基于微型无人机遥感数据的城市水环境信息提取初探

A Preliminary Study on Remote Sensing Information Extraction of Urban Water Environment based on Micro UAV Images

近年来,黑臭水体整治已经成为各级政府的重要工作内容。为更好地开展黑臭水体筛查和监管工作,提出基于多旋翼无人机搭载多光谱相机的遥感监测方案,尝试采用影像特征变换结合面向对象分类方法提取黑臭水体的遥感信息。结果表明:(1)无人机影像上地物具有更加丰富的色彩和纹理;(2)与主成分变换和纹理分析方法相比,光谱指数对典型地物具有较好的判别能力;(3)采用标准最邻近分类法提取城市水环境信息,能够获得较高分类精度。

In recent years,government agencies,from central to local ones,have put treatments on black and odorous（ BO）water body on the top list. To smooth the working process,this article brought in the UAV（ Unmanned Aerial Vehicle）technology in screening and supervising BO water body in cities. We proposed the remote sensing monitoring plan based on multi-rotor UAV with multispectral camera,and tried to use the method of feature transform and combined it with object-oriented classification in order to extract remote sensing information from BO water body. As the results showed：（1） The ground objects in UAV images had more abundant color and texture than that in high resolution satellite remote sensing data;（2） Compared with principal component transform and texture analysis,spectral index had good ability to classify typical objects of UAV image;（3）The method of standard nearest neighbor classification had high precision in extracting remote sensing information of BO water body.