Design And Implementation For Objective Assessment System Of Speech Quality

With the development of voice and audio compress coding technology recent years, It has been an important question for discussion that studying how to estimate the performance of Speech Quality Test system.From the assessment main body, speech quality assessment has two modes: subjective assessment and objective assessment. But it is very difficult to estimate the more and more complicated system and the workload is very huge. So, to develop a speech evaluation system, that is flexible, convenient and reliable, is essential. Objective evaluation system can test the voice quality depending on the apparatus only and apply to different time and positions, the results can be compared directly. It has broad applied foreground.This thesis has introduced the fundamental theory about voice objective assessment, 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. The work of the thesis includes the following several parts:(1) to introduce the core model PESQ and core chip TMS320VC33;(2) to design the system according to the requirement, including the total structure , working modes, module functions and flow chart.(3) data sampling: including preamplifier, pre_filter, post_amplifier and CODEC;(4) data processing: including DSP core, inside modules and the data transmission between DSP and CODEC;(5) data storage: including no-waiting external storage SRAM and FLASH to storage code and code book;(6) other problems about system designation, including power system, scheme, PCB, control circuit and voltage matching.