本文主要研究被摄物与相机之间的相对运动所造成的运动模糊图像的恢复问题。通过分析运动模糊图像的成像原理，采用恢复算法进行恢复处理。运动模糊是成像过程中普遍存在的问题，如在飞机上拍下来的照片，用照相机拍摄高速运动物体的照片等均可能存在这种现象。因此运动模糊图像的恢复是图像恢复中的主要课题之一。 首先着重对已有的基于频域的模糊参数辨识方法进行了改进和修正，给出了自动检测模糊角度和任意方向模糊长度的可行算法。 其次对基于水平运动模糊图像的恢复算法的理论进行了研究，采用逆滤波、维纳滤波、刃边函数法与PSF法四种图像恢复方法。主要对逆滤波与维纳滤波法进行了理论研究及仿真实验对比分析，在有噪声和无噪声两种条件下对其做了运动模糊恢复实验，结果表明：维纳滤波算法对有噪声的运动模糊图像复原时对模糊方向的估计均方误差小及对角度检测精度高，从而复原的效果优于逆滤波、刃边函数法与PSF图像恢复算法。再利用维纳滤波算法对实际采集的运动模糊图像进行恢复时往往会存在振铃效应，通过一系列的计算机仿真实验分析得到，滤波参数会对振铃效应产生影响，本文给出了改进的维纳滤波算法，通过选取适当的参数能够有效抑制恢复图像上暗色条纹，即通过对该传统算法中的参数改变(信噪比)来提高其精度和减少运算耗时，实验结果证明了改进算法的有效性。 最后对基于任意方向的图像复原进行了研究，主要针对的是在重度模糊的影响下，维纳滤波并不能达到人眼所要求的最佳效果，因而采用了刃边函数法，但其算法缺点是对于重度模糊图像的复原效果不够理想，所以对于任意方向的运动模糊图像复原，本文通过采用PSF法来弥补刃边函数法的缺点。实验证明，基于任意方向的图像复原采用刃边函数法与PSF法相结合恢复的效果较好。

Motion-blurring is caused by the relative motion between the camera and the object. It is a ubiquity to our matters, by analyzing the imaging principle of motion blurred images, the image can be recovered using the recovery processing algorithm. This motion-blurred is a common problem in the imaging process, such as the photos taken in aircraft or space craft, with fast-moving objects, and missiles flying in battle. It has realistic significance, which can be applied in many fields. For example, military affairs, traffic, medical images, industry controlling and deceptive field. Motion Blurred Image Restoration based on the following was researched. Firstly, point to the existing fuzzy parameters frequency domain identification method was improved and modified, which given automatically detect the theta and the direction of fuzzy length of the feasible fuzzy algorithm. Secondly, the level of motion blurred images based on the recovery of the theory of algorithms, has been studied, it mainly uses four recovery methods of the inverse filtering, Wiener filtering, edge function method and PSF method. mainly on the inverse filtering and Wiener filtering method for the theoretical study and comparative analysis of simulation experiments, in both noise and non-noise movement under fuzzy recovery experiment done. The results show that Wiener filtering algorithm with motion-blurred image restoration noise, when the direction of the estimated fuzzy Mean Square Error(MSE) is small and the theta of detection is high precision, thus Wiener filtering algorithm of the motion blurred image with noise reduction of the original image can be more effective than the inverse filter, blade edge function method and the PSF algorithm. Generally, there will be ringing effect in Wiener filtering algorithm using the actual acquisition of motion-blurred image restoration, through a series of computer simulation analysis obtained, the filter parameters will affect the ring effect, and therefore choosing appropriate parameters, which can effectively restore the image on the inhibition of dark stripes. The traditional method through the parameters of the change (noise ratio) improves the accuracy and reduces the computation time-consuming. Experimental results show the effectiveness of the improved algorithm. Thirdly, with regard to any direction of the image restoration research, using the blade edge function method, but the disadvantages of algorithm is effect that does not ideal for the severe blurred image restoration. So for any direction of the motion blurred image restoration in this paper, PSF method makes up its blade edge function algorithm shortcomings. Experimental results show that any direction of the image restoration by PSF blade edge function method and the combination resume is better.