Neural Network 3D Segmentation for Robust CAD Model Partitioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing 3D segmentation methods lack efficiency and robustness in partitioning 3D modeled objects into semantically consistent segments, particularly in the context of CAD, CAE, and CAM systems, where explicit definitions of object portions are often required.
Innovation Solution
A machine-learning method utilizing a neural network that learns from a dataset of labeled 3D modeled object portions to determine the extent to which they belong to the same segment, outputting a value that indicates segment coherence, thereby facilitating efficient and robust 3D segmentation without relying on explicit definitions of the object portions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional 3D segmentation methods are used, then segmentation can be performed, but the process lacks efficiency and robustness
Solution Approach 1:
The patent replaces traditional mechanical segmentation algorithms with a neural network-based machine learning system. The neural network learns semantic relationships from training data and automatically performs segmentation, substituting the mechanical rule-based approach with an intelligent system that achieves both higher efficiency and robustness simultaneously.
Solution Approach 2:
The patent transforms the segmentation problem from a geometric processing task into a learning task by changing the parameters from fixed algorithmic rules to adaptive neural network weights. The system learns optimal segmentation parameters from training data, enabling it to adapt to different object types and achieve robust performance across diverse scenarios.
2Manufacturing precision
If explicit definitions of object portions are required, then segmentation precision can be maintained, but the complexity of the system increases
Solution Approach 1:
The neural network performs self-learning from training data to automatically understand object portions and their semantic relationships. Instead of requiring explicit definitions programmed by humans, the system serves itself by learning from examples, thereby maintaining precision while reducing system complexity.
Solution Approach 2:
The system performs preliminary learning during a training phase where the neural network learns semantic relationships from labeled training data. This preliminary action prepares the model to automatically handle segmentation tasks without requiring explicit definitions during actual operation, reducing complexity while maintaining precision.
Data Source
AI summary
A computer-implemented method of machine-learning including obtaining a dataset of training samples. Each training sample includes a pair of 3D modeled object portions labelled with a respective value. The respective value indicates whether or not the two portions belong to a same segment of a 3D modeled object. The method further includes learning a neural network based on the dataset. The neural network takes as input two portions of a 3D modeled object representing a mechanical part and outputs a respective value. The respective value indicates an extent to which the two portions belong to a same segment of the 3D modeled object. The neural network is thereby usable for 3D segmentation. The method constitutes an improved solution for 3D segmentation.