Method, system and vehicle for identifying lane where target object is located

Abstract

The present invention discloses a method for identifying the lane where a target object is located, comprising the following steps: S1, collecting a road image in front of a current vehicle, and fitting a lane line curve in the road image; S2, extracting the target object in the road image, and calculating the distance dy from the target object to the left lane line and the right lane line of the lane r where the current vehicle is located. left ,dy right , and then calculate the distribution probability P of the target object on both sides of the left lane line and the right lane line of lane r L 、P R S3. Calculate the probability P1 that the target object is in lane r. If P1 is greater than the probability threshold, the target object is determined to be in lane r. As the target object approaches the lane line, the probability of the target object being in that lane decreases as the lane line standard deviation increases, reducing the probability of misjudging the target object's lane.

Description

Technical Field

[0001] The present invention belongs to the technical field of vehicle assisted driving, and more specifically, the present invention relates to a method, system and vehicle for identifying a lane where a target object is located. Background Art

[0002] When the ADAS system implements longitudinal control, it needs to know the lane in which the target object in front is located, and adopt different control strategies for the target object in the lane and the target objects in the adjacent left and right lanes. Therefore, the system needs to obtain the lane information of the target object.

[0003] Combine Figure 1 Existing methods for identifying the lane where a target object is located are described. The process for identifying the lane where a target object is located is as follows: obtaining the distances d1 and d2 between the edge of the target object and the corresponding lane line, determining the degree to which the target object is in contact with the lane line, and assuming that the target object is in the current lane when both distances d1 and d2 are greater than a set distance threshold. This method for identifying the lane where a target object is located has the following problems:

[0004] When the target object is changing lanes by crossing the line, there may be a moment when it is neither in the current lane nor in the adjacent lane. The ACC function may believe that there are no obstacles ahead, which may cause a traffic accident. Summary of the Invention

[0005] The present invention provides a method for identifying a lane where a target object is located, aiming to improve the recognition accuracy of the lane where the target object is located.

[0006] The present invention is implemented as follows: a method for identifying a lane where a target object is located, the method comprising the following steps:

[0007] S1, collecting the road image in front of the current vehicle and fitting the lane line curve in the road image;

[0008] S2. Extract the target object in the road image and calculate the distance dy from the target object to the left lane line and the right lane line of the lane r where the current vehicle is located. left ,dy right , and then calculate the distribution probability P of the target object on both sides of the left lane line and the right lane line of lane r L 、P R ;

[0009] S3. Calculate the probability P1 that the target object is in lane r. If the probability P1 is greater than the probability threshold, determine that the target object is in lane r.

[0010] Further, the distance left ,dy right The acquisition process is as follows:

[0011] Determine the coordinates of the target object in the vehicle coordinate system of the current vehicle (x obj ,y obj );

[0012] Calculate the distance dy between the target object and the left lane line and the right lane line of lane r in the vehicle coordinate system of the current vehicle left ,dy right .

[0013] Further, the distance left ,dy right The calculation formula is as follows:

[0014] dy left =y obj -y left (x obj );

[0015] dy right =y obj -y right (x obj );

[0016] Among them, (x obj ,y obj ) is the coordinate of the target object, y left (x obj ) is the left lane line of lane r at x obj The vertical coordinate, y right (x obj ) is the right lane line of lane r at x obj The vertical coordinate of .

[0017] Furthermore, the distribution probability P L The calculation formula is as follows:

[0018]

[0019] σ lm is the standard deviation of lane line recognition by the monocular camera, P L = 0 means the target object is located on the left lane line of lane r, P L = -1 means the target object is completely on the left side of the left lane line of lane r, P L =1 means the target object is completely on the right side of the left lane line of lane r;

[0020] Probability P R The calculation formula is as follows:

[0021]

[0022] P R= 0 means the target object is located on the right lane line of lane r, P R = -1 means the target object is completely on the left side of the right lane line of lane r, P R =1 indicates that the target object is completely to the right of the right lane line of lane r.

[0023] Furthermore, the probability P1 of the target object in lane r is calculated as:

[0024] P1=(P L -P R ) / 2.

[0025] Furthermore, if it is determined that the target object is not in lane r, the probability P2 that the target object is in the lane to the left of lane r, the probability P3 that the target object is in the lane to the right of lane r, the probability P4 that the target object is on the left side of the left adjacent lane, and the probability P5 that the target object is on the right side of the right adjacent lane are calculated. The lane corresponding to the maximum probability value among probabilities P2, P3, P4, and P5 is the lane where the target object is located.

