The brain tumor segmentation on MRI images is a very difficult and important task which is used in surgical and medical planning and assessments. If experts do the segmentation manually with their own medical knowledge, it will be time-consuming. Therefore, researchers propose methods and systems which can do the segmentation automatically and without any interference. In this article, an unsupervised automatic method for brain tumor segmentation on MRI images is presented. In this method, at first in the pre-processing level, the extra parts which are outside the skull and don't have any helpful information are removed and then anisotropic diffusion filter with 8-connected neighborhood is applied to the MRI images to remove noise. By applying the fast bounding box(FBB) algorithm, the tumor area is displayed on the MRI image with a bounding box and the central part is selected as sample points for training of a One Class SVM classifier. A database is also provided by the Zanjan MRI Center. The MRI images are related to 10 patients who have brain tumor. 100 T2-weighted MRI images are used in this study. Experimental results show the high precision and dependability of the proposed algorithm. The results are also highly helpful for specialists and radiologists to easily estimate the size and position of a tumor.