Application of Structure Lines to Surface Construction and 3-Dimensional Analysis
This paper presents a notion and properties of ‘Structure Lines’ Structure lines give intrinsic geometrical features of curved surfaces or pictures based on differential geometry. They are composed of characteristic lines, division lines and edge lines, which are invariant under the rotation and parallel translation of a coordinate system.
To show the usefulness of structure lines, first we apply them to surface construction problem: With the positional information of division lines, our algorithm synthesizes a surface by 1-dimensional interpolation, i.e. combination of cross-sectional interpolation method and contour line interpolation method so that curvatures of contour lines may have the same sign in each divided region. Application to 3-dimensional analysis is also presented: ‘Cusp points’ are defined in terms of structure lines as sharp corners of edges for finding a key information. We present the methods of matching a stereo pair efficiently by means of them and of 3-dimensional feature extraction from the stereo pair.
KeywordsFeature Point Contour Line Stereo Vision Structure Line Stereo Pair
Unable to display preview. Download preview PDF.
- 1.Brooks R.A., Greiner R. and Binford T.O.: The ACRONYM Model-Based Vision System, Proc. of 6th International Joint Conf. on Artificial Intelligence. 1979, pp 105–113.Google Scholar
- 2.Clowes M.B.: On Seeing Things, Artificial Intelligence. Vol. 2. 1971.Google Scholar
- 3.Coons S.A.: Surfaces for Computer-Aided Design of Space Forms, MIT Project MAC TR-41. 1967.Google Scholar
- 4.Enomoto H., Katayama T., Ito M. and Tsurumi Y.: Structure Lines of Surfaces, Technical Report of the Professional Group on Pattern Recognition and Learning of the Institute of Electronics and Communication Engineers of Japan. PRL72–138, March 1973 (in Japanese).Google Scholar
- 5.Enomoto H. and Katayama T.: Structure Lines of Images, Proc. of 3rd International Joint Conf. on Pattern Recognition, 1976.Google Scholar
- 6.Enomoto H., Katayama T. and Yoshida T.: Computer Experiment on Global Properties of Structure Lines of Images using Graphic Display and its Considerations, Information Processing in Japan, Vol.17, No.7, July 1976 (in Japanese).Google Scholar
- 7.Enomoto H., Katayama T. and Kawamoto E.: Searching Extremal Points of Function with Many Variables Using Linking Method, Information Processing in Japan. Vol.17, No.9, September 1976 (in Japanese).Google Scholar
- 8.Enomoto H., Yonezaki N. and Nitta K.: Schematic Representation and Perception of Curved Surface and Picture, Proc. of Eurographics 79. 1979.Google Scholar
- 9.Hosaka M.: Theory of Curve and Surface Synthesis and Their Smooth Fitting, Information Processing in Japan, Vol.10, No.3, May 1969 (in Japanese).Google Scholar
- 10.Hosaka M. and Kimura F.: A Theory and Methods for 3 Dimensional Free Form Shape Construction, Information Processing in Japan. Vol.21, No.5, May 1980 (in Japanese).Google Scholar
- 11.Huffman D.A.: Impossible Objects as Nonsence Sentences, in Meitzer B. and Michie D. (eds.) Machine Intelligence 6. Edinburgh University Press, 1971.Google Scholar
- 13.Watanabe A. and Yoshida T.: Roles of Central and Peripheral Vision in Pattern Perception, NHK Technical Monoghraph. no.21, 1973.Google Scholar
- 14.Watanabe Y., Mugino S. and Enomoto H.: Surface Construction Method by Contour Lines, Technical Report of the Professional Group on Pattern Recognition and Learning; of the Institute of Electronics and Communication Engineers of Japan, PRL80–86, February 1981 (in Japanese).Google Scholar
- 15.Watanabe Y., Imura T. and Enomoto H.: Recognition of 3-Dimensional Objects by Stereo Vision, Technical Report of the Professional Group on Pattern Recog;nition and Learning of the Institute of Electronics and Communication Engineers of Japan. PRL81–3, May 1981 (in Japanese).Google Scholar