Color image authentication for detecting intentional attacks
Hazem Munawer Al-Otum · 2016
Image authentication can be used to detect intentional attacks in grayscale as well as color images. In this work, a novel semi-fragile watermarking technique is proposed to be implemented for detecting and localizing intentional attacks that may be produced on purpose by cyber-users. The proposed technique revealed high robustness against common image processing attacks while it was able to detect minor intentional attacks using two deployed measures to discern between various types of attacks in a modified multi-pyramidal discrete wavelet transform domain. Simulations have shown high performance in locating hacked regions while demonstrating high robustness under moderate to severe common image processing manipulations.