A Study on Reversible Data Hiding Technique Based on Three-Dimensional Prediction-Error Histogram Modification and a Multilayer Perceptron

In the past few years, with development of information technology and focus on security, many studies have gradually been aimed at data hiding technology. The embedding extraction algorithms are mainly used by to hide that requires secret transmission into a multimedia carrier so cannot be realized achieve secure communication. Among them, reversible (RDH) is for applications demand as well original recovery without distortion, such remote medical diagnosis or military transmission. this work, we hypothesize RDH performance can enhanced more accurate pixel value predictor. We propose new scheme prediction-error expansion (PEE) based multilayer perceptron, which an extensively artificial neural network in plenty applications. utilizes correlation between image values their adjacent pixels obtain well-trained perceptron capable achieving prediction results. Our mapping method three-dimensional histogram modification uses all eight octants space embedding. experimental results our show capacity greatly increases quality still maintained.