The abnormal growth of brain cells is root cause of brain tumors. Tumors can take many different forms and are rare, making determining the survival rate challenging. Magnetic Resonance (MR) images are crucial for identifying malignancies, but manual detection is time-taking and can lead to errors. Using computer-assisted techniques is crucial for addressing these limits. Computer-assisted techniques are essential for overcoming these constraints. In medical imaging, Deep Learning (DL) models diagnose brain cancers in MR images. In this study, brain tumor images are efficiently classified using InceptionV3, a Convolutional Neural Network (CNN) with addition of extra layers to enhance performance. Applying filters to images is one way to increase their quality. Data augmentation methods are employed to increase sample numbers in order to enhance the training of the proposed model. The proposed fine-tuned InceptionV3 model achieves 96.88% accuracy in detecting and classifying types of brain tumor. The model's potential for accurate, automated brain tumor categorization is demonstrated by its high accuracy, which can help radiologists identify patients more quickly and accurately, ultimately leading to better patient outcomes.