[0026] Furthermore, the calculation formulas for probability P2, probability P3, probability P4 and probability P5 are as follows:

[0027] P2=(P LL -P L ) / 2;

[0028] P3=(P R -P RR ) / 2;

[0029] P4=(P RBL -P LL ) / 2;

[0030] P5=(P RR -P RBR ) / 2;

[0031] Among them, P RBL =1,P RBR =-1, P LL P represents the distribution probability of the target object on both sides of the next left lane of the left adjacent lane. RR Represents the distribution probability of the target object on both sides of the next right lane of the right adjacent lane.

[0032] Furthermore, the lane line curve fitting method is as follows:

[0033] Lane lines are extracted from the road image, and the lane lines are divided into equal intervals along the vehicle's travel direction to form lane segments. The coordinates of all lane segments in the vehicle coordinate system are obtained, and then the lane line curves of each lane are fitted.

[0034] The present invention is implemented as follows: a system for identifying a lane where a target object is located, the system comprising:

[0035] The monocular camera captures the road image in front of the current vehicle and sends it to the processor. The processor determines the lane where each target object in the road image is located based on the above-mentioned lane recognition method.

[0036] The present invention is achieved by providing a car, wherein the car is integrated with a recognition system for the lane where the target object is located.

[0037] The method for identifying the lane where the target object is located provided by the present invention has the following beneficial effects:

[0038] The lane model where the target object is located is quantified into a one-dimensional finite sample space S = {in the current lane, in the left adjacent lane, in the right adjacent lane, outside the left adjacent lane, outside the right adjacent lane}, and the corresponding probabilities of the target object in each lane are: P1, P2, P3, P4, P5, satisfying: (1) P1+P2+P3+P4+P5=1, which means that there will always be: the probability of a target object in a certain lane is greater than or equal to the probability of the target object in any other lane. Therefore, when the target object is driving on the lane line, the ACC function will not think that there is no obstacle ahead; (2) Probability calculation formula and It is related to the standard deviation of the lane lines. When the target object approaches the lane line, the probability of the target object in that lane decreases as the standard deviation of the lane line increases, which will reduce the probability of misjudging the lane where the target object is located. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 A schematic diagram of lane identification for an existing target object provided by an embodiment of the present invention;

[0040] Figure 2 A flow chart of a method for identifying a lane where a target object is located provided in an embodiment of the present invention;

[0041] Figure 3 A schematic diagram of a vehicle coordinate system provided by an embodiment of the present invention;

[0042] Figure 4 A schematic diagram of a 5-lane scenario provided by an embodiment of the present invention;

[0043] Figure 5 A schematic diagram of the distribution probability of an object within a lane with a lane line standard deviation of 1m provided in an embodiment of the present invention;

[0044] Figure 6 Schematic diagram of the distribution probability of the target lane with a lane line standard deviation of 0.1m provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0045] The specific implementation methods of the present invention will be further explained in detail below by describing the embodiments with reference to the accompanying drawings, so as to help those skilled in the art to have a more complete, accurate and in-depth understanding of the inventive concept and technical solution of the present invention.

[0046] Figure 2 A flow chart of a method for identifying a lane where a target object is located provided in an embodiment of the present invention. The method specifically includes the following steps:

[0047] S1, collecting the road image in front of the current vehicle and fitting the lane line curve in the road image;

[0048] In this embodiment of the present invention, a vehicle coordinate system of the current vehicle is defined, with the midpoint of the rear axle of the current vehicle as the coordinate origin, the vehicle's travel direction as the x-axis, and the vehicle's width as the y-axis. Lane lines are extracted from the road image, and the lane lines are divided into equal intervals along the vehicle's travel direction to form lane segments. The coordinates of all lane segments in the vehicle coordinate system are obtained, and then the lane line curve y is fitted. i [n], its expression is as follows:

[0049] y i [n]=(C0+C1×x i [n]+C2×(x i [n]) 2 +C3×(x i [n]) 3 )

[0050] That is, y(x) = C0 + C1×x + C2×x 2 +C3×x 3

[0051] Among them, x i [n] represents the horizontal coordinate of the midpoint of the nth lane segment in the i-th lane line. The horizontal coordinate represents the distance from the origin of the vehicle coordinate system in the direction of the lane line extension. i [n] represents the ordinate of the midpoint of the nth lane segment in the i-th lane line. The ordinate represents the distance from the origin of the vehicle coordinate system in the lane width direction. C0, C1, C2, and C3 are fitting coefficients.

