Abstract—Prediction Error Expansion is one of Reversible Watermarking (RW) techniques that gives effective results and has been continuously developed over the years. Among those techniques, Sorting Absolute Prediction Error (APE) is an important step used to reduce image distortion; so in this research we focus on developing the technique of sorting APE in ascending order more accurately. We can use local variance values in predicting APE and these values are calculated from means so they are suitable for predicting which areas contains high or low APE values. Consequently, we have improved this parameter by using Expanded Variance Means which can increase efficiency of sorting process and raise PSNR values.

Index Terms—Reversible watermarking, prediction error expansion, variance sorting, lossless data hiding

Aekawit Kotvicha and M. L. Kulthon Kasemsa are with Faculty of Information Technology, Rangsit University, Thailand (e-mail:aek_kot@hotmail.com)
Parinya Sanguansat is with Faculty of Engineering and Technology,Panyapiwat Institute of Management, Thailand

Cite: Aekawit Kotvicha, Parinya Sanguansat, and M. L. Kulthon Kasemsa, "Expand Variance Mean Sorting for Reversible Watermarking," International Journal of Computer and Communication Engineering vol. 1, no. 3, pp. 196-199 , 2012.