Threshold logic applied to character recognition

The fact that many complex counting and decision functions can be realized quite simply with threshold gates suggests that they may be used to considerable advantage in problems of character recognition. A simplified recognition problem is considered involving the identification of any one of 23 alpha-numeric characters when it is superimposed on an m x n matrix. Each character is required to be identifiable under any degree of translation, stretching, and compression within the confines of the matrix. It is shown that the number of threshold gates required increases linearly as the dimensions of the matrix increase linearly. A threshold gate having the necessary output power to drive the required loads for this application is described together with its implementation in a small experimental model. The model is designed to recognize only a select group of characters, the numeric characters. The primary purpose of the model is to demonstrate some of the diverse applications of threshold logic to character recognition.