Research On Scene Image Rotation Motion Deblurring

The camera in the process of taking photos which often be blurredbecause of some accidental factors, these factors include relative motionbetween the camera and photograph objects, like camera shock, opticaldiffraction, scattering, defocusing, atmospheric disturbances, and so on.These factors make some useful information be covered in the image.Therefore, in recent years, Image restoration becomes one of the mostpopular fields in digital image processing. Blurred images mainly bediscussed in this paper are the linear motion blur and rotational motionblur, linear motion blur mainly reflects in linear motion between thecamera and shooting objects, mobile vehicles, such as camera rotationmotion blur is mainly used in airborne, missile, and so on.An algorithm is proposed in this paper, which is based on the R-Literative algorithm optimization of one dimensional linear to rotarymotion to fuzzy algorithm and fuzzy algorithm.Due to the actualparameters of motion blurred image is not controllable, this paperconducted a series of simulation experiments, first, use the simulationimage validation of the proposed algorithm, as research material, after thecorrectness of the algorithm is proved, and the actual image restoring.The key of rotation motion deblur is to determine rotation Angle andthe point spread function, in this paper, hypothesis that the Angle is known to solve the point spread function, on this basis, discussed severalthe possibility of recovery algorithm, finally selected R-L algorithm, thispaper proposes a rotation motion to the deblur method, change thevariable blur into space-variable blur, using a d deblur algorithm to polarcoordinates conversion, then change back to the coordinate system,achieve the goal of image restoration.