Multi-frame dynamic high-precision lane line detection method for automatic driving

Abstract

The invention relates to a multi-frame dynamic high-precision lane line detection method for automatic driving, and the method comprises the steps: M1, building data information of a historical POI point pool of a road, and obtaining the data information of POI points of each frame of the road in real time; and M2, based on the data information of the POI points of each frame of the road and the data information of the historical POI point pool of the road, updating the historical POI point pool of the road by adopting a time sequence updating algorithm of a lane line to obtain the updated data information of the POI point pool of the road, and representing the track of the lane line of the road by adopting a multi-frame data fusion and track smoothing algorithm of the road. According to the method, key lane line information can be reserved and updated in a multi-frame detection result, the space-time consistency of the detection result is remarkably improved, and the method can adapt to changes of different vehicle speeds and lane widths, so that a more stable detection result is realized in expressways and wide lane scenes.

Description

Technical Field

[0001] The present invention relates to the technical field of lane line detection, and in particular to a multi-frame dynamic high-precision lane line detection method for autonomous driving. Background Art

[0002] A core task of an autonomous driving system is to accurately identify road lane lines and maintain the continuity of detection to ensure safe driving of the vehicle and the accuracy of path planning. Traditional lane line detection methods usually rely on a single sensor mode, such as a camera or a lidar. Cameras provide rich texture information and high-resolution data, but are easily affected by lighting conditions. For example, environmental factors such as strong light, shadows, and rain and fog may cause detection errors. In contrast, lidar can provide more accurate depth information and has good anti-interference capabilities, but its resolution is low and it is difficult to capture the subtle details of lane lines, especially when the road surface is blurred. Therefore, multimodal data fusion methods based on cameras and lidars have gradually become the mainstream technology in the field of autonomous driving lane line detection.

[0003] At present, the implementation of multimodal fusion methods mainly includes the following strategies:

[0004] Direct projection method: This method projects the LiDAR point cloud data directly onto the image plane, combining the texture information of the image with the depth information of the point cloud for lane line detection. This method relies on accurate calibration of the sensor, is easily affected by position deviation, and may cause detection incoherence in dynamic scenes.

[0005] Feature layer fusion method: extract features from the image and point cloud respectively, and map them to the same coordinate system for matching and fusion. This method can effectively improve the accuracy of lane line detection, but it has high requirements for feature matching and high computational overhead under complex road conditions, making it difficult to meet real-time requirements.

[0006] Multi-frame data fusion method: This method uses the spatiotemporal consistency of multi-frame detection results to improve the robustness and continuity of detection by combining the data of historical frames with the current frame. This method can effectively improve the consistency of lane line detection in dynamic scenes. However, the existing multi-frame data fusion technology is mainly based on simple geometric consistency detection, which is difficult to effectively deal with detection drift and loss problems in dynamic environments, and it is difficult to maintain the integrity of long-distance lane lines. Summary of the invention

[0007] In view of the above problems, the present invention provides a multi-frame dynamic high-precision lane line detection method for autonomous driving, which can not only retain and update key lane line information in the multi-frame detection results, significantly improving the temporal and spatial consistency of the detection results, but also adapt to changes in different vehicle speeds and lane widths, thereby achieving more stable detection results in highway and wide lane scenarios.

[0008] In order to achieve the above-mentioned purpose and other related purposes, the technical solution provided by the present invention is as follows:

[0009] A multi-frame dynamic high-precision lane line detection method for autonomous driving, the method comprising:

[0010] M1. Establish the data information of the historical POI point pool of the road, and obtain the data information of the POI point of each frame of the road in real time;

[0011] M2. Based on the data information of the POI point of each frame of the road and the data information of the historical POI point pool of the road, the historical POI point pool of the road is updated using the lane line timing update algorithm to obtain the data information of the POI point pool of the updated road;

[0012] M3. Based on the data information of the updated POI point pool of the road and the data information of the historical POI point pool of the road, the trajectory of the lane line of the road is characterized by using the multi-frame data fusion and trajectory smoothing algorithm of the road to obtain the data information of the trajectory of the lane line of the road;

[0013] M4. Based on the data information of the trajectory of the lane line of the road, the lane line is detected using an algorithm of dynamic segmented step size adjustment and large lane clustering expansion to obtain data information of the detection result of the lane line of the road.

