Video moving object tampering evidence obtaining method based on VGG-11 convolutional neural network

Abstract

The invention discloses a video moving object tampering evidence obtaining method based on a VGG-11 convolutional neural network, and the method comprises the steps: calculating a motion residual error between a forged frame and an unforged frame in a video through employing aggregation operation, and classifying the forged frame and the unforged frame; based on the motion residual error, extracting a motion residual error graph feature; constructing a convolutional neural network based on the VGG-11; training the convolutional neural network based on the VGG-11 by using the motion residual image features; and judging whether the video moving object is tampered or not by using the convolutional neural network based on VGG-11. Compared with the prior art, the method has the advantage that the forged frame in the tampered video can be better and automatically recognized.

Description

technical field [0001] The present invention relates to video tampering detection technology, in particular to a video moving object tampering evidence collection method based on VGG-11 convolutional neural network. Background technique [0002] In today's Internet era, with the continuous development of computer multimedia technology, more and more images, audio and video have become network resources shared by netizens. In particular, digital video has become the main information bearing form of the network because of its intuition, convenience, and rich information content, and has also become an important data source for many social networking software. If necessary, these video files will serve as evidence for many important matters in the fields of journalism, politics, insurance claims, defense and legal trials. However, the widespread use of powerful multimedia editing tools such as Adobe Photoshop, Adobe Premiere, Lightworks, Video EditMagic and Cinelerra makes it ...

AI Technical Summary

Problems solved by technology

[0005] However, the above-mentioned forensics algorithms based on target object video tampering are mostly based on traditional image processing and classifier methods, and do not involve deep learning methods. The reason is that there are many objects in the video frame, and the tampered objects are not suitable for exploiting The deep learning network directly performs feature learning, so there is no research on intra-frame video forensics combined with deep learning methods

Images