Since people's writing styles vary in form and orientation, handwritten digit identification is a difficult job. Two steps are essential for the production of effective handwritten digit recognition. The extraction of discriminating features from handwritten images is the first step, and the classification of new digit images is the second. Convolutional Neural Networks (CNN) are at the heart of spectacular developments in deep learning that combine Artificial Neural Networks (ANN) and cutting-edge deep learning techniques. Pattern recognition, sentence classification, speech recognition, face recognition, text categorization, document interpretation, scene recognition, and handwritten digit recognition are only a few of the applications. Our project's aim is to improve detection accuracy by expanding the model depth. This is accomplished by increasing the number of layers and the number of filters. The Modified National Institute of Standards and Technology (MNIST) datasets were used for this performance assessment using CNN.