Intelligent Threshold Selection for Reversible Watermarking of Medical Images

摘要

In this paper we present reversible data hiding of digital images using integer wavelet transform and companding technique. The prime requirement of reversible data hiding is that the cover media should be least distorted. The information embedded in the cover work should be fully extractable and the cover work be restored to its original state as well. In this paper we implement a watermarking technique which takes the advantage of the companding mechanism and threshold based strategy. Genetic Algorithm based intelligent coefficient selection scheme is applied to improve the imperceptibility with high effective payload. Improvement in imperceptibility at the same level of payload is observed by exploiting the learning capabilities of GA.