基于云环境的网络监控视频解码的研究与应用

Research and application of video decoding based on Cloud environment

随着社交网站崛起、通信和多媒体技术的高速发展,视频、图像日益增长并己成为人们传递和获取信息的重要方式,目前H.264和JPEG2000己成为视频和静止图像领域应用较为广泛的压缩标准。如何高效挖掘海量视频的价值已经成为当前研究的热点问题,然而视频解码是发掘海量视频知识的前提。重点研究在分布式平台下对SDV格式的网络监控视频进行解码,利用Xuggler视觉库设计了能在云环境下Hadoop平台上使用的视频数据类型,解决了Hadoop平台上直接分割视频遇到的帧不完整、缺关键帧和少头数据信息的问题,并比较了传统单机解码与分布式解码的优缺点。

Along with the rise of social networking,the rapid development of communications and modern multimedia technology,video and images are growing and have become important ways,in which people obtain and transfer information. The H. 264 and JPEG2000 have become compression standards which has been widely applied in video and still images. How to use the value of large amounts of video effectively has become a hot problem in the current study. However,video decoding is discovering knowledge of massive video data. This paper mainly researches on the distributed platform for SDV format video decoding network monitoring,use Xuggler visual library to design a video data type which can be used on Hadoop platform in Cloud environment,and solves the problems that on Hadoop platforms directly segmenting video results in the frame incomplete and there＇s a lack of key frames and header data information. The advantages and disadvantages of the traditional stand- alone decoding and distributed decoding are compared.