Image adaptive point-spread function estimation and deconvolution for in vivo confocal microscopy

Visualizing deep inside the tissue of a thick biological sample often poses severe constraints on image conditions. Standard restoration techniques (denoising and deconvolution) can then be very useful, allowing one to increase the signal-to-noise ratio and the resolution of the images. In this paper, we consider the problem of obtaining a good determination of the point-spread function (PSF) of a confocal microscope, a prerequisite for applying deconvolution to three-dimensional image stacks acquired with this system. Because of scattering and optical distortion induced by the sample, the PSF has to be acquired anew for each experiment. To tackle this problem, we used a screening approach to estimate the PSF adaptively and automatically from the images. Small PSF-like structures were detected in the images, and a theoretical PSF model reshaped to match the geometric characteristics of these structures. We used numerical experiments to quantify the sensitivity of our detection method, and we demonstrated its usefulness by deconvolving images of the hearing organ acquired in vitro and in vivo.     

Comparison of I5M and 4Pi-microscopy

The axial (z‐) resolution of ～100 nm provided by 4Pi and I⁵M fluorescence microscopy relies on the coherent addition of spherical wavefronts of two opposing high aperture angle lenses. Both microscopes feature a point‐spread function (PSF) with a sharp central spot that is accompanied by axially shifted sidelobes which leads to replication artefacts in the raw image data. In a 4Pi‐microscope the sidelobes are less pronounced than in I⁵M and without relevant lateral (x,y) substructure, making their posterior removal in the image reliable and fast. On the other hand, high speeds of raw data acquisition are more easily gained by I⁵M. Moreover, I⁵M features a stronger signal as compared to the commonly employed two‐photon excitation (2PE) 4Pi‐imaging mode. We investigate here the capability of both techniques to image (aqueous) specimens without artefacts. To this end, we consider the optical transfer function (OTF) of the two microscopes in conjunction with the signal‐to‐noise‐ratio (SNR) of the object to be imaged. The imaging of E. coli bacteria with an interconvertable setup enabled a direct comparison of the two imaging modes. As both systems rely on high aperture angles, water‐immersion lenses of the largest numerical aperture available (NA = 1.2) were employed. The experimental results are corroborated by simulations assuming the signal strength encountered in the experiment. The comparison of the theoretical with the experimental PSFs/OTFs showed that our setup operated close to theory in both imaging modes. Although I⁵M provided about 10 times brighter raw image data as compared to (2PE) 4Pi‐microscopy, the I⁵M data could not be entirely cleared of artefacts. In conclusion, with the current aperture angles and fluorescence signal strengths, it is not advisable to trade in the suppression of the sidelobes for a larger image signal.     

基于APEX算法改进的图像复原算法

在高斯类点扩函数退化图像复原的研究中,提出了一种基于降晰图像频谱特征改进的APEX图像复原算法.该算法采用APEX算法的基本原理,根据图像频谱信息特征,对点扩散函数(PSF)估计过程进行了改进,采用加权最小二乘算法拟合出降晰图像频谱主方向,采用图像频谱主方向上的数据进行PSF估计,以利用更多的有效数据,从而减少PSF的估计误差.针对模拟和实际采集的降晰图像进行实验,采用主观视觉和峰值信噪比进行评价.实验结果表明,改进的算法较使用非主频谱方向上的频谱数据的复原算法在复原效果上有一定的提高.     

基于三维高斯模型的参数盲解卷积算法

提出了一种新的基于三维高斯点扩展函数（PSF）模型的参数盲解卷积（PBD）算法，并将此算法用于显微光学切片的图像复原。由于PBD算法需要在估计样本函数的同时估计PSF的参数，一般采用的PSF的模型较为复杂，计算量大，收敛慢；而基于三维高斯PSF模型的算法只需要估计2个参数，因此计算量大大降低了。经过实验验证，该算法能够较好地复原光学切片的图像，并且估计出PSF的参数。     

