Fast non-global uniform image shaking blur removal algorithm and system thereof

Abstract

The invention provides a fast non-global uniform image shaking blur removal algorithm and a system thereof. The system comprises a camera fixed on a three-axis rotary table, a projector, a spectroscope and a screen. The method comprises the following steps of: establishing a fast non-global uniform image shaking blur removal system; creating an image blur model, and controlling the three-axis turntable to make servo motion in the exposure time of the camera; correcting the blurred image shot by the camera; and performing iterative computation by use of an image non-blind deblurring algorithm, and calculating and updating a clear image. The fast non-global uniform image shaking blur removal system provided by the invention stimulates the real blurring process in the image acquisition of the camera by controlling the motion of the camera in the exposure time, and obtains a blurred image of a clear image on the given camera motion track within very short time. The fast non-global uniform image shaking blur removal method provided by the invention realizes fast non-blind deblurring of the whole global non-uniform image by use of a three-dimensional rotary motion model.

Description

technical field [0001] The invention relates to the technical field of computational photography, in particular to a fast non-globally consistent image shake and blur removal algorithm and a system thereof. Background technique [0002] Image blur caused by camera shake is an important reason for the degradation of captured image quality. With the development of digital camera equipment and the popularization of digital camera equipment, most of these consumer-grade camera equipment lack effective but very expensive optical anti-shake devices, and consumers who use this type of camera equipment generally do not have good As a result, a large number of pictures are blurred due to careless operation during shooting. [0003] At present, there is a kind of image deblurring method called non-blind image deblurring algorithm. This kind of method refers to the method of estimating the clear image according to the blur kernel and the blur image after estimating the blur kernel of ...

AI Technical Summary

Problems solved by technology

However, due to the huge amount of calculation of the deblurring problem itself, the use of a globally inconsistent fuzzy model makes the FFT-based fast convolution method unable to be applied to the iteration of the deblurring algorithm, resulting in a significant increase in computational complexity, and the calculation time often exceeds what people can achieve. acceptable range

Images