Id): In Figure 
rve at the top, 
om the EPI's 
' is indeed a 
t field): The 
nore obvious 
1 6 in Figure 
ectified EPIs 
S are straight 
d): Some of 
trajectories in 
ects get very 
jectory l in 
?PT, while the 
| (see Figure 
hose the first 
ges, and then 
0  softcopy 
ound control 
're measured 
DEMSs was a 
VirtuoZo. A 
ree test fields 
st field, we 
"he variances 
sted in Table 
up to 30% in 
ment rate is 
Checked 
points 
580 
990 
1500 
  
aerial image 
is technique 
lvantages of 
the normal 
of explicit 
s than those 
escribe the 
d, therefore, 
' orientation 
nd practical 
ile mapping 
rated to be 
' Sequences. 
Because EPI analysis technique for extracting DEM not only 
eliminates the need for stereo matching, but also can solve 
depth discontinuities and occlusion problems, which 
photogrammetrists in the traditional photogrammetric 
community have been wrestling with for decades. 
ACKNOWLEDGMENTS 
This research work was funded by Alexander von Humboldt 
Foundation (AvH Stiftung), Germany under contract number: 
IV CHN 1036848 STP. The aerial image video data were 
collaboratively collected with Prof. G., Fisher at Institute of 
Space Science, Free University of Berlin, Germany. 
REFERENCES 
Ackermann, F., 1983. High Precision Digital Image 
Correlation. Schriftenreihe des Instituts fuer 
Photogrammetrie der Universitaet Stuttgart, Heft 9, 231- 
243. 
Bolles, R.C. and H.H. Baker, 1985a. Epipolar-Plane Image 
Analysis: A Technique for Analyzing Motion Sequence, 
Int. Symposium on Robotics Research, Oct. Gouvieux, 
France, 41-48. 
Bolles, R.C. and H.H. Baker, 1985b. Epipolar-Plane Image 
Analysis: A Technique for Analyzing Motion Sequence, 
IEEE third workshop on Computer Vision: Representation 
and Control, Oct. Bellaire, Michigan, 168-176. 
Fóstner, W., 1986. A Feature Based Correspondence Algorithm 
for Image matching, Int. Arch. of Photogrammetry, Vol. 
26-111, Rovaniemi, Finland. 
Hattori, S. and Y. Myint, 1995. Automatic estimation of initial 
approximations of parameters for bundle adjustment, 
Photogrammetric Engineering and Remote Sensing, Vol. 
61, No. 7, July, 909-915. 
Luhmann, H. and T. Altrogge, 1986. Interesting operator for 
Image Matching, Int. Arch. of Photogrammetry and 
Remote Sensing, Vol. 26, Part IV, 459-475. 
Niini, I., 1990. Object Reconstruction Using 2-D Projective 
Transformations. The Photogrammetric Journal of 
Finland, Vol. 12, No. 1, 22-23. 
Lü, Y. and Z. Zhang, 1986. Fast Implementation for Generating 
Epipolar Line Images with One-Dimensional Resampling. 
Int. Arch. of Photogrammetry and Remote Sensing, Vol. 
26, Part IV, 459-475. 
Tang, L. and C. Heipke, 1996. Automatic Relative Orientation 
of Aerial Images, Photogrammetric Engineering and 
Remote Sensing, Vol. 62, No. 1, January, 47-55. 
Zhou, G., J. Albertz and K. Gwinner, 1999. Extracting 3D 
information Using Spatio-Temporal Analysis of Aerial 
Image Sequences, Photogrammetry Engineering and 
Remote Sensing, Vol. 65, No. 7, July, pp. 823-832. 
  
Fig. 4. The first image in the first test field. 
  
N TR 
Fig. 5. The first image in the second test field. 
  
Fig. 7. Extracted feature points from the first image of the third 
test field. 
   
     
Fig. 8. Tracked feature points and accurately refined ones by 
LSM in 20th frame. 
©) 
   
Fe CE ELLE z RN SUN 3 
Fig. 9. The original image (a) and rectified image (b) for the 
400th frame in Berlin city test field. 
ee a 
Fig. 10. The original image (a) and rectified image (b) for the 
300th frame in Schónefeld field. 
—611-