A model structure for hand-writing recognition is presented based on intelligent control and characteristic model. The complex task can be decomposed into 2 levels: Decision-making.Execution level. Decisions are made to choose the recognized method according to the hand-writing characteristic mode. The correction of recognition result can be gained through the definition of the general characters recognition error. The structure is proved to be efficient through experiment.%从智能控制与特征建模方法入手提出一种手写体汉字识别系统结构模型,仿人思维模式将复杂任务系统分解为决策、执行两层.综合目前常用的手写体汉字识别方法特点,针对不同特征汉字采用合适的识别方法,并通过广义误差对识别结果进行反馈,达到校正的目的.实验证明该结构方法有效.
展开▼