EE Seminar: Coherent Clustering and its Use for Image Denoising

~~
Moran Vatelmacher, 
M.Sc. student under the supervision of Prof. Shai Avidan

Monday, June 22, 2015 at 15:00
Room 011, Kitot Bldg., Faculty of Engineering

Coherent Clustering and its Use for Image Denoising

Abstract

Cluster analysis or clustering is the task of grouping a set of objects in such a way that objects in the same group (called a cluster) are more similar (in some sense or another) to each other than to those in other groups (clusters). Today, the classical clustering methods are based on appearance information only. This information is sometimes insufficient, especially in the case of noisy data where there is a need to differentiate inliers from outliers. The human eye can sometimes see that a certain area belongs to a certain cluster even if the appearance distance from another cluster is shorter. This is partially done using context analysis. When using context information it is easier to identify outliers. This kind of analysis can help in image analysis applications such as image denoising - discussed here - for which segmentation of the image into clusters is relevant. Our results show that better clustering improves denoising.