Method and device for estimating vehicle course angle variable quantity and storage medium

Abstract

The invention relates to a method and equipment for estimating a vehicle course angle variable quantity based on camera lane line model data, and a storage medium. The method only depends on the camera, reliable estimation of the vehicle course angle variable quantity is achieved under the controllable calculation force condition, lane line model data provided by the camera are fully utilized, a method is provided for vehicle course angle change calculation, and the method is used as a substitute for attitude estimation through other sensor equipment in a specific scene.

Description

Technical Field

[0001] The present invention relates to the field of intelligent driving technology, and in particular to a method, device and storage medium for estimating a vehicle heading angle change based on camera lane line model data. Background Art

[0002] Cameras have become an indispensable sensor device in the field of intelligent driving due to their low cost, high accuracy at short distances, and large amount of effective information captured. Especially in the field of lane line recognition and application, the technology has been accumulated deeply and is widely used, but currently the most widely used is to identify lane lines, confirm which lane is in, and the distance from the left and right lane lines. It is rare to complete the calculation and application of the vehicle's position and attitude state (mainly the change of heading angle) through continuous observation of lane lines. The main reason is that it is difficult to solve independently. The main influencing factors are:

[0003] 1. Lane lines are generally fitted into cubic polynomials, which are high-order equations.

[0004] 2. The camera's posture changes with the vehicle's posture. The vehicle's trajectory is complex and involves both distance and direction.

[0005] 3. There are accuracy and delay issues in the measurement of vehicle speed and travel time.

[0006] Because of the above factors, the lane line cubic polynomial directly obtained by camera observation is used to independently calculate the vehicle posture change. Since it involves high order and coordinate rotation, it is a nonlinear equation and has no universal solution. Although a high-precision lane line curve fitting expression is provided, it is difficult to solve directly and has no practical application value. Currently, the most commonly used method is to reduce the cubic polynomial expressed by the lane line, such as reducing it to a linear equation, to obtain an approximate solution. However, due to the loss of precision, the approximate solution obtained will have a relatively large error with the true value. Summary of the invention

[0007] In view of the technical problems existing in the prior art, the present invention provides a method, a device and a storage medium for estimating a vehicle heading angle change based on camera lane line model data.

[0008] The technical solution of the present invention to solve the above technical problems is as follows:

[0009] In a first aspect, the present invention provides a method for estimating a vehicle heading angle change, comprising:

[0010] S1, extracting lane curve model parameters from the front and rear two frames of images taken by the vehicle camera, obtaining the time interval between the front and rear two frames of images and the instantaneous vehicle speed corresponding to the front and rear two frames of images;

[0011] S2, the vehicle position in the previous and next two frames of images is the coordinate origin, the vehicle's forward direction is the horizontal axis, and the vertical axis is the direction perpendicular to the vehicle's forward direction. The lane line curve model parameters are used to establish the pre-motion coordinate system srt and the post-motion coordinate system uov, r and o represent the coordinate origin, t and v represent the horizontal axis, s and u represent the vertical axis, and the lane line curve equation based on the pre-motion coordinate system srt, the lane line curve equation based on the post-motion coordinate system uov, and the transformation relationship between the two coordinate systems are obtained;

[0012] S3, approximately estimating the vehicle mileage and the heading angle change according to the lane curve model parameters, the instantaneous vehicle speed and the time interval, and generating initial particles; the initial particles include two elements: the vehicle mileage and the heading angle change;

[0013] S4, keeping the heading angle change amount unchanged, setting the expansion step length, and performing forward and backward expansion on the vehicle mileage of the initial particle to obtain a backward expansion particle and a forward expansion particle;

[0014] S5, generating a corresponding observation point sequence of the target lane line based on the backward expansion particles and the forward expansion particles, and calculating the coordinate value of each observation point in the observation point sequence in the coordinate system before the movement, and calculating the position deviation value of each observation point in the observation point sequence in the two frames before and after the movement by using the transformation relationship between the two coordinate systems;

[0015] S6, taking the absolute values ​​of the position deviation values ​​and summing them up to obtain the absolute deviation cumulative values ​​corresponding to the two observation point sequences;

[0016] S7, comparing the two absolute deviation cumulative values ​​with the preset thresholds respectively, and taking the particle corresponding to the absolute deviation cumulative value not greater than the preset threshold as the optimal particle;

[0017] S8, outputting the heading angle change of the optimal particle as the heading angle change of the vehicle.