[0014] Furthermore, in step M2, the updating of the historical POI point pool of the road using the lane line time sequence updating algorithm includes:

[0015] M21. Based on the data information of the POI point of each frame of the road and the data information of the historical POI point pool of the road, a matching function F of the POI point of each frame of the road is established.

[0016]

[0017] Among them, x t is the data information of the POI point in the tth frame of the road, y t is the data information of the tth frame in the historical POI point pool of the road, d is the preset distance threshold, f is the Euclidean distance function of the POI point of the road, the historical POI point pool of the road is matched to obtain the data information of the POI point pool of the matched road;

[0018] M22. Based on the data information of the POI point pool of the matched road, the number of occurrences of the POI point in each frame is obtained, and a preset number is set. If the number of occurrences of the POI point in each frame is less than the preset number, it is eliminated. If the number of occurrences of the POI point in each frame is greater than the preset number, it is retained, and the data information of the POI point pool of the road after elimination is obtained;

[0019] M23. Based on the data information of the POI point pool of the road after the elimination, a geometric consistency detection function Q of the POI point pool of the road is established,

[0020]

[0021] Among them, A t is the area of ​​the POI point in the current frame, A t-1 is the area of ​​the POI point in the previous frame, and a preset change rate threshold δ is set. If Q>δ, it is eliminated. If Q≤δ, the historical POI point pool of the road is updated to obtain the data information of the POI point pool of the updated road.

[0022] Furthermore, the Euclidean distance function f of the POI point of the road is,

[0023]

[0024] Among them, x t is the data information of the POI point in the tth frame of the road, y t is the data information of the tth frame in the historical POI point pool of the road, and α1, α2 and α3 are the weight coefficients of the POI points.

[0025] Furthermore, the constraints of the weight coefficients α1, α2 and α3 of the POI point are:

[0026] Furthermore, in step M3, the characterization of the trajectory of the lane line of the road by using the multi-frame data fusion and trajectory smoothing algorithm of the road includes:

[0027] M31. Based on the data information of the updated POI point pool of the road and the data information of the historical POI point pool of the road, a multi-frame data fusion function W of the road is established,

[0028]

[0029] Among them, g is the data information of the POI point pool of the updated road, h is the data information of the historical POI point pool of the road, β1, β2 and β3 are the weighting coefficients of the POI point pool of the road, and the data of the POI point pool of the road are fused to obtain the data information of the fused POI point pool of the road;

[0030] M32. Based on the data information of the fused road POI point pool, a multi-frame smoothing function R of the road POI point pool is established.

[0031]

[0032] Among them, z is the data information of the POI point pool of the fused road, δ1, δ2 and δ3 are the smoothing factors of the POI point pool of the road;

[0033] M33. Based on the multi-frame smoothing function R of the POI point pool of the road, the trajectory of the lane line of the road is characterized to obtain data information of the trajectory of the lane line of the road.

[0034] Furthermore, the constraint function r of the smoothing factors δ1, δ2 and δ3 of the POI point pool of the road is,

[0035]

[0036] Among them, the value range of the constraint function r is (1,2).

[0037] Furthermore, in step M4, the lane line detection using the algorithm of dynamic segmentation step size adjustment and large lane clustering expansion includes:

[0038] M41. Based on the data information of the lane line trajectory of the road, a dynamic segmented step length function G of the lane line trajectory is established,

[0039]

[0040] Wherein, a is the data information of the trajectory of the lane line of the road, γ1, γ2 and γ3 are the dynamic division factors of the trajectory of the lane line, and the trajectory of the lane line is divided to obtain the data information of the trajectory of the divided lane line;

[0041] M42. Based on the data information of the trajectory of the divided lane lines, a cluster expansion function H of the lane lines is established.

