AN APPROACH FOR CLASSIFICATION OF IMAGES AS CENTERLINE OR CONTOUR USING SKELETONIZATION ALGORITHMS

Abstract: Skeletonization is transformation of a component of a digital image into a subset of the original component. There are hundreds of publications on different aspects of these transformations. This paper reports contributions in this area with respect to properties of watershed, thinning and skeletonization algorithms. On application of these algorithms on different images, the results show that we can obtain characteristics of centerline or contour depending on the image. In this paper, it is proposed to apply morphological dilation operation with skeletonization to get results that are more effective. The algorithms are tested over many sample images. Hence it may be concluded that the development, choice and modification of such algorithms in practical applications are domain and task dependent, and there is no best method. By these experimental results, the images can be better classified by using properties of the resulting set in existing skeletonization algorithms for centerline or contour.

Keywords: skeletonization, watershed, thinning and shape simplification.