Study of rotary motion-blurred image restoration algorithm

The imaging sensor, which is set on a rotating plat, rotates with the plat at high speed. As a result, the images is be blurred badly. According to the space-variant of the rotary image, the rotary trails are extracted and the rotary length of each trail is gained. The two-dimensional image restoration is transformed into one-dimensional signal recovery. Two restoration methods are proposed. The first method is based on the Weiner filter, and the second methods is based on the diagonal loading arithmetic. Compared to the previous method, such as the minimum-square method and the logarithmic polar coordinates method, the two methods make better performances with little calculating time.