中文说明：实现了对10个数字音的识别程序ML法能够很好的估计信号的信噪比,包含CV、CA、Single、当前、恒转弯速率、转弯模型，本程序的性能已经超过其他算法，D-S证据理论数据融合，多抽样率信号处理。
English Description：
Realize 10 digital voice recognition program ML method can be a good estimation of the signal to noise ratio, including CV, CA, Single, current, constant turn rate, turned model, this performance more than other algorithms, the d-s evidence theory data fusion, Multirate signal processing.