复杂背景运动模糊图像的盲复原算法研究

Research on Blind Restoration Algorithm for Motion Blurred Images with Complex Background

为了有效恢复具有复杂背景的运动模糊图像,提出一种基于正则化策略和共轭梯度优化迭代复原算法;同时为了辨识运动模糊图像的参数,又提出一种基于模糊图像微分自相关函数的点扩展函数辨识算法。为验证算法的有效性,在微机上对提出的算法与现有算法进行了对比实验,结果表明十分有效,也具有较强的鲁棒性。

In order to effectively restore motion blurred image with complex background, the iterative restoration algorithm based on regularization strategy and conjugate gradient is proposed. Meanwhile in order to identify the parameters of motion blurred image, the point spread function recognition algorithm based on differential autocorrelation function with blurred image is also pro- posed. The effect of the algorithm proposed is validated by comparison experiments in microcomputer, and the experimental rcsuhs show that the algorithms proposed are efficient.