Applying Permutations and Cuckoo Search for Obtaining a New Steganography Approach in Spatial Domain

Dieaa I. Nassr ； Sohier M. Khamis

《International Journal of Network Security》 23卷1期 (2021/01) Pp. 67-76

https://doi.org/10.6633/IJNS.202101_23(1).09

摘要

Video Steganography is an art and science of embedding secret information into a carrying video file in such a way that others cannot observe the embedded information. Cuckoo Search (CS) is a meta-heuristic algorithm which has been developed by Xin-She Yang and Suash Deb in 2009. CS is very effective in solving many optimization problems that have been found in previous literature. In this paper, a new efficient approach for embedding a secret image in a digital video is proposed. Generally, any colored image consists of three color components (Red, Green, and Blue). So, an image's pixel has three bytes; each of which belongs to one different color component. For security purposes, each secret image's color component is embedded separately into a selected cover video's frame. The proposed approach is based on the permutations on 3 sections of a secret byte, 3-3-2 bits. These three sections are permuted to obtain five different patterns of a specified secret byte. Then, the population of five different pairs is built; each pair consists of one different pattern repeated twice. Good pixels are so chosen via using CS algorithm to achieve the minimum distortion in carrier pixels due to embedding. The sum of absolute values of sectional differences is used as an objective function to compare all the distances between the 3-3-2 Least Significant Bit (LSB) values of a cover frame's pixel and the generated different patterns of a specified secret byte. Experimental results show that the efficiency of the suggested approach is successful since the Peak Signal to Noise Ratio (PSNR) is above 52 decibels.

關鍵字

Cuckoo Search ； Lévy flights ； LSB ； PSNR ； Video Steganography