A novel video inter-frame forgery detection method based on histogram intersection

Nowadays videos are widely used in every aspect of society such as transport, security, justice identification and so on. Thus, the authenticity and integrity of video are very important. This paper proposes a new method to detect forgeries of video with statics background. In general, adjacent frames in a video with the same background have strong correlation. If the video being tampered, the continuity of the frames correlation will be disturbed. In this method, pixel lines are obtained by intercepting the sequence of video frames in the horizontal or vertical direction. Every four continuous pixel lines make up a pixel belt. Then, by using the histogram intersection method, the correlation between pixel belts will be calculated. The simulations show that if the video tampered, there will be outliers exist in the correlation coefficients. Simulation results demonstrate that the method of this paper can detect the forgery and locate its position.