[0052] S2. Extract the target object in the road image and calculate the distance dy from the target object to the left lane line and the right lane line of the lane r where the current vehicle is located. left ,dy right , and then calculate the distribution probability P of the target object on both sides of the left lane line and the right lane line of lane r L 、P R ;

[0053] In this embodiment of the present invention, the distance dy between the target object and the left lane line and the right lane line of lane r is left ,dy right The specific method of obtaining is as follows:

[0054] First, determine the coordinates of the target object in the vehicle coordinate system of the current vehicle (x obj ,y obj ), calculate the distance dy between the target object and the left lane line and the right lane line of lane r in the vehicle coordinate system of the current vehicle left ,dy right , and its calculation formula is as follows:

[0055] dy left =y obj -y left (x obj );

[0056] dy right =y obj -y right (x obj );

[0057] Among them, (x obj ,y obj ) is the coordinate of the target object in the vehicle coordinate system, y left (x obj ) is the left lane line of lane r at x obj The vertical coordinate, y right (x obj ) is the right lane line of lane r at x obj The vertical coordinate of .

[0058] In the embodiment of the present invention, the right side of the current vehicle is taken as the positive direction of the x-axis and the driving direction is taken as the positive direction of the y-axis. The error function is used to calculate the distribution probability P L 、P R , where the error function is: The probability of a function that obeys a normal distribution from mean μ to a specified x is as follows: dx=|x-μ|, σ is the standard deviation of the function.

[0059] This method can be used to calculate the probability P of the target object on both sides of the lane line on the left side of lane r. L : σ lm is the standard deviation of the lane line recognized by the camera. At this time, P L = 0 means the target object is located on the left lane line of lane r, P L = -1 means the target object is completely on the left side of the left lane line of lane r, P L= 1 means that the target object is completely on the right side of the left lane line of lane r, such as Figure 3 As shown, the gray area is the standard deviation of the lane line (approximately equal to the error). When the target object overlaps with the lane line error area, P L The absolute value of is between 0 and 1, and the maximum recognition (-1 or 1) can only be achieved when the target is completely out of the gray area.

[0060] The same method is used to calculate the probability P of the target object being near the lane line on the right side of lane r. R : At this time, P R = 0 means the target object is located on the right lane line of lane r, P R = -1 means the target object is completely on the left side of the right lane line of lane r, P R =1 indicates that the target object is completely to the right of the right lane line of lane r.

[0061] S3. Calculate the probability P1 that the target object is in lane r. If the probability P1 is greater than the probability threshold, the target object is determined to be in lane r. The probability P1 of the target object in lane r is calculated as follows:

[0062] P1=(P L -P R ) / 2

[0063] When probability P1 is greater than the set probability threshold, the target object is determined to be in lane r. If the target object is determined not to be in lane r, the probability P2 that the target object is in the left adjacent lane of lane r, the probability P3 that the target object is in the right adjacent lane of lane r, the probability P4 on the left side of the left adjacent lane, and the probability P5 on the right side of the right adjacent lane are calculated. The lane corresponding to the maximum probability value among probabilities P2, P3, P4, and P5 is the lane where the target object is located. If probability P4 is the largest, the target object is determined to be on the left side of the left adjacent lane.

[0064] Combine Figure 4 To illustrate the calculation of probability P2, probability P3, probability P4 and probability P5, Figure 4 Taking 5 lanes as an example, the calculation formulas for probability P2, probability P3, probability P4, and probability P5 are as follows:

[0065] P2=(P LL -P L ) / 2;

[0066] P3=(P R -P RR ) / 2;

[0067] P4=(P RBL -P LL ) / 2;

[0068] P5=(P RR -P RBR ) / 2;

[0069] Among them, P RBL =1,P RBR =-1, P LL P represents the distribution probability of the target object on both sides of the next left lane of the left adjacent lane. RR Indicates the distribution probability of the target object on both sides of the next right lane of the right adjacent lane. Probability P1, probability P2, probability P3, probability P4 and probability P5 are normalized. The normalization process is explained using probability P1 as an example. The normalization of other probability values ​​is the same as that of probability P1. The probability after normalization is

[0070] The probability distribution of the target object in the lane is as follows: Figure 5 、 Figure 6 As shown, Figure 5 The standard deviation of lane lines is 1 meter. Figure 6 The lane line standard deviation is 0.1 meters, and the probability threshold is 70%. It can be seen that when the lane line standard deviation is low, the probability distribution is more concentrated, and the judgment result of the target object's lane is more accurate.

