Many image analysis problems require the description of the shape of sequences of pixels. Such pixel sequences are usually contours or skeletons obtained by processing of images. The pixels are usually but not always adjacent. Early work has emphasized polygonal approximations of the pixel sequence . Such approximations provide a certain degree of data compaction and noise removal but they do not go far enough. If the shape is not polygonal to start with the representation can be unwieldy.
KeywordsCompaction Hull Editing
Unable to display preview. Download preview PDF.
- 1.T. Pavlidis, Structural Pattern Recognition Springer Verlag, 1977.Google Scholar
- 3.R. H. Bartels, J. C. Beatty, and B. A. Barsky, An Introduction to the Use of Splines in Computer Graphics, Morgan Kaufmann, Los Altos, Calif., 1987.Google Scholar
- 4.T. Pavlidis, “Curve Fitting with Conic Splines,” ACM Trans, on Graphics, pp. 1-31, 1983.Google Scholar
- 5.G. Farin, in Curves and Surfaces for Computer Aided Geometric Design, Academic Press, 1988.Google Scholar
- 8.H. Hagen, Th. Schreiber, and E. Gschwind, “Methods for Surface Interrogation,” Proc. Visualization’90, pp. 187-193, San Francisco, Calif., October 23–26, 1990.Google Scholar
- 9.Pratt, V., “Techniques for Conic Splines,” Proceedings of SIGGRAPH’85, pp. 151-159, San Francisco, July 22–26, 1985.Google Scholar
- 10.V.S. Nalwa and E. Pauchon, “Edgel-Aggregation and Edge-Description”, CVGIP, vol. 40, pp. 79–94, 1987.Google Scholar