Research On Seamless Splicing Technology Of Transport Streams

Along with the development of Digital Video Technology and the publishment of several correlative standards, Digital Television has been in a practical period and it is inevitable that Analog TV is replaced by Digital TV. In the conversion of them, the TV Stations will be face with several problems and challenge. Programming or Programs Edition, Advertisements Insertion etc, are indispensable parts of transmission digital programs. The Seamless Splicing of the Transport Streams, which are the carriers of the TV programs, is the key technology of them.This paper is based on MPEG-2 standard. As lucubrating the systemic level and the video elementary level, this paper points out the technical difficulties which it will encounter to splice two MPEG-2 Transport Streams seamlessly, discusses the principles and the conditions of seamless splicing, and educes two seamless splicing methods respectively based on systemic level and elementary level. They are validated by relative algorithmic experiments.The Seamless Splicing based on systemic level deals with the Transport Streams just according to MPEG-2 systemic level syntax. It is not necessary to demultiplex, decode, encode and re-multiplex the Input Streams and the Output Streams when doing with the Transport Streams. The Spliced Stream has most of the old streams'data. The key technology is searching for the splicing points, adjusting the relative information of the Transport Streams (include information about PSI and time), and the synchronism of the video and audio. The advantages of this method are rapidness, convenience, and keeping almost all the old streams'data. But as the restriction from the image types and order of frames in the stream, splicing cannot be to the accuracy of frame, but to the accuracy of some special positions that are according with the seamless splicing.In order to achieve the precision in frame in the splicing, this paper brings forward an another method which is based on elementary level. The core of this method is transcoding, whose function is to convert the frames whose image type or order is not according with the splicing to the legal splicing points. In this method, both the Input Stream and the Output Stream must be demultiplexed and re-multiplexed. Near the splicing points, they will be decoded and re-encoded. The main process is completed in Transcoding Station.