Australian Digital Forensics Conference

Document Type

Conference Proceeding

Publisher

SRI Security Research Institute, Edith Cowan University, Perth, Western Australia

Abstract

Digital watermarks have long been considered as a security feature. A watermarking method that involves the diffusion of limited watermark information into a large part of an image’s data has high robustness. The diffused information is summed up to a single component before detecting the watermark. The summing up process eliminates small noises by an averaging effect, which improves the robustness of the embedded watermark against attack. In this field, thus far, only an asymmetrical Chirp transformation with a small block size has been attempted. In this study, a new verification experiment for a large block size of 256 × 256 pixels is conducted. High robustness of the proposed method is revealed. This includes the finding that, in the case of a JPEG compression attack, the proposed system is robust even at strong compression of 1/70. As for a clipping attack, embedded watermarks can be detected with up to seven pixel clipping of an embedded image.

Comments

Originally published in the Proceedings of the 11th Australian Digital Forensics Conference. Held on the 2nd-4th December, 2013 at Edith Cowan University, Perth, Western Australia

Share

 
COinS