Video frame copy-move forgery detection based on Cellular Automata and Local Binary Patterns

Abstract

Copy-move forgery (CMF) is a common image forgery method that implies copying and moving a part of image to a new location in the same image. In video sequences, CMF can be accomplished by copying a set of frames and pasting them to a new location in the same sequence. The result of this process is usually changing of video content. To identify video CMF, it is necessary to develop a robust descriptor for identification of duplicated video frames. This paper presents a novel method where Cellular Automata (CA) and Local Binary Patterns (LBPs) are used as texture descriptors. The main idea is to divide every frame into overlapping blocks and use CA to learn a set of rules for every block in a frame. Those rules appropriately describe the intensity changes in every block so their histogram can be used as a feature for detection of duplicated frames. Experimental testing showed a good performance of a proposed method for detection of video CMF in all tested cases.