Content-Based Picture Retrieval through Color, Texture and Shape Factors

Abstract: The objective of this paper is to depict the retrieval of pictures from a database using texture, shape and color factors of a picture. The size of an output picture is reduced to (64x64) from the input size is of (256x256) through minimum and maximum quantifier. Color Co-occurrence Feature (CCF) is used to extract the color factors. Error Diffusion Block Truncation Coding (EDBTC) and Bit Pattern Feature (BPF) are used to extract the shape of the picture. Gabor wavelet is used to extract the texture of the picture. Pictures are retrieved using similarity measures through Euclidean distances. The accuracy 97% is achieved through the above methods.

Keywords: Color Histogram Feature, Color Co-Occurrence Feature, Error Diffusion Block Truncation Coding, Bit Pattern Feature, Gabor Wavelet, Euclidean Distance.