Enhancement Technology Research On Blurred Image

In the actual environment,the images obtained by the imaging sensor device have a certain change,easily lead to the loss of image information.There are some reasons to cause the blurred images,for example:the relative movement and shake between the camera and the target,device defocus caused by jitter,different lighting environment,weather condition(such as foggy,rainy,cloudy),blurry caused by the loss of the information when the three-dimensional object projection into the two-dimensional image plane.The images obtained will appear blurred.In order to further image processing(such as target centroid extraction,monitoring false alarm rate and leakage rate,etc.),it needs to restore and enhance these images while suppressing unnecessary details to improve the visual effects.Solutions:There are 2 methods.Frequency domain enhancement method is the Fourier transform of the image from the time domain to the frequency domain and then deal with by filter algorithm.Firstly,the Fourier transform of blurring image is carried out;Secondly,convolution with a transfer function.The results are inverse Fourier transformed,and the enhancement image is obtained.Image enhancement method in spatial domain:the definition of a function such as gray conversion,histogram equalization,histogram normalization,image smoothing and sharpening,etc..For the motion blurred image,combined with the classical Wiener filter restoration algorithm,this improved algorithm is proposed to restore the motion blurred image.Wiener filter(so called Mean Square Error minimum filtering method,belong to the frequency domain method)combined with image equalization method is proposed to enhance the blurred image of uniform motion.For the blurred image of low-pixel,the degradation model is established.And blurry principle is analyzed.Blurred image is processed by LIP algorithm.For the blurred image caused by compressing,especially under the condition of high compression ratio,the least squares method based on compression-post-processing is proposed to solve the problem of image blurring.Iterative solution is carried out.And the constrained set mapping and optimization parameters are established.For the optical fouling blurring image,a new image enhancement algorithm model is proposed.For the fog weather condition,we proposed white balance based on dynamic threshold.The contrast of the original fog image is enhanced.And the white balance image and the contrast enhancement image are fused.In HSV space,component V is enhanced and automatic grade processing.For haze,rain and snow,dust and other inclement weather,atmospheric scattering model is established.According to the dark-a priori algorithm,the blurred image is enhanced.Result:for the motion blur,motion blur is added into the ’lena’ and’checkboard’.The Fourier transform is used to obtain the spectrum.The improved Wiener filter is used to enhance the image processing.The experiment shows that the improved method is better than the original Wiener filter and median filter.For low pixel blur,the degradation model is established by analyzing the blurry principle of low pixel.The algorithm based on logarithm image processing(LIP)is used to enhance the target image.For the compression blur,a constrained least squares method post-processing is proposed.The iterative solution is carried out for the constrained optimal problem.The experimental result shows that the image enhancement of the post-processing algorithm under the compression condition.Optical fouling blur model caused by the fog,rain and snow,dust and other inclement weather is established.And then in accordance with the dark color prior algorithm,the blurry image is enhanced.Conclusion:the algorithm proposed in this paper is pertinent.Each algorithm can solve the problems of motion blur,low-pixel blur,compression blur and optical blur.