information on the own vehicle 2 input via the threshold adjustment unit 10 and the information on the detection target (oncoming vehicle 16) T <b> 1 from the collision prediction determination unit 13. When the distance in the width direction of the host vehicle 2 is x and the distance in the traveling direction is y, the current position x, y (0, 0) of the host vehicle 2 is set, and the current position (x, y) of the detection target T1 is Based on this, the position where the own vehicle 2 and the oncoming vehicle 16 pass each other and the time (time until the collision) are predicted, and the distance when the vehicle passes each other is calculated as the offset amount 21.

  When there is a sufficient offset amount 21 and no collision occurs, the collision prediction determination unit 13 keeps the vehicle 2 as it is with the threshold set by the threshold adjustment unit 10 based on the lane width 7, the vehicle speed 18, the curvature 8 of the traveling lane 3, and the like. When the vehicle 2 does not travel and the future travel locus of the host vehicle 2 and the future travel locus of each oncoming vehicle 16 collide, that is, when the offset amount 21 is negative, it is output as a collision to the threshold adjustment unit 10. Then, an alarm output or steering control is performed on the departure prevention control unit 22 via the departure determination unit 11 so as to avoid a collision.

  At this time, if the time until the collision when the collision is predicted exceeds a predetermined value, the threshold adjustment unit 10 sets the threshold of each element such as the change amount of the lateral displacement 6, the azimuth angle 9, and the curvature 8. When there is an oncoming vehicle whose time until the collision is less than or equal to a predetermined value, the threshold condition is automatically adjusted to a safer side so that the collision can be avoided. The predetermined value is set based on the vehicle speed 18 at a time sufficient for the host vehicle 2 to avoid a collision when the oncoming vehicle 16 travels with its future travel locus.

  As shown in FIG. 1, the departure determination unit 11 uses time until the vehicle 2 deviates from the white line 4 based on the future travel locus of the vehicle 2 and the threshold, and each element is a threshold adjustment unit 10. When the adjusted threshold value is exceeded, either warning or steering control or both are performed because there is a possibility of deviation.

  At this time, even if the threshold value is adjusted by the threshold adjustment unit 10 from the information of the collision prediction determination unit 13 and the threshold condition is narrowed down in the driving environment on the above-described general road, the time until the collision is less than a predetermined value If there is an oncoming vehicle, the condition is adjusted in the safe direction by the threshold set in the above condition.

  The departure determination unit 11 performs an ON / OFF switch of the winker or the lane departure prevention device 1 or a steering operation intended by the driver such as a change amount of the lateral displacement 6 up to the white line 4 is more than a certain value. A determination is made to suppress alarm output or steering control based on the threshold.

  Next, a description will be given using the flowchart shown in FIG.

  The collision prediction determination unit 13 detects, as targets T1, T2, and T3, a plurality of objects on the lane 3 on which the host vehicle 2 travels and on an opposite lane (not shown) based on information from the target detection unit 12 ( Step S1).

  The white line recognition processing unit 5 detects the white line 4 from the front video 14 input from the camera, and outputs information on the white line 4 to the threshold adjustment unit 10. The threshold adjustment unit 10 is based on the input white line 4 information. The lateral displacement 6 of the own vehicle 2 up to the white line 4, the lane width 7, the curvature 8 of the traveling lane 3, the azimuth angle 9 of the own vehicle 2 and the like are calculated to recognize the traveling environment of the own vehicle (step S2) and travel The environmental information is output to the collision prediction determination unit 13.

  Further, the threshold adjustment unit 10 acquires vehicle information such as the vehicle speed 18, the yaw rate (tilt angle) 19, and the steering angle 20 (step S <b> 3), and outputs these vehicle information to the collision prediction determination unit 13.

  The collision prediction determination unit 13 performs tracking processing based on a plurality of pieces of target information detected by the target detection unit 12, and ID information for identifying the detection targets T1, T2, and T3, and movement / stop / opposite. The type information is given as an attribute to each of the targets T1, T2, and T3 (step S4).

