DocumentCode
2961824
Title
An automatic method of coarse registration between multi-source satellite images
Author
ZhaoHui, Zhang ; ChunHong, Pan ; SongDe, Ma
Author_Institution
Sch. of Math. & Inf., Hebei Normal Univ., Shijiazhuang, China
fYear
2004
fDate
14-17 Dec. 2004
Firstpage
205
Lastpage
209
Abstract
The paper gives an automatic registration method for coarsely aligning multi-source satellite images. The contributions of the presented method lie in its novel feature matching strategy. Area features are depicted by a shape matrix and principle component analysis (PCA). By fusing the constraints of shape similarity and object orientation consistency of corresponding area objects, the ambiguity of feature matching can be effectively avoided. The registration results for SAR-optical satellite images demonstrate the robustness and efficiency of the given method.
Keywords
feature extraction; geophysical signal processing; image matching; image registration; matrix algebra; principal component analysis; remote sensing by radar; synthetic aperture radar; PCA; SAR images; coarse registration; feature matching strategy; multi-source image registration; object orientation consistency; principle component analysis; satellite image registration; shape matrix; shape similarity; Gravity; Image edge detection; Image recognition; Image registration; Mathematics; Pattern recognition; Principal component analysis; Remote sensing; Satellites; Shape;
fLanguage
English
Publisher
ieee
Conference_Titel
Intelligent Sensors, Sensor Networks and Information Processing Conference, 2004. Proceedings of the 2004
Print_ISBN
0-7803-8894-1
Type
conf
DOI
10.1109/ISSNIP.2004.1417463
Filename
1417463
Link To Document