A Two-dimensional FCM Image Segmentation Method Based on Predator-Prey Particle Swarm Optimization

The traditional fuzzy C-means(FCM) clustering algorithms only make use of gray information of the pixels,and takes no account of the spatial neighbor information in image segmentation,which leads to poor anti-noise performance.In order to overcome limitations of the traditional FCM algorithms,a two-dimensional fuzzy C-means clustering method based on predator-prey particle swarm optimization is proposed for image segmentation.In this method,the image segmentation is converted into an optimization problem.The fitness function containing neighbor information is set up based on the gray information and the neighbor relations between the pixels described by the improved two-dimensional histogram.Using the global optimization ability of the predator-prey particle swarm,the image segmentation can be accomplished by iterative optimization to obtain the optimal cluster center.Simulation results show that the proposed method can effectively avoid getting into local optimum.With its strong anti-noise capability,high clustering accuracy and good segmentation effect,the presented method is an effective algorithm for image segmentation.