Analysis and evaluation of template based methods against geometric attacks: A survey

Abstract

Template based methods have been identified as one of the acknowledged methods in image watermarking consisting the influential techniques to prevent geometric attacks. These are added features in the watermarked image by the aim of synchronization recognition without including sensitive information. However, up to now, there is no activity to narrow down each of these methods. This paper, categorized the solutions for prevention and detection of geometric attacks based on predefined templates considered in digital images. Due to this classification in the proposed solutions, a state of the art investigation is performed to evaluate and compare the improvements in this area during 2003 to 2020. The goal is to find the critical factors to make an optimum balance between robustness, imperceptibility, capacity, and to increase security of the watermarking techniques planned in template based methods in order to assist the future researchers to achieve an optimum watermarking algorithm for protection of images against Geometric attacks. The experimental results prove the efficacy of utilizing Pyramidal Just Noticeable Difference (PJND) with high accuracy to increase robustness against more attacks in comparison to the other methods while employing curvelet to hide templates is appropriate to make the capacity of method independent from robustness and imperceptibility.