Graduate student seminar | The What, Why and How of Image Denoising

Speaker

Dominique Brunet
University of Waterloo

Abstract

I will present a general introduction to image restoration. Starting from the really basics of what is a digital image and what is noise, we will then go through three simple examples of image denoising based on three different approaches: variational methods, Fourier/wavelet decomposition and non-local/fractal image processing. We will then compare the results with a perceptual quality measure: something that models the quality appreciation of the human visual system. At the end of the talk, I will shortly discuss about present/future research I am doing during my PhD at Waterloo.