Multi-Camera Lane Line Fusion with Extrinsic Parameter Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The precision and stability of lane line fusion results are compromised due to camera installation errors, extrinsic parameter calibration errors, and image processing errors when using cameras with different fields of view (FOVs) in intelligent driving systems.
Innovation Solution
A method and apparatus that determine a lane line by correcting the extrinsic parameter of one camera based on the other, aligning pixel point sets in a vehicle local coordinate system, and combining them to reduce lateral deviations, thereby improving the consistency and stability of the fused lane line perception.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If cameras with different FOVs are used for lane line perception, then the coverage area and perception capability are improved, but lateral deviations occur due to camera installation errors and extrinsic parameter calibration errors
Solution Approach 1:
The patent corrects extrinsic parameters (rotation matrix and translation vector) of cameras with different FOVs to eliminate lateral deviations. By adjusting the parameters of the projection matrix and correcting the extrinsic parameter calibration errors, the system achieves precise fusion of lane lines from multiple cameras while maintaining wide coverage area
Solution Approach 2:
The patent introduces an intermediary correction process that transforms lane line pixel points from different FOV cameras into a unified coordinate system. This intermediary transformation step, involving projection matrix correction and coordinate system alignment, enables precise fusion by mediating between the different camera perspectives
2Adaptability or versatility
If extrinsic parameter calibration is performed for multiple cameras, then the fusion capability is improved, but calibration errors and installation errors cause lateral deviations
Solution Approach 1:
The patent implements a feedback mechanism where the corrected extrinsic parameters are used to transform lane line pixel points into a unified coordinate system, and the transformation results are evaluated for consistency. This feedback loop ensures that calibration errors are minimized and the fusion results remain stable and reliable
Solution Approach 2:
The patent performs preliminary correction of extrinsic parameters before the actual lane line fusion process. By pre-calibrating and correcting the projection matrices and extrinsic parameters of all cameras, the system establishes a reliable foundation for subsequent fusion operations, preventing calibration errors from affecting the final results
Data Source
AI summary
A method includes: determining a first image corresponding to a first camera and a second image; determining, based on the first image, a first lane line pixel point set; determining, based on the second image, a second lane line pixel point set; determining, based on the first lane line pixel point set and a first extrinsic parameter of the first camera, a first lane line sampling point set corresponding to the first lane line pixel point set in a vehicle local coordinate system; determining a corrected extrinsic parameter; determining, based on the second lane line pixel point set and the corrected extrinsic parameter, a second lane line sampling point set corresponding to the second lane line pixel point set in the vehicle local coordinate system; and determining the lane line based on the first lane line sampling point set and the second lane line sampling point set.