Edge Detection in Segmented Images Through Mean Shift Iterative Gradient ZnNusing Ring
Esley Torres, Roberto M. Rodríguez, Yasel Garces, Osvaldo Pereira · International Journal of Soft Computing Mathematics and Control · 2015
In this paper, we propose a new method for edge detection in obtained images from the Mean Shift iterative algorithm.The comparable, proportional and symmetrical images are defined and the importance of Ring Theory is explained.A relation of equivalence among proportional images are defined for image groups in equivalent classes.The length of the mean shift vector is used in order to quantify the homogeneity of the neighborhoods.This gives a measure of how much uniform are the regions that compose the image.Edge detection is carried out by using the mean shift gradient based on symmetrical images.The difference among the values of gray levels are accentuated or these are decreased to enhance the interest region contours.The chosen images for the experiments were standard images and real images (cerebral hemorrhage images).The obtained results were compared with the Canny detector, and our results showed a good performance as for the edge continuity.