[0042]

[0043] Among them, b is the data information of the trajectory of the divided lane line, η1, η2 and η3 are the clustering expansion factors of the lane line, and the trajectory of the divided lane line is topologically analyzed to obtain the data information of the trajectory of the topological lane line;

[0044] M43. Based on the data information of the trajectory of the lane line after the topology, establish a lane line detection function S,

[0045]

[0046] Among them, c is the data information of the trajectory of the lane line after topology, ρ1, ρ2 and ρ3 are the detection factors of the lane line, and the lane line is detected to obtain the data information of the detection result of the lane line of the road.

[0047] Furthermore, the constraints of the lane line detection factors ρ1, ρ2 and ρ3 are as follows:

[0048] In order to achieve the above-mentioned purpose and other related purposes, the present invention also provides a multi-frame dynamic high-precision lane line detection system for autonomous driving, including a computer device, which is programmed or configured to execute the steps of any one of the multi-frame dynamic high-precision lane line detection methods for autonomous driving.

[0049] In order to achieve the above-mentioned purpose and other related purposes, the present invention also provides a computer-readable storage medium, which stores a computer program programmed or configured to execute any one of the multi-frame dynamic high-precision lane line detection methods for autonomous driving.

[0050] The present invention has the following positive effects:

[0051] 1. The present invention updates the historical POI point pool of the road by adopting the time-series update algorithm of the lane line, and characterizes the trajectory of the lane line of the road by combining the multi-frame data fusion and trajectory smoothing algorithm of the road. It can not only retain and update the key lane line information in the multi-frame detection results, significantly improving the spatiotemporal consistency of the detection results, but also with the support of multi-frame data, even in complex scenarios such as vehicle turning and ramp driving, the system can still maintain the continuity of the lane line, effectively avoiding detection interruption and false detection.

[0052] 2. The present invention detects lane lines by adopting an algorithm of dynamic segmentation step size adjustment and large lane clustering expansion. It can not only adapt to changes in different vehicle speeds and lane widths, thereby achieving more stable detection results in highway and wide lane scenarios, but also the adaptive dynamic segmentation step size ensures flexible adjustment of lane line detection density under different vehicle speed conditions. At the same time, the clustering expansion method enhances the integrity of long-distance lane lines. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a schematic diagram of the method flow of the present invention;

[0054] Figure 2 A schematic diagram of the flow chart of the lane line timing update algorithm of the present invention;

[0055] Figure 3 It is a schematic diagram of the flow of the multi-frame data fusion and trajectory smoothing algorithm of the road of the present invention;

[0056] Figure 4 It is a flowchart of the algorithm for dynamic segmentation step size adjustment and large lane clustering expansion of the present invention. DETAILED DESCRIPTION

[0057] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0058] Example 1: Figure 1 As shown, a multi-frame dynamic high-precision lane line detection method for autonomous driving, the method comprising:

[0059] M1. Establish the data information of the historical POI point pool of the road, and obtain the data information of the POI point of each frame of the road in real time;

[0060] M2. Based on the data information of the POI point of each frame of the road and the data information of the historical POI point pool of the road, the historical POI point pool of the road is updated using the lane line timing update algorithm to obtain the data information of the POI point pool of the updated road;

[0061] M3. Based on the data information of the updated POI point pool of the road and the data information of the historical POI point pool of the road, the trajectory of the lane line of the road is characterized by using the multi-frame data fusion and trajectory smoothing algorithm of the road to obtain the data information of the trajectory of the lane line of the road;

[0062] M4. Based on the data information of the trajectory of the lane line of the road, the lane line is detected using an algorithm of dynamic segmented step size adjustment and large lane clustering expansion to obtain data information of the detection result of the lane line of the road.

[0063] In this embodiment, if Figure 2 As shown, in step M2, the use of the lane line time sequence update algorithm to update the historical POI point pool of the road includes:

[0064] M21. Based on the data information of the POI point of each frame of the road and the data information of the historical POI point pool of the road, a matching function F of the POI point of each frame of the road is established.

