تطبيق خوارزميات كشف الحواف لاستخلاص هياكل الاشكال في الصور الرقمية

Abstract

The study of edge detection in digital images is an essential topic in the field
of image processing and pattern recognition. It is essentially significant in
solving the problems which represented by computer vision in order to
distinguish the structural forms of those images. This research develops a
program to detect the edges that exist in digital images (pictures) in order to
derive the structure of the forms (shapes) that exist in those images. The
study applies a suggested algorithm to different input images. Then, the
study extracts the edges of the images by using four methods of detection:
Robert, Sobel, Prewitt and Canny. Through the practical application, it has
been noticed that the suggested program is reliable in revealing the
boundaries and the edges of the forms that exist in the examined images to
the degree of a single pixel mainly through depending on the choice of
threshold value (automatically or manually) which determines the existence
of any change through excluding the insignificant values in detection phase
of the study.