论文部分内容阅读
卫星原始遥感图像的应用都需要几何校正，即对遥感图像中的每个像元点都必须进行地理经纬度的定位、太阳及扫描仪的高度角和方位角的计算。通常由已知足够数量的地面控制点或近似的插值法来实现。对于广阔的海洋，很难找到地面控制点，因此以卫星的轨道参数和扫描仪的视场角等参数为基础，研究一套较完整的地理定位法和相关的几何参数法，大大地简化了现有几何校正的方法。经过美国海岸带水色扫描仪资料（ＣＺＣＳ）、甚高分辨率辐射计（ＡＶＨＲＲ）和我国风云一号气象卫星资料（ＦＹ－１Ｂ）反演验证和局部区域的实际地面点对比表明，地理定位精度——纬度ＲＭＳ为０．０００１°，经度ＲＭＳ为０．０００５°，太阳的高度角和方位角的ＲＭＳ为０．０００７°，完全可以满足气象和海洋等环境卫星及陆地资源卫星Ｍｓｓ等资料几何精校正的要求，同时可作为ＴＭＴ和ＳＰＯＴ的粗几何校正。该算法不仅适用于升轨的卫星，而且也适用于降轨的卫星，同时还适用于前、后向倾斜和垂直等三种扫描状态的图像。
The application of satellite remote sensing images requires geometric correction, that is, for each pixel point in the remote sensing image, the location of geo-latitude and longitude, the elevation and azimuth of the sun and the scanner must be calculated. Usually by known a sufficient number of ground control points or approximate interpolation method to achieve. For the vast oceans, it is difficult to find ground control points. Therefore, based on parameters such as orbit parameters of satellites and scanner field of view, a more complete set of geolocation methods and related geometric parameter methods are greatly simplified Existing methods of geometric correction. The inversion of the CZCS, AVHRR and FY-1B data and the comparison of the actual ground points in the local area show that the accuracy of the geolocation - The latitude RMS is 0.0001 °, the longitude RMS is 0.0005 °, and the RMS of the elevation and azimuth of the sun is 0.0007 °, which can fully meet the requirements of the meteorological and oceanographic environmental satellites and land resources satellite Mss and other data geometry Fine correction requirements, and can be used as a rough geometric correction TMT and SPOT. The algorithm is not only applicable to ascending satellites, but also to orbit satellites. It is also suitable for three kinds of scanning images, such as front, back tilting and vertical scanning.