Abstract
Recently, several digital watermarking techniques have been proposed for hiding data in the frequency domain of image signals. However, little attention has been given to the optimal position in the frequency domain for embedding watermarks. We have attempted to improve the performance of both the extraction of watermarks and the quality on an image after compressed by JPEG technique by creating a multicriteria optimization problem for deciding the positions of watermarks in the frequency domain and obtaining an approximately optimum solution to this problem. The approximately optimum solution is obtained using the genetic algorithm. The experimental results show that the proposed method generates watermarked images of good quality and high tolerance to JPEG compression. In addition, the improvement of security has been achieved using the characteristic secret key to embed and extract watermark information. The reported method with the wavelet transforms is used to embed and extract watermark information.
