Recognizing hand-drawn quadrilaterals using genetic algorithms

In this paper an approach to recognize hand-drawn quadrilaterals using genetic algorithms is presented. To validate our approach, we have made several experiments recognizing quadrilaterals from a set of users, which were asked to draw quadrilaterals and other shapes using a Graphics Tablet. A Graphics Tablet was used in order to allow users to draw in a natural way. Our approach has shown to be robust to noise and distortion, being capable of recognizing quadrilaterals even when they are subject to a high degree of distortion. Qualitative tests are presented to show the usefulness of our approach.