The major problems that any lip image analysis method is faced are: low chromatic in lip region, low contrast luminance and overlap between the lip and facial skin color. So we are interested on those methods that can extract the lip region automatically and accuracy. In this paper, we present a new automatic approach for the lip extraction based on using the Haar wavelet. The proposed algorithm uses the color space CIE L*U*V* and CIE L*a*b* in order to improve the contrast between the lip and the other face regions. We execute the algorithm on the images that belong to the four different speakers. While the two basic method, it means the fuzzy c-means clustering (FCM) and the spatial fuzzy c-means clustering (SFCM), could not segment the lip for the all four tested image perfectly, the proposed algorithm based on using the wavelet shows the better performance.