I. Introduction
Recent technology advancement has undoubtedly accelerated the speed and volume of digital content acquisition [1]. An obvious example is the ubiquitous cameras in the variety of circumstances, such traffic monitoring [2], assembly line inspection [3], environmental hazard detection [4]. These cameras capture images in massive volumes. However, the reduction cost of the digital content acquisition in no way compromises the importance of content protection. For instance, in the pursuit of traffic violations, the police should present scene images that are intact or authentic to support these cases [5]. Therefore, the research on relevant methods is of persistent interest.