Improved Fuzzy Connected Image Segmentation Algorithm

Image segmentation is the problem of finding the homogeneous regions (segments) in an image, it is important in image processing. The notion of fuzzy connectedness captures the idea of "hanging-togetherness" of image elements in an object by assigning a strength of connectedness to every possible path between every possible pair of image elements. This concept leads to powerful image segmentation algorithms based on dynamic programming whose effectiveness has been demonstrated on thousands of images.Based on it, this paper presents two new algorithms which are both simpler and faster than it. The main ideas of the algorithms are to use the preprocessor and then compute the connective objects. Here the preprocessor by using the threshold judge the affinities and disconnect the affinities which lower than threshold. Then uses the Reachability Matrix algorithm or Spanning tree algorithm (here the latter is better than the former algorithm) to find the connective region. They have been tested in many medical gray images, all experiments were qualitatively demonstrated the effectiveness of these methods.In recent years, how to find the reference seeds automatically for multiple objects image segmentation and speed up the process of large images segmentation as important issues to us. In this work we present a novel TABU search-based approach to choose the reference seeds adaptively and use a spanning tree method for fuzzy object extraction in image segmentation. This proposed algorithm would be more practical and with a lower computational complexity than relative fuzzy connectedness multiple image segmentation. The results obtained in real images confirm the validity of the proposed approach.