SAR Image Segmentation With Parallel MCMC Algorithm

The scheme of changing the pixels' label variable values which are used to denoting the pixels' class properties successively is usually used,when using MCMC(Markov Chain Monte Carlo)algorithm to simulate the image segmentation model based on statistics.When using the serial MCMC scheme,it usually leads to routine time-consuming,slow speed of convergence.Therefore,a parallel MCMC scheme based on multiple threads updating multiple pixels' labels is proposed in this paper,in order to reduce the running time and enhance the efficiency.In the framework of Bayesian inference,SAR(Synthetic Aperture Radar)image segmentation model is set up on the basis of Gaussian distribution and MRF(Markov Random Field)model.A parallel sampling scheme based on multithreading is designed,in order to calculate the optimum SAR image segmentation quickly.An independent pixel parallel sampling criterion is proposed so as to solve the label correlation problem between the neighborhood pixels' labels in MRF label field;At the same time,the number of parallel threads is limited to ensure the sampling randomness.To verify the proposed parallel MCMC scheme,testing is carried out with real and simulated SAR images respectively by the serial algorithm and the proposed method.The results show that the proposed method behaves a significant reduction of running time without affecting the accuracy of segmentation.