科研之友
微信
新浪微博
Facebook
分享链接
摘要
With the rise of the applications of 3D models to the medical field for diagnostics, we here report a novel sketch-based medical 3D model retrieval approach. Since the skeleton displays the topological structure of the 3D models, a novel sketch-based medical 3D model retrieval algorithm is proposed, which adopts skeleton characteristics as the features to extract the object shape. Firstly, the advanced skeleton strength map (ASSM) algorithm for creating the skeleton is proposed, which computes the skeleton strength map by isotropic diffusion on the gradient vector field, and picks critical points among the skeleton strength map, and then connects the critical points by Kruskal's algorithm. Then, we propose histogram feature comparison algorithm which adopts the radii of the disks at skeleton points and the length of skeleton branch to extract the histogram feature, and compare the similarity between two skeletons using the histogram feature matrix of skeleton endpoints. The experimental result show that our ASSM algorithm has an increase in Precision value is around 29.3%; and our histogram feature comparison algorithm has an increase in Precision value is around 22.4%. Our approach has an increase in Precision value is around 25.4%, indicating that our skeleton algorithm is remarkably superior to several leading sketch-based medical retrieval approaches.
- 出版日期2017-11
- 单位西北大学