磷缓释复混肥的制备及其氮、磷缓释性评价

在实验室条件下制得不同程度磷缓释的复混肥(磷缓释肥),分别测定以水为淋洗剂和以2%柠檬酸溶液为淋洗剂时在特定淋洗装置下水溶性氮、磷及有效磷随时间的溶出率,采用双曲线方程Xt=b a/t对肥料的累计养分溶出率进行分段拟合,在此基础上引入缓释指数评价各养分缓释特性。结果表明:磷缓释肥对磷的缓释作用强于氮,对有效性磷的缓释作用强于水溶性磷;其水溶性氮、磷缓释指数(SR IWN、SR IW P)和有效磷缓释指数(SR IAP)与肥料中加入的缓释剂硫酸镁相对含量(X)具有很好的正相关,其直线相关系数为0.7414、0.8958和0.9829,后两者分别达到显著和极显著水平;2%柠檬酸溶液几乎能溶出肥料中全部有效磷,从而证明了自制磷缓释肥的有效性。     

基于高阶残差和参数共享反馈卷积神经网络的农作物病害识别

当前,大部分农作物病害图像识别方法主要关注于精度而忽略了鲁棒性.在面向实际环境时,由于噪声干扰和环境因素影响导致识别精度不高.为此提出了一种高阶残差和参数共享反馈的卷积神经网络模型以应用于实际环境农作物病害识别.其中,高阶残差子网络为病害表观提供丰富细致的特征表达,以提高模型识别精度;参数共享反馈子网络用来进一步抑制原深层特征中的背景噪声,以提高模型的鲁棒性.实验结果表明,当面向实际环境农作物病害识别时,本文方法在识别精度和鲁棒性上均优于其他方法.     

3‐D PSF fitting for fluorescence microscopy: implementation and localization application

Localization microscopy relies on computationally efficient Gaussian approximations of the point spread function for the calculation of fluorophore positions. Theoretical predictions show that under specific experimental conditions, localization accuracy is significantly improved when the localization is performed using a more realistic model. Here, we show how this can be achieved by considering three‐dimensional (3‐D) point spread function models for the wide field microscope. We introduce a least‐squares point spread function fitting framework that utilizes the Gibson and Lanni model and propose a computationally efficient way for evaluating its derivative functions. We demonstrate the usefulness of the proposed approach with algorithms for particle localization and defocus estimation, both implemented as plugins for ImageJ.     

Measurement and evaluation of display scattering

Abstract— A novel approach is introduced for the simple and effective measurement and evaluation of light scattering in general, and its application to electronic displays is described. The method is based on the measurement of the lateral distribution of reflected luminance with an imaging detector. After numerical transformation of the measured luminance distribution, we obtain the characteristics of the BRDF of the sample with high directional resolution within a limited range of inclination angles centered about the specular direction. The method is easy to implement and to carry out. Due to the absence of moving parts, it exhibits only little sensitivity with respect to variations of the geometry parameters and is thus very robust. This paper summarizes the basics of the approach, establishes the relation to conventional BRDF‐measuring methods, it introduces an implementation and presents some typical results for LCD‐monitors.     

侧窗头罩高速层流流场光学传输效应数值模拟

高速流场光学传输效应是影响红外成像末制导技术在高速导弹上应用的关键。研究了气动光学传输效应中层流流场光学传输分量的计算方法。应用光线追迹法和物理光学方法研究层流流场引起的像偏移和低程度的像模糊，结合典型状态下的流场数据，进行了高速层流流场光学传输效应的数学仿真，得到了仿真计算结果。采用实际计算出的流场数据，得出的计算结果与工程计算的结果相比较，证明所建立的气动光学高速流场光学传输效应理论模型基本正确，可以用来进行仿真计算。     

光栅型成像光谱仪杂散光校正技术研究

成像光谱仪中杂散光的存在会导致光谱测量误差,对杂散光的校正可以得到精确的测量结果。采用He-Ne激光器测出成像光谱仪的点扩展函数,假设成像光谱仪是线性波长不变系统的前提下,构建光谱函数矩阵,确定影响波长权重,再通过带宽校正矩阵对其进行带宽校正,从而对仪器测量中的杂散光误差进行校正。最后,通过实验表明此杂散光的修正方法基本上消除了杂散光对测量信号的影响。