Protrusion Based Segmentation of Complex 3D Shape Models
3D shape models have been used in many fields. Segmentation of 3D model is a technology that allows us to split the model into multiple semantically meaningful parts, for reuse and partial shape retrieval. Our proposed method hypothesizes that any complex 3D shape models consist of “protruded parts and the remaining body. To determine the border between “protruded parts and the remaining body, we need the distance measure. In this paper, we first describe a couple of wellknown distance measures, then we propose a new method for approximating a geodesic distance using “landmarks, which we prove to be much efficient than other distance measures. We also propose a new method for computing boundaries between segments. We have used a segmentation benchmark data set to evaluate the validity of our methods. The results we have obtained so far are very encouraging.
Masaki Aono Shiro Wakida Atsushi Tatsuma
Toyohashi University of Technology, Aichi, Japan
国际会议
2011亚太信号与信息处理协会年度峰会(APSIPAASC 2011)
西安
英文
1-8
2011-10-18（万方平台首次上网日期，不代表论文的发表时间）