语音音质客观评价系统的实现  

System Implementation for Objective Assessment of Speech Quality

作　　者：贾冲[1] 孙浩海[1] 吴其前[1] 张雄伟[1] 

机构地区：[1]解放军理工大学通信工程学院,江苏南京210007

出　　处：《解放军理工大学学报（自然科学版）》2004年第3期33-37,共5页Journal of PLA University of Science and Technology(Natural Science Edition)

摘　　要：语音质量评价始终是语音处理系统研发中极为关心和重视的问题。要对日益庞杂的语音系统进行测评 ,研制灵活、方便、快捷、可靠的语音质量评价系统极其必要。通过介绍语音音质客观评价方法的基本原理 ,分析 ITU- T最新建议标准 P.862中的 PESQ算法 ,最终提出了基于 PESQ算法的语音音质客观评价系统的软硬件实现方案 ,并重点介绍了系统各部分的设计思想。As the speech system is more and more complex, the quality tests are becoming more difficult, so there is a great demand for objective quality measurement methods.In this paper, the research has been done on theory and realization of objective speech quality evaluation. The importance of objective test is explained, and an overview of the components that make up the ITU-T PESQ model is presented. In addition, the applications of P.862 and the system design scheme of objective assessment are introduced.

关 键 词：语音音质 客观评价 PESQ算法 系统实现 

分 类 号：TN912.3[电子电信—通信与信息系统]