An automatic image segmentation algorithm involving shortest path basins
Image segmentation is a process of partitioning input image into meaningful regions. It is a challenging task that is involved in almost every image processing system. Currently lot of methods for image segmentation with different approaches was created. Between all of them the methods based on graph theory are more and more popular nowadays. Segmentation methods could be classified for example to interactive and automatic ones. The further class of methods benefits from a user interaction that provides valuable information about a segmentation problem. The later class of methods doesn’t incorporate any user interaction. Nevertheless fully automatic methods that are both precise and robust are still hard to find. In this paper a new method based on shortest path in a graph is presented. This method automatically places seed points that are further used for image segmentation in the sense of path basins. This method allows segment an input image to a predefined or to an undefined number of image segments. Derived seed points could also be used in other interactive methods instead of a user interaction. Experiments with this method show its potential for segmenting a general class of images.
Keywordsimage segmentation shortest path in a graph
Unable to display preview. Download preview PDF.
- 11.M. Pavan and M. Pelillo, “A new graph-theoretic approach to clustering and segmentation,” in Proc. IEEE Conf. on Computer Vision and Pattern Recognition (Madison, 2003), pp. 145–152.Google Scholar
- 15.C. Tomasi and R. Manduchi, “Bilateral filtering for gray and color images,” in Proc. 6th Int. Conf. on Computer Vision (Bombay, 1998).Google Scholar