Brain tumor has become a primary hassle among living humans amongst in which gliomas is the maximum common difficulty. So, treatment is done through MRI. Magnetic Resonance Imaging is the most used technique to asses those tumors, wherein MRI restricts manual segmentation in a completely low-priced time, restricting using precise quantitative measurements within the scientific observe. So, a dependable segmentation vicinity unit is required. However, the structural changes amongst mind tumors create segmentation a difficult downside. This paper has a tendency to suggest partner degree automatic segmentation methodology and classification primarily based on Artificial Neural Networks. Using GLCM and shape functions make the proposed paintings extra strong for classification features.