문형 인식을 위한 궤적 추적 알고리즘에 관한 연구

A Study on Character Recognition by Trajectory Tracking Algorithm

In this study, by scanning the pattern left on a piece of paper to determine how to enter the characters, how to enter the characters using the keyboard, furthermore, how to convey the character of the input patterns that appear when you write a character to a location, such as the mouse, digitizer input device with Fuzzy-Neural algorithms were tested. Fuzzy-Neural hybrid system, fuzzy logic techniques and neural networks can be viewed as two individual subsystems. They do their own jobs by serving different purposes in a system. By making use of their individual strengths, they incorporate and complement each other to accomplish a desired task. This system makes use of the strength of a neural network in its processing speed and the strength of fuzzy logic in its flexible reasoning capability for decision making and control. To use the mouse in everyday life and to relieve these uncomfortable using an intelligent algorithm to determine the pattern analysis of the movement of the mouse, by letters or symbols that can be entered, judgment and experiment proved its usefulness.