Shape object selection using the chi-square method

Abstract

Abstract This paper is focused on identifying the shape the shape of the object. There are some basic shape objects such as rectangle, circle, triangle, square, hexagonal, an ellipse. The method that used in this paper was chi-square feature selection to select a relevant feature toward those basic shape objects. Then, the proximity calculation between the training image and real image that was already available in the database. The chi-square feature selection process eliminates the irrelevant features in each figure that was in the image database. As final results, the system succeeds to detect an appropriate shape of an object by determining the distance of Euclidean between training image and the real image.