Handwritten Character Recognition Using Principal Component Analysis
Sehn Chun · Mini-micro Systems · 2005
To overcome handwritten character model′s instability caused by different writing styles, a novel approach is proposed in this paper. First, the character image is pre processed by morphological thinning and dilation. This ensures the character strokes to be of similar thickness, and improves character′s local features. Then principal component analysis (PCA) is used to extract character features and estimate character′s reconstruction model. Character recognition is conducted based on reconstructed models′ error analysis. Finally, the algorithm proposed in this paper is tested on all the characters in USPS character database, and the experimental results validate the robustness and accuracy of the proposed algorithm.