Super Voxel Ground Segmentation for Accurate Ground Plane Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for ground segmentation in autonomous vehicles are prone to errors, leading to inaccurate estimation of the ground plane and subsequent control issues.
Innovation Solution
The method involves grouping image data into super voxels based on normal vectors to form a single super voxel, which is then processed to estimate the ground plane, thereby improving accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional segmentation processes are used to segment ground data in captured image data, then the processing can be completed, but the segmentation is subject to error leading to incorrect estimation of the ground plane
Solution Approach 1:
The patent applies segmentation by dividing the captured image data into multiple super voxels based on normal vector analysis. Each super voxel represents a coherent ground region with consistent orientation, allowing for more accurate ground plane estimation within each segment while reducing cumulative segmentation errors through the merging process
Solution Approach 2:
The patent merges multiple super voxels into a single consolidated super voxel that represents the entire ground region. This merging process combines the normal vector information from individual super voxels to generate a unified ground plane estimation, thereby improving overall accuracy by aggregating data from multiple segmented regions
2Measurement precision
If multiple subsets of image data are created for segmentation, then the analysis can be more detailed, but the complexity of processing increases
Solution Approach 1:
The patent segments image data into multiple super voxels based on normal vector consistency, enabling detailed analysis of ground regions with different orientations. This segmentation allows precise identification of ground plane variations while maintaining manageable processing complexity through efficient normal vector-based grouping
Solution Approach 2:
The patent performs preliminary grouping of image data into super voxels using normal vector analysis before final ground plane estimation. This preliminary action organizes the data structure in advance, simplifying subsequent processing steps by pre-establishing coherent ground regions that can be efficiently merged and analyzed
Data Source
AI summary
Aspects of the subject technology relate to systems, methods, and computer-readable media for performing ground segmentation through formation of a single super voxel. Image data tracking a surface as an object moves relative to the surface is accessed. Samples of the image data is grouped into different subsets to form a plurality of subsets of the image data. A normal vector to a corresponding plane defined by each subset of the plurality of subsets of the image data are identified. The samples of the image data are re-grouped into a single subset of the image data based on the corresponding normal vector for each subset of the plurality of subsets of the image data. A surface plane representation of the surface is identified based on the single subset of the image data.