Joint image compression and encryption techniques

Abstract

The implementation of a joint lossy image compression and encryption algorithm, which is fast in compression and is secure enough against most of the attack from cryptanalysts, is studied in this project. First, the input image is processed under a sequence of color space transformation, downsampling, block splitting, discrete cosine transform, and quantization, in order to reduce the redundancy in the input image. Next, the processed data stream is encoded with multiple Huffman coding tables. The table selected and the swapping orders of the Huffman code are kept secret. As a result, a secure and fast joint image encryption and compression algorithm is achieved. In order to further enhance the security against the chosen9plaintext attack and known9plaintext attack, random bits generated by chaos9based generator will be inserted into the encrypted bit stream according to a secret key. Also, the plaintext will be divided into segments and different keys are used for different segments.