Mean shift technique for image segmentation and Modified Canny Edge Detection Algorithm for circle detection
Nupur J. Gandhi, Vandana Shah, Ravindra V. Kshirsagar · 2014
Image segmentation is an important and challenging problem in an image analysis. Segmentation of objects in an image is even more difficult and computationally expensive. In this paper an unsupervised object based image segmentation that is mean shift clustering approach will be studied. One of the most important step is pre-processed image by a standard mean shift based segmentation, which preserves desirable discontinuities present in the image and guarantees over segmentation in the image in their Output. This type of mean shift segmentation technique which clusters the regions instead of image pixels mostly reduces the sensitivity to noise and hence enhances the overall segmentation performance. Detection of circle is very important for initial stage of Mean Shift segmentation. It will first detect a circle with Circular Hough Transform and then with Modified Canny Edge Detection Algorithm. The Modified Canny Edge Detection Algorithm is very fast algorithm to detect circle from the images as compared to Circular Hough Transform.