Title :
Auto-registration of satellite image with the existing vector map
Author :
Lu Luping ; Zhang Yong
Author_Institution :
Sch. of Remote Sensing & Inf. Eng., Wuhan Univ., Wuhan, China
Abstract :
A novel method of registering the satellite images is introduced in this paper. The existing vector road maps are used as control information instead of the ground control points. The whole process is divided into 4 issues: find the initial relationship between the road maps and the satellite image with the help of the initial rough value of RPCs (the Rational Polynomial Coefficients) ;find the corresponding features between the image and the map; eliminate the wrong extracted features with some constraints designed from the characters of road. Such as the consistency of road width, as well as that the roads extracted should have similar direction with vector roads back-projected; build error equations and do adjustment to obtain the orientation parameters, reach the goal of registering the image. The algorithm is "fully automatic". The precision in our experiment is around 1.55 pixels.
Keywords :
image registration; polynomial approximation; roads; vectors; auto-registration; error equations; ground control points; initial rough value; rational polynomial coefficients; satellite image; vector road maps; Educational institutions; Equations; Feature extraction; Image edge detection; Remote sensing; Roads; Satellites; auto-registration; generalized point photogrammetry; high-resolution satellite imagery; vector road map;
Conference_Titel :
Geoscience and Remote Sensing Symposium (IGARSS), 2011 IEEE International
Conference_Location :
Vancouver, BC
Print_ISBN :
978-1-4577-1003-2
DOI :
10.1109/IGARSS.2011.6049193