3D model retrieval method based on mesh segmentation

摘要

In the process of feature description and extraction, current 3D model retrieval algorithms focus on the globalfeatures of 3D models but ignore the combination of global and local features of the model. For this reason, they showless effective performance to the models with similar global shape and different local shape. This paper proposes a novelalgorithm for 3D model retrieval based on mesh segmentation. The key idea is to exact the structure feature and the localshape feature of 3D models, and then to compares the similarities of the two characteristics and the total similaritybetween the models. A system that realizes this approach was built and tested on a database of 200 objects and achievesexpected results. The results show that the proposed algorithm improves the precision and the recall rate effectively.© (2012) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.