A self-organizing tree map approach for image segmentation

In this paper, an efficient image segmentation approach by using a self-organizing tree map (SOTM) is proposed. The SOTM neural network is first employed for the coarse segmentation to obtain the global clustering information of the image. Then, a pixel-based classification scheme that utilizes the local features is used to refine the segmentation. The proposed approach considers both global distributions of the image and local pixel characteristics; experimental results clearly show that images can be segmented into meaningful objects or parts. One of the advantages of the proposed approach is that the features used for the coarse segmentation can still be used to help make the final decision of the segmentation.

[1]  Donald Geman,et al.  Boundary Detection by Constrained Optimization , 1990, IEEE Trans. Pattern Anal. Mach. Intell..

[2]  P.K Sahoo,et al.  A survey of thresholding techniques , 1988, Comput. Vis. Graph. Image Process..

[3]  Philippe Andrey,et al.  Unsupervised Segmentation of Markov Random Field Modeled Textured Images Using Selectionist Relaxation , 1998, IEEE Trans. Pattern Anal. Mach. Intell..

[4]  Haluk Derin,et al.  Modeling and Segmentation of Noisy and Textured Images Using Gibbs Random Fields , 1987, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[5]  Ling Guan,et al.  Self-organizing tree map for eliminating impulse noise with random intensity distributions , 1998, J. Electronic Imaging.

[6]  Donald Geman,et al.  Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images , 1984 .