报 告 人： Ke Chen Professor
The University of Liverpool UK
报告题目：On Some Variational Models and Their Algorithms for Restoration and Segmentation of Images
报告时间：2013年12月30日下午4:00
报告地点：静远楼1506学术报告厅
主办单位：数学与统计学院、科技处
Abstract：
Mathematical imaging is an interesting subject to study as almost any knowledge of mathematics can be applied to an imaging problem.Image analysis techniques involve analysis, optimization, differential geometry and nonlinear partial differential equations, computational algorithms and numerical analysis.
In this talk I first review various models and techniques in the variational framework that are used for restoration and segmentation of images.Then I focus on the particular problem of detecting local features in a large image as it is increasingly useful in a number of applications e.g. in medical imaging. Some of our recent work on this local segmentation problem modelled by non-convex energy minimization and non-linear partial differential equations (PDEs) are shown. The issue of reformulating a non-convex and local functional as a convex one is considered. Various test results are given to illustrate the models presented. Some open challenges are also highlighted.
Ke Chen Professor’s Introduction:
Ke Chen is a computational mathematician, specialising in numerical linear algebra, numerical PDEs, numerical optimization and fast algorithms.All research projects in his group are problems-driven, with recent applications in the emerging fields of imaging sciences.
Ke Chen is a deupty head of Department of Mathematical Sciences, director of CMIT(Centre for Mathematical Imaging Techniques), head of Applied Mathematics. He is an editorial board of Int. J. Computer Mathematics (IJCM)、Numerical Algorithms (Springer)、Journal of Applied Mathematics and Journal of Mathematical Research with Applications. He has been published more than 100 scientific papers in all kinds of magazines.
