【摘要】 为确保遥感影像为森林资源调查和监测提供更好的服务,必须对影像进行图像预处理,而图像的几何校正是图像预处理中关键一步。传统的基于TM影像的几何精校正只是采用其地面分辨率为30m的波段的影像,由于分辨率不高,导致校正精度不高。介绍了一种使用ETM+影像的地面分辨率为15m的Pan波段校正,在后续处理中,用其它波段与校正后的Pan波段几何配准,从而达到几何精校正的效果。仿真试验表明,这种先校正分辨率高的影像,再配准分辨率低的影像的方法,在一定程度上提高了校正精度,降低了控制点选取的难度。更多还原

【Abstract】 In order to provide better sensor imagery for surveying and monitoring forest resources, image preprocessing needs to be carried out, and geometric correction is important in the image preprocessing. Usually, the TM image of 30m resolution causes lower precision. In this article, a new correction method is introduced. ETM+ image of 15m resolution is corrected firstly, and then the other images of the ETM+ are registered with the corrected Pan image to achieve geometric precision correction. The simulation test shows, that the way of correcting the higher resolution images firstly and then making the lower resolution image’s registration will increase the precision and reduce the difficulty of GCP selector.更多还原