Research And Design Of Video Cloud Transcoding System Based On Hadoop

The current market due to the diversification of terminal equipment types,a variety of video media formats and network environment differences,which requires a solution to mass video transcoding,real-time fast processing of video technology.And with the rapid increase in the amount of data,the traditional stand-alone video transcoding system is not enough to achieve the scalability of processing capacity with the linear increase in data requirements.In order to solve this problem,this thesis uses the cloud computing cloud technology Hadoop cloud computing platform,it is an open source framework,while supporting distributed computing,mass data processing and storage.The emergence of the Hadoop open source framework makes it possible to solve real-time video transcoding issues for a variety of end devices,multiple video formats,and multiple stream requirements.The main work of this thesis is to study and design Hadoop-based video cloud transcoding system,the cloud computing technology and video transcoding technology works with each other.Using Hadoop HDFS distributed file system for video file storage,Hadoop Map Reduce parallel computing framework and Avidemux ope n source video slicing tools,FFMPEG video transcoding tools,video merge tool Mencoder to complete the video transcoding.In order to improve the transcoding performance of the system,this thesis proposes an improved method for serializing the video data video Seralizable()and Hadoop cloud platform fault tolerance mechanism.The aim is to improve the transcoding performance of the system and enhance the user experience.Through the test,it is verified that the system can make full use of the resources of the existing computer and realize the distributed transcoding of the large-scale video data on the Hadoop cloud computing platform.Compared with the stand-alone transcoding system,the purpose of improving the transcoding efficiency and reducing the transcoding time is achieved.In this thesis,the optimization of the current situation proposed in the future work will be implemented.