Geometry-Based CAD Search Using PCA Entity Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing CAD systems lack an effective geometry-based search method that can flexibly search for native entities, multiple blocks, and geometric combinations across drawings, often requiring manual searches and failing to accommodate variations in scale, orientation, and naming conventions.
Innovation Solution
A geometry-based search method using Principal Component Analysis (PCA) to generate unique 64-bit codes for geometric entities, enabling efficient and automated searches across 2D and 3D primitives, independent of orientation, scale, and transformation, with a parallelized algorithm for matching and indexing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If entity property-based search methods are used, then search can be performed using defined blocks, but the search is not flexible and misses geometry with same shape but different names or non-block cases
Solution Approach 1:
The patent transforms the search approach from property-based parameters (block names, types) to geometry-based parameters (shape, size, orientation). By encoding geometric entities into unique codes based on their spatial characteristics rather than their naming properties, the system achieves both flexibility in searching any geometric form and reliability in finding all matching geometries regardless of their names or block definitions
Solution Approach 2:
The patent creates a simplified geometric model (64-bit code) that copies only the essential shape characteristics of the original geometric entity, stripping away non-essential properties like names and block definitions. This geometric fingerprint allows the search to focus purely on shape matching, making the search both more flexible and more complete
2Reliability
If manual search methods are used, then all geometric entities can be found by plotting, but the process is time-consuming and inefficient
Solution Approach 1:
The patent performs preliminary encoding of all geometric entities in the drawing into unique 64-bit codes before the actual search is initiated. This pre-processing step creates an indexed database of geometric fingerprints, so when a search is performed, the system can quickly compare and match geometries without time-consuming manual plotting, thereby maintaining accuracy while dramatically improving speed
Solution Approach 2:
The patent replaces the mechanical manual plotting process with an automated computational system that uses geometric encoding and code comparison. Instead of manually tracing and comparing geometric shapes on paper, the system automatically encodes geometries into digital codes and performs rapid algorithmic matching, eliminating the time-consuming manual operation while preserving search accuracy
3Adaptability or versatility
If block-based search is used, then structured entities can be found, but composited entities and native entities cannot be searched effectively
Solution Approach 1:
The patent creates a universal search system that handles all types of geometric entities (native entities, blocks, and composited entities) through a single unified approach. The geometric encoding method works independently of entity type, allowing the same search mechanism to find any geometric form regardless of whether it's a simple line, a named block, or a composite of multiple entities, thereby achieving versatility without requiring separate search mechanisms for each entity type
Data Source
AI summary
A method and system provide for searching a computer-aided design (CAD) drawing. A CAD drawing is obtained and includes vector based geometric entities. For each entity, primitives are extracted. A feature coordinate system is created for each of the entities using the primitives. The primitives are transformed from a world coordinate system to feature coordinates of a feature coordinate system. Geometry data of the transformed entities is encoded into index codes that are utilized in an index table as keys. A target geometric entity is identified and a target index code is determined and used to query the index table to identify instances of the target geometric entity in the CAD drawing.