期刊名称：International Journal of Electronics and Computer Science Engineering
电子版ISSN：2277-1956
出版年度：2012
卷号：1
期号：3
页码：1459-1465
出版社：Buldanshahr : IJECSE
摘要：In recent years, sparse representation of signal has drawn a great interest. The assumption that natural signals like images admit sparse decomposition over a redundant dictionary leads to efficient algorithms for handling such sources of data. Applications of sparse representation are compression, regularization of inverse problems, feature extraction, noise reduction, pattern classification and blind source separation and more. Using a overcomplete dictionary that contains prototype signal atoms, signal are described by sparse linear combination of these atoms. In this paper we propose that K-SVD algorithm is the best among other sparse representation algorithms like MOD and MAP based algorithms. Given a set of training signals, we seek the dictionary that leads to the best dictionary for each member in this set, under strict sparsity constraints. The K-SVD algorithm is flexible and can work with any pursuit method (e.g., basis pursuit, FOCUSS, or matching pursuit). We analyze and demonstrate its results both on synthetic tests and in applications on real image data. In K-SVD algorithm, K stands for clustering process. K-SVD is an iterative method that alternates between sparse coding of the examples based on the current dictionary and a process of updating the dictionary atoms to better fit the data
