The “rubber-mask” technique—I. Pattern measurement and analysis

Specific applications to the analysis of human chromosome images, chromatographic recordings, electrocardiogram waveforms, and electroencephalogram waveforms are illustrated.

Abstract

Template matching is a fundamental technique of pattern recognition. Although this technique is very general, its applicability has been limited because of the difficulty often encountered when fitting templates to natural data. Natural patterns are often distorted, misshapen, stretched in size, fuzzy, rotated, translated, observed at an unusual perspective, etc. Flexible templates (rubber masks) have been devised which, when fitted to natural data, can be used for measurement, data reduction, data smoothing, and classification of highly irregular waveforms and image shapes. These problems had been largely unsolved by existing template matching methods. Specific applications to the analysis of human chromosome images, chromatographic recordings, electrocardiogram waveforms, and electroencephalogram waveforms are illustrated. The rubber-mask technique will probably be usable in a wide variety of scientific applications.