统计力学中的勒让德变换

勒让德变换是理论物理学中一个重要的工具,多变量函数的勒让德变换有明显的对称性.系综是统计力学教学中的核心内容.利用勒让德变换的无量纲性,从微正则系综的特性函数熵的勒让德变换出发,类比的定义了配分函数的勒让德变换.同时得到各种系综的特性函数和配分函数以及它们之间的关系和相应的热力学公式.     

神光-Ⅲ主机黑腔腔形研究与设计

利用三维视角因子软件IRAD3D研究了神光-Ⅲ主机靶丸表面辐射驱动时变对称性和驱动强度随黑腔腔形和排布方式的变化。模拟结果表明:采用椭球形黑腔能显著减小黑腔壁面积从而减小腔壁能量消耗并提高靶丸消耗份额;不同腔形和排布方式下均可通过调节腔长使勒让德不对称性模P2时间积分量较小;采用双环辐照和椭球腔形都利于缩短腔长;四环排布利于调节P4分量,但环向不对称性M4较差;椭球腔双环排布能兼顾极向不对称性P2,P4和环向不对称性M4,且靶丸表面X光驱动强度相对柱腔双环提高25%~27%。     

神光Ⅱ装置上辐射驱动瑞利-泰勒不稳定性实验北大核心CSCD

在神光Ⅱ装置上开展了辐射驱动RT不稳定性的一系列实验,获得了不同初始扰动幅度、不同扰动波长、不同材料样品等条件下辐射烧蚀RT不稳定性增长的高时空分辨背光图像,特别是在大初始扰动幅度样品实验中获得了扰动增长的清晰图像,观察到了扰动增长从线性区到非线性区的过渡过程,二次和三次谐波的产生和发展清楚可见。充实了数值模拟程序考核的实验数据库,对间接驱动ICF点火靶设计和研究具有重要作用。     

神光Ⅱ装置上辐射驱动瑞利-泰勒不稳定性实验

在神光Ⅱ装置上开展了辐射驱动RT不稳定性的一系列实验,获得了不同初始扰动幅度、不同扰动波长、不同材料样品等条件下辐射烧蚀RT不稳定性增长的高时空分辨背光图像,特别是在大初始扰动幅度样品实验中获得了扰动增长的清晰图像,观察到了扰动增长从线性区到非线性区的过渡过程,二次和三次谐波的产生和发展清楚可见。充实了数值模拟程序考核的实验数据库,对间接驱动ICF点火靶设计和研究具有重要作用。     

视角因子方法在间接驱动ICF中的应用

采用视角因子方法,对间接驱动惯性约束聚变(ICF)过程中从腔内壁辐射到靶丸表面的辐射进行了计算,并分析了激光原型装置上的一个实验的对称性环境。间接驱动ICF系统的不对称性一般来自于激光焦斑和腔体的几何结构,利用视角因子方法,计算X光能量在腔体内的强度分布,将靶丸表面的驱动对称性表示成腔体结构和激光焦斑几何参数的函数。结果表明,采用该方法的程序模拟能够评估腔靶设计的主要参数。     

基于Python的牛顿环实验数据拟合及分析方法

牛顿环实验是经典的大学物理实验,其实验数据量大,手工计算非常容易出错。基于Python的第三方库对牛顿环实验数据进行拟合分析,利用numpy进行数据的预处理和不确定度的计算,利用scikit-learn建立线性回归模型拟合数据并求解,最后利用matplotlib绘制图像实现数据的可视化。本方法能快速处理实验数据,绘制拟合曲线并输出相对不确定度,形象而又直观,对牛顿环实验数据的分析和处理有很好的辅助作用。     

广角任意反射面速度干涉仪条纹内爆对称性分析

提出在条纹相机前加载异形光纤面板或环转线光纤传像束实现广角任意反射面速度干涉仪（VISAR）条纹采样的诊断设计,发现采样位置坐标处于靶面某圆上。综合运用坐标变换、傅里叶变换、勒让德展开等方法提取广角VISAR条纹相位实现内爆对称性分析,并通过示例验证了其可行性。针对诊断方法的特点、光路设计、装置研发、数据处理等展开讨论,指出广角VISAR诊断内爆对称性的发展方向。运用该方法记录并分析广角VISAR条纹数据,可使靶丸内爆对称性诊断准确、直观、形象,能为惯性约束聚变中激光等离子体不稳定性、流体不稳定性等研究提供支撑。     

