Robust reversible watermarking using integration of histogram shifting and patchwork algorithm to improve quality and capacity

Abstract

In this paper Histogram Shifting Patch Work method is proposed. It is a hybrid method generated from histogram shifting method and patchwork method. Additional security has been given by chaotic neural network. Quality of Histogram Shifting method is better while capacity is less. Embedding capacity of PW method is better at the cost of quality. CNN adds security in a reversible watermarking. Integration of HS and PW method, grabs advantages such as high quality and high embedding capacity respectively. The chaotic neural network is influenced by several factors, which affect the ability of solving problems. The chaotic random sequence is encrypted with LSBs of image pixels, to get additional security. Robustness of extracted watermark using histogram shifting patchwork method, for geometric attack is verified and PW method found inherently more robust to geometric attack than any other method. Robustness increases if it is passed through distortion correction method. HS method is not suitable for geometric attack. If it is combined with distortion correction then it improves robustness. Ultimately, proposed method becomes more robust to geometric attacks.