高技术通讯2012，Vol.22Issue(4)：368-373,6.DOI:10.3772/j.issn.1002-0470.2012.04.006
基于稀疏分解的置换混叠图像盲分离
Blind separation for permuted alias images based on sparse decomposition
摘要
Abstract
This paper presents a new blind separation algorithm for permuted alias images based on sparse decomposing, aiming at a type of permuted alias images with morphological diversity. The method uses the Contourlet dictionary and the local discrete cosine transform dictionary respectively as the characteristic fields of separation so as to use the spaisity diversity existing in sparse representation of the permuting and permuted regions of a permuted alias image. And then the texture image is separated from the permuted alias image by decomposing the permuted alias image on the two characteristic fields. The simulation results show that the proposed algorithm can separate effectively texture images from the permuted alias images regardless of size, location number and types of texture image for a permuted alias image comprising the piecewise smooth part and the texture part.关键词
置换混叠图像/稀疏表示/冗余字典/形态差异/稀疏分解/盲分离Key words
permuted alias images/ sparse representation/ redundant dictionary/ morphological diversity/ sparse decomposition/ blind separation引用本文复制引用
段新涛,方勇..基于稀疏分解的置换混叠图像盲分离[J].高技术通讯,2012,22(4):368-373,6.基金项目
国家自然科学基金(60872114),上海市科委重点实验室项目(08DZ2231100),上海市重点学科基金(S30108)和上海市教委科研创新重点项目(09ZZ89)资助. （60872114）
