ABSTRACT
Image mosaic image is currently a common method that is tampered with and is often used maliciously as well as modified the true content of the original photographs, mostly using a variety of image editing software from two or multiple dierent target areas in the original image into a mosaic amplitude new synthesized image. How to successfully achieve the splicing operation Blind digital image forensics has become an important topic in research; existing digital image mosaic detection technology, including a light source based on the estimated direction of inconsistency[1], based on the edge of bipolar signals modeling[2] and phase consistency[3], using the noise distribution in the camera CCD or CMOS sensors inherent[4], the consistency of the estimated camera response function at the edge of the area on both sides and other methods. Although these methods can achieve splice detection tasks in simple scenarios, but are needed during the execution of a large number of statistical learning algorithms and human intervention, they cannot be fully automated and rapid and eective. These methods are characterized by the presence of the robustness of the estimated dierence in accuracy rate, lower boundary location, and splicing defects such as lack of precision; in order to overcome the earlier shortcomings, we propose a color filter array image mosaic based on consistency analysis features of the rapid detection algorithm.
