Point Cloud Registration and Completion for Low Overlap 3D Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing appearance model acquisition technologies face challenges in constructing a complete appearance model of an object when the overlap between point clouds from different local areas is small, leading to poor applicability.
Innovation Solution
A method and device that acquire a set of target point clouds, perform registration to align them into a common coordinate system, and completion to add missing points, using registration and completion networks with encoders and decoders to transform and enhance point clouds, respectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If registration is performed on point clouds with large overlap, then complete appearance model can be obtained, but applicability is limited to cases with sufficient overlap
Solution Approach 1:
The patent applies preliminary action by performing completion on point clouds before registration. The completion network predicts missing regions and fills them in advance, transforming incomplete point clouds with low overlap into more complete representations. This preliminary completion enables subsequent registration to succeed even when original point cloud overlaps are minimal, thus resolving the contradiction between requiring overlap for registration and needing to handle low overlap scenarios.
2Measurement precision
If only registration is performed on point clouds, then coordinate alignment is achieved, but complete appearance model cannot be obtained when overlap is small
Solution Approach 1:
The patent merges registration and completion into a unified joint network framework where both operations are performed together. The registration network aligns point clouds in coordinate systems while the completion network simultaneously predicts and fills missing regions. This combination ensures both coordinate alignment precision and appearance model completeness are achieved, even when point cloud overlap is small, resolving the contradiction between these two requirements.
Data Source
AI summary
A method and a device of acquiring an appearance model, a computer device and a storage medium. The method includes acquiring a set of target point clouds, the set of target point clouds including at least two point clouds, each of the point clouds being obtained by sampling a local area of a target object; then performing a registration and a completion for the at least two point clouds to obtain a processed point cloud, wherein the registration is to transform different point clouds into a same coordinate system, and the completion is to add points in other areas of the object according to existing points in a point cloud to be completed; and finally acquiring the appearance model of the target object according to the processed point cloud. The method can improve applicability of the appearance model of the target object acquired only by registration technique.