[0018] Furthermore, the lane curve model parameters include:

[0019] The distance between the vehicle and the target lane line in the two frames before and after: c 0 and c' 0 ,

[0020] The slope of the target lane line in the two frames before and after: c 1 and c' 1 ,

[0021] The curvature of the two frames before and after: c 2 and c' 2 ,

[0022] The curvature change rate c between the two frames of images 3and c' 3 .

[0023] Furthermore, based on the pre-motion coordinate system srt, the lane curve equation is: s = c 3 t 3 +c 2 t 2 +c 1 t+c 0 ;

[0024] Based on the post-motion coordinate system uov, the lane curve equation is: u=c' 3 v 3 +c' 2 v 2 +c' 1 v+c' 0 ;

[0025] The coordinate transformation method of any point in two coordinate systems is shown in equations (1) and (2):

[0026] srt coordinate system to uov coordinate system: v = (t + s * tanθ) * cosθ - I (1)

[0027] From uov coordinate system to srt coordinate system: s = t*tanθ-u / cosθ (2)

[0028] Where I is the straight-line distance between the vehicle positions in the previous and next two frames of images, I = sin(θ / 2)*(V+V')*ΔT / θ; θ is the change in the vehicle heading angle.

[0029] Furthermore, the generating of the corresponding observation point sequence of the target lane line based on the backward expansion particles and the forward expansion particles, and calculating the coordinate value of each observation point in the observation point sequence in the coordinate system before the movement includes:

[0030] Taking the horizontal axis coordinate of the backward extended particle or the forward extended particle as the reference, after adding the preset fixed offset, the horizontal axis coordinate t of each observation point in the observation point sequence of the corresponding target lane line is generated in sequence according to the preset point interval;

[0031] Substitute the horizontal axis coordinate of each observation point into the lane curve equation based on the SRT coordinate system to obtain the vertical axis coordinate s of each observation point in the SRT coordinate system.

[0032] Furthermore, the calculation of the position deviation value of each observation point in the observation point sequence in the two preceding and following frames of images by using the transformation relationship between the two coordinate systems includes:

[0033] Using the coordinate conversion method from the srt coordinate system to the uov coordinate system, based on the ordinate coordinate s of each observation point in the srt coordinate system, the abscissa coordinate v of each observation point in the uov coordinate system is obtained;

[0034] Substitute the horizontal axis coordinate v of each observation point in the uov coordinate system into the lane curve equation based on the uov coordinate system to obtain the vertical axis coordinate u of each observation point in the uov coordinate system;

[0035] Using the coordinate conversion method from the uov coordinate system to the srt coordinate system, based on the ordinate coordinate u of each observation point in the uov coordinate system, the ordinate coordinate s' of each observation point in the srt coordinate system is inversely calculated;

[0036] The difference between the vertical axis coordinates s and s' is the position deviation of the observation point in the two frames of images.

[0037] Furthermore, the step of comparing the two absolute deviation cumulative values ​​with a preset threshold value and taking the particle corresponding to the absolute deviation cumulative value not greater than the preset threshold value as the optimal particle includes:

[0038] The two absolute deviation cumulative values ​​are recorded as: 1,T and δ 2,T , δ 1,T is the cumulative absolute deviation of the observation point sequence corresponding to the backward extended particle, δ 2,T is the accumulated absolute deviation of the observation point sequence corresponding to the forward extended particle; the preset threshold is denoted as ε;

[0039] If δ 1,T ≤ε and δ 2,T ≤ε, then the particle corresponding to the smaller value of the two is taken as the optimal particle;

[0040] If δ 1,T ≤ε and δ 2,T >ε, or,δ 1,T >ε and δ 2,T ≤ε, the particle corresponding to the accumulated absolute deviation value not greater than the preset threshold is taken as the optimal particle;

[0041] If δ 1,T ,δ 2,T If both are greater than ε, a new initial particle is generated, and steps S5 to S7 are executed after forward and backward expansion until the optimal particle is found.

