Simple duplicate frame detection of MJPEG codec for video forensic

One of the most common methods of video forgery is make selected frames duplication from a video to another location in the same video. There are many algorithms that have been developed by other researchers and can be used for video forgery detection. Most of those algorithms use image processing technique which is fairly complex. In this paper, the simpler method for a specific purpose will be proposed and discussed. This method can be used to detect the frames duplication in a video that using MPJPEG codec. One of the simplest ways to detect whether a file is same or not is by calculating its hash value. This simple technique can be used for developing the method to detect the duplicate frames in a video. By splitting the MJPEG video into JPEG image frames, we can calculate the hash value of each frame. After we got all the hash value of each frame, we can compare the hash value each other. Furthermore, we can identify the duplicate frames in that video. For the implementation, we will use Python scripting language with OpenCV library. The output of the implemented program is the frame index number which indicates the location of the duplicate frames from the video files. For testing purpose, several original videos will be intentionally edited using video editing software to duplicate some of its frames. The result of the implementation shows that this method is pretty good, fast, and accurate to detect the frames duplication in a video file with MJPEG codec.