Spatial and spectral characteristics of laser-field synchrotron radiation

The spatial and spectral characteristics of laser-field synchrotron radiation (LSR) are investigated. The results show that LSR is emitted normal to the laser axis and emission in the X-ray wavelength can be obtained by increasing the intensity of the laser pulse and the initial energy of the electron. The unique feature offered by the laser-field synchrotron radiation (LSR) makes them a promising compact source of light at shorter wavelengths.     

Formation of aligned ZnO nanorods on self-grown ZnO template and its enhanced field emission characteristics

We report a novel method for producing aligned ZnO nanorods (ANR) on self-grown ZnO template in a single step process involving growth of ZnO by vapor transport, followed by quenching of growing ZnO flux in liquid nitrogen. In the present study Zn powder turns into ZnO sheet under oxygen flow at ∼900 °C and bottom surface of the sheet acts as template for the growth of ANR. It is revealed from XRD and EDAX analysis that the bottom of the sheet is Zn rich region and acts as self catalyst for the growth of ANR. The grown nanorods have length up to several tens of micrometers with diameters ranging from ∼100 to 150 nm. Microstructural analysis of ANR indicates the fractal like configuration. The field emission properties have been investigated for ANR with fractal geometry using the ANR on self-grown ZnO template as a cathode directly. The turn-on electric field required to draw current density of ∼1.0 μA/cm² has been found to be ∼0.98 V/μm. The field enhancement factor based on Fowler-Nordheim (F-N) plot was found to be ∼7815 for ANR. The fractal geometry of ANR has been shown to be advantageous for achieving improved field emission features. The present investigations of synthesis involving formation of ANR over self-grown ZnO template, together with fractal configuration of the as-synthesized ANR, are first of their type.     

基于视线的眼动跟踪人机交互系统研究

介绍了系统的工作原理,建立了信息交流和控制的系统结构,将面向对象和基于事件驱动的处理方法应用于通信协议的建立和通信控制过程.结合具体通信控制单元三菱PLC和FX1N-232-BD与PC串口RS-232C的通信,编写了算法和通信控制的代码.     

基于Chebyshev法实现LPC与LSF相互转换

在语音信号的处理中,常需要将线性预测系数(LPC)与线谱对频率参数(LSF)相互转换.利用余弦函数特性改进的Chebyshev多项式求根算法,将余弦函数转换为高次幂函数再进行搜索求根来实现语音特征的LPC参数与LSF参数的转化.     

基于谱减算法语音增强的研究

谱减法是处理宽带噪声较为传统和有效的方法,它的运算量较小,容易实时实现,增强效果也较好.比较详细地叙述了谱减法的基本原理和降低音乐噪声的方法.     

基于功率谱纹理分析的指纹密钥算法

由于生物特征的模糊性,传统的自动指纹识别使用存储特征模板和模糊匹配的方法,存储的模板直接含有大量指纹信息,由此带来模板安全隐患问题。文中探讨在无模板存储和匹配时提取稳定的参数描述,创新性地提出了一种从频谱中提取指纹特征并描述为稳定密钥的方案,借鉴Fourier频谱分析方法从指纹中提取了反映脊线频率信息的环特征和描述方向信息的楔特征。最后提出多次采集整合的改进方案,得到了稳定率82%的密钥参数。     

一种针对ISF参数的量化算法

宽带语音编码中普遍使用导抗谱频率描述声道。利用转换分类差矢量分裂矢量量化方法对导抗谱频率进行量化,该方法基于转换分类矢量量化及差值分裂矢量量化。IsF矢量先按照给出的码书分类,然后每一类中的差矢量再进行分裂矢量量化。实验结果表明,该算法可在每帧编码比特数为37时达到透明量化要求,并且码书存储量明显少于StephenSo等人给出的转换分类分裂矢量量化方法。     

基于WBCT与平滑共生矩阵的图像检索

利用WBCT变换良好的稀疏特性及其能准确地捕获图像中边缘信息的特性,分析了纹理图像WBCT系数的统计特征,提出了一种滤波算法。该算法根据纹理图像WBCT系数分布的特点,提取纹理特征。加入在低频子带上提取的灰度—平滑共生矩阵统计量,形成最终的特征向量。仿真实验结果表明,该方法在纹理图像检索上有一定的优越性。     

基于MFCC和VQ码书的说话人识别系统研究

随着信息技术和通信技术的迅速发展,说话人识别越来越受到重视,目前在说话人识别中,要提高识别率需要解决如何选取合适的有效的表征说话人特征的可靠参数和选取恰当的识别模型,文中在matlab实现了一个说话人识别系统,并在此基础上对系统进行了改进。     

基于Hopfield的脱机手写数字识别理论及算法

脱机手写数字识别在很多领域具有广泛的应用前景,国内外学者对此做了大量的研究工作,提出了很多预处理和模式识别的算法,大大提高了手写数字的识别精度。为了提高手写数字识别的精度,文章将Hopfield神经网络应用于脱机手写数字识别分析中,Hopfield神经网络的“能量函数”的能量在网络运行过程中,具有不断地减少最后趋于稳定的平衡状态的特性,而且网络一旦建立即可自动运行,无需要训练。