Cook C A, Koretz J F
Cytometrics, Inc., Philadelphia, Pennsylvania 19106, USA.
J Opt Soc Am A Opt Image Sci Vis. 1998 Jun;15(6):1473-85. doi: 10.1364/josaa.15.001473.
Of the methods developed (e.g., phakometry, magnetic resonance imaging, etc.) for noninvasive measurement of the geometry of the anterior segment of the human eye, Scheimpflug photography offers the best resolution and the highest precision. The primary obstacle encountered with this or any other image-based method has been in obtaining quantitative measurements directly from the images. Image enhancement (gray-scale gradient analysis) and pattern recognition methods (Hough transform and recursive least-squares algorithms) are developed so that parametric representations of lens surfaces and zone boundaries can be obtained directly from the images. Methods to correct for nonlinear Scheimpflug camera reproduction ratios and provide error estimates for geometrical parameters are also developed and will be presented separately. Combined, these techniques yield representations of lens geometry having sufficient precision, to which paraxial ray tracing can be applied to determine lens optical properties by using well-posed optical models with one unknown.
在已开发的用于非侵入性测量人眼前节几何形状的方法（例如，晶状体测量法、磁共振成像等）中，Scheimpflug摄影提供了最佳分辨率和最高精度。使用这种或任何其他基于图像的方法时遇到的主要障碍一直在于直接从图像中获得定量测量结果。已开发出图像增强（灰度梯度分析）和模式识别方法（霍夫变换和递归最小二乘算法），以便可以直接从图像中获得晶状体表面和区域边界的参数表示。还开发了用于校正非线性Scheimpflug相机再现比率并提供几何参数误差估计的方法，这些方法将单独介绍。综合起来，这些技术产生具有足够精度的晶状体几何形状表示，通过使用具有一个未知数的适定光学模型，可以将近轴光线追迹应用于该表示来确定晶状体光学特性。
