Discrete 3D Mechanical Part Profile Detection Under Geometric Noise
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for profile detection in discrete 3D models of mechanical parts are inadequate, particularly failing to identify valid profiles in complex cases such as self-intersecting curves and degenerate surfaces, which cannot be processed by CAD software, and are prone to errors in noisy models from 3D scans.
Innovation Solution
A computer-implemented method that evaluates candidate parameterizations and projections of 3D surfaces by computing a function that penalizes distortion between projections and parameterizations, determining a valid profile if the function's value is below a threshold, ensuring robust detection of continuous curves that can be processed by CAD software.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing profile detection methods are used on discrete 3D models, then processing is simpler, but detection reliability fails in complex cases such as self-intersecting curves and degenerate surfaces
Solution Approach 1:
The method segments the discrete 3D model into individual 3D surfaces and evaluates each surface separately for profile validity. By dividing the complex detection task into surface-level evaluations, the method can reliably identify valid profiles even in complex cases with self-intersecting curves or degenerate surfaces, while maintaining manageable computational complexity through localized analysis.
Solution Approach 2:
The patent introduces an energy function as an intermediary criterion to evaluate candidate profiles. This energy function serves as a mediator that quantifies profile validity by penalizing geometric inconsistencies, allowing reliable detection without requiring complex manual validation of each geometric case.
2Adaptability or versatility
If profile detection is performed on noisy models from 3D scans, then real-world applications are enabled, but detection accuracy decreases due to noise
Solution Approach 1:
The energy function is designed to be robust against noise by penalizing only significant geometric inconsistencies rather than minor variations. This allows the method to convert the harmful effect of noise into a manageable factor, where small noise-induced deviations do not trigger false rejections of valid profiles, while still detecting genuine geometric errors in scanned models.
Solution Approach 2:
The method adjusts the evaluation criteria by using an energy function with configurable threshold parameters that can adapt to different noise levels in scanned models. This allows the detection system to maintain accuracy across varying quality inputs by tuning the sensitivity of the energy function to match the expected noise characteristics of the input data.
3Productivity
If a simple distortion metric is used, then computation is faster, but detection precision is insufficient for complex surfaces
Solution Approach 1:
The energy function evaluates only the most critical geometric inconsistencies that indicate invalid profiles, rather than performing exhaustive validation of all surface properties. This partial evaluation approach maintains high detection speed while achieving sufficient precision by focusing computational effort on the most discriminative features of valid versus invalid profiles.
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
The disclosure notably relates to a computer-implemented method for profile detection in a discrete 3D model. The discrete 3D model represents a mechanical part. The method comprises, for each 3D surface of at least one 3D surface of the 3D model, providing a candidate parameterization of the 3D surface and a candidate projection of the 3D surface into ℝ2. The method further comprises computing a function. The function penalizes, for each couple of points of the 3D surface having neighboring parameter values, a distortion. The distortion is between a disparity between the projection of the points, and a disparity between the parameterization of the points. The method further comprises determining that the candidate parameterization and the candidate projection form a valid profile of the 3D surface if the computed function is smaller than a predefined threshold.