3D Model Retrieval via 2D Projection Descriptors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Searching for and retrieving 2D drawings and 3D models is difficult and time-consuming, often resulting in irrelevant results due to the gap between 2D representations and 3D models, and the inability to effectively compare input parameters with large repositories of 3D models and 2D drawings.
Innovation Solution
The development of methods to determine robust orthogonal orientations for 3D models, generate 2D projections, and create descriptors that allow for similarity comparison between 2D drawings and 3D models, enabling efficient retrieval of similar models based on 2D input, using techniques such as maximum normal distribution and inertial principal axes for pose determination, and spherical harmonics for rotation invariant descriptors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional search methods are used to retrieve 2D drawings and 3D models, then the search process is simple to implement, but the retrieval accuracy is low and results are irrelevant
Solution Approach 1:
The patent introduces 2D projections as an intermediary representation that bridges the gap between 2D drawings and 3D models. By projecting 3D models onto 2D planes and comparing these projections with query drawings, the system enables accurate retrieval without requiring direct 3D-2D matching, thus improving retrieval accuracy while maintaining manageable system complexity
Solution Approach 2:
The patent creates 2D copies (projections) of 3D models that can be directly compared with 2D drawing queries. This copying approach allows the system to handle both 2D and 3D data types uniformly in 2D space, improving retrieval accuracy by enabling direct visual comparison while avoiding the complexity of 3D-2D geometric transformation and matching algorithms
2Loss of time
If 3D models are directly compared with 2D drawings, then complete shape information is preserved, but the comparison process becomes computationally complex and time-consuming
Solution Approach 1:
The patent extracts 2D projection views from 3D models that capture the essential shape information needed for retrieval. By taking out only the relevant 2D silhouette and feature information rather than processing complete 3D geometry, the system reduces computational complexity and search time while preserving sufficient shape information for accurate retrieval
Solution Approach 2:
The patent transforms the 3D-2D comparison problem into a 2D-2D comparison by projecting 3D models onto 2D planes. This dimensionality change reduces the computational complexity from 3D spatial operations to 2D image processing, significantly decreasing search time while maintaining shape information through carefully selected projection views
3Measurement precision
If multiple viewing directions are considered for 3D models, then shape representation becomes more accurate, but the data processing complexity increases
Solution Approach 1:
The patent segments the 3D model representation into multiple 2D projection views from different directions. By dividing the complex 3D shape information into separate 2D views (front, side, top projections), the system achieves accurate shape representation while keeping each individual view computationally simple to process, thus balancing accuracy with manageable complexity
Data Source
AI summary
There are disclosed methods to provide stable pose determinations of various three dimensional shapes. Methods are also disclosed for determining multiple, unique drawing descriptors for two dimensional drawings, and for obtaining intermediate three dimensional representations of two dimensional drawings as one way to determine the descriptor. Methods are also disclosed to provide for searching of two dimensional drawings and three dimensional shapes using user-defined input, which may be a drawing or sketch. User interactivity is provided to further refine search results.