Moment-based local image watermarking via genetic optimization

A totally optimized image watermarking methodology that manages to enhance its local behaviour by applying the Krawtchouk moments is presented through this paper. The introduced technique is making use of a simple genetic algorithm in order to optimize the set of parameters that significantly influences the locality properties alongside with the overall performance of the watermarking procedure. Herein, the watermarking process is tackled as an optimization procedure. In this context, the appropriate set of configuration parameters is being searched for ensuring high quality watermarked images and low bit error rates at the extraction stage. For this purpose, several traditional and newly introduced image quality measures are used in order to quantify the influence of the examined set of parameters, as far as the quality of the watermarked image and the accuracy of the extracted watermark are concerned. The proposed method produces watermarked images of high quality and ensures high detection rates under several non-geometric attacks, by using less prior-knowledge at the detector's side. Extensive experiments have shown that by handling the watermarking process as an optimization problem, more robust and accurate watermarking schemes can be derived.