Multi-Sensor Calibration via Trajectory Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current multi-sensor calibration processes for unmanned driving systems, particularly involving lidar, GPS/IMU, and camera, are complex, time-consuming, and limited in accuracy due to the involvement of only two sensors at a time, leading to error accumulation and reduced precision.
Innovation Solution
A multi-sensor calibration method that acquires data from at least three sensors simultaneously during vehicle travel, determines their trajectories using image recognition and machine learning, and performs joint calibration through trajectory alignment, maximizing the utilization of multi-source information and improving calibration accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If only two sensors are involved in calibration at a time, then the calibration process is simpler to implement, but the calibration accuracy is limited and error accumulation occurs
Solution Approach 1:
The patent merges multiple sensor calibrations into a single unified calibration process. Instead of calibrating sensors pairwise (lidar-GPS, camera-lidar), the system performs joint calibration of at least three sensors simultaneously by aligning their trajectories in a common coordinate system, thereby improving accuracy while maintaining manageable complexity through integrated processing
Solution Approach 2:
The patent introduces trajectory alignment as an intermediary mechanism to facilitate multi-sensor calibration. By converting sensor data into trajectories and aligning these trajectories through coordinate transformations, the system enables accurate multi-sensor calibration without directly complexizing the calibration process itself
2Measurement precision
If multiple sensors are calibrated simultaneously through joint calibration, then calibration accuracy improves and error accumulation is avoided, but the computational complexity increases
Solution Approach 1:
The patent segments the multi-sensor calibration process into distinct operational phases: data acquisition from multiple sensors, trajectory determination from acquired data, and trajectory alignment through coordinate transformation. This segmentation allows the system to handle computational complexity in manageable stages rather than as a single overwhelming computation
Solution Approach 2:
The patent replaces traditional iterative mechanical calibration methods with a direct trajectory alignment approach. By substituting the calibration process with trajectory determination and alignment operations, the system reduces computational pressure while achieving higher accuracy through direct geometric relationships rather than iterative optimization
3Productivity
If traditional two-sensor calibration methods are used, then the calibration process is faster to execute, but the precision is reduced due to error accumulation
Solution Approach 1:
The patent performs preliminary trajectory determination for each sensor before performing the alignment operation. By pre-processing sensor data into trajectory representations, the system enables fast execution of the alignment step while ensuring high precision through the use of all available sensor data, avoiding the need for repeated iterative calibrations
Data Source
AI summary
Embodiments of the present disclosure provide a multi-sensor calibration method, a multi-sensor calibration device, a computer device, a medium and a vehicle. The method includes: acquiring data acquired by each of at least three sensors in a same time period in a traveling process of a vehicle; determining a trajectory of each of the at least three sensors according to the data acquired by each of at least three sensors; and performing a joint calibration on the at least three sensors by performing trajectory alignment on the trajectories of the at least three sensors.