[0042] Furthermore, if δ 1,T ,δ 2,T If both are greater than ε, a new initial particle is generated, and steps S5 to S7 are executed after forward and backward expansion until the optimal particle is found, including:

[0043] S701, estimate the angular residual φ using the abscissa t of the last observation point p in the observation point sequence J in the s-r-t coordinate system and the position deviation value δ J : φ J = arctan(δ 0 / t 0 ); Let θ J = θ J - φ k , k = 1; 0 0 k j,k

[0044] S702, calculate the position deviation value δ of each observation point in the observation point sequence corresponding to the current particle in the front and rear two frames of images according to the adjusted heading angle change θ k , j = 1, 2, 3,..., J; j,k

[0045] S703, take the absolute value of the position deviation value δ and sum it to obtain the absolute deviation cumulative value δ j,k , and estimate the angular residual φ T,k = arctan(δ k / t J,k ); J T,k

[0046] S704, judge the magnitude relationship between δ T,k and the preset threshold ε. If δ T,k ≤ ε, then take the current particle as the optimal particle; if δ T,k > ε, then let θ k+1 = θ k - φ k , k = k + 1, record the angular residual sequence {φ 1 ,..., φ k};

[0047] S705, analyze the angular residual sequence {φ 1 ,..., φ k}; If the angular residual sequence converges and k < K, where K is the upper limit of the iteration times, then jump to step S702; if k = K or the angular residual sequence does not converge, then execute step S706;

[0048] S706, generate a new initial particle Q’ 0 (I’ 0 , θ’ 0 ), where I’ 0 is the vehicle driving mileage of the current particle, and θ 0 ' = θ k-1 - φ k ;

[0049] S707, maintain the heading angle change θ' 0 The vehicle mileage of the new initial particle is forwardly and backwardly extended to obtain a new backward-extended particle and a new forward-extended particle;

[0050] S708, repeat steps S5 to S7 until the optimal particle is found.

[0051] Furthermore, in step S707, the reduced expansion step size is 1 / 2 of the original expansion step size.

[0052] In a second aspect, the present invention provides an electronic device, comprising:

[0053] Memory for storing computer software programs;

[0054] The processor is used to read and execute the computer software program, thereby implementing a method for estimating a vehicle heading angle change as described in the first aspect of the present invention.

[0055] In a third aspect, the present invention provides a non-transitory computer-readable storage medium, in which a computer software program is stored. When the computer software program is executed by a processor, a method for estimating a vehicle heading angle change as described in the first aspect of the present invention is implemented.

[0056] The beneficial effects of the present invention are as follows: the present invention only relies on the camera to achieve reliable estimation of the vehicle heading angle change under controllable computing power conditions, and makes full use of the lane line model data provided by the camera to provide a method for vehicle posture (heading angle change) solution, which can serve as an alternative to using other sensor devices to estimate posture in specific scenarios (for example: replacing IMU to perform dead reckoning in tunnels or GNSS weak signal scenarios, etc.). BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 A schematic flow chart of a method for estimating a change in a vehicle heading angle provided by an embodiment of the present invention;

[0058] Figure 2 This is a schematic diagram of the coordinate system of the two frames before and after the particle movement in the actual lane line environment;

[0059] Figure 3 It is a schematic diagram of the coordinate expression of the observation point in the coordinate system of the two frames before and after the particle movement;

[0060] Figure 4 This is a schematic diagram of the optimal particle screening process;

[0061] Figure 5 A schematic diagram of an electronic device according to an embodiment of the present invention;

[0062] Figure 6 A schematic diagram of an embodiment of a computer-readable storage medium provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0063] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0064] In the description of the present application, the term "for example" is used to mean "used as an example, illustration or description". Any embodiment described as "for example" in the present application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any technician in the field to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes will not be elaborated in detail to avoid unnecessary details to obscure the description of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed in the present application.

[0065] like Figure 1 As shown, an embodiment of the present invention provides a method for estimating a vehicle heading angle change, comprising:

[0066] S1, extracting lane curve model parameters from the front and rear two frames of images taken by the vehicle-mounted camera, obtaining the time interval between the front and rear two frames of images and the instantaneous vehicle speed corresponding to the front and rear two frames of images.

[0067] The lane curve model parameters include:

[0068] The distance between the vehicle and the target lane line in the two frames before and after: c 0 and c' 0 ,

[0069] The slope of the target lane line in the two frames before and after: c 1 and c' 1 ,

[0070] The curvature of the two frames before and after: c 2 and c' 2 ,

[0071] The curvature change rate c between the two frames of images 3 and c' 3 .

[0072] The time interval between two frames of images: ΔT.

[0073] The instantaneous vehicle speed corresponding to the two frames of images: V, V'.

[0074] According to the lane curve model parameters, the initial vehicle heading angle change θ can be calculated, θ = arctan ((c' 1 -c 1 ) / (1+c' 1 *c 1 )).

