Partial Face Encryption Based on CAT Swarm Optimization
Ayat Muhammed Abd Al-Munaf, Abeer Salim Jamil, Nidaa Flaih Hassan · 2022
Videos are vulnerable to file hacking and unauthorized sharing or data loss. Many protect unauthorized viewing of video material via video encryption. The encryption of people's faces is becoming increasingly important. The main aim of this work is to use a partial encryption to encrypt the human face, so as to enhance people's privacy. However, because the complete video encryption is time-consuming, the partial encryption is utilized for efficient and speedy encryption. This paper suggests a partial video encryption by encrypting only the edges in the face region of video frames. Light encryption techniques for video encryption are preferable to be utilized. Thus, the Salsa20 algorithm has been employed to encrypt the edges in faces. In contrast, the Viola-Jon algorithm has been used for face area identification, and the Cat Swarm Optimization (CSO) approach has been utilized for edge detection. The CSO and Canny edge detection operators are used in the edge detection procedure. For evaluation, the PSNR, MSE, correlation, UACI, NPCR, and entropy measurements have been utilized to evaluate the proposed work. In several assessment criteria, the suggested algorithm outperformed the Canny operators, according to the comparison results.