A Novel Hybrid Discrete Cosine Transformation and Visual Cryptographic Technique for Securing Digital Images

Abstract

Security of digital images in today's cyber is a major concern. Transmission of compressed and uncompressed multimedia data is necessary for communication between different devices. Devices of different screen features and resolution need very efficient and fast compression techniques in order to maintain visual features of transmitted media. Robustness is of a key consideration and importance in image compression techniques. Full recovery of images after compression is a challenge for image compression algorithms. It is practically closely impossible for most encrypted-compressed-images or compressed-encryptedimages to be efficiently and fully reconstructed. In this paper we proposed a robust, efficient and fully recoverable encrypted-compressed image based on hybrid discrete cosine transformation and visual cryptographic technique for securing digital images. We first encrypted the image without pixel loss and then compressed the image which resulted into pixel loss during the compression phase. We decrypted the compressed image by obtaining perfect visuals but losses in pixel values. The analysis of the process proved to be an efficient way of encrypting images and compressing them for other channels with less data capacity without the worry of visual loss as well as ensuring security of the data during transmission. The programming and implementation was done using MATLAB. � 2015 IEEE.