Multiple forgery detection in digital video based on inconsistency in video quality assessment attributes

With the enormous development of video capture, sharing, and editing tools, the authenticity and correctness of videos are under threat. Videos captured by a CCTV or any surveillance system are modified or altered for malicious purposes. It is essential to verify the integrity of any digital video prior to representing it in court. Inter-frame and intra-frame forgeries are performed on digital videos for the alternation or removal of important content from the video. Frame deletion is the most typical type of inter-frame forgery which can be performed single or multiple times in a video. Many significant frame deletion detection algorithms have been developed for single frame deletion detection. Methods for multiple forgery detection in a video are not well explored so far. In this paper, we propose a three-step frame deletion detection method to detect single and multiple forgeries in a video. In the first step, input videos are separated into two categories: static and dynamic, using a key frame extraction algorithm. In the second step of feature selection, to achieve higher accuracy and precision, two different sets of video quality assessment attributes are selected for static and dynamic videos using the forward selection method. In the third step, multiple linear regression is applied to a set of attributes to detect outliers using inconsistencies in video quality assessment attributes. The number of outliers determines whether the video is an original, a single tampered, or a multiple tampered frame deletion video. The proposed method is applied to a large video dataset that comprises static and dynamic videos with various activities, like traffic, sports, news, a ball rolling, airports, gardens, highways, zoom in and out, etc. The performance of the proposed method is assessed with the help of performance parameters like accuracy, recall, F1-score, and precision. The experimental results show that the proposed algorithm performs better in detecting single and multiple frame deletion forgery. The proposed method is also compared with the state of the art to validate its effectiveness.