Speech enhancement is used to improve the quality of the speech by reducing and removing the unwanted noise. The ultimate goal of this work is to improve the quality of the speech by increasing the clearness of the voice by reducing the background noise for the purpose of feature extraction. This paper dealing with few existing algorithms and novel proposed approach for noise filtering. Here the noise enhancement is done by two levels, with High Band Pass Filtering and Winner Filtering algorithms. Results are shown feature extraction after two way filtering produces better results than one way filtering approach. © 2017 IEEE.