Improving the performance of histogram-based data hiding method in the video environment

Abstract

The rapid development of information technology has brought security to be a crucial factor. Information being transferred between different parties must be well protected to prevent the file from illegal access. Data hiding, as one of the possible securing methods, has been introduced; nevertheless, the quality of the stego file and the capacity of the secret are still challenging. In this research, we work on these problems by exploring the histogram developed in the frame and the prediction error generated between two consecutive frames of a video file. In the generated stego file, these updated prediction errors do not change the composition and the total number of frames. The experimental results show that this proposed method has achieved a better result than the compared algorithms, which is shown by the Peak Signal to Noise Ratio (PSNR) value for various capacities. Equivalently, the number of bits that can be held is higher than others for the same stego quality. Besides, this method is reversible, in which both the cover and the secret can be fully extracted from the stego file. It means this scheme is appropriate to implement in a system, which requires the existence of those both data.