On the optimizing weighting apparatus in active image processing

本研究以“主动调制图像处理中的优化问题”（科学研究资助项目(C) 1994-1995）的理论研究成果为基础，旨在构建一种通用的优化加权装置，以验证调制光谱及其扩展——主动调制图像处理的优化加权技术的实用性。为此，本研究第一阶段（1997年）设计并构建了一套通用优化的调制光谱加权装置。并搭建了信号源模拟器来代替实际的光谱仪。该优化的加权系统包括一个用于数值加权的变增益放大器、一个信号积分器、一个用于输入信号的实时方差估计器和一个用于停留时间加权的调制控制器。这个系统由两台个人电脑控制。在下一阶段（1998年），信号源模拟器进行了重建和编程，以获得更多的第一响应。在电子能谱的情况下，利用该信号源模拟器对系统信噪比的提高进行了评价。此外，针对优化后的权重，建立了一种用于主动图像处理的基本理论处理方法。我们期望将这种优化的加权技术应用于透射电子显微镜的球差校正。由于在该领域中提高信噪比是非常重要的，因此这种优化加权方法在实际应用中具有重要意义

The present study is based on results of the theoretical study, 'optimizing problems in active modulation image processing ( Grant-in-aid for scientific research (C) 1994-1995 )' The aim of this study is the construction of a universal optimized weighting apparatus to confirm the practicability of the optimized weighting technique for the modulation spectroscopy and its extension, the active modulation image processing.According to this aim, a universal optimized weighting apparatus for the modulation spectroscopy was designed and constructed at the first stage (1997) of this study. A signal source simulator was also constructed as substitute for actual spectrometers. This optimized weighting system contains a variable gain amplifier for the numerical weighting, a signal integrator, a real time variance estimator for the input signal, and a modulation controller for the time of stay weighting. This system was controlled by 2 personal computers.In the next stage (1998), the signal source simulator was reconstructed and programmed to get more first response. Evaluations for the improvement of the signal to noise ratio of this system were tried using this signal source simulator under the situation of the electron spectroscopy. In addition to this, a basic theoretical treatment was established with respect to the optimized weighting for the active image processing. We are expecting to apply this optimized weighting technique for the spherical aberration correction of the transmission electron microscope. Since the improvement of the signal to noise ratio is very important in this field, such optimized weighting has great significance in practice