摘要: Two methods for attacking watermarks in images and videos have been proposed. The first attack is to remove robust by iterative block dropping concealment. Our study suggests that the size chosen a watermark embedding system should be large enough circumventing concealment based attacks. second method forge valid authentication systems on fragile-watermarking. Double watermarking then proposed as one possible countermeasure.