Image denoising based on wavelet transform has become one of the main image denoising methods. At the process of image denoising, the traditional method based on wavelet transform doesn’t take into account the edges in image so that the information of the edges in image has a various degree of loss. That is to say, the phenomenon which the edges in denosing image are dark exists in the effects of denoising image. In this paper, considering the shortage of the traditional method, we propose an image denoising method that based on image edge protection. First of all, we get the edges of the noisy image through an advanced method in order to protect the coefficients of the image edges from using threshold to denoise; secondly, under the premise of edge protection, we use the traditional method to denoise; finally, we use edge protection to reconstruct image which is disposed by the wavelet transform. The result of the experiment shows that the proposed method has better denoising effects and higher PSNR than the traditional ones.