窄带干涉滤光片抗击不同输出方式激光的损伤特性研究

利用Nd∶YAG调Q单脉冲激光和自由脉冲激光对硬膜窄带干涉滤光片进行激光损伤阈值的测试,并且采用表面热透镜技术测量了滤光片的吸收率。实验发现:窄带干涉滤光片的吸收率和激光损伤阈值强烈依赖于辐照激光波长与窄带干涉滤光片通带的相对位置;在调Q单脉冲激光作用下,不同中心波长的滤光片损伤形貌存在明显的差别,而在自由脉冲激光作用下,各滤光片的损伤形貌则趋于相同,均表现为典型的热熔烧蚀破坏。根据实验结果,结合损伤形貌,通过驻波场理论对激光作用下滤光片内电场分布的分析与模拟,探讨了两种激光模式作用下滤光片的损伤特征和损伤机理的不同特点。     

基于衰减 SPECT 投影数据的重建研究

基于 Novikov 反演公式针对扇形束投影得到了一种过滤背投影(FBP)算法,利用欧拉公式预先完成了取实步骤,使结果中不再需要取实,大大简化了后期的数据处理计算。同时采用坐标变换,使该算法能分别从平行束和扇形束几何特征下获取投影数据中重建图像。     

三种降采样成像策略的性能优化以及与传统傅里叶望远镜的比较

为了获得成像质量较好且成像时间较少的新型傅里叶望远镜成像策略,本文比较了三种降采样成像策略(压缩感知方法 (CS)、低频全采样方法 (LF)和变密度随机采样方法 (VD))与传统傅里叶望远镜(FT)在图像质量和成像时间上的差异.分析方法如下:利用传统FT外场实验所获得的目标频谱数据作为基础,三种降采样方法 (LF, VD和CS)分别按照各自的采样模式和重构方法实现目标图像的重构;通过直观观察和Strehl比两种方法比较三种降采样方法与传统FT在图像质量上的差异;通过分析成像时间的组成要素,初步比较三种降采样方法与传统FT在成像时间上的差异.分析表明:1)压缩感知方法的图像质量优于其他两种降采样方法 (LF和VD),但略低于传统成像结果; 2)压缩感知方法在成像质量上略低于传统FT,但在成像时间上却明显小于传统FT; 3)分析中采用的外场数据均含噪声,这说明上述三种降采样重构过程对噪声有较好的鲁棒性.综合上述分析结果可以看出,基于压缩感知的傅里叶望远镜(CS-FT)是在实际含噪情况下可大幅减少成像时间的优良成像策略.     

BrCN分子的电子碰撞离子对解离

离子对解离是一类重要的分子过程,常发生于分子被激发到超激发态. 与光激发的离子对解离实验研究不同,电子碰撞的相关过程研究尚存在实验挑战,特别是在测定其阈值方面. 本文报道了相关的利用单色化电子碰撞分子的实验研究进展. 以BrCN→Br^-+CN⁺离子对解离为例,根据CN⁺离子出现能测定其解离阈值为13.78 eV,同时在16.09 eV获得了CN⁺离子的时间切片速度影像且显示出动量分布的各向异性.     

融合形态学连通域和CV模型的民族服饰图案纹样元素分割方法

对民族服饰图案进行自动分割以提取图案纹样元素,是民族服饰图案素材库构建急需解决的难题。通过融合形态学连通域标记和CV模型（MCC-CV）,提出了一种民族服饰图案自动分割方法,首先对民族服饰图案进行预处理,然后采用形态学连通域标记算法获得待分割目标的位置和大致轮廓信息,对CV模型进行初始化,最后通过CV模型对不同分割目标进行边缘追踪,以实现民族服饰图案纹样元素的自动分割。实验表明,融合形态学连通域和CV模型的民族服饰图案纹样元素自动分割方法在边界召回率（BR）为0.5时,分割准确率为60%,与其他自动分割算法相比,该算法更为有效,满足了民族服饰图案素材库建设对图案纹样元素分割的基本要求。     

The fluorescence interference in Raman spectrum of raw coals and its application for evaluating coal property and combustion characteristics

