A non-rigid registration algorithm for compatible skeletonization
This paper presents a registration framework for fitting one skeleton to another. The algorithm matches and transforms closest points iteratively in a non-rigid manner, so that given a set of skeletons, a compatible deforming sequence can be generated automatically. In our optimization, the variables are transformations on skeletal nodes, and the objective function is a weighted combination of three errors: the Data Error, Smoothness error, and Length-Preserving Error. The registration can be efficiently solved using standard L-BFGS non-linear optimizer. we demonstrate effectiveness of our skeleton matching framework on various 3D models.
non-rigid registration:skeleton matching:deformations
Li Wei Wuyi Yu Maoqing Li Xin Li
Xiamen University Louisiana State University
国际会议
The 5th International Conference on Computer Science & Education(第五届国际计算机新技术与教育学术研讨会 ICCSE10)
合肥
英文
1888-1893
2010-08-24（万方平台首次上网日期，不代表论文的发表时间）