The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Publications Copernicus
Download
Citation
Articles | Volume XLI-B1
https://doi.org/10.5194/isprs-archives-XLI-B1-383-2016
https://doi.org/10.5194/isprs-archives-XLI-B1-383-2016
03 Jun 2016
 | 03 Jun 2016

COMPENSATION LOW-FREQUENCY ERRORS IN TH-1 SATELLITE

Jianrong Wang, Renxiang Wang, and Xin Hu

Keywords: Satellite photogrammetry, Low-frequency errors, Attitude determination system, Exterior orientation elements, Location accuracy, Global coverage

Abstract. The topographic mapping products at 1:50,000 scale can be realized using satellite photogrammetry without ground control points (GCPs), which requires the high accuracy of exterior orientation elements. Usually, the attitudes of exterior orientation elements are obtained from the attitude determination system on the satellite. Based on the theoretical analysis and practice, the attitude determination system exists not only the high-frequency errors, but also the low-frequency errors related to the latitude of satellite orbit and the time. The low-frequency errors would affect the location accuracy without GCPs, especially to the horizontal accuracy. In SPOT5 satellite, the latitudinal model was proposed to correct attitudes using approximately 20 calibration sites data, and the location accuracy was improved. The low-frequency errors are also found in Tian Hui 1 (TH-1) satellite. Then, the method of compensation low-frequency errors is proposed in ground image processing of TH-1, which can detect and compensate the low-frequency errors automatically without using GCPs. This paper deal with the low-frequency errors in TH-1: First, the analysis about low-frequency errors of the attitude determination system is performed. Second, the compensation models are proposed in bundle adjustment. Finally, the verification is tested using data of TH-1. The testing results show: the low-frequency errors of attitude determination system can be compensated during bundle adjustment, which can improve the location accuracy without GCPs and has played an important role in the consistency of global location accuracy.