Lossless Data Hiding with Multi-Round Difference Prediction

Abstract

Lossless data hiding belongs to one major branch in information security researches. The major goal is to retain the lossless property of proposed algorithm, meaning that original image and embedded secret should be perfectly recovered at decoder. Based on the multi-round prediction of differences between original and marked images, enhanced amount of capacity can be obtained while retaining the acceptable quality of output image. Simulation results have demonstrated the better performances over relating methods in literature.