点击次数：

论文类型：期刊论文

发表时间：2015-03-15

发表刊物：Journal of Computational Information Systems

收录刊物：EI、Scopus

卷号：11

期号：6

页面范围：2269-2277

ISSN号：15539105

摘要：With the development of information age, image information has been widely applied in various fields, and content-based image retrieval (CBIR) has become a hot topic to research. However, after extracting the underlying visual image feature, the dimensions of the image feature will be high, which will lead to a higher computation complexity and space complexity in the process of image retrieval. In order to solve this problem, a novel manifold learning algorithm, named globally and locally consistent principal component analysis (GLPCA) algorithm, is proposed. Compared with the traditional dimensionality reduction algorithms such as principal component analysis (PCA) and linear discriminant analysis (LDA), GLPCA considers the global and local characteristics of the data structure comprehensively, and it has better performance on both linear and nonlinear data. The experiment results on Corel image data set show the excellent performance of GLPCA algorithm in image retrieval. Copyright ? 2015 Binary Information Press.