Image denoising algorithm based on Gaussian-pepper noise

As the imaging process can be affected by various external factors, resulting in images containing random noise. Therefore, this paper proposes an image denoising algorithm based on Gaussian-Pepper noise. Firstly, the image is decomposed into high-frequency components and low-frequency components using wavelet transform. Secondly, an improved 3D block matching denoising algorithm is used for the high-frequency component of the image, and an algorithm of bilateral filtering combined with median filtering is used for the low-frequency component. Finally, the processed high-frequency components and low-frequency components are reconstructed to obtain the filtered image. Experiments show that the proposed algorithm is significantly better than some of the current better denoising algorithms in terms of mixed noise image restoration, while retaining the image edge and texture detail features.