Mobile Body Pose Estimation Using Defect-Excluded 3D Point Clouds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating the position and orientation of a mobile body using three-dimensional point cloud data from distance measurement devices, such as cameras or LiDAR, fail when blind spots occur, leading to incorrect estimation results due to defective portions in the data.
Innovation Solution
A control device that generates control commands by executing matching between a modified template point cloud, created by excluding defective portions from the three-dimensional point cloud data, to estimate the position and orientation of the mobile body.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If three-dimensional point cloud data is directly used for estimation processing, then the processing is simple, but correct estimation cannot be obtained when blind spots occur
Solution Approach 1:
The point cloud data is segmented into valid portions and defective portions based on detection results. The estimation processing is then performed separately on the valid portions, excluding the defective portions caused by blind spots. This segmentation allows the system to maintain high estimation accuracy by focusing only on reliable data regions.
Solution Approach 2:
The defective portions corresponding to blind spots are extracted and removed from the point cloud data before estimation processing. By taking out the harmful defective data, the system ensures that estimation is performed only on valid portions, thereby maintaining reliability without requiring complex remediation of the entire dataset.
2Productivity
If defective portions are excluded from template point cloud, then processing speed increases, but may lose information
Solution Approach 1:
The blind spots create defective portions in the point cloud data, which would normally be harmful. However, by detecting these defective portions and excluding them from estimation processing, the system converts this harm into a benefit - the estimation is performed only on valid, reliable data portions, improving both speed and accuracy. The harmful defective data becomes a known quantity that can be deliberately excluded.
3Reliability
If blind spots are not considered, then the system is simple, but estimation results become incorrect
Solution Approach 1:
The system performs preliminary detection of defective portions corresponding to blind spots before the main estimation processing. By identifying and marking these problematic regions in advance, the subsequent estimation can proceed efficiently by simply excluding the pre-identified defective portions, rather than dealing with them during the estimation process itself.
Data Source
AI summary
The present disclosure provides a control device that generates a control command for controlling a mobile body by using three-dimensional point cloud data of the mobile body measured by a distance measurement device. The control device includes an estimation unit configured to, in a case where a defective portion is generated in the three-dimensional point cloud data acquired from the distance measurement device, execute matching between a substantial point cloud portion obtained by excluding a defect corresponding portion corresponding to the defective portion from a template point cloud and the three-dimensional point cloud data, to estimate at least one of a position and an orientation of the mobile body.