A novel segmentation algorithm based on bare bones particle swarm optimization and wavelet mutation

Image segmentation is a difficult and challenging problem in the image processing. Bare bones particle swarm optimization (BBPSO) can not get good optimization performance because it easy to get stuck into local optima. Using wavelet mutation when no fitness improvement is observed, a new segmentation algorithm based on wavelet mutation BBPSO (WMBBPSO) and fuzzy entropy is proposed. The proposed algorithm uses WMBBPSO to explore fuzzy parameters of maximum fuzzy entropy, and to get the optimum fuzzy parameter combination, then obtain the segmentation threshold. According to experiment results of the new algorithm compare with other two algorithms, the proposed algorithm performs good segmentation performance and low time cost. It can be use to real time and precision measure coal dust image.