Lane Line Detection Using Lidar Point Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting lane lines using lidar data in autonomous vehicles are not accurate due to errors in point detection and symmetry analysis, leading to incorrect lane line identification.
Innovation Solution
A method and system that utilize a processor to detect points estimated as lane lines by estimating mathematical model parameters, calculating distances, and assigning scores, with repeated sampling and symmetry analysis to accurately determine the lane line based on lidar data, including weighting points for improved accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If current methods are used to detect lane lines using lidar data, then the detection process is simple, but the accuracy of lane line identification is poor
Solution Approach 1:
The patent segments the lane line detection process into distinct operational phases: point cloud generation from lidar data, point filtering to extract candidate lane line points, mathematical model parameter estimation, score calculation based on geometric constraints, and final lane line determination. This segmentation allows each phase to be optimized independently, improving overall accuracy while managing complexity through modular processing
Solution Approach 2:
The patent implements feedback mechanisms where detected points are evaluated against mathematical models, scores are calculated based on how well points fit expected lane line geometries, and the results feed into iterative refinement processes. The system uses feedback from symmetry analysis and score aggregation to continuously improve lane line detection accuracy through multiple processing passes
2Reliability
If point detection and symmetry analysis are performed without weighting, then the processing is faster, but errors in lane line identification increase
Solution Approach 1:
The patent applies local quality by introducing weighting factors that assign different importance levels to different detected points based on their reliability, position, and contribution to lane line formation. High-weight points (from reliable symmetry analysis and within expected lane line regions) have greater influence on the final lane line determination, while low-weight points are downplayed or discarded, improving reliability without uniformly increasing processing time for all points
Solution Approach 2:
The patent changes parameters by introducing weight values as an additional parameter for each detected point, transforming the detection process from uniform treatment of all points to differential processing based on point quality. This parameter change enables the system to focus computational resources on high-value points while maintaining reasonable processing speeds through selective detailed analysis
3Measurement precision
If mathematical model parameters are estimated using all detected points, then the model is more comprehensive, but the computational complexity increases
Solution Approach 1:
The patent applies partial action by estimating mathematical model parameters using only a selected subset of detected points rather than all points. The system identifies and uses high-quality points that most contribute to accurate lane line modeling, performing comprehensive analysis on these critical points while using simplified or skipped processing for less important points, thereby reducing computational power consumption while maintaining model accuracy
Data Source
AI summary
A method of detecting a lane line based on lidar data can include detecting, by a processor, points each estimated as a lane line in a lidar data, performing, by the processor, an estimation operation of estimating parameters of a mathematical model using the detected points, and performing, by the processor, a setting operation of calculating distances between each of the detected points and the mathematical model in which the parameters are estimated and setting the calculated distances as scores. The method can further include performing, by the processor, a summation operation of summing the scores, and setting, by the processor, the mathematical model determined according to the summation score as a lane line.