Effective Tamper Detection and Recovery of Images after Serious Attacks

Document Type : Research Paper

Authors

1 School of Computing and Mathematics, Charles Sturt University, Bathurst, NSW-2795, Australia

2 Department of Computer Science, University of Mohaghegh Ardabili

Abstract

Confirming the integrity of transmitted sensitive digital content is a significant issue due to the evolution in communication technologies and the accessibility of image processing tools. Watermarking has been a successful method of authentication and integrity verification recently. However, several significant problems remain such as confronting some serious attacks and recovery after higher tampering rates. We propose a hybrid method to enable an image to be recovered successfully after a range of attacks. A blind watermarking approach is adopted which includes fragile authentication but robust recovery references. This is performed by embedding verification code as part of the watermarked data along with key features of the original image into a location that is resistant to the attack. To combat different kinds of attacks, the areas of the image have been investigated to find which area is more likely to be affected in each type of specific attack.