Digital watermarking is a potential method for copyright protection and image authenticity verification. The choice of an appropriate watermarking technique is not an easy task specially because there are several different methods presented in scientific literature. The proposed method aims at exploring texture levels or the amount boundary pixels present in the regions where the information will be inserted. The selection of these regions aims at increasing the amount of inserted information making the method more robust without degradation of the visual quality of the watermarked image. The process consists of splitting the host image into blocks and sub-blocks according to the texture level or the amount of boundary pixels belonging to each block. In order to measure the performance of the proposed method several experiments were performed and a comparison to the Wu and Shih`s method is presented.
|