3D Point Cloud Region Extraction for Moving Object Annotation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Three-dimensional point cloud data is difficult to understand and annotate due to its nature as a collection of points, making it challenging to extract regions corresponding to moving objects for annotation processing.
Innovation Solution
A region extraction method and device that compares three-dimensional point cloud information with reference information acquired under different conditions to detect and extract regions of moving objects by setting voxels, determining presence/absence of point clouds, calculating eigenvalues for shape classification, and using similarity thresholds to identify moving objects within the data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If three-dimensional point cloud data is used for annotation processing, then the ability to capture and represent spatial information is improved, but the difficulty of understanding and extracting regions increases
Solution Approach 1:
The patent segments the three-dimensional point cloud data into multiple two-dimensional image data from different viewpoints. This segmentation transforms the difficult-to-process 3D point cloud into multiple easier-to-annotate 2D images, allowing annotation work to be performed on individual viewpoint images while preserving the overall spatial information through multi-view synthesis
Solution Approach 2:
The patent introduces two-dimensional image data as an intermediary between the three-dimensional point cloud data and the annotation process. By converting 3D point cloud into multiple 2D viewpoint images, the system creates an intermediate representation that is more suitable for human annotation while still capturing the essential spatial relationships of the original 3D data
2Measurement precision
If manual annotation of three-dimensional point cloud data is performed, then labeling accuracy can be maintained, but the time and effort required for annotation increases significantly
Solution Approach 1:
The patent divides the annotation task into multiple independent two-dimensional viewpoint images, allowing annotators to work on smaller, more manageable sections. This segmentation enables parallel processing of different viewpoints and reduces the cognitive load on annotators, thereby maintaining accuracy while reducing overall annotation time
Solution Approach 2:
The patent creates multiple two-dimensional copies or projections of the three-dimensional point cloud data from different viewpoints. These 2D copies serve as simplified representations that retain the essential annotation-worthy features while being much easier and faster to annotate than the original 3D data
Data Source
AI summary
A region extraction method of the present disclosure is executed by an information processing device including an arithmetic circuit to extract a desired region corresponding a moving object from three-dimensional point cloud information. The method includes: by the arithmetic circuit, receiving three-dimensional point cloud information acquired by a three-dimensional point cloud acquisition device; receiving reference information that includes at least a part of a range of the three-dimensional point cloud information and is acquired under a condition different from the acquisition of the three-dimensional point cloud information; and comparing the three-dimensional point cloud information with the reference information, detecting the moving object, and extracting a region of the moving object from the three-dimensional point cloud information.