An image denoising method according to the present invention includes the steps of: sequentially selecting a pixel in an image as a current pixel; dynamically determining a current search block and a strength parameter; transferring the comparison block of each pixel in the current search block to a frequency domain; determining a current frequency basis; obtaining a similarity between each neighborhood pixel and the current pixel in the current search block according to the current frequency basis; determining a weighting of each neighborhood pixel related to the current pixel according to the strength parameter, and a distance and the current pixel in the current search block; and weighted averaging each neighborhood pixel and the current pixel in the current search block according to the weighting so as to obtain a reconstruction value of the current pixel.