Find in Library

Abstract A distributed lossless coding system based on cloud computing is proposed to increase the video transcoding capacity of the new media interactive broadcasting project in the research department. Distributed transcoding is carried out by using the programming ideas of two Hadoop cores: HDFS (Hadoop distributed file system) and map reduce. At the same time, the specific process of distributed transcoding is introduced and designed in detail. The experimental results show that the efficiency of the distributed transcoding scheme is greatly improved. The results show that this experiment selects the 32 m with the best segment size as the benchmark, and calculates that the distributed transcoding time is 60.3% higher than the traditional single machine transcoding. This improvement significantly improves the transcoding efficiency, improves the experience effect of video service, and adapts to the trend of the rapid development of new media in the current era. The design of the system architecture not only greatly improves the transcoding efficiency, but also saves the transcoding time.