[0075] The target lane line may be a lane line on the left side of the vehicle's forward direction, that is, a left lane line, or a right lane line. In this embodiment, the target lane line is a left lane line.

[0076] S2, the vehicle position in the previous and next two frames of images is taken as the coordinate origin, the vehicle's forward direction is the horizontal axis, and the vertical axis is perpendicular to the vehicle's forward direction. The lane line curve model parameters are used to establish the pre-motion coordinate system srt and the post-motion coordinate system uov, r and o represent the coordinate origin, t and v represent the horizontal axis, and s and u represent the vertical axis.

[0077] Among them, based on the pre-motion coordinate system srt, the lane curve equation is: s = c 3 t 3 +c 2 t 2 +c 1 t+c 0 ;

[0078] Based on the post-motion coordinate system uov, the lane curve equation is: u=c' 3 v 3 +c' 2 v 2 +c' 1 v+c' 0 ;like Figure 2 shown.

[0079] The coordinate transformation method of any point in two coordinate systems is shown in equations (1) and (2):

[0080] srt coordinate system to uov coordinate system: v = (t + s * tanθ) * cosθ - I (1)

[0081] From uov coordinate system to srt coordinate system: s = t*tanθ-u / cosθ (2)

[0082] Where I is the straight-line distance between the vehicle positions in the previous and next two frames of images, I = sin(θ / 2)*(V+V')*ΔT / θ; θ is the change in the vehicle heading angle.

[0083] S3, based on the lane curve model parameters, instantaneous vehicle speed and time interval, approximately estimate the vehicle mileage and heading angle change, and generate initial particles; the initial particles contain two elements, namely, Q 0 (I 0 ,θ 0 ), vehicle mileage I 0 =(V+V')*ΔT / 2, heading angle change θ 0 =θ.

[0084] S4, maintain the heading angle change θ 0 The expansion step size is set to expand the vehicle mileage of the initial particle forward and backward to obtain a backward expansion particle Q1 and a forward expansion particle Q2;

[0085] When the vehicle mileage of the initial particle is extended forward and backward, the vehicle mileage of the initial particle needs to be re-estimated. Specifically, the chord length C = 2*(I 0 / θ 0 )*sin(θ 0 / 2), then the re-estimated initial particle Q 0 ((I 0 +C) / 2,θ 0 ), the expanded backward expansion particle Q1(I 1 =(I 0 +C) / 2-D,θ 0 ), forward extension particle Q2(I 2 =(I 0 +C) / 2+D,θ 0 ), D is the expansion step size.

[0086] In this embodiment, the extended step length D is the maximum walking distance measurement error. The main sources of walking distance measurement error are: a) vehicle speed measurement error and accuracy, b) measurement time error and accuracy, and c) error caused by data transmission delay. According to the three error sources, the error range is analyzed and estimated, and the maximum walking distance measurement error is taken as the extended step length D.

[0087] S5, based on the backward expansion particle Q1 and the forward expansion particle Q2, the corresponding observation point sequence of the target lane line is generated, and the coordinate value of each observation point in the observation point sequence in the coordinate system before movement is calculated. By using the transformation relationship between the two coordinate systems, the position deviation value of each observation point in the observation point sequence in the two frames before and after is calculated.

[0088] Here we take particle Q'(I', θ') as an example to further explain this step. Figure 3 shown.

[0089] S501, let t j =I'+A+B*j, generating observation point p in the observation point sequence on the left lane line j The horizontal axis coordinate in the srt coordinate system, where A is the preset fixed offset, B is the preset point interval, j is the number of points, j=1,2,3,...,J, J≥2.

[0090] S502, t j Substitute the lane curve equation in the srt coordinate system in turn: s = c 3 t 3 +c 2 t 2 +c 1 t+c 0 , get the observation point p in the observation point sequence j The vertical axis coordinate s in the srt coordinate system j .

[0091] S503, using the coordinate conversion method from the srt coordinate system to the uov coordinate system, that is, formula (1), obtain the observation point p in the observation point sequence on the left lane line j The horizontal axis coordinate v in the uov coordinate system j .

[0092] S504, v j Substitute the lane curve equation in the uov coordinate system in sequence: u=c' 3 v 3 +c' 2 v 2 +c' 1 v+c' 0 , get the observation point p in the observation point sequence j The horizontal axis coordinate u in the uov coordinate system j .

