Research On Information Hiding And Tampering Detection Technology Based On Monitoring Video

In recent years,digital video has gradually become the necessary and guarantee of social security.The camera,which could be found in public,building or any other area,not only ensure the work of people go through well,but also maintain social stability.And because of that,people are becoming increasingly trustful to digital video.As the crime methods become novel and advanced,the destruction and manipulation of digital video by lawbreakers bring great threat to our society.The destruction and manipulation of digital video has become a hot item which is studying by many experts.Digital video watermarking has been regarded as the best anti-counterfeiting method because of its concealment of information,diversity of embedding methods,selectivity in robust and vulnerability of varied methods.This paper analyses the problems in the application of existing methods of information hiding and tampering detection in monitoring video,has been accomplished the following work:(1)The paper proposes a new method which digital video is encoded in H.264 format.What's more,it uses fragile digital watermark and suits for digital video based on the key frame brightness component.In order to balance the relationship among watermarking embedding capacity,subjective invisibility of human eyes and watermarking vulnerability,the algorithm considers the watermarking embedding capacity,the relative position of watermarking embedded video and the depth of embedding frequency domain to the greatest extent.(2)Converts the watermarking information of one-dimensional sequence into two-dimensional code image information that can be read intuitively,which has better watermarking redundancy.The location of the watermarking embedding point is selected as the part of the key frame which has a large range of brightness change and obvious difference between light and shade.The algorithm first defines a range of brightness eigenvalues,then calculates the relative eigenvalues of each DCT macroblock,and changes the relevant information in the brightness component according to the hidden information contained in the two-dimensional code watermarking,which serves as a basis for the information hiding process.Because the extraction of watermarking does not depend on the original video,the blind detection of video tampering can be realized.(3)The simulates common attacks by using MATLAB experimental platform,tries to use C++language to compile and develop automation system,simplifies the process of information hiding and tampering detection in video surveillance,and simulates the effect of embedding and extracting watermarks.The results show that the information hiding and tampering detection algorithm has certain fragility,low complexity and good real-time performance,and is suitable for monitoring video information hiding and tampering detection.