A reversible steganography using prediction value coding and histogram shifting

Abstract

Steganography aims to embed the secret information in digital media for the purpose of secret communication, so that embedded data is not visible. The reversible steganography ensures that image is completely recovered to its original form after the secret data is extracted out. We are implementing reversible steganography that combines linear prediction error value coding and histogram shifting.The cover image is divided into multiple blocks. The above mentioned methods are applied for each block of image. Linear prediction error coding method calculates the prediction error values of cover image using basic pixel by scanning image blocks in inverse S-order. Histogram is generated for prediction error values to find the two peak points and two zero points. Histogram shifting method embeds the secret data in peak points of histogram of image. The reverse linear prediction error calculated to obtain stego image. The extraction and recovery procedure are done to extract secret message and recover the cover image. In the experimental results the better hiding capacity is obtained for smaller block size with fractional improvement to image quality. Keywords: Steganography, Reversible data hiding, Inverse S-order, Histogram shifting.