Point Cloud Reconstruction via Missing Part Detection and Completion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing point cloud compression techniques using projection and octree-based methods often result in incomplete or under-sampled reconstructed point clouds, leading to compromised visual quality due to missing parts and insufficient sampling resolution.
Innovation Solution
A method that detects missing parts in the inverse-projected point cloud by identifying boundary points and completes them by adding new points along connecting lines, and addresses under-sampled points in the octree-based structure by adding new points in the neighborhood, thereby enhancing the visual quality without increasing computational complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If projection and octree-based compression techniques are used, then compression efficiency is improved, but visual quality deteriorates due to missing parts and under-sampling
Solution Approach 1:
The patent performs preliminary detection of missing parts and under-sampled regions before final reconstruction. By identifying boundary points and analyzing neighborhood completeness in advance, the method prepares a completion map that guides subsequent point addition, ensuring that compression efficiency is maintained while visual quality is improved through targeted reconstruction enhancements.
2Manufacturing precision
If more points are added to complete missing parts, then visual quality is improved, but computational complexity increases
Solution Approach 1:
The patent applies local quality by differentiating between complete and incomplete regions of the point cloud. Instead of uniformly processing all points, the method specifically identifies boundary points and under-sampled neighborhoods, applying completion operations only where necessary. This localized approach improves visual quality in critical areas while minimizing unnecessary computational overhead in already-complete regions.
Solution Approach 2:
The patent implements partial action by adding points selectively rather than comprehensively. The completion process targets only the identified missing parts and under-sampled regions, adding the minimum necessary points to achieve visual completeness. This avoids the excessive computational cost of adding points throughout the entire point cloud, maintaining efficiency while improving quality where it matters most.
Data Source
AI summary
The present disclosure concerns a method for reconstructing a point cloud representing a 3D object from an inverse-projected point cloud obtained by inverse-projecting at least one depth image of an original point cloud, said at least one depth image being obtained by projecting points of the original point cloud onto at least one surface, said method comprising the steps of detecting at least one missing part in the inverse-projected point cloud, and completing said at least one missing part based on points in the neighborhood of said at least one missing part.