微型无人机遥感影像的纠偏与定位  

Geometric Correction and Registration of Optical Remote Sensing Image from Miniature Unmanned Aerial Vehicle

作　　者：马瑞升[1] 孙涵[2] 林宗桂[2] 马轮基[2] 吴朝晖[2] 黄耀[1] 

机构地区：[1]南京农业大学资源与环境科学学院,江苏南京210095 [2]中国气象局国家卫星气象中心遥感应用试验基地,广西南宁530022

出　　处：《南京气象学院学报》2005年第5期632-639,共8页Journal of Nanjing Institute of Meteorology

基　　金：中国气象局新技术应用推广项目(20010003D001)

摘　　要：在分析无人机遥感影像几何变形理论基础上,利用标准图形垂直摄影法标定相机,建立相机误差纠正模型;使用大比例尺地形图作为底图,通过同名点配准纠正航摄影像,全过程不使用外方位元素。最终影像经实际检验,单点定位平均误差为1.0 m左右。The miniature aerial optical remote sensing system,which consists of a digital camera, a miniature robot plane （UAV, unmanned aerial vehicle）, and a ground-based subsystem, is a new technology in remote sensing field. Images from this remote sensing system can not be corrected easily. This status quo blocks the application of the UAV remote sensing technology in various research fields. This paper presents a geometric correction method in terms of the analysis of the UAV remote sensing images. Digital camera calibration and its distortion model are finished by a standard graphic method, and the images registration is completed by large-scale topographical maps. The extrinsic camera parameters are not used in this method. The single point mean error of the corrected image map is about 1.0 meters.

关 键 词：航空遥感 影像处理 相机标定 无人机 

分 类 号：P407[天文地球—大气科学及气象学]