为了提高联机手写维吾尔文字母的正确识别率,根据维吾尔文字母的手写特点,提出了中心距离特征CDF(Center Distance Feature)、并基于CDF进行了一系列识别实验。在实验中,该文采集了400个人的手写字母样本,利用CDF的三种不同的实施方案(CDF-2,CDF-4,CDF-8)分别对维吾尔文字的32个母独立形态和128个所有形态进行了识别实验,并对实验结果进行了分析。实验结果表明,CDF是一种非常适合于维吾尔文字母识别的特征,有待于进一步改进和优化。
In order to improve the on-line handwritten recognition accuracy, according to the uniqueness of Uyghur characters , a series of experiments were made based on Center Distance Feature. Handwritten samples collected from 400 different volun-teers participated into the experiments for 32 isolated forms and all 128 forms of Uyghur characters using three implementations of Center Distance Feature which are CDF-2, CDF-4 and CDF-8. Some analysis on the experiment results were given, too. Experiment results showed that Center Distance Feature is very suitable for Uyghur characters and further modification on this method is of highly worth.