The Construction Of Video Forgery Detection Database And Its Performance Evaluation Using Seven Forgery Detection Algorithms

Multimedia information such as images,audios,videos etc.play a significant role in different aspects of daily people life and sharing this information become easy task due to rapid development of network technology.Multimedia files is being used in news reports,social media,business,and court evidence in digital investigation.Moreover,the multimedia editing tools are becoming more and more powerful which let ordinary users to manipulate and modify media contents.In this context multimedia forensics orientation has become a hot topic specifically in Video forensics.Detecting the authenticity of video contents is not easy task because users can forge them without leaving any sensible traces.The research on video forgery detection,which is used to determine the originality and authenticity of videos,has obtained a lot of attention.A variety of video forgery detection algorithms have been proposed,either inter-frame or intra-frame video forgery detection algorithms.However,there is a lack of standard video database in the field of video forgery detection.The main contributions of this thesis are as follows:Video forgery detection database(VFDD1.0)has effectively solved the issue of the lack of standard experimental video database,but the storage capacity of VFDD1.0 is not large enough.In order to solve this problem,VFDD1.0 is extended by VFDD2.0.The new content includes 1550 original and tampered videos with different cameras.This research introduces the new original videos along with tampered videos.Seven video forgery detection algorithms are used to test the validity of the expanded database.The experimental results show that the VFDD2.0 database has better reflection of the performance of different forgery detection algorithms and is a high quality video forgery detection database.