[0093] S505, using the coordinate conversion method from the uov coordinate system to the srt coordinate system, that is, formula (2), inversely calculate the observation point p in the observation point sequence on the left lane line j The vertical axis coordinate s' in the srt coordinate system j .

[0094] S506, each observation point p in the observation point sequence corresponding to particle Q'(I', θ') j The position deviation value δ between the two frames before and after j =sj -s' j .

[0095] The backward extension particle Q1(I 1 =(I 0 +C) / 2-D,θ 0 ) and the forward extension particle Q2(I 2 =(I 0 +C) / 2+D,θ 0 ), and substitute into step S501 to step S506, and obtain the observation points p in the observation point sequence corresponding to the backward extended particle Q1. 1,j The position deviation value δ between the two frames before and after 1,j ; and, each observation point p in the observation point sequence corresponding to the forward extension particle Q2 2,j The position deviation value δ between the two frames before and after 2,j .

[0096] S6, taking absolute values ​​of the position deviation values ​​and summing them up to obtain the absolute deviation cumulative values ​​corresponding to the two observation point sequences.

[0097] The absolute deviation accumulation value of the observation point sequence corresponding to the backward extended particle Q1

[0098] The cumulative absolute deviation of the observation point sequence corresponding to the forward extended particle Q2

[0099] The absolute deviation cumulative value is used to indicate the overlap of lane curves in two coordinate systems. The smaller the absolute deviation cumulative value, the higher the overlap of lane curves in two coordinate systems. Under completely ideal conditions, the absolute deviation cumulative value is 0, which means that the lane curves in the two coordinate systems completely overlap. Under non-ideal conditions, we need to set a smaller absolute cumulative deviation threshold ε, and select particles based on the comparison result between the absolute deviation cumulative value and the preset threshold.

[0100] S7, comparing the two absolute deviation cumulative values ​​with the preset thresholds respectively, and taking the particle corresponding to the absolute deviation cumulative value not greater than the preset threshold as the optimal particle.

[0101] like Figure 4 As shown in the figure, the two absolute deviation cumulative values ​​are compared with the preset thresholds respectively, and the following results may appear:

[0102] 1)δ 1,T ≤ε and δ 2,T ≤ε;

[0103] 2)δ 1,T ≤ε or δ 2,T ≤ε;

[0104] 3)δ 1,T ,δ 2,T are all greater than ε;

[0105] When the first result occurs, compare δ 1,T ,δ 2,T , and take the particle corresponding to the smaller value of the two as the optimal particle.

[0106] When the second result occurs, the particle corresponding to the absolute deviation cumulative value not greater than the preset threshold is directly taken as the optimal particle.

[0107] When the 3rd result occurs, perform the following steps:

[0108] S701, using the Jth observation point p in the observation point sequence J The horizontal axis coordinate t in the srt coordinate system J And the position deviation value δ J , estimate the angle residual φ 0 :φ 0 =arctan(δ J / t J ); let θk = θ 0 -φ 0 , k = 1;

[0109] S702, based on the adjusted heading angle change θ k , using the method from step S501 to step S506, calculate the position deviation value δ of each observation point in the observation point sequence corresponding to the current particle in the two previous and next frames of images j,k ;

[0110] S703, position deviation value δ j,k Take the absolute value and sum it to get the absolute deviation cumulative value δ T,k , and estimate the angle residual φk=arctan(δ J ,k / t J );

[0111] S704, determine δ T,k and the size of the preset threshold ε, if δ T,k ≤ε, the current particle is taken as the optimal particle; if δ T,k >ε, then let k=k+1, θ k =θ k-1 -φ k-1 , record the angle residual sequence {φ 1 ,...,φ k};

[0112] S705, analyzing the angle residual sequence {φ 1 ,...,φ k}; If the angle residual sequence converges and k < K, where K is the upper limit of the number of iterations, then jump to step S702. If k = K or the angle residual sequence does not converge, then execute step S706;

[0113] S706, generate a new initial particle Q' 0 (I' 0 , θ' 0 ), where I' 0 is the vehicle driving mileage of the current particle, and θ 0 ' = θ k-1 - φ k ;

[0114] S707, keep the heading angle change θ' 0 unchanged, reduce the expansion step size, and perform forward and backward expansion on the vehicle driving mileage of the new initial particle to obtain a backward expansion particle Q'1(I' 0 - d, θ' 0 ) and a forward expansion particle Q'2(I' 0 + d, θ' 0 ), where d = D' / 2; D' is the original expansion step size;

[0115] S708, repeatedly execute steps S5 to S7 until the optimal particle is found.

