Compact Composite Descriptors for Content Based Image Retrieval

This paper focuses on the decrement of computational time, by reducing the size of feature extraction. Thus in this work, we have explored the Compact Composite Descriptors (CCD) approach, which enables to reduce the size of image features without affecting the visual content of the image. The proposed approach has two phases. First is to identify and extract the features, by applying Colour Layout Descriptor (CLD), Local Binary Pattern (LBP) and Fourier Descriptor (FD). The second phase is employed to compare the similarity distance between the query images with the image in database, using the Euclidean Distance. The evaluation results show that, the proposed approach has been able to address the issues mentioned earlier.