The Quick Recovery Algorithm Of Motion Blurred Image

With the development of computer multimedia technology, digital image processing has made a great development. However, due to various reasons, such as the affection appearing in the process of image acquisition, image transmission and image storage,the image quality we get will produce degradation. The noise problem and the motion blur problem are the most common image degradation problems. Image noise is mostly generated in the process of image acquisition and transmission, and image motion blur degradation are mostly caused by the relative motion between the camera and the object under observation. Both of them can greatly affect the quality of the image. In this paper, the main research work is to study the image denoising and the restoration of the motion blur image.Removing image noise is a common preprocessing operation in the image process. For image denoising, this paper describes the mathematical principles as well as the advantages and disadvantages of several classical denoising algorithm. Noise, as a special kind of signal,this paper treat it as the most common signal and apply the independent component’ analysis(ICA) algorithm, a classical signal processing and machine identification method. First,ICA consider that the image and the noise in image are mutually independent, then on the basis of the satistics, it reconstruct the original image,and last base on the independent component analysis model, it propose a new single-channel denoise method. In this paper,the result is good.For the restoration of the motion blurred image, this paper introduces several common and classic motion blurred image recovery algorithms. The mathematical principles as well as the advantages and disadvantages of each method have been introduced in detail. In order to look for the motion blurred direction and the blurred length,on the basis of the frequency domain,according to the calculation of the power spectrum of the image gradient and the image gradient changes, coupled with the association of the power spectrum’s structrue in the frequency domain with the blurred direction and the blurred length, this paper proposes a fast method to find the blur direction and blur length, and the simulation is good. Meanwhile, summarize the whole work and forecast the future work.