Compressed Domain Image Retrieval Using Thumbnails of Images

The emergence of multimedia, the availability of a large digital archives, and the rapid growth of World Wide Web (WWW) have recently attracted research efforts in providing tools for effective retrieval of image database. Content based image retrieval (CBIR) has emerged as a promising approach for image retrieval in last decade. To reduce the storage and transmission requirements in image applications, images are compressed while maintaining acceptable visual quality. In this paper, a unique approach which emphasizes the compressed-domain solutions to image retrieval is explored. A reduced complexity approach for efficient retrieval of images using VQ on thumbnails of images is presented. A VQ codebook is used to represent the image. The similarity is computed in terms of MSE between the thumbnail of query image and all the image thumbnails in the database. The system is tested for various thumbnail sizes. UCID database is used for testing the system.