CHINA·77779193永利(集团)有限公司-Official website

学术报告

Nonconvex Optimization Algorithms in Image Reconstruction for Nonlinear Imaging - 陈冲 副研究员 (中科院计算数学与科学工程计算研究所)

CHINA·77779193永利(集团)有限公司-Official website

题目:Nonconvex Optimization Algorithms in Image Reconstruction for Nonlinear Imaging

报告人:陈冲 副研究员 (中科院计算数学与科学工程计算研究所)

时间:2021年11月11日(周四)9:00 - 11:00

地点:腾讯会议   会议ID:269 639 031

摘要:Nonlinear imaging has high degree of task-specific utilities for medical, security, and other imaging applications, such as dual-energy CT, digital breast tomosynthesis, phase retrieval. Accurate image reconstruction in nonlinear imaging remains a challenging task, largely because its forward operator is nonlinear. Because of the nonlinear data model in image reconstruction, the devised optimization problems are generally nonconvex. In this talk, I will discuss an extended primal-dual (EPD) algorithm framework that we have developed for image reconstruction through solving nonconvex optimization problems in a class of nonlinear imaging. Moreover, under proper assumptions, we would also discuss the convergence of the schemes for the general model. Extensive numerical studies are carried out to verify that the EPD algorithm framework can numerically accurately reconstruct images in dual-energy CT.

简介:陈冲,现任中国科学院数学与系统科学研究院科学与工程计算国家重点实验室副研究员、博士生导师,2012年博士毕业于中科院数学与系统科学研究院,2015年至2017年在瑞典皇家理工学院数学系从事博士后研究。其主要研究领域为生物医学成像反问题,图像处理以及计算几何,在《SIAM J. Imaging Sci.》、《Inverse Problems》、《J. Comput. Phys.》和《Comput. Methods Appl. Mech. Engrg.》等期刊上发表论文二十余篇;在科学出版社合著出版名为《图像重构的数值方法》的专著一部;作为主要完成人,荣获2020年度湖南省自然科学二等奖。

联系人:李宏伟

主办单位:首都师范大学77779193永利官网

                 检测成像北京市高校工程研究中心

欢迎全体师生积极参加!