Handwritten Digit Character Recognition by Model Reconstruction Based on Independent Component Analysis

To overcome handwritten character pattern's instability caused by different writing styles, a novel approach is proposed in this paper. By the approach, firstly, Independent Component Analysis (ICA) is used to extract character features and reconstruct character models. And character recognition is then conducted based on the error analysis of reconstructed models'. The proposed algorithm is tested on the entire USPS character database, and the experimental results validate the robustness and accuracy of the proposed algorithm.