Image Quality Assessment Using Natural Scene Statistics

Mr. Hamid R. Sheikh

Ph.D. Student
Laboratory for Image and Video Engineering (LIVE)
Electrical and Computer Engineering
The University of Texas at Austin
Austin, TX 78712, USA

Friday, November 15th, 10:00 AM, ENS 637

sheikh@ece.utexas.edu

This presentation is the open-to-public part of Mr. Sheikh's Ph.D. Qualifying Examination

Abstract

Digital images are part of our lives now. Measurement of image quality is crucial for many image-processing algorithms, such as acquisition, compression, restoration, enhancement and reproduction. Traditionally, image quality assessment algorithms have focused on measuring image similarity or fidelity, where quality is measured as fidelity with respect to a `reference' or the `perfect' image. This approach (known as Full-Reference quality assessment) assumes that the reference image is available to compare with the image whose quality is being evaluated. However, human beings can evaluate the quality of images without the reference, knowing whether an image is 'good' or 'bad' almost like a reflex action. We are thus motivated to consider the problem of 'blind quality assessment' (also known as No-Reference quality assessment), where a computer tries to evaluate the quality of an image without the reference. In this talk, I will present using Natural Scene Statistics for designing NR image quality assessment algorithms. Images of natural scenes form a tiny subspace in the set of all possible images (most of which are random noise) and researchers have developed models to characterize natural images in terms of their statistics. I propose to use these models for quantifying the quality of natural images.

Biography

Hamid Rahim Sheikh earned his B.Sc. in Electrical Engineering from the University of Engineering and Technology, Lahore, Pakistan. He joined the University of Texas at Austin in Fall 99 and in January 2000 he joined the Laboratory for Image and Video Engineering (LIVE) as a graduate research assisstant. He earned his M.S. in engineering in May 2001, and has been working towards his PhD since then, under the supervision of Dr. Alan C. Bovik. His research interests include image and video quality assessment, human visual system modeling, image and video coding, and embedded implementations of video codecs.

A list of Wireless Networking and Communications Seminars is available at from the ECE department Web pages under "Seminars". The Web address for the Wireless Networking and Communications Seminars is http://signal.ece.utexas.edu/seminars