Multilevel thresholding algorithm based on particle swarm optimization for image segmentation

The Otsu method is a popular non-parametric method in image segmentation. However, the computation time grows exponentially with the number of thresholds when this method extended to multi-level thresholding. This paper presents a hybrid optimization scheme based on a self-adaptive particle swarm optimization algorithm for multilevel thresholding by the criteria of Otsu minimum within-group variance to render the optimal thresholding more effective. The experimental results show that the PSO-Otsu can provide better effectiveness on experiments of image segmentation.