[0116] S8, output the heading angle change of the optimal particle as the vehicle heading angle change.

[0117] Please refer to Figure 5 , Figure 5 which is a schematic diagram of an embodiment of the electronic device provided by an embodiment of the present invention. As Figure 5 shown, an embodiment of the present invention provides an electronic device 500, including a memory 510, a processor 520, and a computer program 511 stored in the memory 510 and executable on the processor 520. When the processor 520 executes the computer program 511, the following steps are implemented:

[0118] S1, extract the lane line curve model parameters from the front and rear two frames of images captured by the in-vehicle camera, obtain the time interval between the front and rear two frames of images, and the instantaneous vehicle speeds corresponding to the front and rear two frames of images;

[0119] S2, with the vehicle position in the front and rear two frames of images as the coordinate origin, the vehicle forward direction as the horizontal axis, and the direction perpendicular to the vehicle traveling direction as the vertical axis, use the lane line curve model parameters to establish a pre-motion coordinate system s-r-t and a post-motion coordinate system u-o-v respectively. r and o represent the coordinate origin, t and v represent the horizontal axis, and s and u represent the vertical axis, to obtain the lane line curve equation based on the pre-motion coordinate system s-r-t, the lane line curve equation based on the post-motion coordinate system u-o-v, and the transformation relationship between the two coordinate systems;

[0120] S3, approximately estimating the vehicle mileage and the heading angle change according to the lane curve model parameters, the instantaneous vehicle speed and the time interval, and generating initial particles; the initial particles include two elements: the vehicle mileage and the heading angle change;

[0121] S4, keeping the heading angle change amount unchanged, setting the expansion step length, and performing forward and backward expansion on the vehicle mileage of the initial particle to obtain a backward expansion particle and a forward expansion particle;

[0122] S5, generating a corresponding observation point sequence of the target lane line based on the backward expansion particles and the forward expansion particles, and calculating the coordinate value of each observation point in the observation point sequence in the coordinate system before the movement, and calculating the position deviation value of each observation point in the observation point sequence in the two frames before and after the movement by using the transformation relationship between the two coordinate systems;

[0123] S6, taking the absolute values ​​of the position deviation values ​​and summing them up to obtain the absolute deviation cumulative values ​​corresponding to the two observation point sequences;

[0124] S7, comparing the two absolute deviation cumulative values ​​with the preset thresholds respectively, and taking the particle corresponding to the absolute deviation cumulative value not greater than the preset threshold as the optimal particle;

[0125] S8, outputting the heading angle change of the optimal particle as the heading angle change of the vehicle.

[0126] See also Figure 6 , Figure 6 A schematic diagram of an embodiment of a computer-readable storage medium provided in an embodiment of the present invention. Figure 6 As shown, this embodiment provides a computer-readable storage medium 600, on which a computer program 611 is stored. When the computer program 611 is executed by a processor, the following steps are implemented:

[0127] S1, extracting lane curve model parameters from the front and rear two frames of images taken by the vehicle camera, obtaining the time interval between the front and rear two frames of images and the instantaneous vehicle speed corresponding to the front and rear two frames of images;

[0128] S2, the vehicle position in the previous and next two frames of images is the coordinate origin, the vehicle's forward direction is the horizontal axis, and the vertical axis is the direction perpendicular to the vehicle's forward direction. The lane line curve model parameters are used to establish the pre-motion coordinate system srt and the post-motion coordinate system uov, r and o represent the coordinate origin, t and v represent the horizontal axis, s and u represent the vertical axis, and the lane line curve equation based on the pre-motion coordinate system srt, the lane line curve equation based on the post-motion coordinate system uov, and the transformation relationship between the two coordinate systems are obtained;

[0129] S3, approximately estimating the vehicle mileage and the heading angle change according to the lane curve model parameters, the instantaneous vehicle speed and the time interval, and generating initial particles; the initial particles include two elements: the vehicle mileage and the heading angle change;

[0130] S4, keeping the heading angle change amount unchanged, setting the expansion step length, and performing forward and backward expansion on the vehicle mileage of the initial particle to obtain a backward expansion particle and a forward expansion particle;

[0131] S5, generating a corresponding observation point sequence of the target lane line based on the backward expansion particles and the forward expansion particles, and calculating the coordinate value of each observation point in the observation point sequence in the coordinate system before the movement, and calculating the position deviation value of each observation point in the observation point sequence in the two frames before and after the movement by using the transformation relationship between the two coordinate systems;

[0132] S6, taking the absolute values ​​of the position deviation values ​​and summing them up to obtain the absolute deviation cumulative values ​​corresponding to the two observation point sequences;

[0133] S7, comparing the two absolute deviation cumulative values ​​with the preset thresholds respectively, and taking the particle corresponding to the absolute deviation cumulative value not greater than the preset threshold as the optimal particle;

[0134] S8, outputting the heading angle change of the optimal particle as the heading angle change of the vehicle.

