Split-and-merge segmentation of aerial photographs (original) (raw)

Abstract

A method of segmenting aerial photographs is described which approximates the image intensity surface by planar facets. This is accomplished using a split-and-merge approach. A combination of an F-test and a mean predicate is used to test the uniformity of regions. When two regions are merged together to form a new region, the nine variables needed to compute the least-squares plane for the new region can be computed by adding the corresponding variables for the individual regions. This leads to an efficient algorithm.

Loading...

Loading Preview

Sorry, preview is currently unavailable. You can download the paper by clicking the button above.

References (10)

  1. T. C. Pong, L. G. Shapiro, L. T. Watson, and R. M. Haralick, Experiments in segmentation using a facet model region grower, Comput. Vision Graphics Image Process. 25, 1984, 1-23.
  2. B. A. Parvin, A split and merge algorithm for segmentation of natural scenes, in Proceedings, 7th Int. Conf. Pattern Recogn., pp. 294-296.
  3. T. Pavlidis, Structural Pattern Recognition, Springer-Vedag, New York, 1977.
  4. R. H. Laprade and M. F. Doherty, Split-and-merge segmentation using an F-test criterion, Proc. SPIE Conf. Image Understanding Man-Mach. Interface 758, 1987, 74-79.
  5. C. R. Rao, Linear Statistical Inference and Its Applications, Wiley, New York, 1973.
  6. D. Harwood, S. Chang, and L. S. Davis, Interpreting Aerial Photographs by Segmentation and Search, Technical Report 223, Center for Automation Research, University of Maryland, 1986.
  7. D. M. McKeown and J. L. Denlinger, Map-Guided Feature Extraction from Aerial Imagery, Technical Report 84-117, Computer Science Department, Carnegie-Mellon University, 1984.
  8. P. J. Besl and R. C. Jain, Segmentation through symbolic surface descriptions, in Proceedings, 1986 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 1986, pp. 77-85.
  9. T. Pavlidis, A critical survey of image analysis methods, in Proceedings, Eighth International Conference on Pattern Recognition, 1986, pp. 502-511.
  10. R. H. Bartels and J. J. Jezioranski, Least-squares fitting using orthogonal multinomials, ACM Trans. Math. Software 11, 1985, 201-217.