Harris Hawks optimization with multikeyhomomorphic encryption scheme for video steganography

Harris Hawks optimization with multikeyhomomorphic encryption scheme for video steganography

연구 분야: Strategies

학회: Multimedia Tools and Applications

Video steganography hides data like images, text, or other videos within the video file by modifying its least significant bit. Usually, the modification is invisible to the human eye but can be extracted and detected using a steganography tool. Video steganography can be used for hiding copyrighted content within the video file or for transmitting secret messages. Encryption and Video steganography are two approaches to protecting sensitive video information from unauthorized access and detection. However, the security of the secret message was not effectively analyzed by the prevailing works for securing confidential user information. Hence, in this work, optimization and encryption-based video steganography are developed to improve video information security. This technique gives an additional layer of protection for the video information, which makes it highly challenging for attackers to detect and access the data. This study introduces a Harris Hawks Optimization with a MultikeyHomomorphic Encryption Scheme for Video Steganography (HHO-MKHEVS) technique. The presented HHO-MKHEVS technique accomplishes security in videos via two processes, namely steganography, and encryption. At the initial stage, the HHO-MKHEVS technique transforms the input video into a set of frames. Secondly, the HHO algorithm is applied for the optimal selection of pixels in the cover video images for hiding secret messages. Next, the encryption of the stego frame takes place using the MKHE technique. Finally, the embedding and extraction processes are performed to achieve security. The experimental results of the HHO-MKHEVS algorithm are examined on a set of videos by comparing with the related existing techniques, such as EBFOA-PSVS, IWOA, WOA, and GWO by using the Python software tool. The two different datasets, namely the UCF101 Human action dataset and Multi-Object Tracking (MOT) 20 dataset are utilized to analyze the different performance metrics, including MSE, SSIM, NCC, Security level, PSNR, encryption time, and decryption time. The outcomes indicated that the security level of 99.85%, SSIM of 99.99%, and 935 ms of encryption time is achieved by the proposed method, which represents the betterment of the HHO-MKHEVS method over other approaches.