Hybrid prediction error and histogram shifting method for reversible data hiding in video system

Abstract

With the rapid advancement of communication technology, ensuring the security of information transmission has become increasingly crucial. In today's digital landscape, video has emerged as one of the most popular media formats. Video Reversible Data Hiding (RDH) involves embedding information or secret data within a video file while preserving the perceived quality of the resulting stego video. This technique enables the receiver to perfectly restore the original video and extract the embedded secret data. Achieving high-quality stego files while maintaining sufficient capacity for secret data remains a significant challenge. This research proposes a novel reversible video data hiding method that utilizes an innovative prediction error algorithm for intra-frame pixel value prediction error calculation. The algorithm generates sharp histograms based on these prediction errors, with sharper histograms corresponding to a higher Peak signal-to-noise ratio (PSNR) of the stego video. Additionally, the method incorporates an adaptive zero-point system to identify which zero point that require minimal histogram shifts, thus achieving adaptive effects within each frame by applying different shifts based on frame characteristics. The proposed prediction error algorithm enhances the histogram shifting effects and addresses the limitation of not embedding data in the first frame while maintaining superior PSNR. Extensive experimental analysis demonstrates that proposed method surpasses various existing techniques in terms of steganographic video quality.