Analysis of Reversible Steganography in Image Histograms

Reversible steganography enables recovery of the cover image from the stego image after extracting the hidden data, making it valuable for applications such as medical imaging, content authentication, and military communication. Existing research explores reversible data hiding in pixel-domain and prediction error histograms. Pixel-domain techniques often enhance hiding capacity by dividing the image, its histogram, or both into non-overlapping blocks prior to embedding. In this study, we analyzed the impact of image and histogram partitioning on hiding capacity and histogram distortion. Our results show that increasing the number of image blocks does not always improve the capacity, as the optimal number of blocks varies across images. Additionally, increasing image blocks tends to make histogram distortions more noticeable. For histogram partitioning, a two-block division consistently yields optimal results. Further, we identify that the existing techniques use a fixed histogram shift for all payloads. We propose a novel algorithm that adapts histogram shifting dynamically based on payload size, improving the Peak Signal-to-Noise Ratio (PSNR).