摘要： 介绍了自动标绘系统的需求和系统流程，并对语音采集、语音识别和自动标绘3 个阶段的技术难点及解决方案进行了详细介绍。重点介绍了采用语音自动重叠技术以减少语音切割失误，提高识别精度。描述了对识别出的数字串运用打分系统进行智能切分，提取数据点以提高标绘精度。系统采用VC.NET 和SQL Server 实现，稳定环境下语音识别精度达到97%以上，标绘正确率达到95%，较好地实现了系统需求。

关键词: 语音识别；隐马尔可夫模型；自动标绘

Abstract: This paper introduces the requirement and system flow of the auto-plotting system, and describes the technique difficulties and solution in three phases, speech gathering, speech recognition and auto plotting. First, adopting the speech auto overlapping technique to decrease wrong separating rate and increase recognition accuracy is presented. Second, by using scoring system, separating speech intelligently, picking up data points to improve plotting accuracy is presented. The system is realized with VC.Net and SQL Server, the recognition accuracy can reach 97%, the plotting accuracy can reach 95%, and it meets the system requirements well

Key words: Speech recognition; Hidden Markov model(HMM); Auto-plot