Development and Evaluation of High-Performance Decorrelation Algorithms for the Nonalternating 3D Wavelet Transform
- PDF / 912,592 Bytes
- 11 Pages / 600.03 x 792 pts Page_size
- 0 Downloads / 183 Views
Research Article Development and Evaluation of High-Performance Decorrelation Algorithms for the Nonalternating 3D Wavelet Transform 1 F. J. Quiles,2 and L. Orozco-Barbosa2 ´ E. Moyano-Avila, 1 Departamento 2 Departamento
de Tecnolog´ıas y Sistemas de Informaci`on, Universidad de Castilla-La Mancha, 45071 Toledo, Spain de Sistemas Tecnolog´ıas, Universidad de Castilla-La Mancha, 02071 Albacete, Spain
Received 2 September 2006; Revised 30 December 2006; Accepted 27 March 2007 Recommended by Erwin De Kock We introduce and evaluate the implementations of three parallel video-sequences decorrelation algorithms. The proposed algorithms are based on the nonalternating classic three-dimensional wavelet transform (3D-WT). The parallel implementations of the algorithms are developed and tested on a shared memory system, an SGI origin 3800 supercomputer making use of a messagepassing paradigm. We evaluate and analyze the performance of the implementations in terms of the response time and speed-up factor by varying the number of processors and various video coding parameters. The key points enabling the development of highly efficient implementations rely on the partitioning of the video sequences into groups of frames and a workload distribution strategy supplemented by the use of parallel I/O primitives, for better exploiting the inherent features of the application and computing platform. We also evaluate the effectiveness of our algorithms in terms of the first-order entropy. ´ Copyright © 2007 E. Moyano-Avila et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
1.
INTRODUCTION
In recent years, the rapid growth in the field of medical imaging has enabled the development of several new classes of digital images, videos, or image sequences. Visual data applications are characterized by their stringent requirements in terms of storage, processing, and transmission. The use of video compression techniques can significantly reduce the storage and communications requirements of video applications. However, the compression of visual data is a computationally intensive. In this context, the use of parallel highperformance computer architectures can prove to be an effective solution particularly when large volumes of visual data have to be processed [1, 2]. In this paper, we evaluate the parallel implementations of three image-sequences decorrelation algorithms having been recently introduced in one of our previous works [3]. The algorithms had been developed around the nonalternating classic wavelet-based transform algorithms versus the Classic method [4]. The development of these novel algorithms has been motivated by the need of coping with the stringent requirements of visual data: huge storage and high processing demands [3]. Due to the inherent characteristics of the
video sequences of our interests, that is, angiograms, we have partitioned the video sequ
Data Loading...
