Image denoising is a frequently encountered problem in image processing, this article through the analysis of the characteristics of denoising algorithm which based on wavelet transform and curvelet transform, focus on the artifacts phenomenon generated by curvelet transform denoising algorithm, and proposes a method based on fast discrete curvelet transform denoising algorithm and the non-sampling wavelet transform denoising algorithm and image penetration technology is applied to image denoising, the experimental results show that the method improves the speed of image processing, should have good practical value. © 2015 SERSC.