Abstract
In the very near future, many products and transaction in our daily life will be done in the digital world. For example, tickets, crime images, and medical images are digital. In order to prevent the images from being tampered with, image authentication is extremely important. A semi-fragile watermarking algorithm based on the image content is proposed to inspect and verify the legitimacy of an image. Simulation result shows that choosing the Zernike moments as features in the low pass wavelet band increases the robustness of image. Experimental result shows that we can tell if an image is maliciously attacked or not and locate which block is tampered. Moreover, the image has high robustness against noise and JPEG compression with the quality factor Q 75. The technique can be used in an surveillance system for image authentication to prove the legality both the image and the surveillance system for digital forensic.