一种离焦模糊图像的盲复原方法

摘要

For out-of-focus blurred image caused by the inaccurate focus of the imaging equipment, we used step edge diffusion statistical property and an improved Grubbs abnormal value detection to locate step edge or similar step edge. Therefore, an adaptive selection method for optimal area is proposed to confirm the line spread function in this paper, and according to the relationship between line spread function and out-of-focus blurred parameters, the out-of-focus blurred parameters are calculated, so the radius of the out-of-focus blur and the point spread function are obtained. Thus, the blurred image can be restored by the blind restoration method. The experiment results show that the method proposed by the paper has better effect than existing other methods.%针对离焦模糊图像提出了一种基于图像阶跃边缘扩散特性的盲复原方法.该方法通过改进Grubbs异常值检测准则,对图像中的阶跃或近似阶跃边缘进行定位.通过自适应选取最佳图像区域的方法计算线扩散函数,进一步利用离焦模糊半径与线扩散函数之间的关系,计算出离焦模糊参数.根据参数得到点扩散函数,最终对离焦模糊图像进行复原.实验结果表明,该方法的复原效果较好,具有一定的应用价值.