说明：  实现维纳滤波用于语音增强。第一步采用频域谱函数计算最佳的维纳滤波器，再对输入带噪信号进行滤波；第二步是在此基础上的改进，用了自适应的思想，及时根据当前帧的状态更新修正滤波器的系统函数，得到更精确的无噪语音信号。第三步是采用自适应LMS方法，在时域迭代计算滤波器系数，迭代次数足够多时可得到较准确的有用信号。
(Implement wiener filtering used in speech enhancement. First step optimum wiener filter is obtained by using the frequency spectrum function, again to input the signal with noise filtering The second step is the improvement on the basis of this, with the idea of adaptive, and in a timely manner according to the current frame updates correction system function of the filter, get a more accurate without noise speech signal. The third step is to use the adaptive LMS method, iterative calculation in time domain filter coefficient, the number of iterations enough to get the useful signal accurately.)