Abstract
This paper presents two novel digital watermarking techniques to protect image copyrights. The first technique, called wavelet based image watermarking using rank order and classification (WIWROC), explores the features and distances among rank orders of the wavelet coefficients in the subbands of a wavelet transformation (WT) image. These features are then utilised to develop a classification algorithm which can divide wavelet blocks of the image into three categories. A set of thresholds is exploited to devise three distinct embedding algorithms and their corresponding extraction algorithms for three categories. Each target block is hidden using its corresponding embedding algorithm. The WIWROC technique also uses the set of thresholds to retrieve watermarks without original images. Subsequently, a genetic algorithm (GA) is applied to optimise the WIWROC technique so as to further enhance its performance. The latter is referred to as the GWIWROC technique. Finally, experimental results reveal that the performance of these two techniques is better than that of existing methods.