Multiple Wavelet Denoising Methods in MATLAB (Soft/Hard Thresholding, Adaptive Thresholding, etc.)

denoise0701.m tests various denoising methods using color images % ==================================================================== % Denoising programs for various methods================================================== den1.m performs image denoising using semi-soft thresholding method den1_5_1.m improved semi-soft thresholding method applying mean filtering to the first-level reconstructed image den1_9.m improved semi-soft thresholding method replacing linear decay function with exponential function den1_10.m improved semi-soft thresholding method applying wavelet threshold denoising again to the first-level reconstructed image den2.m denoising using improved soft/hard thresholding functions den3.m denoising using generalized thresholding function den4.m denoising using adaptive feature thresholding function wdenoise.m denoising using Donoho threshold or Birge-Massart strategy with soft or hard thresholding functions meanfilter.m mean filtering method for denoising (for comparison) medfilter.m median filtering method for denoising (for comparison) denoise_extra_1.m additional denoising method 1 denoise_extra_2.m additional denoising method 2 denoise_extra_3.m additional denoising method 3 % ==================================================================== % Helper functions============================================================ noise.m adds Gaussian noise or salt-and-pepper noise to images Donoho.m calculates Donoho global threshold Birge-Massart.m calculates threshold using Birge-Massart strategy MSE.m calculates mean squared error for grayscale images MSE_color.m calculates mean squared error for color images PSNR.m calculates peak signal-to-noise ratio for grayscale images PSNR_color.m calculates peak signal-to-noise ratio for color images helper_functions.m additional helper functions % ==================================================================== % Test images========================================================== lena.png grayscale test image used in the program lena_color.png color test image used in the program % ==================================================================== File description: denoise0615.m and denoise0701.m are programs for denoising grayscale and color images respectively, which can be run directly. Due to processing large amounts of data, the runtime is relatively long. Running on single-core CPU platforms may cause the machine to become unresponsive for a period of time. denoise0701.m has an estimated runtime of over 1 minute. The implementation includes wavelet decomposition, threshold application using various strategies, and image reconstruction with performance metrics calculation. Each method implements specific thresholding algorithms with different noise adaptation characteristics and optimization approaches for image quality preservation.