Study On Chaos-based Image Encryption Technology

With fast developing of computer network and information technology, a huge amount of data of digital multimedia information, such as audio & video can be expediently and fast transmitted in network. Multimedia communication is becoming an important propagating method among people. As one of the most important expression forms of multimedia information, digital image has many advantages such as visual, dramatic and intuitive performances. When problems of storage and transmission caused by a huge amount of data have been overcome, it is becoming or has become a primary communication method today and in the future. Digital image brings much convenience to us, but at the same time it falls across many security problems. For example, some image information involving secrets of state, business or individual privacy is easily browsed, intercepted, intruded, illegally copied and promulgated. If these images can't be protected, it brings unthinkable affection. So the security of image information is becoming the focus on public interest.Image encryption is one core technology in image security, and also one way for direct and effective image security protecting. Meanwhile, it is indispensable too in the fields of information hiding and digital watermarking. Therefore as the increasing applications to multimedia technology today, to study image encryption is of profound theoretical and practical significance.Based on chaos characteristics, the image encryption technologies have been more deeply studied in this dissertation. Some contributions are as follows:1. After researching and summarizing chaotic conditions for estimating a non-linear system in finite precision, a new method to produce digital chaotic sequence, which is extracted from the binary data quantized by multi-level, is presented. The pseudorandom sequences produced by our algorithm are of better cryptograph performance by testing their statistics. The results also show that the digital chaotic sequences fit the requirements in image encryption system.2. Two kinds of image scrambling algorithms, which called m-sequence transform and modulo map, are respectively presented. Through subjective and objective evaluation, m-sequence transform algorithm is better than others. Its advantages contain better scrambling effectiveness and higher scrambling efficiency. In addition, a general image scrambling method based on pseudorandom sequences is put forward after intensively analyzing it. And we validate the method by SINE map. According to the scrambling effect, scrambling efficiency, key space and free degree of the designed algorithm, the method is a very effective algorithm.3. After analyzing the disadvantage of the image encrypted by generic OR-operation, an improved OR-operating encryption method is put forward based on eight neighbors. This method can overcome the disadvantage that the former method is easily attacked by selective plaintext. Further considering that the states of the shifters in m-sequence generator are of pseudo-random property and ergodicity (except for all zero-states), a new way to encrypt image by m-sequence integer modulation is put forward, its fast realizing algorithm is also proposed. The method can resist on selective plaintext attack, and also has advantages of good encryption effect, fast encrypting speed.4. An improved scheme of wavelet-based image encryption algorithm is put forward. The experiment results show that the improved method is effective. Compared with other methods in the existing references, the proposed algorithm has improved the encryption security. Furthermore, a simulation test of image encryption based of fractional Fourier transform is implemented and the experiment results show that it is feasible.5. The chaotic neural network is elementary studied and discussed. An image cryptography combined Hopfield neural network with chaos is presented, and some good encryption effects are got and shown in this dissertation.6. To aim at particular demands to protect information in sensitive area on the TV interview, a video encryption scheme is successful designed by applying video tracking algorithm based mean-shift and image encryption method. In our scheme, we firstly track on the sensitive area using mean-shift tracking algorithm, then encrypt it by an improved m-sequence transform, and we get famous result. Compared with mosaic processing on TV, our algorithm has two advantages: one is that it can automatically real-time track on the sensitive area; the other is that its effect is better than the mosaic style. This is a successful image encryption scheme in the specific application combined image encryption with video tracking. And there hasn't been heard any report of this kind of video encryption work in our country, even in the world.To sum up the above contributions, chaos theory and application in image encryption are mainly studied in this dissertation. The image encryption in spatial domain is intensively studied. Some encryption algorithms such as m-sequence transform, one general method of image scrambling and m-sequence integer modulation are separately presented. An improved scheme of wavelet-based image encryption algorithm is also put forward, and a video encryption scheme is successful designed by using improved m-sequence transform. This scheme solves the problem that tracking sensitive area and encrypting the moving pictures to protect the information on the TV interview.