SEARCHING QUERY BY COLOR CONTENT OF AN IMAGE USING INDEPENDENT COMPONENT ANALYSIS

Content-based image retrieval (CBIR), also known as query by image content (QBIC) and content-based visual information retrieval (CBV IR) is the application of computer vision to the image retrieval problem, that is, the problem o f searching for digital images in large databases. In CBIR each image which is stored in th e database has its features extracted and compared to the features of the query image. The fe atur s that are to be used by the computer should correspond directly to routine notions of vi sion like color, texture, pattern and shape. In Content-based the search will analyze the actual co ntents of the image based on various parameters like color, shape, texture, or any other information which can be derived from the image itself. A major problem of feature-based char acterizations of visual data is the high dimensionality of the feature spaces. The feature s pace becomes increasingly difficult to index efficiently with increased dimensionality. If the f atures are properly chosen, they may lend well to a natural hierarchy in indexing, or be construct ed from a more advantageous space, which can be efficiently indexed. Many indexing techniques are based on global featu r s distribution such as Gabor Wavelets. [1]. In this paper we present an approach for global fe atur extraction using an technique known as Independent Component Analysis (ICA). A comparative study is done between ICA feature vectors and Gabor feature vectors for 180 different t x ure and natural images in a databank. Result analysis show that extracting color and text ur information by ICA provides significantly improved results in terms of retrieval accuracy, co mputational complexity and storage space of feature vectors as compared to Gabor approaches.