3D Lidar Mapping Using Voxel Eigenvalues for Real-Time Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for building three-dimensional maps using three-dimensional lidar struggle with data distortion and slow processing times when the lidar is attached to a moving robot, leading to inaccurate and time-consuming map generation.
Innovation Solution
A real-time three-dimensional map building method using three-dimensional lidar that represents map data as voxels, calculates eigenvalues and eigenvectors, detects corresponding points, and applies rotation and translation transformations to minimize error, allowing for continuous motion and accurate map updates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the ICP algorithm is used to build three-dimensional maps by stopping at each location, then mapping accuracy can be improved, but the mapping time becomes too long and real-time performance is lost
Solution Approach 1:
The patent pre-calculates and stores eigenvalues and eigenvectors for each voxel in the three-dimensional map during the mapping process. This preliminary computation allows the system to quickly identify corresponding points between old and new lidar data without performing complex iterative calculations in real-time, thereby reducing mapping time while maintaining accuracy
Solution Approach 2:
The patent extracts and utilizes only the essential geometric features (eigenvalues and eigenvectors) from the three-dimensional lidar data rather than processing all raw point cloud data. This extraction of key features reduces computational complexity and enables real-time processing while preserving the necessary information for accurate map building
2Productivity
If the lidar is attached to a moving robot to enable real-time mapping, then mapping speed is improved, but data distortion occurs and mapping accuracy deteriorates
Solution Approach 1:
The patent dynamically adjusts the mapping process by continuously updating the three-dimensional map with new lidar data acquired during robot movement. The system uses real-time eigenvalue and eigenvector calculations to adapt to changing environments and maintain accuracy despite the robot's motion, transforming the static ICP approach into a dynamic real-time process
Solution Approach 2:
The patent introduces eigenvalues and eigenvectors as intermediary elements that mediate between the raw lidar data and the final map reconstruction. These intermediaries provide a stable mathematical framework that compensates for motion-induced distortions, enabling accurate mapping even when the lidar is mounted on a moving platform
3Measurement precision
If many feature vectors are used to define two three-dimensional points for scan matching, then alignment accuracy is improved, but algorithm execution time increases significantly
Solution Approach 1:
The patent extracts only the critical geometric characteristics (eigenvalues and eigenvectors) from the feature vectors that define three-dimensional points. By taking out only the essential alignment information rather than using all feature vector components, the system achieves sufficient alignment accuracy with significantly reduced computational time
Solution Approach 2:
The patent uses a partial set of feature information (eigenvalues and eigenvectors) rather than the complete set of feature vectors. This partial action provides just enough information for accurate scan matching while avoiding the computational burden of processing all available feature data, achieving a balance between accuracy and speed
Data Source
AI summary
Real-time three-dimensional (3D) map building method and device using a 3D lidar includes representing 3D map data of a surrounding environment acquired by using a 3D lidar attached to a moving object as voxels, acquiring an eigenvalue and an eigenvector for each voxel based on all 3D points in a 3D map represented as the voxels, detecting a 3D corresponding point in the voxel corresponding to all the 3D points of 3D data newly acquired by using the 3D lidar while the moving object travels, calculating a rotation transformation and a translation transformation for minimizing an error by minimizing an inner product value between the eigenvector weighted by the eigenvalue of the voxel to which the 3D corresponding point belongs and a vector generated from a 3D corresponding point, and updating the 3D map data based on the rotation transformation and the translation transformation.