This project is focused on single object recognition and classification. The MNIST database is used as the image dataset. A single layer CNN is developed which consists of a convolutional layer and a pooling layer. The results obtained for this database is test accuracy of around 97.44% which is comparable with the results found in the literature. In the next step, the single layer CNN is expanded to a three-layer CNN for the CIFAR-10 database in order to classify each image to one of the 10 known categories.