Segmentation Method And Research Of Medical Image Fusion Based On Markov Random Field,Fuzzy Theory And D-S Theory

Medical image segmentation plays an important role in the research and application of clinical medicine.By means of the medical image segmentation,So that clinicians can more directly,more clearly,more convenient to treat the disease.Because the medical image is affected by the noise,the algorithm,the equipment,the display technology and so on,Resulting in medical images with ambiguity.In order to solve the problems of image segmentation accuracy and segmentation efficiency in medical image processing,using MRF theory,GRF theory,knowledge,fuzzy theory,D-S theory and other fusion algorithms for three-dimensional image segmentation,Achieve the best results of medical image segmentation.Medical image segmentation is the process of image extraction,segmentation,restoration,reconstruction,fusion,recognition and so on.Comprehensive use of multiple imaging or a variety of imaging equipment graphics information,limitations of data loss,local data imprecision and ambiguity,using the MRF model to make the image segmentation more accurate.Analysis of the specific application effect in the brain NMRI image segmentation.Make use of the fuzzy clustering segmentation algorithm.Using grayscale information and adding spatial image information.No need to manually set parameters,NDFCM method achieve a better image segmentation effect than FCM method,with stronger anti noise perfornance and high medical value.The use of the latest FCM classification of MRF complex to obtain a higher accuracy of the brain tissue image,then,Markov and fuzzy clustering are used to segment the brain image,which is regarded as the basic probability of D-S evidence theory.Finally,the D-S theory is used to fuse the brain tissue and improve the speed and quality of medical image segmentation.Adopt the latest image fuzzy C mean(FCM)method is used to segment medical images more accurately.Innovative and efficient two-dimensional distance measurement technology.Utilising the corresponding two-dimensional histogram to define neighborhood correlation.This paper presents the idea that the center of the cluster centers on the same value of the pixel value and the neighborhood pixel value,Achieved Two dimensional FCM segmentation algorithm based on neighborhood space image information.When the D-S theory is used to fuse two or more image information,we use the existing multi touch model to solve the problem of data loss,partial data inaccuracy,fuzzy or unclear factors.According to the image fuzzy C mean(FCM)theory and MRF theory.According to the conditional probability obeys the Gauss signal distribution and prior probability.Seting the probability value M.Take advantage of the D-S theory is used to fuse several kinds of data,and the D-S decision criterion is used to make decision segmentation,which improves the accuracy of image segmentation.Medical image segmentation methods are mainly based on intelligent,high precision,complexity,anti noise ability,high speed,robustness and adaptive ability.The development trend of the traditional medical image segmentation technology is the combination of traditional segmentation technology and the latest advanced segmentation technology.