Elastic block scrambling and RGB channel mixed image encryption algorithm based on 2D-SCLH chaotic system

Abstract In view of the shortcomings of current color image encryption schemes in terms of resistance to statistical attacks and scrambling adequacy, this study proposes a highly secure image encryption scheme. The scheme is constructed based on an extended two-dimensional sine-cosine-logistic-Hennon (2D-SCLH) chaotic map.Through bifurcation diagrams, Lyapunov exponents, sample entropy and trajectory diagram analysis, it is verified that the chaotic system has the characteristics of wide chaos range and high complexity. As a dynamic chaotic sequence source, it provides key support for the encryption process,lays the foundation for the security of the scheme, and further improves the color image encryption effect. Based on this, this study proposes a two-stage scrambling-diffusion mechanism: adaptive block scrambling and three-channel mixed scrambling: decompose the color image into RGB channels, dynamically determine the number of blocks according to the image size, and use the chaotic sequence to independently scramble, and then nonlinearly fuse the pixels of the three channels of R, G, and B to break the independence of the channels. Dynamic selective diffusion: divide the scrambled image into four quadrants; dynamically select different chaotic sequence subsets according to the mean of the four corner pixel values of each channel in each quadrant; combine the selected chaotic sequence through channel-level XOR operation to realize the diffusion modification of pixel values, significantly changing the statistical characteristics of pixel values. Experimental results show that the scheme can effectively process color images of any size and effectively destroy the correlation between adjacent pixels. The correlation coefficients of adjacent pixels in the horizontal, vertical and diagonal directions are close to 0. It significantly improves the ability to resist statistical and differential attacks. The NPCR and UACI values are close to the ideal values of 99.61% and 33.46% respectively, which is suitable for high-security scenarios of color-coded images.