基于COSM的三维显微图像复原算法研究  

Study of 3D Microscopical Image Restoration Algorithms Using COSM

作　　者：周宾[1] 郝英立[1] 王式民[1] 

机构地区：[1]东南大学空间科学与技术研究院,江苏南京210096

出　　处：《工程热物理学报》2011年第4期683-686,共4页Journal of Engineering Thermophysics

基　　金：国家自然科学基金资助项目(No.50906013);教育部新教师基金资助项目(No.20090092120064)

摘　　要：光学分层显微成像技术(COSM)不仅可用于去除非聚焦层杂散光的影响以及噪声的干扰,而且能复原样本的清晰三维像,将成为提取微流体的三维温度和速度场分布信息的有效工具。本文研究了基于已知点扩展函数的6种COSM成像算法,详细讨论了6种COSM成像算法的原理和实现,构造了三维样本图像和点扩展函数模型,并对复原结果进行了分析。实验结果表明逆滤波、正则逆滤波算法复原效果较好。迭代类算法中的Landweber算法具有速度快,抗噪声干扰能力强的优势,图像的复原质量也较高。The technology of Computational Optical Sectioning Microscopy can be used to reduce the interference of stray light from the non-focused layer and suppress the affection of noise and restore the three dimensional image of the original objects clearly,which should be a useful tool in reconstruction of 3-D temperature and velocity field of micro-flow.In this paper,the three dimensional model and point spread function are constructed,and the numerical simulation results based on six algorithms are evaluated.The numerical experiments show that the inversion results of Inverse Filtering and Regularized Inverse Filtering are satisfactory.For the category of iteration algorithms,Landweber algorithm are of the advantages of fast inversion,high noise immunity,and high quality in image restoration.

关 键 词：微流动 光学分层显微成像技术 图像复原 点扩展函数 

分 类 号：TN911.73[电子电信—通信与信息系统]