Recently, sparse and redundant representation modeling has received extensive attention and shown to be quite effective in signal and image processing. This framework allows for the successful reconstruction of the true image from severely corrupted or undersampled data.
In this talk, we will introduce an image reconstruction scheme based on this framework and show the effectiveness of such a scheme in various applications.