Fuse, Takashi 
  
4.3 Application to Different Time Interval Successive Image 
Furthermore, we also applied the proposed method to some successive images at different time intervals, which were 
0.5s, 1.5s, 2.5s and 4.0s. The data of the images are same as those in Section 4.2. 
Table 3 shows the results. When the time intervals were less than 1.5 seconds, the correct rates were good. On the 
contrary, when the time intervals were over 2.5 seconds, the correct rates were less than 80%. When the time interval 
is less than 1.5 seconds, the accuracy of vehicle tracking is sufficient in this case. It is note that the relation between 
time interval and accuracy is much affected by states of traffic flow. 
Table 3: Correct Rates to Different Time Interval Images. 
  
  
  
  
  
  
t (seconds) Correct rates 
0.5 100.0% 
1.5 97.9% 
2.5 78.9% 
4.0 70.5% 
  
  
  
For all the methods mentioned above (Section 4.1 to 4.3), the processing time is about 60 seconds with a usual 
computer (CPU: Pentium 266MHz, RAM: 128MB). 
S. CONCLUSION 
The conclusions of this paper are as follows: 
(1) We proposed a new method by improving probabilistic relaxation method; 
(2) We confirmed the effectiveness of the proposed method through applications to simulated data, sample images and 
some different time interval successive images. 
The future works are as follows: 
(1) Application to more complex traffic flow; 
(2) Comparison with other matching method; 
(2) Development of automatic vehicle detection method; 
(3) Unification of detection method and tracking method as vehicle tracking system. 
REFERENCES 
Barnard, S.T. and Thompson, W.B., 1980. Disparity Analysis of Images. IEEE Transaction of Pattern Analysis and 
Machine Intelligence, PAMI-2 (4), pp.333-340. 
Ohmi, K. and Yu, L.H., 1998. Performance of the Relaxation Method PTV on the Basis of PIV Standard Images. 
Journal of Visualization Society of Japan, 18 (2), pp.113-114. 
Peleg, S., 1980. A New Probabilistic Relaxation Scheme. IEEE Transactions of Pattern Analysis and Machine 
Intelligence, PAMI-2 (4), pp.362-369. 
Sakamoto, M., Uchida, O. and Wang, P., 1998. Automatic High Accuracy Tie Points Detection in Stereo Images 
Using Relaxation Method and Gradient Method. Journal of the Japan Society of Photogrammetry and Remote Sensing, 
37 (5), pp.35-46. 
Takagi, M. and Shimoda, H., 1991. Handbook of Image Analysis. The University of Tokyo, Tokyo, pp.707-746. 
Zucker, S.W., Hummel, R.A. and Rosenfeld, A., 1977. An Application of Relaxation Labeling to Line and Curve 
Enhancement. IEEE Transactions of Comput., C-26 (4), pp.394-403. 
  
284 International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B5. Amsterdam 2000.