讲座题目：3D multiscale approach for diffeomorphic image registration

主办单位：三峡数学研究中心/理学院

报告专家：韩欢 副教授（武汉理工大学）

报告时间：2021年11月18日（周四） 上午（9:00—12: 00）

报告地点：腾讯会议（ID： 829 258 683）

专家简介：韩欢，武汉理工大学数学系副教授，硕导，长期从事图像配准、去噪相关的变分理论与应用研究；在微分同胚图像配准理论方法与数值计算方面做出了系列结果，发表在SIAM J. Multiscale Model. Sim., SIAM J. Imag. Sci.，Inverse Probl. Imaging 等重要数学、应用数学国际主流期刊上。担任SIAM J. Imag. Sci., J. Sci. Comput., Appl. Math. Model.等多个国际主流数学期刊审稿人；主持国家自然科学基金和国家重点实验室开放项目各1项，参与国家重点研发项目1项。

报告摘要： Many 3D image registration models are proposed during the last decades. All these models achieve only the minimizer of the cost functional with some prior regularization. In addition, physical mesh folding phenomenon is not taken into consideration in most models. This raises a question of whether one can find a global smooth minimizer of cost functional without regularization and mesh folding. To give an answer to this question and eliminate mesh folding, we proposed a multi-scale approach for 3D conformal image registration. This approach achieves a smooth minimizer for cost functional without regularization. Furthermore, a numerical algorithm for this multi-scale approach is proposed and the convergence of the numerical algorithm is proved. Moreover, several numerical tests are also listed to show the good performance of the proposed algorithm.