复杂环境下变电站运动目标跟踪技术研究  

The Moving Target Tracking Technology Research about Substation in Intricacy Environment

作　　者：顾波[1] 刘新宇[1] 张红涛[1] 

机构地区：[1]华北水利水电学院,郑州450011

出　　处：《电测与仪表》2012年第1期63-66,共4页Electrical Measurement & Instrumentation

基　　金：国家自然科学基金资助项目(30871449);河南省教育厅资助项目(2007470004);华北水利水电学院高层次人才科研启动项目(201118)

摘　　要：背景光线变化、树木摇动、下雨下雪等自然环境变化会给运动运动目标检测带来的很大影响。本文介绍一种背景动态生成算法,利用此算法实时动态地生成背景图像,在一定程度上解决了自然环境变化给运动目标检测带来的影响。然后使用背景差分法来检测运动目标,当存在运动目标时,根据连续2帧图像与背景图像差分的结果来确定运动目标的运动方向、运动目标的位置坐标。最后给出了一种云台转动角度的计算公式,根据此计算公式计算出云台的水平旋转角度和垂直旋转角度,实现运动目标检测和自动跟踪功能。The influence of environment changing, e.g., background beam change, trees waggle, and rain or snow day for moving target detecting is very great. This paper introduced an algorithm for dynamic building background image, and to use the algorithm to dynamically construct background image, it subdued the influence of environment changing a certain extent for moving target detecting. It used background subtraction technique to detect moving target, when the moving target exists, to make certain the direction and coordinate of moving target according to the result of concatenation 2 frames subtract background. At last, the paper gave an algorithm for calculating the turn angle of pan/tilt/zoom, according the pan/tilt/zoom angle algorithm to calculate the level turn angle and plumb turn angle of pan/tilt/zoom. These algorithms realized the moving target detecting and tracking.

关 键 词：运动目标 背景差分法 帧间差分法 云台 

分 类 号：TM63[电气工程—电力系统及自动化] TP391[自动化与计算机技术—计算机应用技术]