FCM Algorithm Based on Improved Automatic Segmentation of MRI Brain Images

Traditional fuzzy C means is widely used in image segmentation. It is a classic method of fuzzy clustering analysis, but the FCM algorithm requires logarithmic or using non-parametric model of the partial field for MR data. Also we can use a continuous spatial distribution of brain tissuepriori constraint model. The computational algorithm have the problems of large amount of calculation, complex parameter estimation, segmentation of noise-sensitive and sensitive to the shortcomings of the initialization algorithm. So we put forward an improved FCM algorithm, which use the bias field parameter model and neighbor to simultaneously complete the pixel domain constraintsimage segmentation and bias field estimates. The experiment can show that the improved algorithm has a good segmentation, while robust to noise.