Multi-View Light Field Images Compression and Encryption Using Enhanced 3D Chaotic System and Pixel-Bit-Scrambling

Light field images, rich in spatial and angular domain information, have garnered increasing interest in fields like three-dimonsions (3D) imaging and virtual reality. Despite advancements in natural image encryption, encryption techniques for light field images remain nascent. This paper proposes a hierarchical compression encryption scheme based on light field multi-view array, using Zu Chongzhi algorithm (ZUC), chaotic systems, and pixel permutation and diffusion. Initially, the view array undergoes processing via our proposed prediction structure to yield residual and central view. Subsequently, the residual views are segmented into few layers and individually encoded using high efficiency video coding (HEVC) encoder to produce bitstreams, each of which is sequentially chain-encrypted using the ZUC algorithm. Finally, an enhanced three-dimensional chaos system is employed to generate a key, which is applied to the pixel-shift, bit-shift, and pixel substitution operation of the central color image. Experimental results demonstrate that our compression encryption scheme not only achieves significant compression effects but also surpasses the existing schemes in terms of anti-statistical attack ability, anti-differential attack ability, and keys sensitivity.