Method and apparatus of matching lane line data, device and storage medium

Patent metadata

Field	Value
Publication number	US-12008820-B2
Application number	US-202117351832-A
Country	US
Kind code	B2
Filing date	Jun 18, 2021
Priority date	Dec 15, 2020
Publication date	Jun 11, 2024
Grant date	Jun 11, 2024

Abstract

The present disclosure provides a method of matching lane line data, which relates to a field of automatic driving, and in particular to a field of locating an autonomous vehicle. The method of matching lane line data includes: obtaining a first data indicating sensed lane lines and a second data indicating map lane lines respectively; constructing a correlation weight matrix for the first data and the second data, based on differences between the sensed lane lines and the map lane lines; and matching the first data with the second data according to the correlation weight matrix, so as to obtain a third data having an optimal match with the first data from the second data. The present disclosure further discloses an apparatus of matching lane line data, a device, a storage medium and a computer program product.

First claim

We claim:

1. A method of matching lane line data, the method comprising:
obtaining a first data indicating sensed lane lines and a second data indicating map lane lines respectively;
constructing a correlation weight matrix for the first data and the second data, based on differences between the sensed lane lines and the map lane lines, the constructing the correlation weight matrix comprising:
determining, based on differences in horizontal intercept between the sensed lane lines and the map lane lines, a first difference matrix,
determining, based on differences in relative yaw angle between the sensed lane lines and the map lane lines, a second difference matrix,
determining, based on control point offsets between the sensed lane lines and the map lane lines, a third difference matrix, and
constructing, according to the first difference matrix, the second difference matrix and the third difference matrix, the correlation weight matrix for the first data and the second data; and

matching the first data with the second data according to the correlation weight matrix, so as to obtain a third data having an optimal match with the first data from the second data.

2. The method of claim 1 , wherein the determining a first difference matrix comprises:
obtaining a first intercept of each of the sensed lane lines on a horizontal axis in a vehicle body coordinate system;
obtaining a second intercept of each of the map lane lines on the horizontal axis in the vehicle body coordinate system; and
determining the first difference matrix, according to absolute values of differences between the first intercept of each sensed lane line and the second intercept of each map lane line.

3. The method of claim 1 , wherein the determining a second difference matrix comprises:
obtaining a first yaw angle of each of the sensed lane lines relative to a horizontal axis in a vehicle body coordinate system;
obtaining a second yaw angle of each of the map lane lines relative to the horizontal axis in the vehicle body coordinate system; and
determining the second difference matrix, according to absolute values of differences between the first yaw angle of each sensed lane line and the second yaw angle of each map lane line.

4. The method of claim 1 , wherein the determining a third difference matrix comprises:
selecting at least one control point on each of the sensed lane lines;
determining distances from the control point on each of the sensed lane lines to each of the map lane lines, and setting the determined distances as control point offsets; and
determining the third difference matrix, according to the control point offsets.

5. The method of claim 1 , wherein the constructing the correlation weight matrix comprises:
constructing a sensed lane line set according to the sensed lane lines related to the first difference matrix, the second difference matrix and the third difference matrix, and constructing a map lane line set according to the map lane lines related to the first difference matrix, the second difference matrix and the third difference matrix;
determining correlation weights between each sensed lane line in the sensed lane line set and each map lane line in the map lane line set, according to the first difference matrix, the second difference matrix and the third difference matrix; and
constructing the correlation weight matrix for the first data and the second data, according to the correlation weights between each sensed lane line in the sensed lane line set and each map lane line in the map lane line set.

6. The method of claim 5 , wherein the determining correlation weights comprises:
obtaining first differences between each sensed lane line in the sensed lane line set and each map lane line in the map lane line set from the first difference matrix, obtaining second differences between each sensed lane line in the sensed lane line set and each map lane line in the map lane line set from the second difference matrix, obtaining third differences between each sensed lane line in the sensed lane line set and each map lane line in the map lane line set from the third difference matrix;
normalizing the first differences, the second differences and the third differences; and
determining the correlation weights between each sensed lane line and each map lane line according to:
weight i =c i *Σ j (1−error i,j ),

wherein, weight i is a correlation weight of an i-th matched pair of the sensed lane lines and the map lane lines, c i is a correction coefficient determined according to a position of the sensed lane lines, error_norm i,j is a normalized j-th difference of an i-th matched pair of the sensed lane lines and the map lane lines, i and j are natural numbers, 1≤i≤m*n, 1≤j≤3, m is a number of sensed lane lines in the sensed lane line set, and n is a number of map lane lines in the map lane line set.

7. The method of claim 1 , wherein the matching the first data with the second data comprises:
determining matched pairs having a highest correlation weight value for each of the sensed lane lines respectively from matched pairs related to said each of the sensed lane line, according to the correlation weight matrix;
setting the matched pairs having the highest correlation weight value for each of the sensed lane lines respectively as a match between the first data and the second data, in response to map lane lines related to the matched pairs having the highest correlation weight value being different from each other; and
determining an optimal match between the sensed lane lines and the map lane lines for the correlation weight matrix based on a maximum matching set of a weighted bipartite graph, and setting the optimal match as the match between the first data and the second data, in response to same map lane lines existing in the matched pairs having the highest correlation weight value.

8. The method of claim 1 , further comprising, before the constructing a correlation weight matrix:
comparing a distance between sensed boundary lines with a predetermined boundary distance, in response to determining that the obtained sensed lane lines include the sensed boundary lines;
modifying the sensed boundary lines to outermost sensed lane lines, in response to an absolute value of a difference between the distance between the sensed boundary lines and the predetermined boundary distance being smaller than or equal to a first threshold; and
discarding the sensed boundary lines, in response to the absolute value of the difference between the distance between the sensed boundary lines and the predetermined boundary distance being greater than the first threshold.

9. The method of claim 8 , further comprising, before the constructing a correlation weight matrix, selecting a set of sensed lane lines, according to horizontal intercepts of the sensed lane lines on a horizontal axis in a vehicle body coordinate system and yaw angles of the sensed lane lines relative to the horizontal axis in the vehicle body coordinate system, so as to obtain a candidate sensed lane line set.

10. The method of claim 9 , wherein the constructing a correlation weight matrix comprises constructing the correlation weight matrix for the first data and the second data, based on differences between sensed lane lines included in the candidate sensed lane line set and the map lane lines.

11. An electronic device, comprising:
at least one processor; and
a memory communicatively connected to the at least one processor,
wherein the memory stores instructions executable by the at least one processor, and the instructions, when executed by the at least one processor, cause the at least one processor t

Assignees

• Beijing Baidu Netcom Sci & Tech Co Ltd

Inventors

• Yang Pengbin
• Yang Yang
• Zhang Jianxu
• Chen Wenlong

Classifications

• G05D1/81

  Handing over between on-board automatic and on-board manual control · CPC title

• G05D1/689

  Pointing payloads towards fixed or moving targets (positioning towed, pushed or suspended implements G05D1/672) · CPC title

• G05D1/249

  from positioning sensors located off-board the vehicle, e.g. from cameras · CPC title

• G06V2201/07

  Target detection · CPC title

• G05D1/0246

  using a video camera in combination with image processing means · CPC title