[0071] The present invention also provides a system for identifying a lane where a target object is located, the system comprising:

[0072] The monocular camera captures the road image in front of the current vehicle and sends it to the processor. The processor determines the lane where each target object in the road image is located based on the above-mentioned lane recognition method.

[0073] The present invention also provides a car, which is integrated with the above-mentioned target object lane recognition system, wherein the monocular camera is integrated into the front windshield, and the processor is separately provided in the vehicle or integrated into the vehicle controller of the vehicle.

[0074] The present invention has been described exemplarily. Obviously, the specific implementation of the present invention is not limited to the above-mentioned method. As long as various non-substantial improvements are made using the method concept and technical solution of the present invention, or the concept and technical solution of the present invention are directly applied to other occasions without improvement, they are all within the scope of protection of the present invention.

Claims

1. A method for identifying a lane where a target object is located, characterized in that: The method comprises the following steps: S1, collecting the road image in front of the current vehicle and fitting the lane line curve in the road image; S2. Extract the target object in the road image and calculate the distance dy from the target object to the left lane line and the right lane line of the lane r where the current vehicle is located. left ,dy right , and then calculate the distribution probability P of the target object on both sides of the left lane line and the right lane line of lane r L 、P R ; S3. Calculate the probability P1 that the target object is in lane r. If the probability P1 is greater than the probability threshold, determine that the target object is in lane r. Distribution probability P L The calculation formula is as follows: σ lm is the standard deviation of lane line recognition by the monocular camera, P L = 0 means the target object is located on the left lane line of lane r, P L = -1 means the target object is completely on the left side of the left lane line of lane r, P L =1 means the target object is completely on the right side of the left lane line of lane r; Probability P R The calculation formula is as follows: P R = 0 means the target object is located on the right lane line of lane r, P R = -1 means the target object is completely on the left side of the right lane line of lane r, P R =1 means the target object is completely to the right of the right lane line of lane r; The probability P1 of the target object in lane r is calculated as: P1=(P L -P R ) / 2; If the target object is determined not to be in lane r, calculate the probability P2 that the target object is in the lane to the left of lane r, the probability P3 that the target object is in the lane to the right of lane r, the probability P4 that the target object is on the left side of the left adjacent lane, and the probability P5 that the target object is on the right side of the right adjacent lane. The lane corresponding to the maximum probability value among probabilities P2, P3, P4, and P5 is the lane where the target object is located. The calculation formulas for probability P2, probability P3, probability P4 and probability P5 are as follows: P2=(P LL -P L ) / 2; P3=(P R -P RR ) / 2; P4=(P RBL -P LL ) / 2; P5=(P RR -P RBR ) / 2; Among them, P RBL =1,P RBR =-1, P LL P represents the distribution probability of the target object on both sides of the next left lane of the left adjacent lane. RR Represents the distribution probability of the target object on both sides of the next right lane of the right adjacent lane.

2. The method for identifying the lane where the target object is located according to claim 1, characterized in that: Distance dy left ,dy right The acquisition process is as follows: Determine the coordinates of the target object in the vehicle coordinate system of the current vehicle (x obj ,y obj ); Calculate the distance dy between the target object and the left lane line and the right lane line of lane r in the vehicle coordinate system of the current vehicle left ,dy right .

3. The method for identifying the lane where the target object is located according to claim 2, characterized in that: Distance dy left ,dy right The calculation formula is as follows: of left =y obj -the left (x obj ): of right =y obj -the right (x obj ): Among them, (x obj ,y obj ) is the coordinate of the target object, y left (x obj ) is the left lane line of lane r at x obj The vertical coordinate, y right (x obj ) is the right lane line of lane r at x obj The vertical coordinate of .

4. The method for identifying the lane where the target object is located according to claim 1, characterized in that: The lane line curve fitting method is as follows: Lane lines are extracted from the road image, and the lane lines are divided into equal intervals along the vehicle's travel direction to form lane segments. The coordinates of all lane segments in the vehicle coordinate system are obtained, and then the lane line curves of each lane are fitted.

5. A lane identification system for a target object, characterized in that: The system comprises: A monocular camera captures an image of the road ahead of the current vehicle and sends it to a processor. The processor determines the lane where each target object in the road image is located based on the lane identification method of any one of claims 1 to 4.

6. A car, characterized in that: The vehicle is integrated with a lane identification system for the target object as claimed in claim 4.

Images