视频图像理解的一般性框架研究

Research on common framework of video image understanding

视频图像理解侧重于对视频序列进行解释,既涉及到图像的空间特性,也涉及到视频序列的时间特性,是目前计算机视觉领域的一个研究热点。回顾了视频图像理解方法的研究现状,提出视频图像理解的一般性框架,包括层次结构、涉及的技术领域和应用的系统结构,并以一个实际应用作为示例解释该框架的层次结构。

Video image understanding technology places emphasis on interpreting video sequences, including spatial and temporal features of video sequences. It is a hot spot in the field of computer vision. This paper first reviewed the current status of research on video image understanding methods, and then proposed a common framework for understanding video sequences, including the layered structure, the correlative fields and the common system structure. Finally, discussed a real application for interpreting the layered structure of this framework.