The goal of the present study is the restoration in real space of degraded or blurred images. The present method requires a pre-knowledge of the degradation point spread function PSF. This PSF is estimated by curve fitting to a sample values of the amount of blur present in some gray values of the degraded image. After that the coefficients of the polynomial which represent the PSF curve is calculated Before working toward this goal, we have first blurred an unblurred image by a known to us PSF. This degradation PSF estimated by curve fitting to a sample values of the amount of blur required. The rest of the procedure is identical to that presented in the above paragraph. In the present study the arithmetic mean is calculated only for the nonnoisy pixels within the mask. The suggested procedure is to calculate a threshold value T from the variance of the image then calculate the absolute differences D. , where i = 1,2,….., n2-1, where nxn is the mask, size, in the gray levels between the central pixel and each other pixel within the same mask. This is followed by calculating an average value from the summation of the gray values of only the pixels with D; > T. The last step in this procedure is to replace the central! gray value by this average. This described procedure is somehow similar to the procedures followed in ordered statistics, where the gray values are arranged in ascending or descending order. After this ordered arrangement, the extreme gray values are excluded from the averaging operation.