The Research On Data Recovery Technology For Live FLV Streaming

To date,the concentration of video forensics is few in online video.So,in this paper,we aim at restoring online live video files efficiently and consider FLV as our research objective.The main work of this paper is as following:Firstly,we posit the situation of no packet loss during network data capturing process and use the features of application layer protocol proposed.This method first locates all the relevant data packets according to the characteristics of the transmission protocol and then extracts the relevant data of the video in light of the characteristics of the application layer protocol.After that,the extracted video data are merged to generate a new video file.Then,we carried out an experiment showing that the proposed recovery method can recover the live video in network data without any loss of network data,and the recall rate and F-value can reach more than 93%.Secondly,we think about the situation of packet loss or corruption during the network data capture process,therefore we propose an method based on FLV video tag structure.This method extracts the FLV video data by using tag structural features,then combining the extracted video data and generating a new video file.Experimental results show that this method achieves a high recall rate and F-value at more than 95% and can successfully carve out video files?Finally,we advance a FLV video stream adaptive recovery method regarding the different integrity of network data.The cost function is used to recover the FLV video data adaptively by choosing the appropriate recovery method.Experimental results show that the adaptive FLV video carving method can achieve a good balance between the average recall and recovery time.