This paper proposes a digital video tampering method which uses textural features of video frames. For a video with still background or slow and continuous movement, the correlation of interframe textural features is large and it would become smaller to some extent if tampered. This paper finds out the outliers according to the correlation of the interframe GLCM. This method is unlimited for video format, applied to any kind of video format. Experimental results show that this method can effectively detect video frame tampering including frame insertion and frame substitution of diverse-sourced in the similar background.