Improved Way Of Watershed Based On Marked Using In The Segmentation Of Brain Tumor Image Of MRI

Brain tumor is a disease that harm the health of mankind badly.The symptoms of the disease influence the life of the patient seriously.The location and quickly spreading of the tumor makes a critical problem in the treatment of it,thus,the segmentation has became the hot topic of the research of image in recent years.The algorithm of watershed has so many merits,the most important one is its precision in segmentation,and a closed wire on object as well.At the same time,it is sensitive to the noises formed in the images extremely.As a algorithm sensitive to noises,and be influenced by the noises and the errors in quantification and the texture details in image,many minimum values will be produced in the course of the imaging. Thus,it is so easy to get so many meaningless areas in using it in Magnetic Resonance Imaging directly,and the results can not be used in clinic.This paper discuss the problem of using watershed algorithm in segmentation of brain tumor,and stress the improved watershed algorithm to solve the over-segmentation in using it in brain tumor directly.This paper making use of extraction of brain tissue to carry out the segmentation of watershed method innovatly.It can avoid the uniformity of gray distribution and cerebral-vascular light spots of catastrophe point of gray,avoiding the over-segmentation in the place of joint of skull and brain tissue,and it can get better result in segmentation.