C++ Neural Networks and Fuzzy Logic: Preface

C++ Neural Networks and Fuzzy Logic

by Valluru B. Rao

MTBooks, IDG Books Worldwide, Inc.

ISBN: 1558515526   Pub Date: 06/01/95

Previous Table of Contents Next

Chapter 12

Application to Pattern Recognition

Using the Kohonen Feature Map

In this chapter, you will use the Kohonen program developed in Chapter 11 to recognize patterns. You will

modify the Kohonen program for the display of patterns.

An Example Problem: Character Recognition

The problem that is presented in this chapter is to recognize or categorize alphabetic characters. You will

input various alphabetic characters to a Kohonen map and train the network to recognize these as separate

categories. This program can be used to try other experiments that will be discussed at the end of this chapter.

Representing Characters

Each character is represented by a 5×7 grid of pixels. We use the graphical printing characters of the IBM

extended ASCII character set to show a grayscale output for each pixel. To represent the letter A, for example,

you could use the pattern shown in Figure 12.1. Here the blackened boxes represent value 1, while empty

boxes represent a zero. You can represent all characters this way, with a binary map of 35 pixel values.

Figure 12.1

  Representation of the letter A with a 5×7 pattern.

The letter A is represented by the values:

0 0 1 0 0

0 1 0 1 0