[0135] It should be noted that in the above embodiments, the description of each embodiment has its own emphasis, and for parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0136] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0137] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0138] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0139] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0140] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0141] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for estimating a vehicle heading angle change, characterized in that: include: S1, extracting lane curve model parameters from the front and rear two frames of images taken by the vehicle camera, obtaining the time interval between the front and rear two frames of images and the instantaneous vehicle speed corresponding to the front and rear two frames of images; S2, the vehicle position in the previous and next two frames of images is the coordinate origin, the vehicle's forward direction is the horizontal axis, and the vertical axis is the direction perpendicular to the vehicle's forward direction. The lane line curve model parameters are used to establish the pre-motion coordinate system srt and the post-motion coordinate system uov, r and o represent the coordinate origin, t and v represent the horizontal axis, s and u represent the vertical axis, and the lane line curve equation based on the pre-motion coordinate system srt, the lane line curve equation based on the post-motion coordinate system uov, and the transformation relationship between the two coordinate systems are obtained; S3, approximately estimating the vehicle mileage and the heading angle change according to the lane curve model parameters, the instantaneous vehicle speed and the time interval, and generating initial particles; the initial particles include two elements: the vehicle mileage and the heading angle change; S4, keeping the heading angle change amount unchanged, setting the expansion step length, and performing forward and backward expansion on the vehicle mileage of the initial particle to obtain a backward expansion particle and a forward expansion particle; S5, generating a corresponding observation point sequence of the target lane line based on the backward expansion particles and the forward expansion particles, and calculating the coordinate value of each observation point in the observation point sequence in the coordinate system before the movement, and calculating the position deviation value of each observation point in the observation point sequence in the two frames before and after the movement by using the transformation relationship between the two coordinate systems; S6, taking the absolute values ​​of the position deviation values ​​and summing them up to obtain the absolute deviation cumulative values ​​corresponding to the two observation point sequences; S7, comparing the two absolute deviation cumulative values ​​with the preset thresholds respectively, and taking the particle corresponding to the absolute deviation cumulative value not greater than the preset threshold as the optimal particle; S8, outputting the heading angle change of the optimal particle as the heading angle change of the vehicle.

2. The method according to claim 1, characterized in that The lane curve model parameters include: The distance between the vehicle and the target lane line in the two frames before and after: c0 and c'0, The slope of the target lane line in the two frames before and after: c1 and c'1, The curvatures in the two frames before and after: c2 and c'2, The curvature change rates c3 and c'3 in the two frames.

3. The method according to claim 2, characterized in that Based on the pre-motion coordinate system srt, the lane curve equation is: s = c3t 3 +c2t 2 +c1t+c0; Based on the post-motion coordinate system uov, the lane curve equation is: u=c'3v 3 +c'2v 2 +c'1v+c'0; The coordinate transformation method of any point in two coordinate systems is shown in equations (1) and (2): srt coordinate system to uov coordinate system: v = (t + s * tanθ) * cosθ - I (1) From uov coordinate system to srt coordinate system: s = t*tanθ-u / cosθ(2) Where I is the straight-line distance between the vehicle positions in the previous and next two frames of images, I = sin(θ / 2)*(V+V')*ΔT / θ; θ is the change in the vehicle heading angle.

4. The method according to claim 1, characterized in that: The generating of the corresponding observation point sequence of the target lane line based on the backward expansion particles and the forward expansion particles, and calculating the coordinate value of each observation point in the observation point sequence in the coordinate system before the movement, includes: Taking the horizontal axis coordinate of the backward extended particle or the forward extended particle as the reference, after adding the preset fixed offset, the horizontal axis coordinate t of each observation point in the observation point sequence of the corresponding target lane line is generated in sequence according to the preset point interval; Substitute the horizontal axis coordinate of each observation point into the lane curve equation based on the SRT coordinate system to obtain the vertical axis coordinate s of each observation point in the SRT coordinate system.

