期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2008
卷号：XXXVII Part B1
页码：761-768
出版社：Copernicus Publications
摘要：In order to acquire the largest possible coverage for environmental monitoring, small overlapping for satellite images is often the case. Hence, the traditional 3-D bundle adjustment as used in photogrammetry may not be directly employed for orientation modeling. In this investigation, we compare three approaches of block adjustment methods for satellite images that are with weak convergent geometry by using Digital Elevation Model (DEM) as an elevation control. The first one is a revised version of traditional bundle adjustment. The second one is based on the direct georeferencing approach. The third one is a rational function model (RFM) with sensor-oriented rational polynomial coefficients (RPC). These approaches integrate the collocation technique to improve the positioning accuracy. In the bundle adjustment, we calculate the orientation parameters by using the collinearity equations. Meanwhile, the DEM is used as an elevation control. Then, we collocate the orientation parameters by using least squares filtering. For the direct georeferencing, we adjust the orientation parameters from the satellite's ephemeris data, then, we calculate the error vectors for each tie point using DEM and refine the orientation parameters using the least squares filtering. In the implementation of elevation control in RFM, the block adjustment deals with the coordinate transformation and the observation equations of 3-D ground coordinates. In the adjustment, an iterative procedure is employed to combine the adjustment with DEM elevation control. The experimental results indicate that the proposed method using DEM as elevation control could significantly improve the geometric accuracy as well as the geometric discrepancies
关键词：Orientation; orthorectification; georeferencing; bundle adjustment; rational function model