Origin9.0软件在微波技术实验中的应用探讨

以近代物理实验中的微波技术实验为例,详细介绍了用Origin9.0软件进行绘图和非线性拟合处理实验数据的方法,对实验测定的数据进行绘图、计算、拟合。该软件的应用,能够有效避免人工计算处理所带来的误差,大大简化实验数据处理过程,获得人工绘图难以获得的更多的信息量,且运用软件结果高效准确。     

1.06μm 激光直接驱动烧蚀靶内爆压缩特性

介绍了直接驱动内爆压缩特性实验研究的实验靶型和测量方法.在神光Ⅱ激光装置上采用多种高时空分辨和高能谱分辨的组合型诊断设备获取了内爆压缩相关信息，得到了靶丸辐照均匀性、内爆压缩对称性、径向会聚比、内爆速度以及靶丸球壳平均电子温度，获得的 结果可作为优化靶结构与参数及激光辐照条件的依据.给出了实验中获得的典型结果，并对结果进行了简要的分析讨论. 关键词： 激光烧蚀 直接驱动 诊断技术     

用勒让德级数处理Abel变换

提出了一种处理弱非中心对称Abel变换的新方法.把测量的积分值用Yasutomo的方法分成奇、偶两部分.非对称的原函数被表达为一个权函数和一个中心对称函数之积.对称函数用勒让德级数展开.级数的系数通过用最小二乘法来决定 关键词： Abel变换 勒让德级数 测试函数 对称 非对称     

Offline estimation of 2D crystal lattice parameters by processing the electron diffraction image

Electron diffraction provides useful information about the internal composition of materials and has been in the use of material scientists for more than fifty years. In order to extract useful information from offline diffraction images, they are manually analyzed by using some photometric technique. Manual analysis is however a cumbersome, laborious and difficult task. To reduce the labors of material scientists one can employ image processing techniques to perform automated analysis, due to the well established popularity and clear evidence of widely used image processing techniques. In this work an image processing technique is being proposed for the extraction of 2D unit cell information from diffraction images on one hand and finding the 2D point group contained by the lattices on the other. The technique employs a morphological shrinking operation to find the center of each spot in the underlying preprocessed diffraction image. This is followed by the extraction of eight points with reference to the spot produced by the transmitted electron beam. The resultant nine points, i.e. the extracted eight plus the reference spot generated by the transmitted electron beam, are then subjected to symmetry operations, rotation symmetry and mirror symmetry, in polar coordinate system, to classify the point group of the lattice produced by the electron diffraction. One of the difficult task, even in manual analysis, is to ascertain the exact spot where the transmitted electron beam hit the sample at the time of realization of the image. This has been accurately and intuitively done by employing the notion that the transmitted spot must have greater number of pixels, with the highest gray value, among the diffracted spots. The proposed strategy has been applied to a sample set of various images and the results shows that the technique is efficient in determining the unit cell in 2D and classify the point group with good accuracy.     

一种新的适用于彩色图像处理的彩色空间变换方法

从人眼的视学特性角度出发，在ＣＩＥ国际颜色标准系统和孟塞尔系统基础上，提出了一种适用于人彩色图像处理的色空间转换模式，并利用此方法对医用胃镜输出的图像进行了不破坏色调前提下的图像增强处理，明显地提高了彩色图像的细节分辨，实践证明该方法不失为一种具有实用意义的彩色空间变及图像处理方法。     

刚体微位移非接触式光电检测系统

提出用标定法测量刚体多维微位移，该方法具有结构简单，性能可靠，集成化程度高，工业上易实现和价格便宜等优点。本文采用一个小功率激光器为光源及一种新的ＣＭＯＳ图像传感器为光电传感器件，通过相关分析法和数据拟合等图像处理技术，提高整套系统的精度和分辨率。     

