Recognition of Scanned Handwritten digits using Deep Learning

The capability of ML (Machine Learning) algorithms to recognize images of handwritten numerals is known as HDR (Handwritten Digit Recognition). Because handwritten numerals are imperfect and can be generated with a variety of flavors, it is a difficult work for the machine. The answer to this issue is handwritten digit recognition, which uses an image of a digit to identify the digit that is contained in the image. Additionally, there may be various visual artifacts including intensity changes, blurred and noisy effects makes this process more difficult. A CNN (Convolutional Neural Network)-based deep learning technique for HDR is proposed to get beyond the aforementioned restrictions. The input photos are initially transformed to a grayscale image for the area of interest using a pixel resolution ratio of 0-255. For classification of written digits is trained for recognition. The proposed model has been tested using MNIST dataset to show its effectiveness, and it achieved higher accuracy value of 99%. The suggested method can recognize HDR images with diverse writing styles and variations of sizes numbers, with prediction in plots and predicted digits in results. Tensorflow with Keras module of Python was imported for training and implementation. Scaling the training data it gives 98.7% of accuracy.