Planar Feature Extraction for 3D LiDAR Motion Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional vision-based navigation systems using point features struggle with 3D data from LiDAR images, as they often have insufficient identifiable features and are unstable due to surface discontinuities and occluding objects, making it difficult to accurately estimate motion.
Innovation Solution
A method is developed to extract and match planar features from 3D data points, determining motion through rotation and translation, using a system that includes a sensor and processing unit to identify normal vectors and orthogonal distances, and group data points into cells to find local maxima, which are then refined to extract planar features for motion estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If point feature detectors are applied to LiDAR images, then feature detection can be performed, but the features are insufficient and unstable for reliable motion estimation
Solution Approach 1:
The patent transitions from detecting point features in 2D LiDAR range grids to extracting planar features in 3D space. By utilizing the full 3D coordinate information (x, y, z) from LiDAR sensors and analyzing planar surfaces rather than isolated points, the system increases the dimensionality of feature representation. This dimensional expansion provides significantly more identifiable features while improving their stability for motion estimation.
2Measurement precision
If point features are used from LiDAR data, then motion estimation can be attempted, but the features are unstable due to surface discontinuity and occluding objects
Solution Approach 1:
The patent segments the 3D point cloud data into distinct planar surfaces by analyzing local geometric properties and grouping points that lie on the same plane. This segmentation approach transforms unstable individual point features into stable planar feature representations. Each planar feature is defined by multiple points fitting a plane equation, making it robust against surface discontinuities and occlusions that would destabilize individual point features.
Solution Approach 2:
The patent creates composite planar features by combining multiple 3D points that belong to the same planar surface. Instead of relying on single point features that are vulnerable to noise and occlusion, the system synthesizes planar features from collections of points, similar to creating composite materials for enhanced properties. This composite approach stabilizes the features while maintaining measurement precision for motion estimation.
Data Source
AI summary
A method of controlling an actuator based on a set of three-dimensional (3D) data points is provided. The method includes obtaining a first set of 3D data points for a scene and a second set of 3D data points for a scene with a sensor. At least a first set of planar features is extracted from the first set of 3D data point. At least a second set of planar features is extracted from the second set of 3D data points. A motion is determined between the first set of 3D data points and the second set of 3D data points based on a rotation and a translation from the at least a first set to the at least a second set. At least one actuator is controlled based on the motion.