Based On The Digital Fingerprint Of Ngb Network Video Content Security Monitoring Technology Research

With the rapid development of next generation broadcasting network（NGB），toconstruct a set of reliable security control system is particularly important. Videoservice, as the main body of NGB network business, and its security will be not onlydirectly determine the quality of NGB, but also influence the vital interests of the people.At present, however, how to control network video effectively is becoming the biggestobstacle of video business development, because that it still based on artificial judgmentand still can not realize automation. Therefore, this paper focuses on NGB networkvideo security broadcast control system, which based on video fingerprinting (VF) andrelated technology.First of all, to analysis the security problem of NGB network video and potentialtechnical solutions, we get digital fingerprinting (DF) used in video monitoring field isfeasible and excellent.Then, to ensure the accuracy and to further strengthen the real-time of the videomonitoring, the paper analyzes and demonstrates the principle video broadcast controlsystem based on VF, moreover study the algorithm and related technologies, which isgood for improving the efficiency of VF extraction and matching.In order to meet the real-time of video monitoring system and to control networkvideo content efficiently. Firstly, the paper puts forward the key-frames selectionmechanism based on the minimum vertex cover principle. In the mechanism, we designa new measurement method of the distance between adjacent video frame, and a quickkey frames selection method, which plays an important role for reducing the extractiontime effectively of VF. Secondly, the paper also proposes a fast video fingerprintinggeneration algorithm, which effectively combines temporal characteristics and imagecharacteristics of the video frame, and finally being in binary flow form thatconvenience for computer processing. Not only enhance differentiating solvent andrepresentation, but also further narrowing the time complexity, because of omittingvideo decoding process.Finally, according to the actual demand of the NGB network environment, this paper constructs a kind of model about audio-video mix fingerprinting (AVMF) anddesigns the corresponding experimental platform. The platform is used to analysis thefeasibility about synchronous monitoring of audio and video, at the same time, to verifythe accuracy and robustness, especially real-time of the system. At last, we summarizethe main influence factors or parameters about the performance of the system anddiscuss the direction of further research work.