Detecting brain tumors in medical images is a tough nut to crack. These tumors vary widely in shape and texture, making them hard to pinpoint. They originate from different cell types that also shed light on their characteristics, severity, and how rare they are. The tumor's location can hint at what type of cells are forming it, aiding further diagnosis. Common issues like poor lighting in digital photos compound the difficulty of spotting brain tumors. The similarities in brightness between tumor and non-tumor areas add another layer of challenge, as it confuses models trying to distinguish based on the raw images alone. To overcome these limitations, this work introduces a new method for spotting tumors in various brain scans by initially applying certain image processing tricks like opening and histogram equalization and then using a convolutional neural network (CNN). Experimental studies reveal the proposed model provides an impressive recall of 98.55% on the training set, 99.73% on the testing set which is very compelling and, which outclasses other existing frameworks by giving the adequate accurracy of 97.94%.