Vehicle self-calibration method and system based on natural scene data
By collecting lidar and camera data in natural scenes and combining image edge features and point cloud projection analysis, the problem of sensor acquisition time difference is solved, the sensor acquisition time difference is corrected, and the stability and accuracy of the calibration system are improved.
CN120672869APending Publication Date: 2025-09-19SHENZHEN YOUBIKANG TECH CO LTD
Patent Information
- Application Number
- CN202510764455.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-09
- Publication Date
- 2025-09-19
Smart Images
Figure CN120672869A_ABST
Abstract
The invention discloses a vehicle self-calibration method and system based on natural scene data, belongs to the technical field of vehicle calibration, and provides a time difference recognition mechanism combining space alignment error and dynamic target velocity vector offset analysis, and introduces a point cloud correction method based on prediction compensation. According to the method, accurate estimation and correction of the acquisition time difference of the sensors are realized, and through a projection consistency verification mechanism, it is ensured that a calibration result still has high robustness and accuracy in a complex dynamic scene, so that the stability and sensing precision of multi-sensor fusion in an automatic driving system are effectively improved.
Need to check novelty before this filing date? Find Prior Art
Citation Information
Cited By
Robot scene sensing method and system based on multi-sensor fusion
CN121541216A