一种视频对象分割算法和实现  

A Video Segmentation Algorithm And Its Implementation

作　　者：郑浩[1] 沈廷根[1] 宋雪桦[1] 何鸿[1] 方明阳[1] 

机构地区：[1]镇江江苏大学通信工程系

出　　处：《微计算机信息》2006年第10X期278-280,共3页Control & Automation

基　　金：江苏省自然科学基金(BK2004059)资助课题;江苏省镇江市产学研(200473)资助课题

摘　　要：多媒体通信发展在出现了MPEG-1和MPEG-2后,MPEG委员会又提出了MPEG-4的标准。这种标准是基于对象的编解码技术。其关键是视频对象(VO)的分割和提取。因此,视频对象的提取在MPEG-4中起着重要的作用。本文在一种基于对称差分的运算检测方法基础上,提出改进方法,对每个运动对象加矩形框。实验表明该算法可以快速清晰的分割出多个运动视频对象。The technique of multimedia developed quickly, after the standard MPEG-1, MPEG-2, a new standard named MPEG-4 was put forward by the MPEG committee. The key technical of the standard is the Video Object Segmented from the video frame. So, video segmentation plays an important role in the MPEG-4. In this paper, we make some progress based on the symmetrical DFD algorithm. Every moved video object was segmented by the form of the rectangle. The test shows that we can acquire more than one video obiect fast and clearly.

关 键 词：MPEG-4 视频目标 对象分割 对称差分法 宏块 

分 类 号：TP37[自动化与计算机技术—计算机系统结构]