In this paper, benchmarking results of watermarking techniques are presented. The benchmark includes evaluation of the watermark robustness and the subjective visual image quality. Four different algorithms are compared, and exhaustively tested. One goal of these tests is to evaluate the feasibility of a Common Functional Model (CFM) developed in the European Project OCTALIS and determine parameters of this model, such as the length of one watermark. This model solves the problem of image trading over an insecure network, such as Internet, and employs hybrid watermarking. Another goal is to evaluate the resistance of the watermarking techniques when subjected to a set of attacks. Results show that the tested techniques do not have the same behavior and that no tested methods has optimal characteristics. A last conclusion is that, as for the evaluation of compression techniques, clear guidelines are necessary to evaluate and compare watermarking techniques.

Date Published: 9 April 1999
PDF: 11 pages
Proc. SPIE 3657, Security and Watermarking of Multimedia Contents, (9 April 1999); doi: 10.1117/12.344673

Published in SPIE Proceedings Vol. 3657:
Security and Watermarking of Multimedia Contents
Ping Wah Wong; Edward J. Delp III, Editor(s)