There are several methods used for image denoising, depending on the type and characteristics of the noise present in the image. Here are some commonly used methods:

  1. Gaussian and Median Filters: Gaussian filter is a widely used linear filter that reduces Gaussian noise by convolving the image with a kernel. Median filter, on the other hand, replaces each pixel value with the median value in its neighborhood, which is effective in reducing impulsive noise.

  2. Total Variation Denoising: Total variation denoising is a nonlinear denoising method that minimizes the total variation of the image, while preserving edges and details. It is particularly effective in reducing salt-and-pepper noise.

  3. Non-local Means Denoising: Non-local means (NLM) denoising exploits the redundancy in natural images by averaging multiple similar patches from different parts of the image. It effectively reduces Gaussian noise while preserving details and textures.

  4. Wavelet Denoising: Wavelet denoising applies wavelet transforms to decompose the image into different frequency bands and then applies denoising in each band. This method is effective in reducing both Gaussian and impulsive noise.

  5. Bilateral Filtering: Bilateral filter applies a weighted average of nearby pixels, where both spatial and intensity distances are considered. It is useful in preserving edges while reducing noise.

  6. Patch-Based Denoising: Patch-based denoising methods exploit the self-similarity of images, where similar patches are iteratively averaged to reduce noise. These methods are effective in reducing various types of noise, including Gaussian, salt-and-pepper, and Poisson noise.

  7. Sparse Representation Denoising: Sparse representation denoising models the noisy image as a linear combination of a set of basis patches and exploits the sparsity in the representation to estimate the clean image. This method is particularly effective in reducing additive noise.

  It is worth noting that different denoising methods have their strengths and weaknesses, and the choice of method depends on the specific characteristics of the noise and the desired balance between noise reduction and preservation of image details.