An Automatic Registration Algorithm for 3D Maxillofacial Model

详细信息    查看全文

• 作者：Luwen Qiu ; Zhongwei Zhou ; Jixiang Guo ; Jiancheng Lv
• 刊名：3D Research
• 出版年：2016
• 出版时间：September 2016
• 年：2016
• 卷：7
• 期：3
• 全文大小：1,018 KB
• 刊物类别：Engineering
• 刊物主题：Computer Imaging, Vision, Pattern Recognition and Graphics
  Signal,Image and Speech Processing
  
• 出版者：3D Display Research Center, co-published with Springer
• ISSN：2092-6731
• 卷排序：7

文摘

3D image registration aims at aligning two 3D data sets in a common coordinate system, which has been widely used in computer vision, pattern recognition and computer assisted surgery. One challenging problem in 3D registration is that point-wise correspondences between two point sets are often unknown apriori. In this work, we develop an automatic algorithm for 3D maxillofacial models registration including facial surface model and skull model. Our proposed registration algorithm can achieve a good alignment result between partial and whole maxillofacial model in spite of ambiguous matching, which has a potential application in the oral and maxillofacial reparative and reconstructive surgery. The proposed algorithm includes three steps: (1) 3D-SIFT features extraction and FPFH descriptors construction; (2) feature matching using SAC-IA; (3) coarse rigid alignment and refinement by ICP. Experiments on facial surfaces and mandible skull models demonstrate the efficiency and robustness of our algorithm.Keywords3D-SIFTFPFHSAC-IARegistration3D maxillofacial model