Fluorescence interference in Raman spectrum is a big barrier for rapid and precise analysis of coal structures by Raman spectroscopy. Dealing with fluorescence interference suitably is one of the key tasks before efficient application of Raman spectroscopy in coal assessment. In this study, Raman spectra and coal combustion characteristics of 32 kinds of Chinese coals were respectively obtained in a micro-Raman spectrometer and Thermal Gravimetric Analyzer. The degree of fluorescence interference in Raman spectrum was firstly defined and quantified as the drift coefficient α using a simple method without curve-fitting the spectrum. The correlations between the degree of fluorescence interference and coal property, coal combustion characteristics were set up and multivariable analysis was done. The results indicate that the degree of fluorescence interference is well related to the coal structures, and it is synthetically determined by volatile, moisture and ash content in coal. With the increase of volatile, moisture content in coal, the fluorescence interference increases continuously, and it can be reduced but not eliminated by drying the moisture in coals. Significant mathematical relations between the drift coefficient α and volatile, moisture content, coal combustion characteristic temperatures have been found. Coal with more evident fluorescence interference in Raman spectrum usually has lower degree of coalification, more polar functional groups, and burns at a lower temperature. The drift coefficient α can act as an efficient probe for coal property and coal combustion characteristics. This study provided a new and simple approach for evaluating coal property and coal combustion characteristics by fluorescence interference in Raman spectrum.     

基于机器视觉的量子点STM形貌图像识别研究

为减轻量子点表面形貌分析过程中的人工工作,使量子点的STM图像分析更加自动化,基于机器视觉对衬底的斜切角及量子点的形貌特性展开研究.利用腐蚀和边缘检测提取台阶形状,并通过反三角变换计算斜切角.利用二值化和阈值下降对量子点的数量与空间坐标进行提取,在此基础上,通过邻域密度计算分析其均匀性,并在解决图像中的粘连问题后找出量子点的尺寸.实验结果显示,与人工统计相比,斜切角、量子点计数及尺寸的平均误差分别为5.02%, 0.7788%及1.12%;并实现量子点均匀性的自动化统计与分析.基于机器视觉算法的自动识别过程,对协助研究者分析量子点表面形貌有实际意义.     

基于主成分分析和分层树集合划分的Huffman算法图像压缩研究

互联网的飞速发展,产生了大量的图像信息.为了减少图片占用的存储空间,提高图像质量,提出了一种将主成分分析（PCA）和分层树集合划分（SPIHT）压缩算法相结合的有损图像压缩算法.首先对图像进行主成分分解,选取主要特征值进行压缩,再利用SPIHT算法将图像分解成不同子带的小波系数进行压缩,对SPIHT压缩系数进行哈夫曼编码,实现图像二级压缩.将本文提出的算法与SPIHT、SPIHT的哈夫曼编码、JEPG2000、PCA压缩算法进行了比较,结果表明本算法较其他压缩算法具有更好的性能,在压缩比相同的情况下能获得更高的PNSR和SSIM.     

一种基于特征分解的图像融合方法

一幅图像可以分解成几何特征不同的纹理部分和卡通部分,基于这两大特征提出了一种图像融合方法.利用卡通和纹理特征的差异,通过学习分别得到卡通字典和纹理字典.在融合过程中,分别利用特定的卡通和纹理字典对源图像的卡通和纹理部分进行融合,融合后的卡通和纹理部分经简单相加得到融合图像.实验结果表明,所提方法是有效的.     

Experimental investigation on the wake interference among wind turbines sited in atmospheric boundary layer winds

We examined experimentally the effects of incom-ing surface wind on the turbine wake and the wake interfer-ence among upstream and downstream wind turbines sited in atmospheric boundary layer (ABL) winds. The experi-ment was conducted in a large-scale ABL wind tunnel with scaled wind turbine models mounted in different incom-ing surface winds simulating the ABL winds over typical offshore/onshore wind farms. Power outputs and dynamic loadings acting on the turbine models and the wake flow char-acteristics behind the turbine models were quantified. The results revealed that the incoming surface winds significantly affect the turbine wake characteristics and wake interference between the upstream and downstream turbines. The velocity deficits in the turbine wakes recover faster in the incoming surface winds with relatively high turbulence levels. Varia-tions of the power outputs and dynamic wind loadings acting on the downstream turbines sited in the wakes of upstream turbines are correlated well with the turbine wakes charac-teristics. At the same downstream locations, the downstream turbines have higher power outputs and experience greater static and fatigue loadings in the inflow with relatively high turbulence level, suggesting a smaller effect of wake inter-ference for the turbines sited in onshore wind farms.