A high-speed moving image deblurring method based on gray-scale probability consistency of image columns

Abstract

A high-speed motion image deblurring method based on the consistency of image column grayscale probability, including the following steps: (1) modeling the camera forward motion blur process; (2) spatial transformation to reduce the spatial dimension; (3) based on column grayscale consistent image deblurring. The present invention has a better effect on the recovery process of the image blur caused by the one-way movement of the camera perpendicular to the optical axis, especially the long blur problem caused by the high-speed movement of the high-definition camera, and the space transformation greatly reduces the degree of freedom for solving the problem , the dual goals make the recovery process preserve a certain degree of difference and overall consistency.

Description

technical field [0001] The invention relates to the fields of image processing and probability statistics calculation, in particular to a high-speed moving image deblurring method based on the consistency of gray scale probability of image columns. Background technique [0002] Image acquisition has become an important source of information in the information age, but in the process of machine vision acquisition, due to the high-speed movement of the camera relative to the acquisition target in the long exposure time or the acquisition state of relatively weak light, it may cause the acquisition of The image appears blurred. There are many reasons for blurring. There are many types of image blurring, such as motion blur, out-of-focus blur, and Gaussian blur. The more traditional image deblurring methods include RL deconvolution, Wiener filtering, regularization methods, and other image blindness methods. The method of deconvolution. Most of these methods are aimed at diffe...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to solve the image blurring problem of camera fast forward motion acquisition, based on the probabilistic and statistical characteristics of the co-existence of gray scale distribution consistency and difference in natural image columns, a method for removing motion blur is proposed, which can overcome the existing algorithm Can not handle high-resolution images and their shortcomings of severe motion blur, improve the restoration quality of moving images

Images