The Restoration Of Motion And Defocus Blurred Image

The paper studies the restoration of motion blurred image, defocus blurred image and motion and defocus mixed blurred image mainly. The paper analyzes systematically the formation theory of the blurred image, establishes the models of degenerating and restoration and the establishment of Point Spread Function and its estimated precision is an important reason to impact image restoration. At the same time, the paper also analyzes the defection of the traditional blurred image restoration whose Point Spread Function's model is limited, such as uniform motion in a straight line, gauss defocus model and pillbox defocus model, on this basis, in order to improve the estimating accuracy of Point Spread Function, the paper integrates the optical theory with the method of the digital image processing and uses the computer as a tool to restore the blurred image. This paper proves theoretically and experimentally that the method and conclusion are suitable to the restoration of the blurred image by the computer simulation and real blurred image. This paper mainly does studying work as follows:Firstly, a new method is presented to distinguish defocus blurred and motion blurred images robustly and accurately in this paper, this method is based on Hough transform and compared the number of highlights in the Hough matrix, it has high accuracy, up to 100%.Secondly, this paper improves the method which was identification of motion blur direction from motion blurred image by direction derivation method to estimate the motion direction of the motion blurred image, proposed a two derivative method to estimate the motion direction. First do line direction different ion to the image and get gradient image, then using the old algorithm to gradient image. This method improves the estimation accuracy of motion direction.Thirdly, based on the optical knowledge such as Line Spread Function, Edge Spread Function, Modulation Transfer Function and Point Spread Function and discarded concrete Point Spread Function model, an improved method is presented to calculate the edge spread function using the improved Prewitt operator and Fermi function, then attained the modulation transfer function, and then get the point spread function by transforming, and then restore the blurred image using the improved Wiener Filter. Some experiments were performed to validate the performance of our method and experimental results show that our algorithm not only has effectiveness and strong resistance to the noise, but also can work on noisy images with low SNR. When SNR is 20dB, it also can work robustly. And compared with the traditional image restoration method and model, the image restoration results are improved significantly using our method.At last, we use one method which is based on fuzzy projection onto convex sets (POCS) to deblur the mixed blurred image. This method can improve the quality by using multiple blurred image of the same scene. This method does not estimate the Point Spread Function of blurred image. The mutually different pieces of information from the multiple blurred images of a scene are merged in the frequency domain remove image blur. At the same time, we prove the validation of the algorithm by experiment, and the results show that this method can improve the quality of the blurred image.