Research On Restoration Algorithm Of Motion Blurred Image In Low Light Environment

The quality of the captured image depends not only on the quality of the camera,but also on the external environment.The contrast of the image under low light conditions is poor,and the details are not obvious enough.In addition,if there is a relatively fast movement between the lens and the imaging target during the imaging exposure period,the moving image will be blurred.Motion blur not only affects the visual effect of the whole image,but also causes the loss of its important details.It also seriously affects the further analysis of the image.Therefore,the fuzzy image restoration algorithm is one of the hot issues in the field of image processing and has important Scientific significance and research value.At present,the method of motion blur parameter estimation mainly includes airspace method and frequency domain method.In the frequency domain method,orthogonal crossing of noise and frequency domain is an insurmountable problem.However,the space domain method is prone to errors when estimating parameters,and the amount of calculation is very large.To solve this problem,this paper first proposes a multi-scale image enhancement algorithm based on PLIP model to stretch the overall contrast and edge information of motion-blurred images in a low-light environment.Compared with the commonly used enhancement methods such as grayscale correction and image sharpening,on the one hand,the PLIP model enhancement is more in line with the nonlinear characteristics of the image operation and the saturation curve of the human eye;on the other hand,multi-scale image enhancement The method can well suppress the image noise under weak light,so after the processing,the image becomes brighter overall,the contrast is improved,the detail information is more obvious,butthe motion blur still exists.Secondly,based on the Minimum Directional Differential Method(MDD)in the spatial domain,a new wavelet analysis method and improved spatial domain direction differential method are proposed to estimate the fuzzy parameters.Firstly,the enhanced image is decomposed by wavelet;because the fuzzy information of the image is mostly concentrated in the low-frequency coefficients,the low-frequency coefficient image can completely reflect the original image motion blur information,and then the multi-scale linear weighted coefficients are used to estimate the spatial domain direction differential method.A fuzzy direction angle is used,and the motion blur length is estimated using a differential autocorrelation curve method.The simulation results show that the proposed method outperforms other methods in estimating the fuzzy parameters.Compared with the frequency domain method,this method avoids the complicated frequency domain transformation and reduces the computational complexity compared with the spatial domain.Finally,in order to verify the effectiveness of the algorithm,a true uniform motion camera is used to obtain images under low light conditions.After image enhancement and motion blur parameter estimation,the L-R algorithm was used to recover the blurred motion images in low-light environment and achieve a good recovery effect.