Assistant Professor, Mathematics Department, Clarkson University
Abstract: Many mathematical problems are multiscale in nature. In particular, images consist of features of varying scales. Thus, multiscale image processing techniques are extremely valuable for image processing. Such techniques are especially useful for medical images where various organs in the body have different scales. In this talk, we will discuss inverse-scale and forward-scale image processing techniques based on variational methods. These techniques give rise to integro-differential equations, and TV-flow equation. We will discuss the applications to 'real-time' denoising, inpainting, and image registration.