基于MHMM的脱机手写体字符识别

Handwritten character recognition based on MHMM

对隐马尔可夫模型(HMM)的训练方法及模型参数的选取进行了探讨,并将HMM理论用于脱机手写体识别中,建立了一种基于字符投影变换图像的边界链码特征的多重隐马尔可夫模型(MHMM).实验结果表明,该方法是可行的,且具有良好的兼容性和灵活性,可应用于手写体字符的自动识别中.

The HMM （hidden Markov model） is applied to handwritten character recognition,the training methods and the parameter selection of the HMM are studied in this paper. A multiple hidden Markov model （MHMM） based On the characteristic of boundary chain code of character projection transformation image is created. Experimental results show that this method has good compatibility and flexibility, and it can be applied to handwritten character recognition.