Passive detection of doctored JPEG image via BAG mismatching
基于BAG误匹配的JPEG窜改图像的被动检测算法

When JPEG image is tampered with, researchers could distinguish a detected image which may be changed or not according to the trace of BAG mismatching. But this algorithm produce high false alarm rate. This paper proposed a method which could improve original algorithm. Firstly, it detected an image based on the principle of BAG mismatching. Secondly, it could make use of the edge of adjacent BAG in vertical and horizontal direction to make a distinction between doctored region and accidental noisy point in doubtful region of the image. Finally, in order to reduce more noise, morphology processing was performed on the image. The experiment demonstrates that the algorithm can lower false alarm rate with effective detection and will be applied widely in the region of passive detection of digital image.