Abstract: To alleviate the problem of current two dimensional Qian methods slow running time, a fast implementation of two dimensional Qian methods is proposed. And it is applied to image denoising for the first time. First, the inner product calculation of the two dimensional Qian method is optimized by using two one-dimensional integral to calculate the two-dimensional integral. Then the inner product calculation process is placed into the GPU computing unit. CPU and GPU parallel computing is used to improve running speed. For the strip ripples which are generated after the decomposition and reconstruction of the image by using Qian method, the method of extending the image boundary is used to improve the quality of the image. Finally, two dimensional Qian method to image denoising is applied. Experimental results show that, the running time of the improved scheme of two dimensional Qian method proposed in this paper is significantly improved. Especially when the step size is 0.5, it can be used in real time. So it is possible to solve the practical problems by using the theory of two dimensional Qian method. And it can be seen in the experimental results of image denoising in two dimensional Qian method, the denoising algorithm is simple and can obtain a good visual effect. And the denoised image could get a better effect of removing the noise and preserve image details at the same time. The effect of denoising is ideal.