Research On Copy-Move Forgery Of Digital Video Passive Forensics

With the rapid development of digital multimedia technology, digital video has been widely used in various fields, and plays an increasingly important role in our daily life. In the meantime, video-editing software with powerful-function making non-professional people can easily tamper a video. Once these carefully doctored video are malicious used in some important fields, the society order will be seriously disrupted. So it is of important practical significance to verify the video’s integrity and authenticity.At present, the digital video passive forensics research is still in its infancy, all the studies focus on specific forgery, then corresponding detection algorithm is put forward. Copy-move forgery is one of the most common types of video forgeries, it can be divided into:inter-frame tampering and intra-frame tampering. Through in-depth analysis and studies on copy-move forgery detection, three algorithms are proposed to detect the forged videos. The main work includes the following three aspects:(1) To detect video inter copy-move forgery, a detection algorithm based on Tamura textural features is proposed. The algorithm firstly extracted all texture features in video frames, sorting the corresponding eigenvector matrix constituted by texture feature, finally calculating the difference between the adjacent feature vectors, and then locating the forgery sequence through the distance threshold of frame number. Experimental results show that the algorithm not only has improved greatly in the detection accuracy, but also can pinpoint the location of tampering.(2) In order to further improve the time efficiency of the detection algorithm, a detection algorithm based on the geometric mean decomposition and structural similarity is put forward. Firstly, a block-based search strategy is proposed to locate the starting frame in the duplicated sequences. In addition, using the structural similarity to recheck the starting frame in duplicated sequences. Since the value of similarity between duplicated frames is higher than that between the normal inter-frames, a coarse-to-fine method based on structural similarity is put forward to locate the tail frame. Finally, recover the video. Experimental results show that the algorithm can quickly and accurately locate the tampered sequence. Besides, different encoding formats of video can be effectively detected.(3) To detect intra copy-move forgery, a blind detection approach based on abnormal hue rate and SIFT is proposed. Firstly, calculating the abnormal hue rate of each frame in video sequence. Then, determining the location of tampered frame by analyzing the abnormal hue rate. Finally, the tampered areas are confirmed by similarity matching. The proposed method has a great improvement in terms of precision and computation time.