资源三号影像中城市高大地物阴影检测方法  

Shadow detection for urban tall standing objects based on ZY-3 satellite imagery

作　　者：杨兴旺[1] 杨树文[1] 刘正军[2] 姚花琴 李轶鲲[1] 

机构地区：[1]兰州交通大学测绘与地理信息学院,兰州730070 [2]中国测绘科学研究院,北京100830

出　　处：《测绘科学》2015年第9期98-101,共4页Science of Surveying and Mapping

基　　金：中国测绘科学研究院的科研专项(513157);甘肃省高等学校基本科研业务费项目(212091);中国博士后科学基金资助项目(2014M552558XB)

摘　　要：针对遥感影像处理中阴影检测和信息补偿不准确的问题,该文在研究已有阴影检测算法的基础上,结合资源三号(ZY-3)影像数据的特性,构建了阴影检测方法:首先对原始图像分别做差值运算和主成分变换,并利用多峰阈值自动提取算法检测出阴影区域;其次将差值运算提取的粗阴影区域与主成分变换提取的阴影区域做并运算生成一个新的阴影区域;然后判断影像中是否含有水体,如果含有水体则利用多峰阈值自动提取算法检测出水体并与新合并的阴影区域影像做布尔运算得到完整的阴影区域,反之则新合并的区域即为完整的阴影区域。实验结果表明该方法针对ZY-3具有较好的普适性、较高的提取精度和提取效率。Aiming at the problem of inaccuracy of shadow detection and information compensation in the processing of remote sensing images, this paper built a shadow detection method combining the characteristics of ZY-3 data based on the existing shadow detection algorithms： firstly, Differential Operation （DO） and Principal Component Transform （PCT） were applied to the raw data, and shadow areas were detected by Multi-Peak Threshold （MPT） extraction algorithm; secondly, a new shadow area was generated through a union operation between the shadow areas extracted from IO and PCT methods; and then if the image was judged containing water, the MPT method was used to detect the water bodies, and the complete shadow region was generated by experiencing Boolean operation with the newly merged shadow area; whereas the newly merged area should be the complete shadow region. Finally, experimental result showed that the method would have good universality, high extraction accuracy and efficiency for ZY-3 satellite imagery.

关 键 词：资源三号 阴影检测 差值运算 主成分变换 

分 类 号：P237.9[天文地球—摄影测量与遥感]