摘要：Relative movement between the objects and the camera may leads to the occurrence of motion blur. Restoration of the motion blurred image plays an important role in such areas as astronomy and military. The Point Spread Functions (PSF) of blurred images must be known in order that classic restoration algorithms can apply to image restoration. Since the point spread functions of the blurred images shaped during uniform linear motion is determined by the direction and the displacement of motion, this study proposes a method in combination with Bresenham algorithm for straight line to sum up the gray values of the spectrogram of blurred images via twice Fourier transform and to determine the direction of motion. The experiments demonstrate the algorithm proposed in this study can make relatively accurate judgment on the direction of motion of blurred images in a simple and fast way, especially with higher accuracy of judgment for 45°, 90° and more special angles. Moreover, the proposed algorithm can process blurred images with different height-width scales. Having the aid of the proposed algorithm, the restoration effect of the blurred image is satisfying.
