SPATIALLY ADAPTIVE WAVELET-BASED MULTISCALE IMAGE-RESTORATION

In this paper, we present a new spatially adaptive approach to the res toration of noisy blurred images, which is particularly effective at p roducing sharp deconvolution while suppressing the noise in the flat r egions of an image, This is accomplished through a multiscale Kalman s moothing filter applied to a prefiltered observed image in the discret e, separable, 2-D wavelet domain, The prefiltering step involves const rained least-squares filtering based on optimal choices for the regula rization parameter, This leads to a reduction in the support of the re quired state vectors of the multiscale restoration filter in the wavel et domain and improvement in the computational efficiency of the multi scale filter, The proposed method has the benefit that the majority of the regularization, or noise suppression, of the restoration is accom plished by the efficient multiscale filtering of wavelet detail coeffi cients ordered on quadtrees, Not only does this lead to potential para llel implementation schemes, but it permits adaptivity to the local ed ge information in the image, In particular, this method changes filter parameters depending on scale, local signal-to-noise ratio (SNR), and orientation, Because the wavelet detail coefficients are a manifestat ion of the multiscale edge information in an image, this algorithm may be viewed as an ''edge-adaptive'' multiscale restoration approach.