Use of Ant Colony Optimization for Finding Neighborhoods in Image Non-stationary Markov Random Field Classification
In global classifications using Markov Random Field (MRF) modeling, the neighborhood form is generally considered as independent of its location in the image. Such an approach may lead to classification errors for pixels located at the segment borders. The solution proposed here consists in relaxing the assumption of fixed-form neighborhood. Here we propose to use the Ant Colony Optimization (ACO) and to exploit its ability of self-organization. Modeling upon the behavior of social insects for computing strategies, the ACO ants collect information through the image, from one pixel to the others. The choice of the path is a function of the pixel label, favoring paths within a same image segment. We show that this corresponds to an automatic adaptation of the neighborhood to the segment form. Performance of this new approach is illustrated on a simulated image and on actual remote sensing images SPOT4/HRV.
KeywordsMarkov Random Fields Label Image Find Neighborhood Markov Random Fields Modeling Pixel Label
- 2.Geman, D., Geman, S., Graffigne, C., Dong, P.: Boundary detection by constrained optimization. IEEE Trans. on PAMI 12, 609–628 (1990)Google Scholar
- 3.Geman, S., Reynolds, G.: Constrained restoration and recovery of discontinuities. IEEE Trans. on PAMI 14, 367–383 (1992)Google Scholar
- 6.Yao, X.: Evolutionary Computation: Theory and Applications. World Scientific, Singapore (1999)Google Scholar
- 11.Sigel, E., Denby, B., Le Hégarat-Mascle, S.: Application of ant colony optimization to adaptive routing in a leo telecommunications satellite network. Ann. of Telecommunications 57, 520–539 (2002)Google Scholar