Paper Details

Content based image retrieval (CBIR) great deals with the problem of searching for digitalimages from large databases. CBIR is also known as Query By Image Content (QBIC). CBIR workson basis of extracting low level features of an image such as color, texture and shape and comparessimilarity measures with feature database for finding relevant image retrieval from large database. Inthis paper, we have proposed a novel approach of image retrieval on the basis of multi featureextraction using HSV histogram, log-gabor filter, wavelet transformation and SVM (Support VectorMachine) classification. In the proposed system texture feature extracted using log gabor forinvariant scale and rotation image retrieval. Wavelet transformation is used to extract the localfeature of images for most important part of the image. A SVM (Support Vector Machine) c lassifiercan reduce the semantic gap between low level image features and high level image concept andlearned from training data of relevance images and irrelevance images marked by user. Inexperimental result we have demonstrate the Graphical and statically result that our proposed methodoutperform the other existing methods.

Bhavik M. Patel, Prof. U. K. Jaliya, Prof. Kanubhai G. Patel, “A Novel Approach For CBIR Using Multi-feature Extraction And SVM Classification”, International Journal of Advance Engineering and Research Development (IJAERD), Vol. 1, Issue 5, pp. 464-470, May 2014.