论文部分内容阅读
隐Ｍａｒｋｏｖ模型（ＨＭＭ）是在语音识别中得到成功应用的一种统计建模工具．该研究首次将 ＨＭＭ技术引人数字的离散信号滤噪研究并提出了ＨＭＭ综合平均法，经对加噪谐波信号、方波信号、 斜波信号的计算机重建仿真实验后，表明利用ＨＭＭ进行此项工作有着比较好的效果。
Hidden Markov Model (HMM) is a statistical modeling tool successfully used in speech recognition. The study for the first time introduced the HMM technology to digital discrete signal filtering noise research and proposed the HMM integrated averaging method. After the computer simulation of noise-added harmonic signal, square wave signal and ramp signal was simulated, it was shown that using HMM The work has a good effect.