Chaotic Digital Image Encryption Algorithm Based on Compressive Sensing

Current chaotic encryption algorithms for digital images are difficult to eliminate the strong correlation between neighboring pixels, and have problems such as poor attack resistance and low efficiency. In this paper, a chaotic digital image encryption algorithm based on compression perception is proposed. The algorithm extracts the RGB color components of the image by normalization, performs DCT transform after rotational fusion, and processes the image information by using two-dimensional discrete cosine transform and adaptive sparsity compression perception. The chaotic encryption model is constructed by adaptive random phase coding and compressed sensing optimization. The experimental results show that the algorithm can maintain high clarity and consistency of the decrypted image and the detail information is effectively preserved, no matter facing shear attack or destruction attack. The pixel correlation of the algorithm is less than 0.0125, and the encryption and decryption times are both less than 549 ms. It can be seen that the proposed method of the research significantly improves the security and processing efficiency of image transmission, and it can provide a new kind of support for the technological development in this field. These contributions are of great significance in addressing current image security challenges, enhancing the complexity and attack resistance of image encryption, and are suitable for high-security application scenarios such as real-time video surveillance and online image transmission.