激光靶测量图象的在线处理与应用

介绍了图象在线处理系统及其被用于激光靶测量的工作。自行建立了图象处理的自动测量分析系统，研制了适应该系统的图象测量分析的专用软件ＹＡＮＧ．ＥＸＥ。实际测量结果表明，与以前方法比较，本系统测量数据重复性好，实测精度好于以前手工方法，并且可以对靶表面缺陷进行定量测量。已经逐步用于ＩＣＦ靶与ＸＲＬ靶的实际测量。     

Fringe analysis by automatic image fitting and its application to speckle pattern photography

A new image processing technique is described that has been found well suited for the fully automatic evaluation of Young's fringes in speckle pattern photography. The basic idea is fringe pattern analysis by creating synthetic images and fitting them to the image picked up by a TV camera. High speed fringe generation and comparison is achieved using a videoprocessor. A theoretical discussion and experimental results show the efficiency of this method for subfringe evaluation, i.e. the measurement of displacements well below the mean speckle size.     

Circle detection using scan lines and histograms

Circle detection is significant in image processing and pattern recognition. We present a new algorithm for detecting circles, which is based on the global geometric symmetry of circles. First, the horizontal and vertical midpoint histograms of the edge image are obtained by using scan lines. Then, we apply the peak-finding algorithm to the midpoint histograms to look for the center of the circle. The normalized radius histogram is finally used to verify the existence of the circle and extract its radius. Synthetic images with different levels of pepper noise and real images containing several circles have been taken to test the performance. Experimental results demonstrate that the proposed algorithm has the advantage of computational efficiency as compared with the randomized Hough transform and some other algorithms.     

小波变换在图像压缩中的应用研究

小波变换理论是近几年发展起来一个新的数学分支。由于它克服了传统傅里叶变换的缺陷，具有良好的时、频局部化性能，从而使得小波理论在图像处理领域得到广泛的应用。首先对图像进行小波塔式分解，然后依据小波系数的统计特性和分布特点，对不同的子图像块采用不同的量化、编码方法。并在计算机上进行模拟实验，结果表明所提出的方案可以在保证重构图像质量良好的情况下获得较大的压缩比。     

A general method for automated analysis of caustics

In most published work on the optical method of caustics, experimental results were obtained by manually measuring some characteristic dimensions of the shadow pattern and, therefore, used only a limited amount of data from the caustic. However, some previous work has been done using digital image processing and light intensity plots combined with least-squares techniques to analyse the shape of a caustic obtained from a particular problem. All of these methods have been developed to solve only one specific problem. This paper highlights the essential guidelines of an automated method, which has the novelty of generalising the whole analysis, such that only the parametric equations of a particular problem are required to solve it. In the proposed method the image is digitised, the central line of the caustic pattern is identified, and the parameters in the mapping equation of the caustic are determined. The method has been validated, and the aim of this paper is to describe the key features of the method, i.e. the optical arrangement, digital image processing, and numerical analysis.     

A novel frame shift and integral technique for processing low-light-level moving image sequence

A new method for processing low-light-level moving image sequence is proposed, in which (1) a novel algorithm based on difference processing has been firstly developed to determine the motion parameters of moving image sequence, (2) the spatial relativity of the frames within moving image sequence can be then well established after frame shifting according to motion parameters obtained above, (3) finally, frame integration method can be applied to each processed frame, thereby increasing the signal-to-noise ratio of low-light-level moving image sequence. Experiments have been carried out to verify the validity of the proposed method, which show that the parameters obtained by the developed algorithm coincide well with the actual values, and the SNRs of the moving image sequences were effectively increased with the proposed method, indicating that the method offers significant advantages in the enhancement of low-light-level moving image sequence.     

基于VC++和MATLAB的氢原子光谱数据处理系统

在氢原子光谱实验中,由色散规律求氢原子光谱波长的数据处理十分复杂。本文利用3次样条插值和误差随机化方法来进行数据处理。利用VC++和MATLAB的混合编程,建立氢原子光谱数据处理系统。该系统充分发挥了VC++在界面设计上的优势,同时利用MATLAB处理复杂数据和表达图像的能力,因此该系统结构合理、使用方便。