权益分类	功能权益	普通用户	{{item.name}}会员
{{category.name}}	{{benefitItem.name}}

Research on a Structure Analysis of Materials with a Maximum Entropy Restoration of Electron Microscope Images.

电子显微镜图像最大熵恢复材料结构分析研究。

基本信息

批准号：
09650063
负责人：
HANAI Takaaki
金额：
$ 2.05万
依托单位：
Nagoya University
依托单位国家：
日本
项目类别：
Grant-in-Aid for Scientific Research (C)
财政年份：
1997
资助国家：
日本
起止时间：
1997 至 1998
项目状态：
已结题

来源：
https://kaken.nii.ac.jp/grant/KAKENHI-PROJECT-09650063/
关键词：
Maximum Entropy Method Image Restoration Electron Microscopy Random Noise Constraint Quantum Noise Structure Analysis Contrast Transfer Function Complex Wave Function ランダム空間分布制約条件

项目摘要

1.Implementation of Maximum Entropy (ME) Restoration Programs : Computer programs for the ME image restoration constrained by the spatially random distribution of quantum noise were developed and implemented on PCs. The programs allows visualization of changes in the image during the process of restorations and the difference between the modified ME restoration and conventional kai-squared constrained method. It also enables us to study the effect of the ME restoration in the Fourier space.2.Restoration of TEM Images of Weak-phase Objects : The modified ME method was applied to deconvolution of largely defocused TEM images of ferritin particles. The cores of ferritin particles, which were blurred with reversed contrast due to the severely oscillating contrast transfer function (CTF) in the defocused image, are recovered to the original size with the correct contrast. The Fourier analysis showed that the frequency regions of reversed contrast in the defocused image is automatically reconstructed by the phase shift of pi as a result of the ME restoration even around zeros in the CTF.3.Implementation of the Imaging of Crystals : For application of the ME restoration to crystal specimens, simulation of TEM imaging is required to calculate the image intensity from a simulated complex wave function just below the specimen and to compare the image intensity with the observed image. An image simulation program was, therefore, implemented with a C++ language and incorporated into ME restoration programs. The results of the simulation were in good agreement with those calculated with a standard program.4.Development of ME Restoration of High-resolution Crystal Images : Existing ME methods cannot be applied to crystal images since they are designed. to treat only real number while the electron wave function is complex. To solve this problem, the algorism was extended to deal with complex function.

1.最大熵（ME）恢复程序的实现：在量子噪声空间随机分布约束下，开发并实现了最大熵图像恢复程序。该程序允许在恢复过程中可视化图像的变化，以及改进的ME恢复与传统的凯平方约束方法之间的差异。这也使我们能够在傅里叶空间中研究ME恢复的效果。弱相位物体TEM图像的恢复：将改进的ME方法应用于铁蛋白颗粒的大散焦TEM图像的反卷积。在散焦图像中，由于对比度传递函数（CTF）的严重振荡，对比度被反向模糊的铁蛋白颗粒的核心在正确的对比度下恢复到原始尺寸。傅里叶分析表明，即使在CTF.3中零点附近，散焦图像中反向对比度的频率区域也会由于ME恢复而被pi的相移自动重建。晶体成像的实现：为了将ME恢复应用于晶体样品，需要模拟TEM成像，从样品下方的模拟复波函数计算图像强度，并将图像强度与观测图像进行比较。因此，用c++语言实现了一个图像模拟程序，并将其合并到ME恢复程序中。模拟结果与标准程序计算结果吻合较好。高分辨率晶体图像的ME复原研究进展：现有的ME复原方法由于设计的原因，无法应用于晶体图像。当电子波函数是复数时，只处理实数。为了解决这一问题，将该算法扩展到处理复杂函数。