LiDAR Odometry Using Directed Geometric Points and Sparse Frames
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing LiDAR odometry methods based on frame-to-map matching suffer from low efficiency due to the large quantity of points stored in the map, affecting real-time performance.
Innovation Solution
A LiDAR odometry method using directed geometric points and sparse frames, where planar points are selected, clustered, and fitted to create directed geometric point sets, which are then converted and optimized using the Levenberg-Marquardt algorithm to reduce the number of points in the map, improving efficiency by omitting plane fitting and accumulating only sparse frames.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the map is constructed by accumulating point clouds of all frames before the current frame, then the pose estimation accuracy is improved, but the quantity of points in the map increases, which seriously affects the efficiency of frame-to-map matching
Solution Approach 1:
The patent extracts only geometric points from point cloud data to construct the map, rather than accumulating all point cloud information. This extraction principle reduces the quantity of points in the map while maintaining pose estimation accuracy, thereby improving frame-to-map matching efficiency
Solution Approach 2:
The patent applies different processing strategies to different parts of the point cloud data. Geometric points are extracted and processed with high precision to ensure accurate pose estimation, while other point cloud data is handled more efficiently to maintain real-time performance
2Productivity
If the map is constructed by accumulating only geometric points of each frame, then the quantity of points in the map is reduced, but plane fitting is required in the process of frame-to-map matching, thus it is still necessary to maintain a high-density point cloud distribution
Solution Approach 1:
The patent performs plane fitting in advance during map construction rather than during frame-to-map matching. By pre-processing the geometric points and fitting planes beforehand, the system eliminates the need for real-time plane fitting during matching operations, thereby improving efficiency without sacrificing accuracy
Solution Approach 2:
The patent performs plane fitting on a subset of geometric points during map construction rather than on all points. This partial action approach reduces computational complexity while maintaining sufficient accuracy for pose estimation
3Productivity
If the map is constructed by accumulating point cloud information of limited consecutive frames before the current frame, then the quantity of points in the map is reduced to a certain extent, but it is still difficult to ensure real-time performance of frame-to-map matching
Solution Approach 1:
The patent changes the parameter of map construction from accumulating point cloud information of limited consecutive frames to accumulating only extracted geometric points from all frames. This parameter change reduces the quantity of points while maintaining real-time performance through efficient data structure organization and processing
Data Source
AI summary
A LiDAR odometry method based on the directed geometric point and sparse frame includes: obtaining original three-dimensional point cloud data of a surrounding environment in the coordinate system of the current frame of the LiDAR sensor; constructing and converting the first directed geometric point set and first environmental directed geometric point set into the second directed geometric point set and second environmental directed geometric point set in the world coordinate system, respectively; calculating candidate associated points in the second environmental directed geometric point set in association with each point in the second directed geometric point set, and determining the best associated point of each point in the second directed geometric point set; constructing and solving a pose optimization function to obtain an optimized pose from the coordinate system of the current frame of the LiDAR sensor to the world coordinate system; and updating directed geometric point sets of sparse frames.