3D Model Slicing for ML-Based Dimensionality Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for processing and analyzing 3D data are resource-intensive and prone to human error, particularly in fields like dental care and medical imaging, due to the high dimensionality of the tasks, which makes accurate measurement and diagnosis challenging.
Innovation Solution
Computing a central axis for a 3D model and generating multiple 2D slices that intersect this axis, using a trained machine learning model to process each slice and convert the 2D information back to 3D data, maintaining spatial accuracy through 2D-3D point correspondence.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If 3D data is processed directly using traditional methods, then comprehensive spatial information is captured, but computational resources are excessively consumed and processing time increases
Solution Approach 1:
The patent segments the 3D data into multiple 2D slices that intersect along a central axis. Each slice is processed independently through the ML model, dividing the computationally intensive 3D processing task into smaller, more manageable 2D tasks that consume fewer computational resources while maintaining comprehensive spatial coverage.
Solution Approach 2:
The patent transforms the 3D data processing problem into a 2D problem by creating slices at different angular orientations around the central axis. This dimensionality reduction from 3D to 2D significantly decreases computational complexity and resource requirements while the multi-angular approach ensures comprehensive 3D coverage is maintained.
2Measurement precision
If manual analysis methods are used for 3D data, then human expertise can be applied, but human error increases and processing time is excessive
Solution Approach 1:
The patent replaces manual human analysis with an automated ML model that processes the 2D slices. This substitution eliminates human error and significantly reduces processing time while maintaining or improving measurement precision through consistent, data-driven analysis of the segmented 3D structures.
Solution Approach 2:
The system enables self-service analysis where the ML model automatically processes the 2D slices and generates diagnostic information without requiring manual human intervention. The model independently extracts features, performs measurements, and produces results, eliminating both human error and time loss associated with manual analysis.
3Productivity
If fewer 2D slices are used for processing, then computational resources are saved, but coverage and detail of the 3D model decrease
Solution Approach 1:
The patent employs dynamic adjustment of the number of slices based on the specific processing needs and computational resources available. The system can adaptively select an optimal number of angular orientations for slicing, balancing processing speed with information retention. This dynamic approach allows flexible optimization between productivity and information completeness depending on the application context.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Computing a central axis for a three-dimensional (3D) model, generating a number of two-dimensional (2D) slices from the 3D model, the central axis passing through each of the 2D slices, and 2D points of the plurality of 2D slices corresponding to 3D points from the 3D model via a 2D point - 3D point correspondence. The method also includes computing 3D information about the 3D model by proposing for each 2D slice of the plurality of 2D slices, using a trained ML model, 2D information about the 2D slice using the 2D slice as input, to obtain a plurality of 2D information for the plurality of 2D slices, and converting the plurality of 2D information to the 3D information about the 3D model based on the 2D point - 3D point correspondence.