Using genetic programming on digital watermarking

A watermark has to be recognized and cannot cause serious impact on the unwatermarked data. For example, hold an American twenty-dollar bill against the light, a watermark can be found and indicates that U.S. government made them. The concept of a watermark is simple: if there is dirt on the paper, we can still see the content of the paper through the dirt. If we consider the watermark as the dirt and the un-watermarked data as the content of the paper, then we should be able to see the un-watermarked data with the watermark. Calculating the watermarked image individually is one way to have the quality of the watermarked image. The other way is having the performance of the quality of the watermarked image with corresponding parameters. Once the performance is found, the quality of every watermarked image becomes predictive. In this research, we use Structured Costa Scheme to generate watermarks and use Genetic Programming to find the performance. Keyword: Genetic Algorithm, Genetic Programming, Costa Scheme, Structured Costa Scheme, Digital Watermarkin