Restoration of Blind Image based on FPNMF

By using NMF to separate the blind images, a new method of blind image restoration, Fixed Point NMF (FPNMF), was proposed. This technique avoids the prior knowledge of Point Spread Function (PSF) used in the processing of the classical linear or non-linear deconvolution. It features identifications of the real images is implemented by the observed images directly and also as in the blind source separation (BSS). By processing the natural gray images in the simulation, the results show the efficiency and robustness of this approach we made. It has to be noted that this method developed in the paper takes on some functions of the identification in facial positions and expressions.