Research On Image Uniform Scrambling Method Based On Affine Module Transformation

With the rapid development and application of computer network and multimedia techniques, the security of image information comes to be an increasing important problem. Digital image has many characteristics, such as large data, high redundancy and so on, which are different form text information. These characteristics make traditional encryption methods are not suitable to image information. In recent years, people develop image scrambling technology by the features of digital image. There are many mature digital image scrambling method just like the method based on position scrambling, the method based on the value of pixel, and combine with chaotic map, and so on.The digital image can be seen as a matrix. The principle of scrambling encryption is disrupting the matrix through some kind of algorithm and change the value in the matrix. This not only get a"beyond recognition"of image, but also change the image histogram. Affine modular transformation is an easily applied method of image encryption nowadays, and it has same periodicity as Cat transformation but it has some disadvantages such as small key quantities and insecurity. Its anti-attack capability is weak, and the effect of few times scrambling is bad. Logistic chaotic sequence has many merits such as randomicity, strong sensibility to initial condition, ergodicity etc, it is usually combined with other methods using in image encryption. But Logistic chaotic sequence is not satisfied even distribution.A concept of uniform block encryption is proposed in this paper from entropy and enhancing image information capacity. We make theoretical analysis for an ideal state of uniform scrambling and sub-block size by mathematics, and then make improvements of Logistic chaotic map to meet the uniform distribution. We make position scrambling by classic affine module transformation combine to modified Logistic chaotic map, and get scrambling based on the value of pixel by spread factor combine with chaotic map to get the purpose that change the image histogram. The algorithm finally take the effect of double encryption.With MATLAB software, we do substantial experiment, theoretical analysis and the experimental results show that the algorithm in this paper can get better encryption effect in one time and reach high-safety. Moreover, the proposed algorithm also can resist cut attack and noise attack. In this paper, quantitative indicators and detailed analysis of the experimental results, so the results demonstrate the feasibility practicality and effectiveness of algorithm. Finally we use the VC++ 6.0 compile the software can run independently. Practice has fully proved that the software has the practicality.