一种图像DCT域的混沌加密方法研究

Research on Chaos Encryption Method in Image DCT Domain

摘要: 文中针对JPEG图像压缩的特点，提出了一种图像DCT域的混沌加密方法。该算法通过一维Logistic映射迭代产生的混沌矩阵来置乱经过量化的DCT系数矩阵，根据JPEG图像压缩标准，以8 × 8像素块为单位对DCT系数矩阵进行位置置乱，从而使DCT低频成分仍然保留在矩阵的左上角。在混沌序列与置乱后的DCT系数矩阵进行异或操作时，采用只对感兴趣DCT系数加密的方案。利用以上两种方式，不仅提高了图像的加密速度，而且避免了在图像DCT域置乱和异或所导致的图像压缩率降低这一现象。最后利用混沌符号矩阵来扰乱DCT系数的符号位。仿真实验表明该算法具有良好的加密效果，速度快，安全性高，易于硬件实现。

Abstract: In this paper, we suggest one chaos image encryption method in DCT domain according to the characteristics of JPEG image compression. The chaotic models used in this algorithm are Logistic mapping and Chebyshev mapping. The one-dimensional Logistic mapping is used to generate a chaotic sequence which is considered as control matrix for scrambling DCT coefficient matrix. Ac-cording to the JPEG image compression standard, scrambling the DCT coefficient matrix by 8 × 8 pixels in blocks can make the low-frequency composition remain in the upper-left corner of the DCT matrix. When we implement XOR operation between chaotic sequence and DCT coefficient matrix, we only encrypt interested DCT coefficients. Using the above two methods, we not only improve the encryption speed of the image, but also avoid low image compression rates due to scrambling the image encryption method. At last, we use the two chaotic models to generate sign matrix. The simulation results show that the algorithm has good encryption effects, fast encryption speed, and high security.