Aiming at the problem of photographic motion-blurred image restoration,the characteristic differences between photographic image and simulated blurred image are analyzed.A method for deterioration parameter estimation of photographic image is proposed.Firstly,we calculate the cepstrum of degraded image,choose the absolute gray value of the minimum cepstrum to be a threshold,and transform the cepstrum to a binary image.And then remove the bright cross of the center.Secondly,the distance from a point to a straight line is used to calculate the bright stripe direction in the binary image (the direction of motion blur).Finally,the center of degradation image is taken as a rotation axis,the direction of motion blur is rotated to the horizontal direction,then the blur scale is calculated by the differential autocorrelation method.The estimated point spread function (PSF) is put into the Wiener filter algorithm to recovery the photographic images.The result shows that estimating parameters are correct.