We show how a deconvolution can be performed from the significant structures of an image in the wavelet space. These significant structures are computed from the wavelet coefficients of the data. The wavelettransform algorithm that is used is a new one based on the fast Fourier transform. This approach is first studied for interferometric images, simulations are done, and then the method is generalized to any kind of data.
© 1994 Optical Society of America
Jean-Luc Starck and Albert Bijaoui, "Multiresolution deconvolution," J. Opt. Soc. Am. A 11, 1580-1588 (1994)