Wavelet image denoising has been well acknowledged as an important method of image denoising. Recently, most of the research efforts about wavelet image denoising focus on how to select the threshold. Edge information is the most primary high frequency information of an image. Therefore maintenance of more edge information is more important in the denoising process. According to the idea of this paper, an image denoising method based on wavelet transform and mathematic morphology is proposed. The coefficients of wavelet transform of image are manipulated by morphological operator, so as to remove the noise whose support are small area or no area at all and preserve the edge and the small features whose support are large or consecutive area. The experimental results show that, compared with the usually used wavelet threshold denoising method and median filtered method, the proposed method can keep images edges from damaging and increase PSNR to 3～5 dB and SNR to 6～10 dB respectively.