Content Based Image Retrieval by Using Daubechies Wavelet Transform

In this paper, a new algorithm for image indexing and retrieval using Daubechies Wavelet Transform (DWT) is presented. In this DWT has been implemented by using wavelet sub-bands of R, G and B components of images from Coral-1000 database. The database images are decomposed up to third level decomposition. The features of database images are stored in matrix. The query image is decomposed up to third level decomposition by same method and its features are compared with saved database features by using Cityblock and Euclidean distances and top-20 similar images are retrieved. The extracted wavelet features and the retrieval results demonstrate significant improvement in precision and average retrieval rate. From experimental results it is revealed that Dinosaur class gives excellent results. KEYWORDS—CBIR, Image database, Wavelet Transform, Daubechies Wavelet Transform, Similarity measure.