[0065]

[0066] Among them, x t is the data information of the POI point in the tth frame of the road, y t is the data information of the tth frame in the historical POI point pool of the road, d is the preset distance threshold, f is the Euclidean distance function of the POI point of the road, the historical POI point pool of the road is matched to obtain the data information of the POI point pool of the matched road;

[0067] M22. Based on the data information of the POI point pool of the matched road, the number of occurrences of the POI point in each frame is obtained, and a preset number is set. If the number of occurrences of the POI point in each frame is less than the preset number, it is eliminated. If the number of occurrences of the POI point in each frame is greater than the preset number, it is retained, and the data information of the POI point pool of the road after elimination is obtained;

[0068] M23. Based on the data information of the POI point pool of the road after the elimination, a geometric consistency detection function Q of the POI point pool of the road is established,

[0069]

[0070] Among them, A t is the area of ​​the POI point in the current frame, A t-1 is the area of ​​the POI point in the previous frame, and a preset change rate threshold δ is set. If Q>δ, it is eliminated. If Q≤δ, the historical POI point pool of the road is updated to obtain the data information of the POI point pool of the updated road.

[0071] In this embodiment, the Euclidean distance function f of the POI point of the road is:

[0072]

[0073] Among them, x t is the data information of the POI point in the tth frame of the road, y t is the data information of the tth frame in the historical POI point pool of the road, and α1, α2 and α3 are the weight coefficients of the POI points.

[0074] In this embodiment, the constraints of the weight coefficients α1, α2 and α3 of the POI point are:

[0075]

[0076] In this embodiment, if Figure 3 As shown, in step M3, the characterization of the trajectory of the lane line of the road by using the multi-frame data fusion and trajectory smoothing algorithm of the road includes:

[0077] M31. Based on the data information of the updated POI point pool of the road and the data information of the historical POI point pool of the road, a multi-frame data fusion function W of the road is established,

[0078]

[0079] Among them, g is the data information of the POI point pool of the updated road, h is the data information of the historical POI point pool of the road, β1, β2 and β3 are the weighting coefficients of the POI point pool of the road, and the data of the POI point pool of the road are fused to obtain the data information of the fused POI point pool of the road;

[0080] M32. Based on the data information of the fused road POI point pool, a multi-frame smoothing function R of the road POI point pool is established.

[0081]

[0082] Among them, z is the data information of the POI point pool of the fused road, δ1, δ2 and δ3 are the smoothing factors of the POI point pool of the road;

[0083] M33. Based on the multi-frame smoothing function R of the POI point pool of the road, the trajectory of the lane line of the road is characterized to obtain data information of the trajectory of the lane line of the road.

[0084] In this embodiment, the constraint function r of the smoothing factors δ1, δ2 and δ3 of the POI point pool of the road is,

[0085]

[0086] Among them, the value range of the constraint function r is (1,2).

[0087] Example 2: Based on the multi-frame dynamic high-precision lane line detection method for autonomous driving in Example 1, the present invention is further illustrated and described below.

[0088] like Figure 1 As shown, a multi-frame dynamic high-precision lane line detection method for autonomous driving, the method comprising:

[0089] M1. Establish the data information of the historical POI point pool of the road, and obtain the data information of the POI point of each frame of the road in real time;

[0090] M2. Based on the data information of the POI point of each frame of the road and the data information of the historical POI point pool of the road, the historical POI point pool of the road is updated using the lane line timing update algorithm to obtain the data information of the POI point pool of the updated road;

[0091] M3. Based on the data information of the updated POI point pool of the road and the data information of the historical POI point pool of the road, the trajectory of the lane line of the road is characterized by using the multi-frame data fusion and trajectory smoothing algorithm of the road to obtain the data information of the trajectory of the lane line of the road;

[0092] M4. Based on the data information of the trajectory of the lane line of the road, the lane line is detected using an algorithm of dynamic segmented step size adjustment and large lane clustering expansion to obtain data information of the detection result of the lane line of the road.

