Secure fragile watermarking method for image authentication with improved tampering localisation and self-recovery capabilities

In this paper, we present a new fragile watermarking method aimed at providing improved tampering localisation and self-recovery capabilities. First, a secure block-wise mechanism, resilient to cropping, is used to localise distorted blocks of pixels. Subsequently, the tampering localisation accuracy is refined by means of an iterative pixel-wise mechanism, which estimates the original watermarked pixels in altered regions. Extensive tests, using a large data-set, are presented to show to what extent the size of the tampered area influences the performance of our method. Moreover, results show that the proposed scheme compares favourably with a method presented in recent literature, in terms of tampering localisation and self-recovery performance. In fact, we show that our method is capable of partially reconstructing a cover image when 5% of its total number of pixels has been cropped.