线阵列卫星传感器定向方法的研究

Determination of Satellite Sensor Orientation

介绍了确定线阵列卫星影像外方位元素的几种常用方法 。

Successful satellite development and launch have provided numerous applications in photogrammetry,GIS,and remote sensing.For example,SPOT offered the first stereo satellite imaging capability of the earth surface,though at only moderate ground resolution,namely,10m.MOMS_02 now provides along_track stereo imagery with 4.5m and 13.5m resolution.Within next few years,high_resolution earth observation satellites (HREOS) with spatial resolutions of 1～5m are expected to usher in a new era of satellite photogrammetry,remote sensing and GIS data acquisition.Both SPOT and HREOS employ linear CCD sensors to acquire imagery.The geometry of the linear CCD sensors is a“line central projection” which,however,is different from area matrix array sensors or cameras which are often used in aerial and close_range photogrammetry.The perspective geometry is only valid on a sensor line,whereas it is close to a parallel projection in the orbit direction.For linear CCD imagery on SPOT and HREOS,every sensor line has its own orientation which is different from but highly dependent upon that of its neighbouring scan lines.The traditional photogrammetric bundle adjustment therefore can not be used to acquire the orientation elements of each sensor line.In recognition of the relatively stable dynamics of the flight attitude of space platforms,however,polynomial functions can be employed to interpolate the orientation parameters of each sensor line.Additional orbital constraints are also imposed in some cases. In this paper,various determination strategies for exterior orientation of the linear CCD sensors used in SPOT and MOMS are investigated.The strategies include Bingo approach,Krathy's solution,orbital parameter approaches,quadratic polynomials,Lagrange polynomials and affine model.Eventually,we present our own approach for the orientation determination of satellite imagery.By testing MOMS data,the accuracy acquired is satisfied with 1∶50 000 topographic mapping.Also,the approach can be extended to the best models for new steerable high_resolution pushbroom scanning systems (eg.,IKONOS by space imaging/EOSAT,which was launched in Sept.1999).