Handwritten Digit Recognition Using Rotations

Handwritten digit recognition is a subproblem of the well-known optical recognition topic. In this work, we propose a new feature extraction method for offline handwritten digit recognition. The method combines basic image processing techniques such as rotations and edge filtering in order to extract digit characteristics. As classifiers, we use k-NN (k Nearest Neighbor) and Support Vector Machines (SVM). The methods are tested on a commonly employed database of handwritten digits' images, MNIST (Mixed National Institute of Standards and Technology) on which the classification rate is over 99%.