摘要： 提出一种多聚焦散焦图像复原算法。基于物体成像的点扩散函数退化模型，通过同一场景(包括前景和后景)中任意2幅对焦不同的模糊图像，得到该场景比较清晰的图像。同时，利用点扩散函数的模糊半径与物体深度信息之间的对应关系，估计物体的深度信息。实验结果证明了该算法的有效性。

关键词: 图像复原, 散焦, 点扩散函数, 模糊估计, 深度估计

Abstract: This paper proposes a multi-focus defocus image restoration algorithm. Basing on the degraded Point Spread Function(PSF), it generates all-focus image from two arbitrarily focused images on the same scene(foreground and background). It can estimate the depth of the object from the relationship between the blur radii and depth. Experimental results prove the validity of the algorithm.

Key words: image restoration, defocus, Point Spread Function(PSF), blur estimation, depth estimation

中图分类号: 

• TN919