CN-122024188-A - Lane marking fitting method, device, equipment, medium and product

CN122024188ACN 122024188 ACN122024188 ACN 122024188ACN-122024188-A

Abstract

The application discloses a lane marking fitting method, a lane marking fitting device, lane marking fitting equipment, lane marking fitting medium and lane marking fitting products, wherein each marking vector is determined according to acquired multi-source data; the method comprises the steps of calculating a mean azimuth angle and a marking centroid according to each marking vector, determining a datum line, dividing each grid by taking the datum line as a symmetry axis of the grid according to a preset interval distance and a preset grid width, and respectively calculating different parts of the lane marking according to marking vectors of areas where each grid is located, so that the lane marking is obtained through combination. According to the method, the lane line dividing grids are calculated in a zoning mode based on the datum lines, the lane lines are converged to data which relatively accord with the real lane lines, interference of abnormal and jump data is reduced, and finally the lane lines with relatively higher fitting generation precision and quality are generated.

Inventors

HUANG XIAOMIN
CHEN LU
ZHAO QIHENG
HUANG WENJUN
LIU YANG
HU MIN
ZHANG JIN
PAN LEI
HE WEIJIANG
YANG KANG
CHEN YANLEI
HAN YUNBO
YU ZHITAO
WANG ZHUN
Chen Jicang
WANG LEYI
CAI GUILIN
CAO JIAQING

Assignees

中国移动通信集团设计院有限公司
中国移动通信集团有限公司

Dates

Publication Date: 20260512
Application Date: 20241112

Claims (13)

1. A lane marking fitting method, the method comprising: determining each marking vector according to the acquired multi-source data; calculating a mean azimuth angle and a marking centroid according to each marking vector, and determining a datum line; Dividing each grid by taking the datum line as a symmetry axis of the grid according to a preset interval distance and a preset grid width; and respectively calculating different parts of the lane marking according to marking vectors of the areas where the grids are positioned, and combining to obtain the lane marking.
2. The lane-marking fitting method according to claim 1, wherein calculating different portions of the lane-marking from the marking vectors of the areas where the respective grids are located, and combining to obtain the lane-marking, comprises: Splitting reticle vectors spanning different grids into different reticle vectors; calculating the grid vector of each grid according to the distributed line vector in each grid; and generating the lane marking according to each grid vector fitting.
3. The lane-marking fitting method of claim 1, wherein generating a fitted lane-marking from each grid vector comprises: Sequentially starting and ending the vector of the grid vectors of two adjacent grids; Starting from the first grid vector, sequentially connecting the vector starting point of the next grid vector with the vector ending point of the next grid vector to form a continuous marking; and thinning and smoothing the continuous marked lines to generate the lane marked lines.
4. The lane-marking fitting method of claim 3 wherein calculating a grid vector for each grid from the distributed marking vectors within each grid further comprises: Calculating according to the vector starting points of the marking vectors distributed in each grid, and determining the vector starting points of the grid vectors of each grid; and calculating according to the vector end points of the line vectors distributed in each grid, and determining the vector end points of the grid vectors of each grid.
5. The lane-marking fitting method according to claim 1, wherein before dividing each grid with the reference line as the symmetry axis of the grid according to a preset separation distance and a preset grid width, the method further comprises: calculating the transverse distance from each marking vector to the datum line; According to the transverse distance of each marking vector, determining a peak window with the largest distribution quantity of marking vectors in a preset statistical window; and eliminating the marked line vector outside the peak value window.
6. The lane-marking fitting method according to claim 5, wherein calculating the lateral distance of the respective marking vector to the reference line comprises: calculating the midpoint of each marking vector according to the starting point and the ending point of each marking vector; and calculating the vertical line distance from the middle point of each marking vector to the datum line as the transverse distance of each marking vector.
7. The lane-marking fitting method of claim 1, further comprising: Calculating coverage rate of the line vectors in the area range determined by different widths according to the datum line as a center; And taking the width reaching the preset first coverage rate as the width of the grid.
8. The lane-marking fitting method of claim 1 wherein calculating a mean azimuth and a marking centroid from the respective marking vectors, determining a fiducial line comprises: Calculating azimuth angles of the marking vectors according to the starting points and the ending points of the marking vectors; calculating the mean azimuth according to the azimuth of each marking vector; calculating the marking centroids of all marking vectors according to the starting points and the ending points of the marking vectors; and determining an extension line passing through the center of mass of the marking and presenting the mean azimuth as the datum line.
9. The lane-marking fitting method according to claim 8, wherein after calculating the mean azimuth from the azimuth of the respective marking vector, the method further comprises: Calculating standard deviation of azimuth angles of the marking vectors and the mean azimuth angle; Removing abnormal marked line vectors with standard deviation larger than a preset first standard deviation threshold value; and re-calculating the mean azimuth according to the rest marked line vectors.
10. A lane marking fitting apparatus, the apparatus comprising: The vector acquisition module is used for determining each marking vector according to the acquired multi-source data; The datum line determining module is used for calculating a mean azimuth angle and a datum line centroid according to each datum line vector and determining a datum line; The grid dividing module is used for dividing each grid by taking the datum line as a symmetry axis of the grid according to a preset interval distance and a preset grid width; And the marking fitting module is used for respectively calculating different parts of the lane marking according to marking vectors of the areas where the grids are positioned, and combining to obtain the lane marking.
11. A terminal device comprising a processor, a memory and a computer program stored in the memory and configured to be executed by the processor, the processor implementing the lane-marking fitting method according to any one of claims 1 to 9 when executing the computer program.
12. A computer readable storage medium, characterized in that the computer readable storage medium comprises a stored computer program, wherein the computer program, when run, controls a device in which the computer readable storage medium is located to perform the lane marking fitting method according to any one of claims 1 to 9.
13. A computer program product comprising computer programs/instructions which, when executed by a processor, implement the steps of the method of any of claims 1 to 9.

