中文说明：
摘 要 盲图像恢复的主要困难是信息不足，而为了恢复图像和确定点扩散函数需要适当的先验知识。解决这个问题的法、 法以及正则化方法等。但是这些方法的计算量都太大，针对上述方法的不足，文章提出了一种恢复图方法有图像的新算法，它通过恢复残差的最小化和后验概率的最大化来估计参数和恢复图像。其中，巧妙地利用了最陡梯度法和 共轭梯度法的迭代求解。对由于运动造成的模糊图像，可以明显地改善图像的质量，实验结果证明，在对模糊操作没有严格限制的情况下，仍可得到较好的恢复图像。
English Description：
The main difficulty of blind image restoration is the lack of information, and appropriate prior knowledge is needed to restore the image and determine the point spread function. Methods, methods and regularization methods to solve this problem. However, the amount of computation of these methods is too large. In view of the shortcomings of the above methods, this paper proposes a new image restoration algorithm, which estimates the parameters and restores the image by minimizing the restoration residual and maximizing the a posteriori probability. Among them, the steepest gradient method and