The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B4. Beijing 2008 
Table 2. Assessment results of crater DTMs (crater diameter>8 
pixels). DP : detection percentage compared with automated 
verification and manual verification standards, BF : branching 
factor, QP ; quality percentage. See equations in section 3.1 
According to these assessment result, the resolution limit of 3D 
crater extraction algorithms is radius >500m and depth <70m 
with HRSC stereo images. 
Finally to check the quality of the constructed crater DTM, a 
few of them were compared with CTX (10m) stereo DTMs. The 
method to build CTX is described in Kim and Muller (2007). 
Even though the availability of such DTMs is very limited at 
present, it provides a good insight into the reliability of the 
crater matching system as shown in Figure 8 . 
CTX 10m 
DTM 
a I 
eiEOfm 
-3202 -310' 
CTX 10m 
DTM 
HRSC 20m 
DTM orbit 2099 
(r=0.98km) 
a ! 
EIZEOKm 
-3JGÏ 
HRSC 20m 
DTM orbit 2099 
(r=0.41km) 
A 
/ 
Blue : CTX, Red: HRSC 
DTM profile 
Blue : CTX, Red: HRSC 
DTM profile 
Figure 8. The DTM comparison with CTX (note stereo height 
difference inferred to come from the influence of illumination 
angle) 
4. CONCLUSIONS & FUTURE WORK 
We showed an implementation of an automated and semi- 
automated crater GIS construction. There have been several 
previous studies on automatic impact crater detection 
algorithms but we believe this is the first case to incorporate 
crater detection algorithms into a practical crater GIS data 
generation systems. As shown in the 2D GIS assessment, a 
completely reliable and automated crater detection method is 
not available yet. This is believed to be mainly due to the 
accuracy of the verification method being insufficient. We 
described the software tools for editing and merging to 
compensate for any weakness in the construction of crater GIS. 
Moreover, we demonstrated a comprehensive system developed 
for the automated 3D crater DTM. The assessment shows that 
the data fusion method produced reliable 3D information on 
individual craters. The 2D crater GIS was produced over very 
extensive areas including the whole of Iani and Elysium. It is 
expected that the processing software will provide a very 
powerful tool for research in surface dating and 
geomorphological analysis. On the other hand, the general 
approach to update the performance of this system will be 
continuously explored through enhanced 3D construction 
algorithms such as shape from shading. In addition, the 
employment of HiRISE and CTX stereo data sets for creating 
finer resolution 3D crater GIS data is now being explored to 
allow finer-scale detail to be obtained for testing different 
models of crater size frequency distributions and surface age to 
be tested. 
ACKNOWLEDGEMENTS 
We thank STFC under grant PP/C502630/1 for supporting this 
study. 
Kim, 
REFERENCES 
Barlow, N. G. et al, 2003. Utilizing GIS in Martian impact 
crater studies. International Society for Photogrammetry 
and Remote Sensing Working Group IV/9: 
Extraterrestrial Mapping Workshop, "Advances in 
Planetary Mapping 2003”, 22, March, 2003, Houston, 
Tx, USA, 
http://astrogeology.usgs.gov/Projects/ISPRS/MEETING 
S/Houston2003/ahstracts/BarlowJsprs_mar03.pdf 
Cheng, Y. et al, 2003. Optical landmark detection for spacecraft 
navigation. Proceedings of the 13th AAS/AIAA Space 
Flight Mechanics Meeting. Ponce, Puerto Rico, A AS. 
Fitzgibbon, A. W., Pilu, M., and Fisher, R. B. 1996. Direct 
Least Squares Fitting of Ellipses. IEEE Transactions on 
Pattern Analysis and Machine Intelligence. 21(5), pp. 
476-480. 
Gruen, A. 1985. Adaptive Least Squares Correlation: A 
powerful Image Matching Technique. South Africa 
Journal of Photogrammetry’, Remote Sensing and 
Cartography. 14(3), pp. 175-187. 
Hartmann, W. K., 1966. Martian cratering. Icarus 5, pp. 565- 
576. 
Kanatani, K., Ohta, N., 2004. Automatic detection of circular 
objects by ellipse growing, Int. J. Image Graphics, 4(1), 
pp. 35-50. 
J. R., 2005. Landscape Object Detection and 
Reconstruction by Multi-sensor Data Fusion. PhD. 
Department of Geomatic Engineering, University 
College London (unpublished). 
Kim, J. R. et al, 2005. Automated Crater Detection, A New 
Tool for Mars Cartography and Chronology. 
Photogrammetric Engineering & Remote Sensing, 71 
(10), pp. 1205-1217. 
Kim, J. R. and Muller, J.-P., 2007. Very high resolution DTM 
Extraction from HiRISE Stereo Imagery, European 
Mars Science and Exploration Conference, 12-16, Nov, 
2007, Noordwijk, Netherlands, 
http://sci.esa.int/science- 
e/www/object/index.cfm?fobjectid=41364&fbodylongid 
=1991 
Salamuniccar, G. and Loncaric, S., 2007. Open framework for 
objective evaluation of crater detection algorithms with 
first test-field subsystem based on MOLA data. 
Advances in Space Research, In Press, Corrected Proof. 
Scholten, F.,et al 2005. Mars Express HRSC Data Processing- 
Methods and Operational Aspects, Photogrammetric 
Engineering & Remote Sensing, 71 (10), pp. 1143-1153. 
Shufelt, J. A., 1999. Performance evaluation and 
analysis of monocular building extraction from aerial 
imagery. Ieee Transactions on Pattern Analysis and 
Machine Intelligence, 21, pp. 311-326. 
Turk, M., and Pentland, A. 1991. Eigenfaces for Recognition. 
Journal of Cognitive Neuroscience, 3, pp. 71-76. 
Vinogradova, T. et al, 2002. Training of a crater detection 
algorithm for Mars crater imagery. Aerospace 
Conference Proceedings, 2002. IEEE, 7, pp. 3201-3211 
1054 
-