3D Geometric Data Unification and Part Hiding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional CAD models face challenges in real-time 3D visualization due to geometry defects like incorrectly oriented faces and normals, leading to performance barriers and unacceptable visualization quality, as industry-standard algorithms like ray-casting are computationally expensive and impractical for large models.
Innovation Solution
The technology employs multi-scale visibility patches and adjacency heuristics to restore proper face and normals orientation at a per-object and per-polygon level, targeting ray-casting metrics for refined visibility determination, thereby improving rendering efficiency and quality by reducing the need for brute-force ray-casting.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If ray-casting algorithm is used to determine normal orientation, then measurement precision is improved, but productivity deteriorates due to computational expense
Solution Approach 1:
The patent segments the problem of normal orientation determination by dividing it into two stages: (1) a fast heuristic phase that processes all triangles to identify candidate inverted faces, and (2) a selective ray-casting phase that only processes identified candidates. This segmentation allows the computationally expensive ray-casting to be applied only where necessary, rather than to all triangles in the model.
Solution Approach 2:
The patent applies local quality by using different methods for different parts of the data: fast adjacency-based heuristics are applied to all triangles, while expensive ray-casting is applied only to specific candidate triangles that are likely to be inverted. This localized application of high-precision methods optimizes the balance between accuracy and performance.
2Reliability
If all CAD data is retained for 3D visualization, then reliability is improved, but productivity deteriorates due to rendering cost of redundant data
Solution Approach 1:
The patent extracts and removes internal parts and hidden features from the CAD model before rendering. By identifying and eliminating geometry that cannot be seen from any external viewing angle (internal parts) or is occluded by other geometry (hidden features), the patent reduces the rendering workload while preserving all visually relevant information.
Solution Approach 2:
The patent performs preliminary processing to remove internal parts and hidden features before the actual rendering operation. This advance elimination of redundant geometry ensures that the rendering pipeline only processes visible, relevant features, significantly improving rendering performance without affecting visualization quality.
3Adaptability or versatility
If tesselation process is used to create 3D model, then adaptability is improved, but reliability deteriorates due to normal inversion errors
Solution Approach 1:
The patent implements a self-correcting system where the adjacency-based heuristic analysis automatically identifies triangles with inverted normals by examining local geometric relationships. The system uses the inherent structural information in the mesh data (adjacency relationships, face orientations) to detect and flag errors without requiring external reference information, enabling automatic correction of tesselation-induced normal inversions.
Data Source
AI summary
Technology is disclosed for improving the rendering of a three-dimensional (3D) object. In one aspect, the technology applies multi-scale visibility patches and adjacency heuristics to restore proper face and normals orientation at a per-object and per-polygon level; and target where best to apply ray-casting metrics to refine visibility determination.