Optimizations using the genetic algorithm for reversible watermarking

Important requirements for reversible data hiding techniques: the embedding capacity should be large and distortion should be low. This paper represents a high performance reversible watermarking technique which involves adaptable predictor and sorting parameter to suit each image and each payload in order get lowest image distortion. Our proposed method relies on a well-known prediction error (PE) expansion technique. Having small PE values and a harmonious PE sorting parameter will greatly decrease distortion. In order to get adaptable tools, Gaussian weight predictor and expanded variance mean were used as parameters in this work. A genetic algorithm has also been introduced to optimize all parameters and produce the best results possible. Implementation showed a significantly improved result compared to previous work.