[0093] In this embodiment, if Figure 4 As shown, in step M4, the lane line detection using the algorithm of dynamic segmentation step size adjustment and large lane clustering expansion includes:

[0094] M41. Based on the data information of the lane line trajectory of the road, a dynamic segmented step length function G of the lane line trajectory is established,

[0095]

[0096] Wherein, a is the data information of the trajectory of the lane line of the road, γ1, γ2 and γ3 are the dynamic division factors of the trajectory of the lane line, and the trajectory of the lane line is divided to obtain the data information of the trajectory of the divided lane line;

[0097] M42. Based on the data information of the trajectory of the divided lane lines, a cluster expansion function H of the lane lines is established.

[0098]

[0099] Among them, b is the data information of the trajectory of the divided lane line, η1, η2 and η3 are the clustering expansion factors of the lane line, and the trajectory of the divided lane line is topologically analyzed to obtain the data information of the trajectory of the topological lane line;

[0100] M43. Based on the data information of the trajectory of the lane line after the topology, establish a lane line detection function S,

[0101]

[0102] Among them, c is the data information of the trajectory of the lane line after topology, ρ1, ρ2 and ρ3 are the detection factors of the lane line, and the lane line is detected to obtain the data information of the detection result of the lane line of the road.

[0103] In this embodiment, the constraints of the lane line detection factors ρ1, ρ2 and ρ3 are:

[0104]

[0105] In this embodiment, the present invention provides a multi-frame dynamic high-precision lane line detection system for autonomous driving, including a computer device that is programmed or configured to execute any one of the steps of the multi-frame dynamic high-precision lane line detection method for autonomous driving.

[0106] In this embodiment, the present invention provides a computer-readable storage medium, on which is stored a computer program programmed or configured to execute any one of the multi-frame dynamic high-precision lane line detection methods for autonomous driving.

[0107] Any reference to memory, storage, database or other media used in the embodiments provided in the present application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM) and memory bus dynamic RAM (RDRAM).

[0108] In summary, the present invention can not only retain and update key lane line information in multi-frame detection results, significantly improving the spatiotemporal consistency of detection results, but also adapt to changes in different vehicle speeds and lane widths, thereby achieving more stable detection results in highway and wide lane scenarios.

[0109] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A multi-frame dynamic high-precision lane line detection method for autonomous driving, characterized in that: The method comprises: M1. Establish the data information of the historical POI point pool of the road, and obtain the data information of the POI point of each frame of the road in real time; M2. Based on the data information of the POI point of each frame of the road and the data information of the historical POI point pool of the road, the historical POI point pool of the road is updated using the lane line timing update algorithm to obtain the data information of the POI point pool of the updated road; M3. Based on the data information of the updated POI point pool of the road and the data information of the historical POI point pool of the road, the trajectory of the lane line of the road is characterized by using the multi-frame data fusion and trajectory smoothing algorithm of the road to obtain the data information of the trajectory of the lane line of the road; M4. Based on the data information of the trajectory of the lane line of the road, the lane line is detected using an algorithm of dynamic segmented step size adjustment and large lane clustering expansion to obtain data information of the detection result of the lane line of the road.

2. The multi-frame dynamic high-precision lane line detection method for autonomous driving according to claim 1, characterized in that: In step M2, the updating of the historical POI point pool of the road using the lane line time sequence updating algorithm includes: M21. Based on the data information of the POI point of each frame of the road and the data information of the historical POI point pool of the road, a matching function F of the POI point of each frame of the road is established. Among them, x t is the data information of the POI point in the tth frame of the road, y t is the data information of the tth frame in the historical POI point pool of the road, d is the preset distance threshold, f is the Euclidean distance function of the POI point of the road, the historical POI point pool of the road is matched to obtain the data information of the POI point pool of the matched road; M22. Based on the data information of the POI point pool of the matched road, the number of occurrences of the POI point in each frame is obtained, and a preset number is set. If the number of occurrences of the POI point in each frame is less than the preset number, it is eliminated. If the number of occurrences of the POI point in each frame is greater than the preset number, it is retained, and the data information of the POI point pool of the road after elimination is obtained; M23. Based on the data information of the POI point pool of the road after the elimination, a geometric consistency detection function Q of the POI point pool of the road is established, Among them, A t is the area of ​​the POI point in the current frame, A t-1 is the area of ​​the POI point in the previous frame, and a preset change rate threshold δ is set. If Q>δ, it is eliminated. If Q≤δ, the historical POI point pool of the road is updated to obtain the data information of the POI point pool of the updated road.

