Study on Gabor Filter-based Handwritten Chinese Character Feature Extraction
手写汉字识别弹性网格Gabor特征提取方法的研究*

New elastic meshing based on Gabor feature extraction for handwritten Chinese character recognition is proposed in this paper. And the effect of different nonlinear normalization methods and different meshing techniques for Gabor feature extraction approach is studied. From experiments, it is found that:The proposed elastic meshing based on Gabor feature is better than conventional directional feature.Elastic meshing Gabor feature is much better and more robust than nonlinear normalization based on Gabor features.Normalization size plays important role in Gabor feature extraction.