The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Voi. XXXVII. Part Bl. Beijing 2008 
329 
5(a) 5(b) 5(c) 
Fig.5 results of three methods for elevation computation 
5(a) represents algorithm based on PGM; 5(b) stands for method of AGM; 5(c) is the output of EV-InSAR 
Assuming that J_DEM data reflected the real elevation of this 
area, we used this data to examine precision of different 
methods. After randomly choosing 100 points, we compared 
errors in three DEM. However, the errors in DEM are not 
satisfactory regardless of which method is used. In 
consequence, we knew that introduction of Control Points (CPs) 
is a must. We separately established quadratic and cubic 
polynomials by using different number of control points 
summarized in Tab.2. We reached a conclusion that more CPs 
brings better precision. 
Classification of elevation computation 
dMax (m) 
dMin (m) 
dMean (m) 
dStd (m) 
Quadratic polynomial (8CPs) 
-62.953526 
0.396539 
2.155458 
29.704037 
Quadratic polynomial (16CPs) 
-38.400963 
0.163755 
2.351149 
11.341672 
Cubic polynomial (16CPs) 
-32.556606 
-0.140959 
2.323029 
11.253101 
Tab. 2 outputs of three polynomials by using control points 
4. CONCLUSION 
This paper has systematically analyzed the disadvantages of 
Approximate Geometry Model (AGM). To achieve higher 
precision especially when SAR images a wider area than it 
usually does, we proposed Precise Geometry Model (PGM) to 
develop algorithms of flattening phase, baseline estimation and 
elevation computation for SAR interferometry. In addition, we 
implemented these steps in Visual Studio.Net and compared the 
experiment results among different approaches. Then we 
proposed a method of elevation revision by utilizing 
polynomials and control points. Some suggestions were also be 
made for future research on PGM in more depth. 
REFERENCES 
[ 1 ] Paul A. Rosen, Scott Hensley, Ian R. Joughin, Fuk K. Li, 
Soren N. Madsen, Ernesto Rodriguez, Richard M. 
Goldstein, 2000. Synthetic aperture radar interferometry. 
Proceedings of the IEEE, 88(3), pp 333-381 
[2] Massonnet D., Rossi M, Camoma C., 1993. The 
displacement field of the landers earthquake mapped by 
rader interferometry. Nature, 36(4), pp 138-142 3 
[3] Goldstein R. M., Zebker H. A., Werner C. L, 1998. 
Satellite radar interferometry: two-dimensional phase 
unwrapping. Radio Science, 23(4), pp 4993-4999 
[4] Zheng Fang, Wu Junting, Ma Debao, 2005. Baseline 
estimation based on track errors and elevation precision 
images. Proceeding of the Information Engineering 
Univeristy, 6(1), pp 77-79