A hybrid chaos and fuzzy C-means clustering technique for watermarking authentication

Abstract

In the last two decades, different watermarking schemes are proposed for image authentication and copyright protection. The accuracy of both tamper detection and localization are considered as two important aspects of authentication of watermarking schemes. In this paper, a novel hybrid Chaos and Fuzzy C-Means (FCM) clustering technique is proposed for watermarking authentication. The proposed technique is used to improve the security of the watermarking system and it can be applied to any image with different sizes. The Chaotic maps are sensitive to initial values, so watermark can’t be established without the same initial values. Therefore, it gives high security to scheme. The Fuzzy C-Means clustering technique is used to make the watermark dependent on the plain image. The experimental results show that the proposed scheme achieves superior tamper detection and localization accuracies under different types of image attacks.