The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part Bl. Beijing 2008
image point, y is the row coordinate of corresponding image
point.
In this way, as far as any point in left image is concerned, its
corresponding point in the ground can be found easily,
meanwhile, left satellite position at that time can be determined.
Then according to matched right image, corresponding right
satellite position of homologous image point may be found.
And thus, oblique distance, baseline and the angle of incidence
can be calculated easily. Certainly, absolute precision is low,
but for the oblique distance of more than 800 kilometers,
calculation result is enough to remove the effect of flat earth.
Fig (2) indicates the interferogram before removing the phase of
flat earth and the one after removing the phase of flat earth
respectively.
4. CONCLUSION AND DISCUSSION
Because spacebome radar has only a pair of antenna, when it
receives scattering signals of same points on the ground at
certain time interval, correlation of two signals received is
reduced due to the natural condition change, at the same time ,
the complexity of geometric deformation caused by oblique
distance projection, especially in the mountainous area, makes
the correlation of the signal received further reduced, which
results in lower phase correlation of two signals in the complex
images, and greater difference between homologous image
points in grey images. The two points bring extreme difficulty
to precise registration of images. Data applied in the research
are ERS-1, ERS-2 complex images of Zhang Bei mountainous
area, and their time interval is about 24 hours. Based on the data
source, the precision of registration accomplished by method
described in the paper can reach 0.3 pixels or so. The high
precision and the analysis of interferogram show the method put
forward in the paper is basically feasible.
REFERENCE
Baatz, M. and Schape, A., 2000. Multiresolution segmentation:
an optimization approach for high quality multi-scale image
segmentation. Angewandte Geographische Informations
Verarbeitung, pp. 12-23.
Benz, U., Hofmann, P., Willhauck, G., Lingenfelder, I., and
Heynen, M., 2004. Multi-resolution, object-oriented fuzzy
analysis of remote sensing data for GIS-ready information.
ISPRS Journal of Photogrammetry and Remote Sensing, 58,
pp.239-258.
Blaschke,T. and Strobl, J., 2001.What’s wrong with pixels?
Some recent developments interfacing remote sensing and GIS.
GeoBIT/GIS 6. pp,12-17.Zhang, Z.X. and Zhang, J.Q., 1996.
Digital photogrammetry. Wuhan. Publishing house of Wuhan
Technology University Surveying & Mapping, pp.58-60.
Zhou, Y.Q. and Shu, N., 1997. ERS-1 radar interferometry
principle -monochrome and frequency spectrum displacement
method. Translation version of Wuhan Technology University
Surveying & Mapping, pp.23-25.
Liao, M.S., 2000. The key technology on automatically
producing interferogram by high precision INSAR images.
Doctor's academic dissertation of Wuhan Technology
University Surveying & Mapping, pp.85-89.
Zheng, Z.B., 2000. Precision position of images characteristic
applied in Feature registration. Journal of Wuhan Technology
University Surveying & Mapping, pp. 63-66.
Shi, S.P., 2000. Using ERS-1/2 data to produce DEM. Journal
of Surveying and mapping, pp. 35-37.
Rutino, G., Moccia, A. and Esposito S., 1998. DEM Generation
by Means of ERS Tandem Data. ISPRS Journal of
Photogrammetry and Remote Sensing, pp.l 12-115.
Paillou, P., Gelautz, M., 2000. Relief Reconstruction from SAR
Stereo Pairs The “Optimal Gradient” Registration Method.
ISPRS Journal of Photogrammetry and Remote Sensing.
pp.226-228.
ACKNOWLEDGEMENTS
This work was funded by Natural science fund project of
university of Jiangsu province (Item: 07KJD420215) and
Natural science fund project of Xu Zhou Normal University
(Item: 07PYL09).
Fig.(2). (a)Original interferogram(only multiple vision
processing),(b)Interferogram of removal of the phase of
flat earth(processing by direction filer).