Pattern Recognition by Fourier Series Transformations

摘要

The purpose of this investigation is to find out what are the characteristics of pattern recognition systems which utilize Fourier Series transformations. Pertinent mathematical properties of this transform are presented, and several mechanisms for obtaining it are discussed. A means by which this transform may be computed as a part of the human recognition process is developed. Recognition systems which utilize the Fourier transform in different ways are simulated with the use of a digital computer. The recognition capabilities of these systems are tested using alphabetic pattern representations. The superiority of these systems to pattern correlation systems is clearly shown. (Author)