Xixia Characters Recognition Based on Neural Networks

摘要

The recognition of Xixia characters has been an area of great interest for many years. There are several major problems with Xixia character recognition: the character size is very large and many structurally similar characters exist in the character set. This article presents a new technique for the recognition of Xixia characters using neural networks. The system was tested with 2000 characters(each character has 10 samples), and the rate of recognition obtained was 89%. This strongly supports the usefulness of the proposed measure for Xixia character recognition.