Title: Digital Image Processing using the Energy Model.
Thời gian: Lúc 14:30 thứ 6, ngày 10/12/2010
Địa điểm: phòng I11G (lầu 11, dãy nhà I)
Người trình bày: GS Tien Dai BUI, ĐH Concordia, Canada.
Abstract:
This talk will present some recent work in our Image Processing Research Lab. I will introduce some fundamental problems in image modeling and propose solutions with particular emphasis on their applications to handwritten document processing, medical imaging, biometrics, and fine art applications. Recently there has been a number of models for digital image enhancement and processing. These models are based on various forms of energy functional, for example, the Mumford-Shah model, active contours, snakes, geodesics and many others. This talk will address some limitations of these models, and will propose a new alternative. Many new results in digital image processing will be presented including image inpainting, image and video enhancement, textline and word detection in handwritten documents, and face age estimation and progression.
AILab.