Shape distribution-based retrieval of 3D CAD models at different levels of detail

Abstract

With increasing design reuse in modern products, accurate and efficient 3D CAD model retrieval methods are required. To improve retrieval capability, a 3D shape comparison method is often utilized. In this method, the shapes of 3D CAD models in a database are compared with the shape given by a user. Meanwhile, for rapid generation of query models, a freehand sketch-based modeling method is adopted for retrieval systems. This method creates a low-level-of-detail (low-LOD) 3D CAD model with abbreviated exterior shapes. On the other hand, the target 3D CAD model in the database is a high-LOD 3D CAD model including detailed shapes of a product or components of a product. Considering different LODs of query and target models, we propose a new 3D CAD model retrieval method consisting of a 3D CAD model simplification system and a shape distribution-based shape comparison engine that compares multi-resolution models in a database to improve retrieval accuracy using a query model with simple shape. Experiment is conducted on 64 LOD models generated from 8 test cases and 8 query models generated by freehand sketch-based modelling method. Result shows a 200 % improvement on retrieval success rate for lower LOD models (100 %) compared with source models (50 %). Moreover, the proposed method has an advantage on efficiency, due to the simple calculation method and short computation time.