Over a past decade many people are suffering from Brain tumour which leads to death because of not providing treatment at right time. Early detection of Brain tumour and providing proper treatment will improve the survival rate of the patient which can be possible with the radiologist. Tumour will arise due to aging of cells in the brain which are not destroyed in proper time will leads to Brain tumour. One of the Major technique to visualise this brain tumour by using Magnetic resonance imaging in order to identify in which area of the brain is located, where its shape and size is also identified. It can be analysed with the help of machine learning, deep learning and Convolution neural network, in this paper we are using AASSD Algorithm for segmentation and fully CI dense net to classify Brain tumour. Here we are using 3064 T1 MRI images of 233 patients where 60% of this images are going for training and remaining 40% for testing. Here we are comparing different segmentation algorithms with different techniques in order to provide better Accuracy of 96%, specificity of 96%, sensitivity of 96%, precision of 93%, F1 score of 0.94.