Introduction

One of the greatest advantages of the ubiquitous Internet, the free and easy transmission of information, is also one of its greatest weaknesses, leading to the copying and outright theft of information. Of particular interest to the readers of this journal is the theft of images and videos. A content provider may want to publish information without having that information stolen and misrepresented as another’s. Legally, the content provider needs to be able to enforce a copyright and convince an impartial arbitrator (e.g., a judge) that the content is his and not the other person’s.

This special issue contains five papers on digital watermarking. A digital watermark is a signal embedded in digital data that can be used to establish ownership or to ensure the integrity of the data. The term digital watermark is in analogy to the watermarks that are used in the publishing industry. Most currencies, for example, have embedded watermarks that are difficult to reproduce but are visible in the correct viewing or lighting conditions.

There are two kinds of digital watermarks: visible and invisible. The translucent logos that television stations transmit (usually in a corner of the picture) are examples of visible watermarks. Invisible watermarks, in contrast, are designed to be as unobtrusive as possible and are used in many applications. Digital watermarks should be difficult to remove; resist a variety of signal processing attacks including compression, cropping, printing, transmission, rotation, scaling, etc.; and be difficult to forge.

The authors of these papers are experts in this new field and were chosen by the editors to illustrate some of the issues. The first paper, “Intellectual Property Protection Systems and Digital Watermarking,” by Lacy et al. presents a nice survey of the issues in watermarking and a new technique for embedding watermarks in a perceptually based compressor. The second paper, “Improved robust watermarking through attack characterization,” by Kundur and Hatzinakos looks at possible attacks and incorporates this knowledge in designing a two-part watermark. The “fragile” part locates changes in the data and these locations help the “robust” part reliably extract the identifying signal. The next two papers, “Wavelet-based digital image watermarking,” by Wang et al. and “Wavelet transform based watermark for digital images,” by Xia et al. present new watermarking schemes based on wavelet transforms. The last paper, “Image watermarking using bock site selection and DCT domain constraints,” by Bors and Pitas presents a watermark specifically designed to be resistant to JPEG compression.

Digital watermarking is an exciting new research area. New uses, new watermarks, and new attacks are being developed everyday. The editors and authors hope to convey to the readers some of the excitement we feel and welcome new contributors and new contributions to this growing field!