Institute of Information Theory and Automation

Mathematical Methods for Resolution Enhancement of Digital Images

Project leader: Prof. Ing. Jan Flusser, DrSc.
Department: ZOI
Supported by (ID): ASCR and CSIC (Spain) bilateral
Duration: 2007 - 2008
Publications at UTIA: list

Abstract:

The project proposal is a continuation of our current joint project “Multifocus and multimodal image fusion techniques for biomedical applications” and the previous one “New developments on multimodality image fusion and spatial variant image processing”, which both have been very successful. They led to several joint publications (see the list above), fruitful short-term visits, and attracted several PhD. students that joined our research groups. Thanks to the first project, Dr. Sroubek from UTIA received a scholarship from Spanish Ministry of Education, Culture and Sport in the frame of the “National program for the mobility of Spanish and foreign university professors and researchers”. The scholarship allowed him to spend 18 months in the Instituto de Optica, Madrid, and to work further on this topic, which was crowned with a press conference held for nationwide Spanish newspapers and a Spanish national television TVE. The new proposal is concerned with superresolution (SR) of digital images. The term superresolution (sometimes superresolution imaging or superresolution fusion) means in general methods and tools for improvement spatial resolution of digital images beyond the physical limit of the sensor. A general approach to SR consists of combining several low-resolution images (video frames) of the same scene together in order to obtain one high-resolution image. The goal of this new project is to solve the superresolution problem in a very general form that would enable us to cope with all possible radiometric (blurs) and geometric deformations. We aim to develop a robust method that can be utilized in a wide variety of applications, such as security and surveillance cameras, cell phones, web cameras, low-cost video cameras, etc. Preliminary results indicate that combining blind deconvolution with superreolution is a correct approach to follow.
Responsible for information: ZOI
Last modification: 18.04.2014
Institute of Information Theory and Automation