Video motion and Decryption. A Comparative Study for Compressed Videos
Anastasiia Voitenko, Mourad Oussalah · 2024
Transmission of video data is known to cause high privacy leak risk due to effect of data compression rates and disparate standards, which challenge encryption/decryption methods. This paper presents a comparison of AES cipher, Arnold Cat chaotic map and Logistic chaotic map in the context of various video motion types (slow, medium and fast). The study aims to distinguish the video motion types according to decryption result. For this purpose, several performance metrics were tested, which include unified averaged changed intensity (UACI), number of changing pixel rate, peak signal-to-noise ratio, correlation coefficient (CC) and histogram of intensity. A simulation setting is devised where each video undergoes encryption phase, followed by lossy compression using discrete cosine transform and followed by image decryption phase. Simulation results demonstrate that AES and Logistic chaotic map present relatively close performance outperformance that of Arnold cat map. On the other hand, correlation coefficient appears to be the only performance metric that presents interesting capabilities to distinguish slow-motion video from medium and fast-motion types.