Quasar > Image Processing > Restoration > Wavelets

Wavelets


An overview of the different functions and classes:

autocor_to_covmtxConversion from a spatial autocorrelation function to spatial covariance matrix with a local window of the specified size
bkf_priorCalculates the probability density function for the hidden multiplier of Bessel K Form prior for a Gaussian Scale Mixture
compute_autocorParallel computation of a spatial autocorrelation function i.e. assuming spatial stationarity
compute_covmtxParallel computation of a spatial covariance matrix without assuming spatial stationarity
compute_covmtx_spat_stationaryComputation of the covariance matrix under the assumption of spatial stationarity
correct_eigenvaluesMake: sure the eigenvalues of the symmetric matrix are positive. All negative eigenvalues are replaced by a small positive constant
denoise_band_blsgsmDenoising of a wavelet subband using the BLS-GSM method
denoise_band_vectorprobshrinkDenoising of a wavelet subband using the vector-Probshrink method.
denoise_image_blsgsmDenoises an image using the method of BLS-GSM. Both grayscale and RGB color images are supported.
denoise_image_vectorprobshrinkDenoises an image using the vector Probshrink method. Both grayscale and RGB color images are supported.
denoise_subband_bayesshrinkDenoising of a subband using Bayesshrink. This method is just an educational example of how wavelet soft thresholding may work. Note that this estimator is actually the MAP estimator for a wavelet subband, assuming Laplacian distributed noise-free wavelet coefficients with additive stationary Gaussian noise.
jeffreys_priorCalculates the probability density function for the hidden multiplier of Jeffrey's non-informative prior for a Gaussian Scale Mixture
symsqrt_and_invComputation of the symmetric square root of a matrix and its inverse