A Novel Content Based Image Retrieval using YUV Color Space and Texture Features

The escalating quantity of multimedia data leads to a requirement for new sophisticated technique to extract the focused data. The traditional method – manual annotation – alone cannot keep up with the fast increase of accessible information anymore. As a result, Content-Based Image Retrieval (CBIR) systems developed to index the media documents according to the characteristics gathered from their content rather than by textual query based retrieval. Some of the features that can be considered for image retrieval are color, shape, texture, objects layout, edge direction, etc. This paper focuses on using the color feature called YUV along with the texture features using wavelet transform for image retrieval. In this paper, used YUV color space and wavelet transform approach for feature extraction. Firstly, the color space is quantified in non-equal intervals, then constructed one dimension feature vector and represented the color feature. Similarly, the texture feature extraction is obtained by using wavelet. Finally, color feature and texture feature are combined based on wavelet transform. Experimental result shows that the proposed image retrieval system results in better retrieval of images than the existing techniques.