Point Cloud Denoising via Adaptive Local Coordinate Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current point cloud denoising methods often fail to accurately preserve sharp features and edges while reducing noise, leading to oversmoothing or inadequate noise suppression in 3D point clouds.
Innovation Solution
The method involves determining a local coordinate system for each point in the point cloud, establishing adaptive directional neighborhoods, and using a local polynomial approximation with confidence intervals to compute local estimates, followed by weighted order statistics filtering to refine neighborhood sizes and aggregate estimates, ensuring effective noise suppression and feature preservation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If conventional point cloud denoising methods are applied, then noise is reduced, but sharp features and edges are lost due to oversmoothing
Solution Approach 1:
The patent applies local quality by adapting the neighborhood shape and size to each point's local geometry. For points on sharp edges or corners, the method creates anisotropic neighborhoods that extend more in directions parallel to the surface and less in directions perpendicular to sharp features. This allows noise reduction in smooth regions while preserving sharp features by using directionally-sensitive filtering that adapts to local geometric properties.
Solution Approach 2:
The patent implements dynamics by making the neighborhood parameters dynamic rather than fixed. The neighborhood shape and extent are computed adaptively for each point based on local point cloud density and geometric characteristics. This dynamic adaptation allows the filtering process to respond to local variations in the data, preserving sharp features while effectively denoising smooth regions.
2Device complexity
If fixed-size neighborhoods are used for filtering, then processing is simpler, but spatial accuracy is reduced due to inability to adapt to local variations
Solution Approach 1:
The patent applies preliminary action by pre-computing the local coordinate systems and neighborhood parameters for each point before performing the actual filtering operation. This preparation step establishes the adaptive neighborhood structures that will be used during filtering, allowing the method to achieve high spatial accuracy without excessive complexity during the main processing phase. The preliminary computation of geometric properties enables subsequent efficient filtering.
Solution Approach 2:
The patent implements segmentation by dividing the point cloud into local neighborhoods with distinct geometric characteristics. Each point is associated with its own adaptive neighborhood defined in a local coordinate system, allowing different filtering behaviors in different regions. This segmentation approach enables the method to handle local variations in geometry and density, achieving high spatial accuracy through localized processing.
Data Source
AI summary
Techniques are disclosed for point cloud denoising systems and methods. In one example, a method includes determining a respective local coordinate system for each point of a point cloud. The method further includes determining a respective first adaptive-shape neighborhood for each point of the point cloud based on each respective local coordinate system. The method further includes performing filtering associated with each respective first adaptive-shape neighborhood to obtain a respective second adaptive-shape neighborhood for each point of the point cloud. The method further includes determining local estimates for points inside each of the second adaptive-shape neighborhoods. The method further includes aggregating the local estimates for each point of the point cloud to obtain a denoised point cloud. Related devices and systems are also provided.