Suitable Methods for Watermarking

Patchwork

The patchwork algorithm allows for the detection of a single, specific bit in an image. Patchwork will embed in a host image a specific statistic, a small watermark that tells whether a larger watermark is embedded within an image. In short, patchwork is an indicator that tells a program that the rest of the watermark is present. While this method by itself works quite well, there have been a number of performance improvements made to the patchwork process, including treating patches at several points rather than just one, and using visibility masks to avoid putting patches where they would be easily noticed.

Patchwork is robust in its resistance to cropping and tone scale corrections. Patchwork is not robust against translation, rotation, or scaling. Because patchwork is nearly invisible, its robustness and low data rate is very suitable for digital watermarking.

Spread Spectrum

The spread-spectrum technique for encoding information can work just as well in watermarking as it does in steganography. Spread spectrum takes the information to be hidden; embeds it within the noise of the image, video, or sound file; and then inserts the modified noise back into the cover. This technique is very robust, difficult to detect, and has a good payload capability.

Orthogonal Projection

Orthogonal projection is particularly suited for digital watermarking. Its method for hiding is rather ingenious. Essentially, the image is manipulated in such a way that the hidden information is stored in the most important parts of the data and is less likely to be damaged or destroyed by compression or normal manipulation.

Because this method of encoding uses the meat of the image to its benefit, orthogonal projection is resistant to attacks such as lossy compression, photo manipulation programs, and programs such as StirMark and Unzign.