Spectrally Adaptive System for Microscopic Image Enhancement

这个小型企业创新研究第一阶段项目将应用新兴的多光谱声光成像技术和专门设计的无监督学习神经网络来增强显微成像。利用荧光、吸收和拉曼效应的各种光谱系统目前广泛用于生物结构的微观研究。物理光学公司的研发部门提出开发一种新的光谱自适应系统，以提高图像质量，并改善与单线滤波相比，由成像相机收集的光量。同时，由于系统的光谱响应将自动与感兴趣对象的光谱相匹配，因此信噪比将大幅降低。所提出的系统还将减少所需的观察时间和被测样品的照明，允许实时跟踪样品中的动态事件，而不会被强烈的照明损坏。

This Small Business Innovation Research Phase I project will apply emerging multispectral acousto-optic imaging technology and a specially devised unsupervised learning neural network to enhance microscopic imaging. Various spectral systems employing fluorescence, absorption, and Raman effect are currently in wide use in the microscopic investigation of biological structures. Physical Optics Corporation's R&D Division proposes to develop a novel spectrally adaptive system to enhance image quality and to improve the amount of light collected by an imaging camera as compared with single line filtering. At the same time, the signal-to-noise ratio will be drastically reduced since the system's spectral response will automatically be matched to the spectrum of the object of interest. The proposed system will also reduce the required observation time and the illumination of the sample under test, permitting real-time tracking of dynamic events in the sample without its being damaged by intensive illumination.