邻域法在三维显微图像复原中的应用研究  

Application Study of Near-neighbor Algorithm on Three-dimensional Microscope Image Restoration

作　　者：王霞[1] 姚晋丽[1] 金伟其[1] 陈华[1] 苏秉华[1] 

机构地区：[1]北京理工大学信息科学技术学院光电工程系,北京100081

出　　处：《生命科学仪器》2007年第7期34-38,共5页Life Science Instruments

基　　金：高等学校博士学科点专项科研基金资助项目(20020007006)

摘　　要：三维显微图像复原方法通常分为两类:邻域法和三维去卷积法。本文详细介绍了邻域法的基本原理;对邻域法公式中参数c,α的不同选择和相应的复原结果进行了评价,通过仿真实验得出了考虑不同切片数时c,α的最佳选择以及不同显微成像系统所需考虑切片数M的最佳选择。与三维去卷积法相比,邻域法具有简单、计算量小,容易实现等优点,适合于观察活的生物和进行实时连续观察。The near-neighbor algorithm is one of main methods of three-dimensional （3D） microscope image restoration. In this paper, firstly the fundamental of near-neighbor algorithm is introduced. Then the best choice of coefficients c and ？ in the near-neighbor algorithm under different slice number is obtained by comparing the restoration quality with different choices of the coefficients. And we also obtain the reasonable selection of slice number M. Compared with 3D deconvolution algorithm, the near-neighbor algorithm is simple, easy to program and suitable to observe living specimen.

关 键 词：图像复原 邻域法 数字共焦显微 图像退化 图像质量评价 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]