Performance-Analysis-Based Acceleration of Image Quality Assessment
Laboratory of Computational Perception and Image Quality, School of Electrical and Computer Engineering, Oklahoma State University, Stillwater, OK 74078 USA
IEEE Southwest Symposium on Image Analysis and Interpretation, 2012
@article{phan2012performance,
title={Performance-Analysis-Based Acceleration of Image Quality Assessment},
author={Phan, T. and Sohoni, S. and Chandler, D.M. and Larson, E.C.},
year={2012}
}
Two stages are commonly employed in modern algorithms of image/video quality assessment (QA): (1) a local frequency-based decomposition, and (2) block-based statistical comparisons between the frequency coefficients of the reference and distorted images. This paper presents a performance analysis of and techniques for accelerating these stages. We specifically analyze and accelerate one representative QA algorithm recently developed by the authors (Larson and Chandler, 2010). We identify the bottlenecks from the abovementioned stages, and we present methods of acceleration using integral images, inline expansion, a GPGPU implementation, and other code modifications. We show how a combination of these approaches can yield a speedup of 47x.
May 29, 2012 by hgpu
