The Algorithm Research Of Spatial Resolution Transcoding Based On The H.264/AVC

With the abroad application of multimedia and the rapid development of the internet, many video coding standards have been established for various video applications, such as MPEG-1 for video storage and CD player, MPEG-2 for digital TV broadcast, H.26X for lower bit-rate video communication, and more popular standard MPEG-4 and H.264. These standards solve the limitations of video storage and net bandwidth. However, kinds of video coding standards are different from each other in syntax as well as in compressing-rate, bit-rate and picture format. They are designed for diverse networks, decoders and monitors. When the incoming coded video stream with a specific format is not fit for the current network or guest requirement, we need to transfer the coded stream from one format to the required format seamlessly, which is known as the transcoding.In this article the spatial domain transcoding based on the H.264/AVC is researched. Besides the key technologies for the spatial domain transcoding, such as video transcoding frame, decreasing the resolution, motion re-estimation, macroblock mode decision, interpolation algorithm and etc are also researched. To summarize, the researches include the following aspects:Firstly, a brief review of the scalable coding for decreasing bandwidth is given for the theory foundation. By analyzing the demerit of the scalable coding, the necessity of the video transcoding is put forward in this article. A new video transcoding frame based on the research of traditional scalable coding is proposed. This frame decreases the complexity of transcoding.One key technology of transcoding is decreasing the resolution. Its validity will influence the excursion error after transcoding. In order to obtain the best optical effect and satisfy our research environment we put forward the minimal sum algorithm of absolute difference in four pixels. The result proves this algorithm is effective.In the process of decreasing the resolution the re-estimation of the motion vector should also be done. It will influence the excursion error. The best method is chosen from previous methods as our method. This method not only obtains the better result but also has lower complexity.For the recovery of video after decreasing the resolution, we put forward the direct pixels and median filtering algorithm based on the research of traditional interpolation method. This algorithm obtains the higher signal-noise ratio and the better optical effect.Finally, we make a conclusion and propose the future research directions in this field.