Soft Classification for 3D Model Segmentation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional three-dimensional digital modeling systems face challenges in accurately identifying and manipulating segments of diverse digital models, requiring significant resources and often resulting in inefficient processing and potential data corruption.

Innovation Solution

The implementation of a digital segmentation system that determines a soft classification of three-dimensional digital models, selects appropriate segmentation algorithms, and combines parameters to accurately identify segments, improving accuracy and efficiency by customizing algorithms to specific model features.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional modeling systems use fixed segmentation algorithms for specific model categories, then segmentation accuracy is improved for those categories, but the system cannot accurately identify segments across diverse model types

Engineering Contradiction:

Improvesegmentation accuracyVSAvoidmodel type coverage

Core Design Contradiction:

Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The system dynamically selects and adjusts segmentation algorithms based on the characteristics of each input model. Instead of using a fixed algorithm, the system adapts its approach by analyzing model features and choosing appropriate segmentation strategies, enabling accurate segmentation across diverse model types including industrial parts, organic shapes, and abstract forms

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system changes key parameters such as curvature thresholds, feature sensitivity, and algorithm selection based on the detected model category. By adjusting these parameters according to the specific model type (e.g., industrial vs. organic vs. abstract), the system maintains high segmentation accuracy across different domains without requiring manual reconfiguration

Inventive Principle:

Principle #35Parameter changes

2Reliability

If conventional systems use inefficient algorithms to identify segments, then comprehensive analysis is performed, but significant processing time and computational resources are required

Engineering Contradiction:

Improvesegment identification completenessVSAvoidprocessing time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The system performs preliminary classification of the input model to determine its category (industrial, organic, abstract, etc.) before applying segmentation algorithms. This preliminary action enables the system to select the most appropriate and efficient algorithm for the specific model type, avoiding the use of overly complex or inappropriate algorithms that would waste computational resources and time

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system segments the overall processing task into distinct phases: model classification, algorithm selection, and execution. By dividing the workflow and applying specialized algorithms to specific model categories, the system achieves comprehensive and reliable segment identification while minimizing processing time through targeted rather than exhaustive analysis

Inventive Principle:

Principle #1Segmentation

3Measurement precision

If conventional systems repeatedly apply segmentation analysis to achieve useful results, then accurate segmentation is eventually obtained, but excessive computing resources are consumed

Engineering Contradiction:

Improvesegmentation accuracyVSAvoidcomputational resource consumption

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The system performs preliminary classification to determine the model category before executing segmentation algorithms. This upfront classification prevents repeated trial-and-error segmentation attempts by ensuring the correct algorithm is selected from the beginning, thereby reducing computational resource consumption while maintaining high segmentation accuracy

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system uses pre-trained classification models and pre-configured algorithm profiles for different model categories. By leveraging these pre-prepared resources, the system avoids the need to develop and test multiple segmentation approaches during runtime, significantly reducing computational overhead while achieving accurate and efficient segmentation results

Inventive Principle:

Principle #26Copying

Data Source

PatentUS11823391B2Utilizing soft classifications to select input parameters for segmentation algorithms and identify segments of three-dimensional digital models

Publication Date: 2023.11.21 ADOBE INC

AI summary

The present disclosure includes methods and systems for identifying and manipulating a segment of a three-dimensional digital model based on soft classification of the three-dimensional digital model. In particular, one or more embodiments of the disclosed systems and methods identify a soft classification of a digital model and utilize the soft classification to tune segmentation algorithms. For example, the disclosed systems and methods can utilize a soft classification to select a segmentation algorithm from a plurality of segmentation algorithms, to combine segmentation parameters from a plurality of segmentation algorithms, and/or to identify input parameters for a segmentation algorithm. The disclosed systems and methods can utilize the tuned segmentation algorithms to accurately and efficiently identify a segment of a three-dimensional digital model.