论文部分内容阅读
数据压缩技术主要用于减小视频数据的大小 ,以满足存储和传输的需要 ,但现有的图象处理技术必须先将视频压缩流完全解压后才能进行处理 .本文讨论了不需要完全解压缩或仅需很小的解压就能直接对视频压缩流进行处理的方法 ,例如 MPEG压缩视频流中 P帧和 B帧以及常用的压缩域处理方法以及存在的问题 ,最后证明了在半象素运动矢量情况下 MPEG简化帧抽取的准确性问题 .视频压缩流在压缩域处理的特点是不仅减少了解压缩所需要的计算时间 ,而且还节省了相应的存储空间
Data compression technology is mainly used to reduce the size of video data to meet the storage and transmission needs, but the existing image processing technology must be completely decompressed video compression stream before it can be processed.This paper discusses the need to fully decompress Or directly decompressing the video stream, such as P-frames and B-frames in the MPEG-compressed video stream, as well as the commonly used compression domain processing methods and the existing problems. Finally, it is proved that the half-pixel motion MPEG simplifies the accuracy of the vector frame extraction problem.Video compression stream processing in the compressed domain is characterized not only reduce the computational time needed to decompress, but also saves the corresponding storage space