International Journal of Information Technology Vol.12 No.7 2006 Prashan Premaratne 1 , Liyanage DeSilva 2 , and Ian Burnett 1 Low Frequency Component-Based Watermarking Scheme using 2D Datamatrix prashan@uow.edu.au 1 School of Electrical, Computer & Telecommunications Engineering University of Wollongong, North Wollongong, NSW Australia. 2 Institute of Information Sciences & Technology Massey University, Palmerston, North, New Zealand. Abstract A robust watermark should be able to withstand most image manipulations such as cropping, additive noise and JPEG type compressions. It should also be able to hold telltale signs of attacks that are unknown to the user. Both of these aims can be fulfilled by embedding two watermarks, one fragile and one robust in different domains where watermarks are not vulnerable to the aforementioned manipulations. It is a well known fact that watermarking of DC components of discrete cosine transform (DCT) stand against JPEG compression. We build on this concept by inserting a watermark in the transform domain to further enhance the robustness. We use a 2D Datamatrix as the watermark as this has error correction capabilities and we show that this can be used to insert data imperceptibly into the host image. A second Datamatrix watermark is embedded in the low pass component of any wavelet decomposition at a specific level only known to the author. This improves the resistance of the watermarking scheme to attack. Our experimental results indicate that these invisible watermarks can carry significant information and are robust to many image manipulations. Keyword: Datamatrix, Low frequency components, Transform domain, Error correction, Wavelet decomposition, Watermark attacks, DC components, Robustness, Image watermarking. I. Introduction The burst of growth in the Internet and the overall development of digital technologies in the past decade have significantly increased the availability of digital content. One of the great advantages of digital data is that it can be reproduced without loss of quality. However, this has prompted and increase in illicit activities such as unauthorized duplication, distribution and piracy of copyrighted digital material and products. Digital watermarking is intended to encode secret or copyright information into host digital data to demonstrate and protect the ownership of these products. The three main digital data types in wide usage are audio/speech, images and video. The research reported here is specifically aimed at images and video copyright protection related to the Human Visual System (HVS). Copyright or the ownership of any digital image is aimed at individuals or 1