基于共振峰谐波特征和支持向量机的VDR人声检测方法  

Speech detection method for VDR based on formant-consonance characteristic and SVM

作　　者：李正友[1,2] 李天伟[1] 黄谦[1] 

机构地区：[1]大连舰艇学院航海系 [2]中国人民解放军69029部队

出　　处：《舰船科学技术》2013年第2期99-102,共4页Ship Science and Technology

摘　　要：针对VDR人声检测的特殊要求,从语音的产生机理出发,结合驾驶室环境下各种声音特点,提出了基于共振峰谐波特征和支持向量机的人声检测方法。首先采用线性预测分析方法提取共振峰频率,再从线性预测分析的残差信号中提取低频段谐波个数和谐波频率间隔的方差,作为人声检测的主要特征参数,然后用支持向量机分类器进行人声检测,以在驾驶室内实际采集的多种语音和非语音构成语料库进行实验,得到了较高的检测正确率。Speech detection for voyage data recorder （VDR） has some special requires. After the analysis of speech generation process and the character of all kinds of sounds in steer house on ships, a method based on formant-consonanee characteristic and support vector machines（SVM） was developed. First formant frequency was obtained by linear prediction analysis （LPA） , and the number of eonsonance and standard deviation of consonance frequency intervals were obtained from residual error of LPA. These were the main characters for speech detection. The SVM was chosen to classify the speech and non-speech. We test the method in experiments by the speech and non-speech signals collected in steer house, and high accuracy rate of detection was obtained.

关 键 词：共振峰 谐波 支持向量机 VDR 人声检测 

分 类 号：U664.7[交通运输工程—船舶及航道工程]