基于遗传自适应的维纳滤波图像去模糊算法

The Adaptive Wiener Filtering Deblurring Based on the Genetic Algorithm

为了去除图像在拍摄过程中由于相机的抖动或物体本身运动产生的匀速直线运动模糊,在维纳滤波图像复原的基础上,结合遗传算法,本文提出一种基于遗传自适应的维纳滤波去模糊算法。此算法用遗传算法自适应优化维纳滤波的正则项k值,可达到所选自适应度函数的最佳图像复原效果。实验结果表明,该算法能够通过选择合适的自适应度函数,尽可能地去除实拍图像的运动模糊,还原图像性能优于维纳滤波算法,具有一定的实际意义。

In order to remove the blur of uniform linear motion caused by the camera shake or the motion of the object itself,on the basis of the theory of Wiener filter image restoration and genetic algorithm,an adaptive Wiener filter based on genetic algorithm for deblurring is proposed.The genetic algorithm is used to adaptively optimize the k value of the regularization term,which could achieve the best image restoration effect from the fitness function.The experimental results show that this method could select the appropriate adaptive-degree-function,remove the motion blur of the real image as much as possible,and the effect of image restoration is better than that of Wiener filtering algorithm,which has some practical significance.