Sparsity in signal and image processing: from modeling and representation to reconstruction and processing
Student No.:50
Time:16:00-17:00, Dec. 10(Tue.)
Instructor:Ali Mohammad-Djafari  [Université Paris Sud]
Place:Conference Room 4, floor 2, Jin Chun Yuan West Building
Starting Date:2013-12-10
Ending Date:2013-12-10







Sparse signal and image representation and modeling has recently been the focus of many researchers in many applications and has been used in signal and image reconstruction from direct sparse samples.

The main idea is to use an over-complete basis with the desired properties and project the interested signal or image on that basis in the optimal way of keeping the least number of coefficients.

Sparse signal and image representation also has been used as prior modeling in inverse problems arising in different signal and image processing, in particular, medical or industrial imaging systems. As it is well known, in inverse problems, the main difficulties are the fact that we do not have direct samples (observations) and the ill-posedness of the inversion. Regularization methods have been proposed to introduce prior information and in particular the sparsity of the solution. The Bayesian estimation approach with the sparsity enforcing of the solutions to inverse problems is the focus of this tutorial.

This tutorial talk is an overview and synthetic presentation of these methods and techniques both in signal and image representation and in inverse problems such as signal deconvolution, image restoration and image reconstruction in Computed Tomography.


Contact person: Jian Sun