Multi-LiDAR Extrinsic Calibration for Non-Overlapping Fields of View
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Calibrating multiple lidar devices with non-overlapping fields of view is challenging, as existing methods rely on overlapping point clouds to determine calibration parameters, which are not applicable when fields of view do not overlap, leading to inconsistent and inaccurate 3D representations of the environment.
Innovation Solution
A method and system that involve moving a vehicle with mounted lidar devices along a predetermined path to generate overlapping fields of view over time, allowing for the determination of extrinsic calibration transformations between lidar devices, enabling consistent and coherent views of the environment even when their fields of view do not overlap initially.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If multiple lidar devices are positioned on a vehicle to obtain a 3D representation of the environment, then the coverage area and detection capability are improved, but the calibration difficulty increases when their fields of view do not overlap
Solution Approach 1:
The patent applies the dynamics principle by transitioning from a static calibration approach (requiring overlapping fields of view at a fixed position) to a dynamic approach where the vehicle moves along a predetermined path. This motion enables the fields of view of multiple lidar devices to overlap at different positions in time, allowing calibration to be performed even when the devices are positioned to cover different areas. The calibration system processes lidar returns collected during motion to determine extrinsic calibration transformations.
2Area of stationary object
If lidar devices are positioned to have non-overlapping fields of view to maximize coverage, then the area coverage is improved, but the ability to perform traditional calibration methods is lost
Solution Approach 1:
The patent applies the preliminary action principle by having the vehicle follow a predetermined path pattern before calibration data is collected. This pre-planned motion trajectory ensures that the lidar devices will capture overlapping point clouds of the same environmental features at different positions, creating the necessary conditions for calibration. The predetermined path is designed in advance to guarantee sufficient overlap opportunities for accurate calibration parameter determination.
Solution Approach 2:
The patent uses environmental features (such as walls, poles, or other stationary objects in the calibration environment) as intermediaries to establish correspondence between lidar devices. These environmental features serve as reference points that appear in the point clouds of multiple lidar devices at different positions, enabling the calculation of extrinsic calibration transformations even when the devices' fields of view do not overlap simultaneously.
3Measurement precision
If traditional calibration methods are used with overlapping fields of view, then calibration accuracy is improved, but the device configuration flexibility is reduced
Solution Approach 1:
The patent applies the universality principle by creating a calibration method that works for both overlapping and non-overlapping field of view configurations. The same calibration system and algorithm can handle various lidar device arrangements on the vehicle, making the calibration process universally applicable regardless of the specific positioning or orientation of the lidar devices. This multi-functional approach maintains calibration accuracy while accommodating different device configurations.
Data Source
AI summary
A method and a system of calibration of a first lidar device and a second lidar device are described. A subset of pairs of positions of a vehicle as the vehicle moves along a path according to a predetermined path pattern are determined. Each pair from the subset of pairs includes a first position and a second position of the vehicle. A first field of view of the first lidar device when the vehicle is located at the first position overlaps with a second field of view of the second lidar device when the vehicle is located at the second position at a second time. Based on the subset of pairs, an extrinsic calibration transformation that transforms a position of the second lidar device into a calibrated position that allows to obtain a consistent view of the world between the first and the second lidar devices is determined.