938 
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B5. Beijing 2008 
Table 2 shows the results of variance homogeneity tests for some 
representative 2.5-D LST tasks. Obviously, the VC-matrices are 
homogeneous; this means that no significant variations occur and 
the VC-matrix estimated by VCE can be accepted. In other words, 
when using the same sensor, the measurements of one group of 
observations (intensity or range) is subject to the same stochastic 
errors. Thus, an aggregation of all observations of one group with 
one weight is acceptable. An a-posteriori weighting by robust 
VC-matrix estimation is not necessary. Due to its low computa 
tional effort, the estimation of robust VC-matrices is still prac 
ticable and maybe useful in the case of VCE modeling failures 
within the VC-matrix. 
The experiment shows that the consideration of two groups of 
observations in a VCE is sufficient and that there is no significant 
variation of the precision of observations within the groups of 
observations. 
6 CONCLUSIONS AND OUTLOOK 
In this article, a least squares tracking approach based on 3-D 
camera intensity and range data was proposed. The presented 
functional model combines the transformation parameters for in 
tensity and range images and has been proven by various ex 
periments with synthetic and real data. It could be shown that 
an increase in accuracy, stability and reliability can be reached 
for least squares matching and tracking by the integrated treat 
ment of intensity and range information. The stochastic model 
has been designed by using a variance component estimation ap 
proach as well as robust variance covariance matrix estimation. It 
could be shown that a separation of the heterogeneous data into 
two groups of observations is sufficient for the accuracy of the 
stochastic model, and that there is no significant variation of pre 
cision within the groups of observations. As an additional prod 
uct, the procedure delivers information on the precision of 3-D 
camera range and intensity measurements. 
So far, the affine scale parameters are modeled through the ad 
ditional range information. Future work will address other ge 
ometric patch transformation parameters, which are not consid 
ered by the 2-D affine transformation. In particular, the keystone 
distortion caused by an inclination between the sensor plane and 
the captured object can be expressed through the RIM depth off 
set parameter. Beyond this, the effect and elimination of outliers 
should be addressed. Robust estimation procedures with respect 
to outlier phenomena will be more appropriate than dealing with 
certain HS-pattems. Also, likelihood approaches with a heavy 
tailed distribution could be used for a Gaussian distributed error 
term within the GLSE framework. 
References 
Ackermann, R, 1984. High precision digital image correlation. In: Proceedings of 
the 39th Photogrammetric Week, Schriftenreihe der Universität Stuttgart, Vol. 9, 
pp. 231-243. 
Baltsavias, E. R, 1991. Multiphoto Geometrically Constrained Matching. Ph.d. 
thesis, Institute of Geodesy and Photogrammetry. ETH Zurich. Switzerland. 
Breuer, P, Eckes, C. and Muller, S., 2007. Hand gesture recognition with a novel ir 
time-of-flight range camera: A pilot study. In: MIRAGE07, pp. 247-260. 
Du, H., Oggier, T., Lustenberger, F. and Charbon, E., 2005. A Virtual Keyboard 
Based on True-3D Optical Ranging. In: British Machine Vision Conference 
2005, Vol. 1, pp. 220-229. 
Förstner, W., 1984. Quality assessment of object location and point transfer using 
digital image correlation techniques. In: International Archives of Photogram 
metry, Vol. 25-111, A3a, pp. 197-217. 
Göktürk, S. B. and Tomasi, C„ 2004. 3d head tracking based on recognition and 
interpolation using a time-of-flight depth sensor. In: CVPR04, pp. II: 211-217. 
Greene, W. H., 2007. Econometric Analysis. 6th edn, Prentice Hall. ISBN 978- 
0130661890. 
Grün, A., 1985. Adaptive least squares correlation - a powerful image matching 
technique. In: South African Journal of Photogrammetry, Remote Sensing and 
Cartography, Vol. 14, 3, pp. 175-187. 
Grün, A. and Akca, D., 2004. Least squares 3d surface matching. In: International 
Archives of the Photogrammetry, Remote Sensing and Spatial Information Sci 
ences, 5, Vol. XXXIV, ISPRS ’’Panoramic Photogrammetry Workshop”. 
Gudmundsson, S. A., 2006. Robot vision applications using the CSEM swissranger 
camera. Master’s thesis, Informatics and Mathematical Modelling, Technical 
University of Denmark, DTU. 
Hempel, M., 2006. Validierung der Genauigkeit und des Einsatzpotentials einer dis 
tanzmessenden Kamera. Unveröffentlichte Diplomarbeit, Technische Universität 
Dresden Institut für Photogrammetrie und Fernerkundung. 
Hempel, R. and Westfeld, P, 2008. Statistical modeling of interpersonal distance 
with range imaging data. In: Accepted at COST 2102 training school on ’’Multi 
modal Signals: Cognitive and Algorithmic Issuses”. Vietri sul Mare, Italy, 21st- 
26th April 2008. 
Holte, M. B. and Moeslund, T. B., 2007. Gesture recognition using a range camera. 
Technical Report CVMT-07-01 ISSN 1601-3646, Aalborg University, Denmark. 
Isard, M. and Blake, A., 1998. Condensation - conditional density propagation for 
visual tracking. International Journal of Computer Vision 29(1), pp. 5-28. 
