Motion Blur Image Motion Parameter Estimation And Restoration Research

Motion-blurring is caused by the large enough relative motion between a camera and the shoot object. The research of motion-blurring has realistic significance, and has been applied in many fields, such as in industry, traffic, military, and medical images.The usual process of image restoration is to search the degradation model of an image, and then to find the original image by taking the inverse process. The two key parameters of the degradation model are the motion-direction and the motion-scales. Then, the main point of this paper is how to get the two important parameters. The main content of the study is:Firstly, it introduces the parameter estimate of a motion-blurred image and objective of the image restoration research. It also describes the basic theory of motion-blur especially in the case of linear and unchanged speed.Secondly, after summarizing the image noisy, median filter and quantity estimation of an image, we give am improved median filter based on the condition judgment, and applied in the process of parameter estimate.Thirdly, based on the research of the usual parameter calculation algorithm and the zero value of the dark line of the spectrum image, an improved motion-direction estimate algorithm is found out. This paper also research the sample curve of any row of the spectrum image, summarize that the line made up of all max. that all row and motion-direction are vertical, and get another improved motion-direction estimate algorithm.Fourthly, this paper study some motion-blur restoration algorithms, and take many experiments and quantity estimation for each algorithm.Lastly, we summarize the main content of this dissertation and point out the shortages and the further research directions.