Segmentation of range images: A neural network approach

  • W. P. Cheung
  • C. K. Lee
  • K. C. Li
Neural Networks for Perception
Part of the Lecture Notes in Computer Science book series (LNCS, volume 930)


In this paper we present a neural computation model for histogram based range image segmentation. An optimal thresholding vector for the range histogram is determined. The number of elements in the vector is characterized by the histogram. Since our model is the parallel implementation of maximum interclass variance thresholding, the time for convergence will be much faster. Together with a real-time histogram builder, real time adaptive range image segmentation can be achieved.

The multithresholding criterion is derived from maximizing the interclass variance and hence the average of the c.g. (center of gravity) of two neighboring class pixel values should be equal to the interclass threshold value. The learning (weight matrix evolution) procedure of the neural model is developed based on the above condition. We use a three-layer neural network with binary weight synapses. The number of neurons in the first layer equals to that of the level of the range image and complex number inputs are used because the arguments of second layer outputs represent the c.g. of the class. The third layer neurons receive the argument output of the second layer and give an indication of the reach of the optimum condition.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    P.K.Sahoo, S.Soltani, A.K.C. Wong and Y.C.Chen, “A survey of thresholding techniques,” CVGIP vol. 41, pp233–260, 1988.Google Scholar
  2. 2.
    D. Bhandari, N.R.Pal and D.D.Majumder, “Fuzzy divergence, probability measure of fuzzy events and image thresholding,” Pattern Recognition Letter, vol.13, pp857–867, 1992.Google Scholar
  3. 3.
    J.Kitter and J.Illingworth, “Minimum error thresholding,” Pattern Recognition, vol.19, No.1, pp41–47, 1986.Google Scholar
  4. 4.
    T.Kurita, N.Otsu and N. Abdelmalek, “Maximum likelihood thresholding based on population mixture models,” Pattern Recognition, vol. 25, No.10, pp1231–1240, 1992.Google Scholar
  5. 5.
    R.Brunelli, “Optimal histogram partitioning using a simulated annealing technique,” Pattern Recognition Letter, vol.13, pp581–586, 1992.Google Scholar
  6. 6.
    A. Rosenfeld and P.De La Torre, “Histogram concavity analysis as an aid in threshold selection,” IEEE Trans SMC, vol.13, pp231–235, 1983.Google Scholar
  7. 7.
    N. Otsu, “A threshold selection method from gray level histograms,” IEEE Trans SMC, vol-9, No.1, pp62–66, 1979.Google Scholar
  8. 8.
    S. Ghosal, R. Mehrotra, “Application of neural networks in segmentation of range images,” IJCNN, vol.3, p297–302, 1992.Google Scholar
  9. 9.
    R. Hoffman and A.K.Jain, “Segmentation and Classification of range images,” IEEE Trans. PAMI, vol.9, no. 5, pp608–620, 1987.Google Scholar
  10. 10.
    W.P.Cheung, C.K.Lee and K.C.Li., “Multithresholding for arithmetic logic pattern matching,” Electronics Letters, Vol.29 No.5 pp. 481–483, 1993.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1995

Authors and Affiliations

  • W. P. Cheung
    • 1
  • C. K. Lee
    • 1
  • K. C. Li
    • 1
  1. 1.Department of Electronic EngineeringHong Kong Polytechnic UniversityKowloonHong Kong

Personalised recommendations