DocumentCode :
3042574
Title :
A research on geometric precision correction with high accuracy and high performance
Author :
Zhao, Lingjun ; Ma, Yan ; Liu, Dingsheng ; Zhang, Wanfeng ; Wang, Jiang
Author_Institution :
Center for Earth Obs. & Digital Earth, Chinese Acad. of Sci., Beijing, China
fYear :
2011
fDate :
26-28 July 2011
Firstpage :
3370
Lastpage :
3373
Abstract :
As an important step in the process of remote sensing,geometric precision correction, also known as the most complicated and intensive part in calculation, rectifies the geometric distortion of original remote sensing images caused by various factors by using ground control points. Therefore, the accuracy and processing speed of geometric precision correction have direct effect on the subsequent flows in most remote sensing applications. As for accuracy, the polynomial model and the multi-quadric function model are respectively introduced and then Gopfert model which is the smart combination of these two models is put forward. Speaking of capability, after analyzing various links of geometric precision correction and conventional parallel mode, the asynchronous parallel mode based on parallel I/O and I/O hidden strategy is given to optimize the parallel of Gopfert calculation on the cluster environment. All tests that were done indicate that the method mentioned above greatly improves both the accuracy and the function of geometric precision correction, which can be extensively used in project processing.
Keywords :
geophysical image processing; polynomials; remote sensing; Gopfert calculation; Gopfert model; I/O hidden strategy; asynchronous parallel mode; geometric distortion; geometric precision correction; ground control point; multiquadric function model; parallel I/O; polynomial model; remote sensing; Accuracy; Computational modeling; Interpolation; Load modeling; Mathematical model; Remote sensing; Satellites; Geometric Precision Correction; Gopfert; I/O hidden strategy; parallel I/O; parallel computing;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Multimedia Technology (ICMT), 2011 International Conference on
Conference_Location :
Hangzhou
Print_ISBN :
978-1-61284-771-9
Type :
conf
DOI :
10.1109/ICMT.2011.6002693
Filename :
6002693
Link To Document :
بازگشت