turkmath.org

Recent years have witnessed the emergence of a new class of image smoothing techniques which have provided new insights and raised new questions about the nature of this well-studied problem. These new models interestingly aim at separating a given image into its structure and texture layers by utilizing non-gradient based definitions for edges or special measures that can distinguish edges from texture. In this talk, I will first talk about how region statistics can be used as a patch descriptor, and then describe a new smoothing approach which employs pixel affinities defined over this patch descriptor. The experimental results show that the proposed approach is very effective for structure extraction from texture, leading to better image decompositions as compared to the state-of-the-art. I will also demonstrate several applications of the proposed model on some image editing and manipulation tasks such as edge detection, image abstraction, texture and detail enhancement, etc. This is a joint work with Levent Karacan and Aykut Erdem, Department of Computer Engineering, Hacettepe University.