area 2. 
area 3 can 
vavelength 
top of the 
rain. It is 
DSM). As 
[uation are 
rom three 
hod in the 
calculated, 
n methods 
histograms 
d has the 
eristics of 
d also has 
1 baseline 
ig whereas 
  
ds 
| shown in 
green, red, 
> methods. 
range of 
meter. We 
ribution of 
of test area 
th the dual 
e the true 
rates dual- 
r elevation 
    
  
  
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XXXIX-B7, 2012 
XXII ISPRS Congress, 25 August — 01 September 2012, Melbourne, Australia 
            
    
2 d -50 
(a) MLE (b) ESM (c)Proposed method height (m 
Figure.9. DSM mapping results using three methods. 
Jaan Praks, Elise Colin-Koeniguer, and MarttiT.Hallikainen, 
2009. Alternatives to Target Entropy and Alpha Angle in SAR 
  
Proposed 
  
  
  
  
  
Ubxwr E ITT Tweed Polarimetry, IEEE Trans. Geosci.Remote Sens., vol.47, no7, 
LOXIC $1 fi 3 pp.2262-2274. 
S4 + A i S RCI : nar 
: x 7 .R.Cloude and E. Pottier, 1997. Anentropy based classification 
6.0x10° f 3 scheme for land applications of polarimetricSAR," IEEE Trans. 
4x1 | V x 3 Geosci. Remote Sens., vol.35, no.1, pp.68-78. 
swe X jp S. Sauer, L. Ferro-Famil, A. Reigber, and E. Pottier, 2007. 
9 = = EN = Multibaseline POL-InSAR analysis of urban scenes for 3D 
height(m) modeling and physical feature retrieval at L-band.Proc. 
Figure. 10. Surface height histograms of three methods. IGARSS, pp. 1098-1101. 
ACKNOWLEDGEMENTS 
6. CONCLUSION : 
The work was supported by national 863 project (Grant. 
: ; ; No.2011AA120401). The authors would like to thank CECT-38 
The  Multi-mode-XSARdataset is applicable for land for providing the PolInSAR data. 
classification object detectionand DSM mapping. First, 
polarimetric analysis has shown that X-band can provide a good 
discrimination between the different land types. Next, the 
experiments employing the selected polarimetric descriptors for 
land classification and man-made objects detection show the 
higher accuracy results. In addition, concerned with the 
characteristics of the Multi-mode-XSAR datasets, another 
experiment of DSM mapping employing the proposed dual- 
baseline polarimetric interferometry method has been proved to 
have the potentials of promotion of elevation accuracy. 
REFERENCES 
B. E. Boser, I. Guyon, and V. Vapnik, 1992. A training 
algorithm for optimal margin classifers. In Proceedings of the 
Fifth Annual Workshop on Computational Learning Theory, pp. 
144-152.ACM Press. 
C.-C. Chang and C.-J.Lin., 2011. LIBSVM -- A Library for 
Support Vector Machines. 
http://www.csie.ntu.edu.tw/cjlin/libsvm(S Nov, 2011) 
Corr,D.G., Walker, A., Benz, U.,Lingenfelder, I., Rodrigues, A., 
2003. Classification of urban SAR imagery using object 
oriented techniques.Proc. IGARSS, pp.198-190. 
E.Colin-Koeniguer, N.Trouvé, 2010. A review about 
alternatives to classical Polarimetric SAR parameters. 2010 gh 
European conference on Synthetic Aperture Radar, Germany. 
pp.414-417. 
Elise Colin, Cecil Titin-Schnaider, and WalidTabbara, 2006.An 
Interferometric Coherence Optimization Method in Radar 
Polarimetry for High-Resolution Imagery, IEEE Trans. Geosci. 
Remote Sens., vol. 44, no. 1, pp. 167-175.