摘要： 针对人体运动目标的自动实时跟踪，设计一种新的基于P89v51内核的运动摄像头云台控制系统。对人体运动的图像检测与跟踪，提出一种快速的模板匹配方法。用改进Surendra算法自适应地获取背景图像以提取匹配模板，通过摄像头运动前后帧相同的运动区域来缩小匹配区域，进行快速目标匹配，达到实时性要求。该系统已应用在视频教学及会议中，实现自动人体运动目标跟踪与摄像。

关键词: 自动跟踪平台, 小区域目标匹配, 人体运动跟踪

Abstract: This paper designs a new system of motion video tracking motion human body, and the tracking platform which is based on P89v51 is simple and stable. A fast model matching way is presented on motion object detection and tracking. In the complex motion background, it adopts the adapting Surendra background updating algorithm for capturing the background image and subtracting the target template, then decreas the model matching area by the same area of before and after frame. It is realized in video automatic tracking of motion human body at real-time video meeting.

Key words: automatic tracking platform, small area object matching, people motion tracking

中图分类号: 

• TP391.41