星空运动背景中目标识别算法研究  

The algorithmic research for target detection of moving background by microspacecraft target detector

作　　者：刘辉[1] 周锦松[1] 胡柄樑[1] 张文喜[1] 

机构地区：[1]中国科学院西安光学精密机械研究所

出　　处：《微计算机信息》2007年第03S期285-287,共3页Control & Automation

基　　金：国家自然科学基金资助项目(60532070)

摘　　要：针对微型航天探测器在星空运动背景下对目标识别的要求,提出一种基于图像配准与边缘提取的差分算法。该算法将采集的连续两帧图像配准后差分,将差分图像二值化后与后一帧边缘提取后的图像进行逻辑乘,提取出目标物。计算机仿真结果表明该算法计算量小,识别效果好。In order to meet the needs of detecting the target under the moving space background by microspacecraft target detector . This paper proposed an algorithm-the image edge extract and deference base on image registration. The algorithm use image difference to the two collected and seriate image after image registration, then multiply the difference image after binarizing and the second image after edge extract by pixel, then detect the target. The result of simulation indicate that computational effort is low and detected result is good.

关 键 词：微型航天探测器 差分 图像配准 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]