Handwriting Digit Recognition Based on Fractal Edge Feature and BP Neural Net

摘要

Handwritten digit recognition (HDR) is one of the difficult research areas on pattern recognition, Hence, evaluation a performance of algorithms on HDR problem is of great importance. In this paper, a method of the recognition of the handwriting digits based on fractal edge feature and BP neural net is proposed. First, extracting the edge feature of the character by using the fractal edge detect method; then, combining the fractal edge feature with three other features (ring zones, projection histograms, moments); finally, taking BP neutral net as the classifier and getting the result. The experimental results show that the proposed method has good anti-noise and high accuracy performance.