LiDAR-Assisted Wheel Encoder to Camera Calibration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for calibrating a wheel encoder to a camera are cumbersome or lack robust estimation, particularly in hand-eye calibration processes.
Innovation Solution
The method involves using LiDAR signals to separate the calibration process into camera to LiDAR calibration and wheel encoder to LiDAR calibration, allowing for the determination of scaling factors and estimating rotations and translations without Z-axis movement, thereby calibrating the camera and wheel encoder signals effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional wheel encoder to camera calibration methods are used, then the calibration can be performed, but the process is cumbersome and lacks robust estimation
Solution Approach 1:
The patent introduces LiDAR as an intermediary sensor to bridge the calibration between wheel encoder and camera. The LiDAR provides accurate 3D point cloud data that serves as a reliable reference for establishing the transformation relationship, thereby improving robust estimation while maintaining a manageable calibration process through intermediate reference data
2Measurement precision
If LiDAR assisted calibration is used, then robust hand-eye calibration estimation is achieved, but additional LiDAR signal processing is required
Solution Approach 1:
The patent segments the calibration process into distinct stages: LiDAR-to-world calibration, camera-to-world calibration, and wheel encoder-to-world calibration. Each sensor type is calibrated independently against a common world coordinate system, which improves measurement precision while organizing the signal processing into manageable, modular segments
3Manufacturing precision
If multi-sensor calibration is performed, then accurate sensor integration is enabled, but the calibration time and computational resources increase
Solution Approach 1:
The patent performs preliminary calibration of each sensor (LiDAR, camera, wheel encoder) to a common world coordinate system before final integration. This preliminary action establishes stable reference frames in advance, enabling accurate sensor integration while reducing the computational burden and time required for the final multi-sensor calibration
Data Source
AI summary
A method of wheel encoder to camera calibration, including receiving a LiDAR (Light Detection and Ranging) signal, receiving a camera signal, receiving a wheel encoder signal, calibrating the camera signal to the LiDAR signal, calibrating the wheel encoder signal to the LiDAR signal and calibrating the camera signal to the wheel encoder signal based on the calibration of the camera signal to the LiDAR signal and the wheel encoder signal to the LiDAR signal.