A brain tumour is an abnormal growth of tissue caused by the rapid division of abnormal cells. It invades the skull and prevents the brain from doing what it's supposed to do. Because the tumour may develop into cancer, it is crucial that it be found while it is still relatively tiny utilising MRI or CT scanned pictures. In this study, we suggest a technique for using magnetic resonance imaging (MRI) scans of patients to identify and localise the precise location of any brain tumours that may already be present. Pre-processing, edge detection, and segmentation are the three steps that make up the suggested technique. A grayscale version of the original picture is created, and any visible or undetected noise is eliminated, all during the pre-processing phase. Next, we apply picture enhancing methods to the results of our edge detection utilising the Sobel, Prewitt, and Canny algorithms. The MRI images of the tumour are then segmented so that the damaged area may be seen clearly. At last, the kmeans algorithm is used to group similar pixels together. With this case, we developed the project in MATLAB 2021a. Glioma brain images provide a difficult challenge for tumour area recognition owing to their low sensitivity border pixels. In this work, we improve upon the original brain scan using Non-Sub sampled Contourlet Transform (NSCT), and then we extract texture characteristics from that improved scan. These characteristics are then used in an ANFIS-based training and classification process to determine whether or not a given brain picture is a Glioma. After that, morphological functions are used to segment the tumour sections in the Glioma brain picture.