Motion Blurred Image Restoration In Low Illumination Scenes

In recent years,there are two aspects of degradation in the images taken outdoors.On the one hand,the resulting images are dark due to insufficient ambient light.On the other hand,relative motion occurs between the photographing device and the scene,which may cause the photographed image to produce a blur effect,reduce the image quality,and at the same time make the detail of the photographed image unclear.This article will deal with the low-light scene motion blurred image processing,so that the image clarity.The research in this paper mainly focuses on the following two parts.This paper mainly focuses on the field of color fusion,and proposes an improved low-light intensity enhancement algorithm based on the principle of joint color filtering as a pre-processing process of motion blurred images in low illumination scenes.The use of HSV color space to produce twelve solid color standard,plus a total of 14 kinds of black and white colors,the V value is lowered to produce a set of test color,the test color and the solid color number of several fusion found more and more color The result is closer to the R,RG,G,GB,B,and BR channels.The color is refined and brightened.This principle is used to enhance the low-light image processing.In the process of motion blur image restoration,this paper improves the estimation algorithm of PSF function parameters,and proposes an improved PSF parameter estimation algorithm based on Radon transform for motion blurred images.Through the morphological closed operation of the spectrogram of the blurred image,the interference of the cross fringes in the spectrogram on the parameter estimation is reduced.However,due to the difference in the size of the fuzzy scale,the width of the main lobe of the spectrogram in the center of the spectrogram will be different,and eventually the influence will also be affected.For the estimation of fuzzy parameters,this paper proposes to set a threshold value for this problem,and perform multiple corrosion operations on the binarized and closed spectrum map until the width is less than the threshold,and then estimate the fuzzy parameters.When the background color of the blurred image is relatively single and the area of this area is relatively large,the estimated fuzzy parameter value error is also large.In response to this problem,the fuzzy image block processing is proposed to find the image block that best reflects the image blur characteristics..Specifically,edge detection is performed on the image,and the blocks with more contour information are retained and corresponding to the original image.The fuzzy image parameters are estimated separately for these block images,and two groups of similar data are averaged to obtain the final blur angle and blur scale.After estimating the PSF parameters,use the Lucy-Richardson algorithm to recover.Finally,this paper combines the actual image restoration,the image belongs to the low illumination image and has the motion blurring.The image is preprocessed first,that is,the image is enhanced by the improved low-illuminance image enhancement algorithm based on the improved joint filtering principle,and then the improved image is used.The Radon transform-based motion blurring image PSF parameter estimation algorithm estimates the blur angle and blur scale,and finally uses the Lucy-Richardson algorithm to recover.The results show that the proposed algorithm can effectively enhance and recover motion blurred images in low illumination scenes,and has good robustness and practicability when dealing with low-illumination motion blurred images.