Video Steganography System Based on Optical Flow for Object Detection

Steganography is the art of covert communication; the secret message is hidden in a coverless file. Steganography and steganalysis systems become more important in many areas of information security. In contract with cryptography, it is the existence of the message that hidden and not its meaning. In this paper, a coverless video steganography method is proposed to embed and extract the hidden data based on the motion estimation of optical flow. The proposed method is based upon the selection of macroblocks with the highest optical flow magnitude based on Gunnar Farneback's algorithm. The potential of this method is to produce a minimum distortion in the stego video as it evaluated using multiple factors of the imperceptibility of a successful Coverless video steganography, the proposed system has been tested and evaluated in 8 videos where the results illustrate the efficacity of the proposed modelling scheme.