Aiming at the scaling of the noise after muhiwavelet decomposition and the characteristic of image, a multiwavelet adaptive denoising method was proposed based on genetic algorithm. This method can adaptivly look for the least RMSE of the denoised image according to genetic algorithm. The experimental results show that this algorithm is superior to the traditional methods. It can not only remove the noise of image, hut remain the better edge of image, and this method has more perfect denoising effects.