Study And Cross-platform Implementation Of Speech Enhancement Algorithm

In the wireless short-wave communication systems, the short-wave signal is corrupted acoustically by noise during the transmission. And the quality of the signal in the receive part is affected in different degrees. Intelligibility of the voice signal will be significantly reduced as the impact of the noise. In the context, in order to improve the intelligibility of the signal, we add an speech enhancement system at the receiver part of the communication system. We introduce the real-time speech enhancement system based on Qt, a cross-platform develop tool, in this paper including the algorithm research and the implementation.The major job of this paper is mainly divided into two parts.In the first part of this paper we introduce the characteristics of the speech signal and noise, and then describe the Power Spectrum Subtraction and MMSE algorithm based on short-time spectrum. In the end of the first part, the function of the two algorithm are given based on the Matlab simulation results. And then we assess the performance of this two algorithms.The second major job of the design and implementation of the system include the design of the system’s diagram. According to this diagram, firstly, we introduce the specification of voice card, the signal generator of the system, and show how to use this card. Secondly, we state the basic feature of the Qt and it development tool Qt-creator. Thirdly, we introduce the Linux operating system and it sound architecture. ALSA and OSS. and we also talk about the concept of preemptive kernel. Finally, the data structures and multi-threaded tech using in the system is present in this paper. A picture shot of the application is also present as the end of the whole paper.