Single threshold segmentation for noisy image based on fuzzy ant colony algorithm
Ye Chen1, Xiaoqun Qin1, Xinmin Zhou2
1School of Information Science and Engineering, Hunan International Economics University, Changsha 410205, Hunan, China
2School of Computer Science and Information Engineering, Hunan University of Commerce, Changsha 410205, Hunan, China
Firstly, this paper pre-processes the image to be segmented through grey-scale morphological method. Then, based on the in-depth analysis of basic ant colony algorithm, it explains the shortcomings of this algorithm; proposes the improved strategy of ant colony algorithm, namely fuzzy ant colony algorithm, which designs the fitness function of artificial ant colony algorithm with minimum cross entropy and applies the improved fuzzy ant colony algorithm in the spatial-domain noisy image single segmentation. Finally, starting from the segmentation results and convergence, it compares the performances of the improved ant colony algorithm and the basic ant colony algorithm, GA algorithm and AFS algorithm.