Optimized video steganography using Cuckoo Search algorithm

摘要

Steganography is a process of concealing the secret information into cover object such as text, audio, image, and video. In this paper, we propose a new treatment of Cuckoo Search (CS) algorithm to solve the problem of video steganography. CS is a metaheuristic search algorithm which was recently devolved by Xin-She Yang and Suash Deb in 2009, inspired by the cuckoo bird breeding behavior. The suggested algorithm is based on taking secret data byte by byte. The bits of each byte are arranged to obtain five different forms. The next step focuses on searching about the best carrier pixel in the cover frame. The best pixel is determined using Euclidian distance which evaluate the similarity between the pixels and different byte forms. The random move from pixel to another is achieved using Lévy flight random walk. Finally, the suitable carrier pixel is detected and embedded in its RGB components using the 3-3-2 Least Significant Bit (LSB) replacement technique. In addition, each secret image's color component is embedded separately into a selected cover video's frame. Results show that CS is superior to Genetic Algorithm (GA) and the base technique in term of Peak Signal to Noise Ratio (PSNR).