  Thereafter, the collision prediction determination unit 13 performs the collision prediction according to the attribute information for all the ID targets T1, T2, and T3 (step S5), extracts the ID target having the highest possibility of collision, The ID target information is output to the threshold adjustment unit 10 and the like. At this time, in the collision prediction for the oncoming vehicle, it is determined whether or not the own vehicle 2 and the oncoming vehicle 16 collide from the relative relationship between the future traveling locus of the own vehicle 2 and the future traveling locus of the oncoming vehicle 16. In case of a collision, the time until the collision is calculated. In addition, the ID target extracted by the collision prediction determination unit 13 is basically one in which a collision in the shortest time is predicted among the ID targets that are likely to collide.

  Thereafter, collision prevention control corresponding to the type is performed based on information from the collision prediction determination unit 13 and the threshold adjustment unit 10 (step S6), and lane departure prevention control is performed (step S7).

  Specifically, when the ID target extracted by the collision prediction determination unit 13 is an oncoming vehicle, it is determined whether or not the time until the collision is a predetermined value or less, and the time until the collision is a predetermined value When the vehicle speed exceeds the threshold, the vehicle travels based on a threshold automatically adjusted based on the lane width, the vehicle speed, the lane curvature, and the like. When the time until the collision is equal to or less than a predetermined value, the collision prevention control is performed under the threshold condition set by the threshold adjustment unit 10, and the alarm output or the steering control is performed as the lane departure prevention control.

  In this way, the future travel trajectory of the host vehicle and the future travel trajectory of the oncoming vehicle are compared, and when a collision is predicted based on these trajectories, the time until the collision is calculated and sufficient time for avoiding the collision ( (Predetermined value) is set, and the vehicle runs at the adjusted threshold until it reaches the predetermined value, so that the alarm does not occur frequently and becomes less than the predetermined value. When a collision is unavoidable, an alarm is sounded and the anti-collision control is steered based on a safer threshold, so that it is possible to perform an appropriate alarm and control, and it is safer. A highly reliable lane departure prevention device can be provided.

DESCRIPTION OF SYMBOLS 1 Lane departure prevention apparatus 5 White line recognition process part 10 Threshold adjustment part 11 Deviation judgment part 12 Target detection means 13 Collision prediction judgment part

Claims (4)

  A white line recognition processing unit that detects white lines on both sides of the lane in which the host vehicle is traveling, a target detection unit that detects a plurality of objects on the lane in which the host vehicle is traveling or on opposite lanes, and detection by the target detection unit A collision prediction determination unit that performs a target tracking process to provide ID information and type information such as movement / stop / opposite to each target and performs collision prediction for all ID targets, and the white line recognition processing unit Information is input, the lateral displacement of the vehicle to the white line, the lane width, the curvature of the traveling lane, the azimuth angle of the vehicle is estimated, and a threshold to departure is set based on these, and the collision prediction determination A threshold adjustment unit that adjusts the threshold based on information from the unit, and an alarm or steering control indicating that there is a risk of deviation when the threshold set by the threshold adjustment unit is exceeded. Lane departure prevention apparatus being characterized in that a deviation determination unit for performing.   The threshold adjustment unit adjusts a threshold for avoiding a collision and outputs a warning at a departure control unit when a collision is predicted by the collision prediction determination unit and the time until the collision is a predetermined value or less. The lane departure prevention apparatus according to claim 1, wherein the adjusted threshold value is output to the departure control unit to avoid a collision.   2. The lane departure according to claim 1, wherein the threshold adjustment unit controls the vehicle to travel with the adjusted threshold when a collision is not predicted by the collision prediction determination unit or when a time until the collision exceeds a predetermined value. Prevention device.   The target detection unit includes a millimeter wave radar, a laser radar, or a camera, and the collision prediction determination unit determines an oncoming vehicle by comparing relative speed information from the target detection unit with speed information of the own vehicle. Item 3. A lane departure prevention apparatus according to item 1 or 2.

Images