Handwriting recognition using supervised neural networks
B.V.S. Murthy · 2003
We proposed a method for person authentication based on his/her handwriting. We used multilayer feedforward neural network with backpropagation learning for the task. The method is based on the observation that there exists a relationship between the heights and widths of the alphabets written by an individual which is unique and specific to him/her. For classification 100% accuracy in person authentication was achieved using a database consisting of 10 people.