Denoising algorithm based on bilateral filtering and improved wavelet threshold function

In order to effectively remove the Gaussian noise in images while preserving the detailed information in images, a novel algorithm combines bilateral filter and image's method noise thresholding by using wavelet transform is proposed. Method noise can be obtained after subtracting bilateral filtering image from noisy image. And the method noise which is considered to be the sum of noise and image's details will be transferred to the wavelet domain. The improved threshold equation is used to raise the high frequency coefficient as a whole, and it is easy to separate the Gaussian noise component and retain effective information. After inverse wavelet transform, details can be acquired from the information. By adding the resulting details to the denoised image, we obtain the final image. The experimental results show that the proposed method can effectively remove noise and has rich details, which is visually superior to other existing filtering algorithms.