Neural Network OCR.

There are many different approaches to optical character recognition problem. One of the most common and popular approaches is based on neural networks, which can be applied to different tasks, such as pattern recognition, time series prediction, function approximation, clustering, etc.



In this article, I ll try to review some approaches for optical character recognition using artificial neural networks. The attached project is aimed as a research project, so don t try to find here a ready solution for scanned document processing.