Kahlmann, T. and Ingensand, H., 2006. Calibration of the fast range imaging camera 
swissranger for use in the surveillance of the environment. In: Electro-Optical 
Remote Sensing II. Edited by Kamerman, Gary W.; Willetts, David V; Steinvall, 
Ove K.. Proceedings of the SPIE, Volume 6396, pp. 639605 (2006)., Presented 
at the Society of Photo-Optical Instrumentation Engineers (SPIE) Conference, 
Vol. 6396. 
Kahlmann, T., Remondino, F. and Guillaume, S., 2007. Range imaging technology: 
new developments and applications for people identification and tracking. In: 
A. Beraldin, F. Remondino and M. R. Shortis (eds), Proceedings of Videometrics 
IX - SPIE-IS&T Electronic Imaging, Vol. 6491. 
Kahlmann, T., Remondino, F. and Ingensand, H., 2006. Calibration for increased 
accuracy of the range imaging camera swissrangertm. In: H.-G. Maas and 
D. Schneider (eds), ISPRS Image Engineering and Vision Metrology, Vol. 
XXXVI, 5, pp. 136-141. ISSN 1682-1750. 
Koch, K.-R., 2004. Parameterschätzung und Hypothesentests in linearen Modellen. 
4. edn, Ferd. Dümmlers Verlag, Bonn. 
Koller-Meier, E. B., 2000. Extending the condensation algorithm for tracking mul 
tiple objects in range image sequences. PhD thesis, Technische Wissenschaften 
ETH Zrich, Nr. 13548. 
Lange, R., 2000. 3D Time-of-Flight Distance Measurement with Custom Solid- 
State Image Sensors in CMOS/CCD-Technology. Phd thesis, Department of 
Electrical Engineering and Computer Science at University of Siegen. 
Maas, H.-G., 2000. Least-squares matching with airborne laserscanning data in a 
tin structure. International Archives of Photogrammetry and Remote Sensing 
33(3a), pp. 548-555. 
Maas, H.-G., 2002. Methods for measuring height and planimetry discrepancies 
in airborne laserscanner data. Photogrammetric Engineering & Remote Sensing 
68(9), pp. 933-940. 
Maas, H.-G., Stefanidis, A. and Grün, A., 1994. From pixels to voxels - tracking 
volume elements in sequences of 3-d digital images. In: International Archives 
of Photogrammetry and Remote Sensing, 3/2, Vol. 30. 
Newey, W. K. and West, K. D., 1986. A simple, positive semi-definite, heteroskedas- 
ticity and autocorrelationconsistent covariance matrix. NBER Technical Work 
ing Papers 0055, National Bureau of Economic Research, Inc. 
Oggier, T., Lehmann, M., Kaufmann, R„ Schweizer, M., Richter, M., Metzler, R, 
Lang, G., Lustenberger, F. and Blanc, N., 2004. An all-solid-state optical range 
camera for 3D real-time imaging with sub-centimeter depth resolution (Swiss- 
Ranger). In: L. Mazuray, P. J. Rogers and R. Wartmann (eds), Optical Design 
and Engineering, Presented at the Society of Photo-Optical Instrumentation En 
gineers (SPIE) Conference, Vol. 5249, pp. 534—545. 
Rounault, A., 2007. Asymptotic behavior of random determinants in laguerre, gram 
and jacobi ensembles. Latin American Journal of Probability and Mathematical 
Statistics III, pp. 181-230. 
Schneider, D. and Maas, H., 2007. Integrated bundle adjustment with variance com 
ponent estimation: Fusion of terrestrial laser scanner data, panoramic and central 
perspective image data. In: Proceedings ISPRS Workshop Laser Scanning and 
SilviLaser 2007, Helsinki, International Archives of Photogrammetry, Remote 
Sensing and Spatial Information Sciences, Vol. XXXVInumber 3/W52, p. 373. 
Schwarte, R., 1997. Verfahren und Vorrichtung zur Bestimmung der Phasen- 
und/oder Amplitudeninformation einer elektromagnetischen Welle. Patentschrift 
DE 19704496, Inhaber Schwarte GBR. 
Seber, G. A. F., 2003. Linear Regression Analysis (Wiley Series in Probability and 
Statistics). 1st edn, John Wiley & Sons, Inc. ISBN 978-0471415404. 
Westfeld, R, 2007a. Ansätze zur Kalibrierung des Range-Imaging-Sensors SR- 
3000 unter simultaner Verwendung von Intensitäts- und Entferungsbildem. In: 
T. Luhmann (ed.), Photogrammetrie - Laserscanning - Optische 3D-Messtechnik 
(Beiträge Oldenburger 3D-Tage 2007), Herbert Wichmann Verlag Heidelberg, 
pp. 137-146. 
Westfeld, R, 2007b. Development of approaches for 3-d human motion behaviour 
analysis based on range imaging data. In: A. Gruen and H. Kahmen (eds), Op 
tical 3-D Measurement Techniques VIII, Vol. II, Institute of Geodesy and Pho 
togrammetry, ETH Zürich, pp. 393-402. 
White, H., 1980. A heteroskedasticity-consistent covariance matrix estimator and a 
direct test for heteroskedasticity. Econometrica 48(4), pp. 817-838. 
Zywitza, F., Massen, J., Brunn, M., Lang, C. and Gömig, T., 2005. One-to-three- 
dimensional ranging for future automotive safety systems. In: H. Ingensand and 
T. Kahlmann (eds), Proceedings of the 1st Range Imaging Research Day at ETH 
Zurich in Switzerland. ISBN 3-906467-57-0.