A method for two-dimensional pattern recognition, applicable to particle shape and size determination, is presented which employs fractal geometric analysis. Fractal contour transformation presents several advantages over spatial harmonic analysis. Coordinate transformation is evaluated with respect to its effect upon feature discrimination in polar-mapped multivariable contours. A variety of non-differentiable contours are analyzed. Results are presented in terms of correlation, signal-to-noise ration, and computational load.
One print or electronic copy may be made for personal use only. Systematic reproduction and distribution, duplication of any material in this paper for a fee or for commercial purposes, or modification of the content of the paper are prohibited. This manuscript is an author version and may be cited as: Caimi, F. M., & Schmalz, M. S. (1985). Geometric properties of non-differentiable contours: concurrent spatial harmonic and fractal analyses. In N. A. Chigier & G. W. Stewart (Eds.), Particle sizing and spray analysis, August 21, 1985, San Diego, California. Proceedings of the SPIE, 573 (pp. 21-29). Bellingham, WA, USA.
Florida Atlantic University. Harbor Branch Oceanographic Institute contribution #401.