3. The multi-frame dynamic high-precision lane line detection method for autonomous driving according to claim 2, characterized in that: The Euclidean distance function f of the POI point of the road is, Among them, x t is the data information of the POI point in the tth frame of the road, y t is the data information of the tth frame in the historical POI point pool of the road, and α1, α2 and α3 are the weight coefficients of the POI points.

4. The multi-frame dynamic high-precision lane line detection method for autonomous driving according to claim 3, characterized in that: The constraints of the weight coefficients α1, α2 and α3 of the POI point are:

5. The multi-frame dynamic high-precision lane line detection method for autonomous driving according to claim 1, characterized in that: In step M3, the characterization of the trajectory of the lane line of the road by using the multi-frame data fusion and trajectory smoothing algorithm of the road includes: M31. Based on the data information of the updated POI point pool of the road and the data information of the historical POI point pool of the road, a multi-frame data fusion function W of the road is established, Among them, g is the data information of the POI point pool of the updated road, h is the data information of the historical POI point pool of the road, β1, β2 and β3 are the weighting coefficients of the POI point pool of the road, and the data of the POI point pool of the road are fused to obtain the data information of the fused POI point pool of the road; M32. Based on the data information of the fused road POI point pool, a multi-frame smoothing function R of the road POI point pool is established. Among them, z is the data information of the POI point pool of the fused road, δ1, δ2 and δ3 are the smoothing factors of the POI point pool of the road; M33. Based on the multi-frame smoothing function R of the POI point pool of the road, the trajectory of the lane line of the road is characterized to obtain data information of the trajectory of the lane line of the road.

6. The multi-frame dynamic high-precision lane line detection method for autonomous driving according to claim 5, characterized in that: The constraint function r of the smoothing factors δ1, δ2 and δ3 of the POI point pool of the road is, Among them, the value range of the constraint function r is (1,2).

7. The multi-frame dynamic high-precision lane line detection method for autonomous driving according to claim 1, characterized in that: In step M4, the lane line detection using the dynamic segmentation step size adjustment and large lane clustering expansion algorithm includes: M41. Based on the data information of the lane line trajectory of the road, a dynamic segmented step length function G of the lane line trajectory is established, Wherein, a is the data information of the trajectory of the lane line of the road, γ1, γ2 and γ3 are the dynamic division factors of the trajectory of the lane line, and the trajectory of the lane line is divided to obtain the data information of the trajectory of the divided lane line; M42. Based on the data information of the trajectory of the divided lane lines, a cluster expansion function H of the lane lines is established. Among them, b is the data information of the trajectory of the divided lane line, η1, η2 and η3 are the clustering expansion factors of the lane line, and the trajectory of the divided lane line is topologically analyzed to obtain the data information of the trajectory of the topological lane line; M43. Based on the data information of the trajectory of the lane line after the topology, establish a lane line detection function S, Among them, c is the data information of the trajectory of the lane line after topology, ρ1, ρ2 and ρ3 are the detection factors of the lane line, and the lane line is detected to obtain the data information of the detection result of the lane line of the road.

8. The multi-frame dynamic high-precision lane line detection method for autonomous driving according to claim 7, characterized in that: The constraints of the lane line detection factors ρ1, ρ2 and ρ3 are:

9. A multi-frame dynamic high-precision lane line detection system for autonomous driving, comprising a computer device, characterized in that: The computer device is programmed or configured to execute the steps of the multi-frame dynamic high-precision lane line detection method for autonomous driving as described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program that is programmed or configured to execute the multi-frame dynamic high-precision lane line detection method for autonomous driving as described in any one of claims 1 to 8.

Images