文章基本信息

标题：Parallelization of Image Similarity Analysis
本地全文：下载
作者：Naiouf M. ; Tarrio D ; De Giusti A 等
期刊名称：Journal of Computer Science and Technology
印刷版ISSN：1666-6046
电子版ISSN：1666-6038
出版年度：2001
卷号：1
期号：5
出版社：Iberoamerican Science & Technology Education Consortium
摘要：The algorithmical architecture and structure is presented for the parallelization ofimage similarity analysis, based on obtaining multiple digital signatures for each image,in which each "signature" is composed by the most representative coefficients of thewavelet transform of the corresponding image area.In the present paper, image representation by wavelet transform coefficients isanalyzed, as well as the convenience/necessity of using multiple coefficients for thestudy of similarity of images which may have transferred components, with change ofsizes, color or texture.The complexity of the involved computation justifies parallelization, and thesuggested solution constitutes a combination of a multiprocessors "pipelining", beingeach of them an homogeneous parallel architecture which obtains signature coefficients(wavelet). Partial reusability of computations for successive signatures makes thesearchitectures pipelining compulsory.
关键词：Parallel Algorithms. Image similarity analysis. Pattern recognition. Wavelet Transform.;Parallel architectures