5. The method according to claim 4, characterized in that The method of calculating the position deviation value of each observation point in the observation point sequence in the two preceding and following frames of images by using the transformation relationship between the two coordinate systems includes: Using the coordinate conversion method from the srt coordinate system to the uov coordinate system, based on the ordinate coordinate s of each observation point in the srt coordinate system, the abscissa coordinate v of each observation point in the uov coordinate system is obtained; Substitute the horizontal axis coordinate v of each observation point in the uov coordinate system into the lane curve equation based on the uov coordinate system to obtain the vertical axis coordinate u of each observation point in the uov coordinate system; Using the coordinate conversion method from the uov coordinate system to the srt coordinate system, based on the ordinate coordinate u of each observation point in the uov coordinate system, the ordinate coordinate s' of each observation point in the srt coordinate system is inversely calculated; The difference between the vertical axis coordinates s and s' is the position deviation of the observation point in the two frames of images.

6. The method according to claim 1, characterized in that The step of comparing the two absolute deviation cumulative values ​​with the preset threshold values ​​and taking the particle corresponding to the absolute deviation cumulative value not greater than the preset threshold value as the optimal particle comprises: The two absolute deviation cumulative values ​​are recorded as: 1,T and δ 2,T , δ 1,T is the cumulative absolute deviation of the observation point sequence corresponding to the backward extended particle, δ 2,T is the accumulated absolute deviation of the observation point sequence corresponding to the forward extended particle; the preset threshold is denoted as ε; If δ 1,T ≤ε and δ 2,T ≤ε, then the particle corresponding to the smaller value of the two is taken as the optimal particle; If δ 1,T ≤ε and δ 2,T >ε, or,δ 1,T >ε and δ 2,T ≤ε, the particle corresponding to the accumulated absolute deviation value not greater than the preset threshold is taken as the optimal particle; If δ 1,T , δ 2,T If both are greater than ε, a new initial particle is generated, and steps S5 to S7 are executed after forward and backward expansion until the optimal particle is found.

7. The method according to claim 6, characterized in that If δ 1,T , δ 2,T If both are greater than ε, a new initial particle is generated, and steps S5 to S7 are executed after forward and backward expansion until the optimal particle is found, including: S701, using the last observation point p in the observation point sequence J The horizontal axis coordinate t in the srt coordinate system J And the position deviation value δ J , estimate the angle residual φ0: φ0 = arctan(δ J / t J );Let θ k =θ0-φ0, k=1; S702, based on the adjusted heading angle change θ k , calculate the position deviation value δ of each observation point in the observation point sequence corresponding to the current particle in the two frames before and after j,k , j = 1, 2, 3, ..., J; S703, position deviation value δ j,k Take the absolute value and sum it to get the absolute deviation cumulative value δ T,k , and estimate the angle residual φ k =arctan(δ J,k / t J ); S704, determine δ T,k and the size of the preset threshold ε, if δ T,k ≤ε, the current particle is taken as the optimal particle; if δ T,k >ε, then let θ k+1 =θ k -φ k , k=k+1, record the angle residual sequence {φ1,...,φ k }; S705, analyze the angular residual sequence {φ1,..., φ k}; if the angular residual sequence converges and k < K, where K is the upper limit of the number of iterations, then jump to step S702; if k = K or the angular residual sequence does not converge, then execute step S706; S706, generate a new initial particle Q'0 (I'0, θ'0), I'0 is the vehicle mileage of the current particle, θ0' = θ k-1 -φ k ; S707, keeping the heading angle change θ'0 unchanged, reducing the expansion step, and performing forward and backward expansion on the vehicle mileage of the new initial particle to obtain a new backward expansion particle and a forward expansion particle; S708, repeat steps S5 to S7 until the optimal particle is found.

8. The method according to claim 7, characterized in that In step S707, the reduced expansion step size is 1 / 2 of the original expansion step size.

9. An electronic device, characterized in that: include: Memory for storing computer software programs; A processor is used to read and execute the computer software program, thereby implementing a method for estimating a vehicle heading angle change as described in any one of claims 1-8.

10. A non-transitory computer-readable storage medium, characterized in that: The storage medium stores a computer software program, and when the computer software program is executed by a processor, the method for estimating a vehicle heading angle change amount as described in any one of claims 1 to 8 is implemented.

Images