Hu Y, Frieden B R
Appl Opt. 1988 Jan 15;27(2):414-8. doi: 10.1364/AO.27.000414.
In this paper, a method of restoring longitudinal images is developed. By using the transfer function for longitudinal objects, and inverse filtering, a longitudinal image may be restored. The Fourier theory and sampling theorems for transverse images cannot be used directly in the longitudinal case. A modification and reasonable approximation are introduced. We have numerically established a necessary relationship between just-resolved longitudinal separation (after inverse filtering), noise level, and the taking conditions of object distance and lens diameter. An empirical formula is also found to well-fit the computed results. This formula may be of use for designing optical systems which are to image longitudinal details, such as in robotics or microscopy.
本文提出了一种恢复纵向图像的方法。通过使用纵向物体的传递函数并进行逆滤波，可以恢复纵向图像。横向图像的傅里叶理论和采样定理不能直接用于纵向情况。为此引入了一种修正和合理近似。我们通过数值方法建立了刚分辨出的纵向间距（逆滤波后）、噪声水平与物距和透镜直径的拍摄条件之间的必要关系。还找到了一个经验公式，它能很好地拟合计算结果。该公式可用于设计对纵向细节成像的光学系统，如机器人技术或显微镜中的光学系统。