Description

Lane marking fitting method, device, equipment, medium and product Technical Field The invention relates to the technical field of GIS (geographic information system), in particular to a lane marking fitting method, a lane marking fitting device, lane marking fitting equipment, lane marking fitting media and lane marking fitting products. Background With the development of map updating technology, the updating mode of directly automatically generating map elements through sensing data of multi-source hardware equipment replaces the previous mode of updating map elements through manual drawing operation in terms of cost and efficiency, the element data sensed by various equipment of different manufacturers are random in precision, sensing difference, algorithm identification and the like, the final result data are slightly different due to the fact that sensing conditions are different at the time and sensing algorithms are different, the data jump mode is more various, the deviation between the result and quality after the fitting of the existing scheme and a real scene is large, and the fitting precision is poor. Disclosure of Invention In order to solve the problems, the invention provides a lane marking fitting method, a lane marking fitting device, lane marking fitting equipment, a lane marking fitting medium and a lane marking fitting product, which are used for realizing accurate fitting of lane markings of sensing data of multi-source equipment. The embodiment of the invention provides a lane marking fitting method, which comprises the following steps: determining each marking vector according to the acquired multi-source data; calculating a mean azimuth angle and a marking centroid according to each marking vector, and determining a datum line; Dividing each grid by taking the datum line as a symmetry axis of the grid according to a preset interval distance and a preset grid width; and respectively calculating different parts of the lane marking according to marking vectors of the areas where the grids are positioned, and combining to obtain the lane marking. Preferably, calculating different parts of the lane markings according to the marking vectors of the areas where the grids are located, and combining to obtain the lane markings includes: Splitting reticle vectors spanning different grids into different reticle vectors; calculating the grid vector of each grid according to the distributed line vector in each grid; and generating the lane marking according to each grid vector fitting. As a preferred aspect, generating a fitted lane-marking from each grid vector includes: Sequentially starting and ending the vector of the grid vectors of two adjacent grids; Starting from the first grid vector, sequentially connecting the vector starting point of the next grid vector with the vector ending point of the next grid vector to form a continuous marking; and thinning and smoothing the continuous marked lines to generate the lane marked lines. Further, calculating a grid vector of each grid according to the distributed reticle vector in each grid, further includes: Calculating according to the vector starting points of the marking vectors distributed in each grid, and determining the vector starting points of the grid vectors of each grid; and calculating according to the vector end points of the line vectors distributed in each grid, and determining the vector end points of the grid vectors of each grid. Preferably, before dividing each grid according to the preset interval distance and the preset grid width by taking the reference line as the symmetry axis of the grid, the method further comprises: calculating the transverse distance from each marking vector to the datum line; According to the transverse distance of each marking vector, determining a peak window with the largest distribution quantity of marking vectors in a preset statistical window; and eliminating the marked line vector outside the peak value window. Further, calculating a lateral distance of the respective reticle vector to the fiducial line includes: calculating the midpoint of each marking vector according to the starting point and the ending point of each marking vector; and calculating the vertical line distance from the middle point of each marking vector to the datum line as the transverse distance of each marking vector. Preferably, the method further comprises: Calculating coverage rate of the line vectors in the area range determined by different widths according to the datum line as a center; And taking the width reaching the preset first coverage rate as the width of the grid. As a preferred scheme, calculating a mean azimuth and a reticle centroid according to the respective reticle vectors, and determining a datum line includes: Calculating azimuth angles of the marking vectors according to the starting points and the ending points of the marking vectors; calculating the mean azimuth according to the azimuth of each mar