Advertisement

Multi-level dynamic programming for axial motion stereo line matching

  • Raymond K. K. Yip
Poster Session B: Active Vision, Motion, Shape, Stereo
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1310)

Abstract

In this paper, a multi-level dynamic programming approach is used to solve the line segment based correspondence problem in axial motion stereo. In this method, a Local Similarity Measure is calculated for each line segment pair between the Front and Back images. In level 1, the matching probability between line segments is represented by their Local Similarity Measure (LSM). Line segment pair that have a matching probability larger than a threshold T1 is selected as potential matching pair. T1 is set to a relative high value so that the probability of correct match of level 1 is very high. Dynamic programming is then used to search for their best match. Based on the geometric properties between the matched and the unmatched line segments, a Global Similarity Measure (GSM)is calculated for each unmatched line segment pair. An overall Similarity Measure (matching probability) is then obtained by the LSM and the GSM. Then, the algorithm begin the second match but with a slightly lower threshold T2. The new matched results are then used to modify the GSM and the overall Similarity Measure. These processes are repeated until a predefined level n stop (or a predefined condition) is reached. By using the GSM and multi-level searching technique, the proposed technique increases the matching accuracy and reduce the number of unmatched line segment due to misordering when dynamic programming is used for axial motion stereo matching.

Keywords

Line Segment Similarity Measure Stereo Match Correct Match Matching Accuracy 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

References

  1. [1]
    Nicolas Alvertos, Dragana Brzakovic and Rafael C. Gonzalez, “Camera Geometries for Image Matching in 3-D Machine Vision”, IEEE Trans. on PAMI, Vol. 11, No. 9, pp. 897–915, Sept. 1989.Google Scholar
  2. [2]
    U.R. Dhond and J.K. Aggarwal, “Structure from stereo-A Review”, IEEE Trans. Syst. Man Cybern., Vol. 19, No., 6, pp. 1489–1510, Nov./Dec. 1989.Google Scholar
  3. [3]
    X.Y. Jiang and H. Bunke, “Line Segment Based Axial Motion Stereo”, Pattern Recognition, Vol. 28, no. 4, pp. 553–562, 1995.Google Scholar
  4. [4]
    S. H. Lee and J. J. Leou, “A dynamic programming approach to line segment matching in stereo vision”, Pattern Recognition, vol. 27, no. 8, pp. 961–986, 1994.Google Scholar
  5. [5]
    Z. N. Li, “Stereo correspondence based on line matching in Hough space using dynamic programming”, IEEE Trans. Syst. Man Cybern., vol. 24, no. 1, pp. 144–152, 1994.Google Scholar
  6. [6]
    R.K.K. Yip and W.P. Ho, “Multi-level Based Stereo Line Matching with Structure Information Using Dynamic Programming”, Proceeding of the 1996 International Conference on Image Processing, Lausanne, Switzerland, September 16–19 1996.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1997

Authors and Affiliations

  • Raymond K. K. Yip
    • 1
  1. 1.Department of Electronic EngineeringCity University of Hong KongKowloonHong Kong

Personalised recommendations