基于PCA的红外图像区域特征选择方法

特征提取在模式识别和分类中起着关键的作用,本文针对红外目标的准确分类识别问题,围绕红外图像特征提取和特征选择进行研究,提出了基于PCA的区域特征选择方法。该方法首先采用基于数学形态学的滤波技术对红外图像进行预处理,有效地增强了目标区域,便于目标特征的提取;其次,本文研究了区域形状特征提取及基于PCA的特征选择方法,通过对区域特征进行优化选择,构造准确描述目标特性且维数较低的特征。实验结果表明,本文提出的方法有效地提取红外目标的特征,可用于红外目标的分类且有利于提高算法的效率。     

FLAC3D前处理程序的开发及仿真效果检验

在FLAC^3D软件提供的初始单元模型（primitive mesh)的基础上，采用Fortan语言编写了FLAC^3D前处理程序，在工程实践中，实现更准确，快捷地建立复杂的FLAD^3D计算模型，并探讨了其基本思路，计算方法以及该程序的的突出特点，最后，以工程实例检验了此程序的处理功能和仿真效果。     

多层导电结构缺陷电涡流扫描检测信号预处理技术研究

应用小波多分辨分解及重构技术和小波包分析技术进行检测信号预处理(噪声和干扰信号分离与去除)的基本原理,结合电涡流扫描检测实验,采用离散小波变换强制消噪法、软阈值消噪法和不同熵准则的小波包分析消噪法对检测信号进行预处理,并以SNR和RMSE为判断消噪效果好坏的标准,进行了效果的比较和优选.从理论分析和实验研究结果可知,分离提离等干扰信号,可采用强制法;消除高频噪声,采用基于Shannon熵准则的WPA法,效果最好.     

一种扩展水下扩频信号多普勒容限方法

扩频信号的模糊函数图呈图钉型,对多普勒频移比较敏感,影响了扩频信号的使用性能.本文分析了扩频信号的多普勒敏感性问题,提出了一种基于回波预处理的方法.首先将回波信号经过正交混频得到零中频I,Q通道信号,然后将零中频I,Q通道信号送入多普勒预处理系统进行处理,通过时间延迟处理方法得到的2个输出信号,分别经过匹配滤波器后再进行合并处理,脉压输出相关函数中不含多普勒频移和时间乘积,消除了多普勒频移对匹配滤波器输出的影响.数值仿真证明了该方法的有效性,且具有实现简单,运算量小的优点.     

一种高性能低功耗的运动估计的设计

基于运动估计的全搜索算法,人们提出了有预处理的脉动阵列硬件结构,相较于无功耗处理的原始脉动阵列减少了12.3%。在此基础上,进一步根据逻辑产生门控时钟以及停止寄存器链对电路进行管理,取得了69.2%的功耗节省。     

海量道路路面测量数据的若干预处理方法研究

回顾了道路路面数据的测量系统的构成和测试方法。针对海量的道路路面试验数据,将若干信号处理的预处理方法,包括粗大误差的甄别与剔除、信号趋势项的提取、分段试验数据的平滑过渡连接、路面数据随机性检验和车速异常时刻的处理等应用于海量道路路面数据的分析处理中,并进行了相应的改进。在分析这些预处理方法原理的基础上,使用实际采集的试验路面数据作为例子,充分显示了这些方法的处理效果。实例分析结果表明,这些信号预处理方法用来处理和修正道路路面试验数据,使用简单方便,处理效果明显,可在道路路面数据处理及其它工程振动信号预处理中推广应用。     

A computer-based system to support forensic studies on handwritten documents

Computer-based forensic handwriting analysis requires sophisticated methods for the pre-processing of digitized paper documents, in order to provide high-quality digitized handwriting, which represents the original handwritten product as accurately as possible. Due to the requirement of processing a huge amount of different document types, neither a standardized queue of processing stages, fixed parameter sets nor fixed image operations are qualified for such pre-processing methods. Thus, we present an open layered framework that covers adaptation abilities at the parameter, operator, and algorithm levels. Moreover, an embedded module, which uses genetic programming, might generate specific filters for background removal on-the-fly. The framework is understood as an assistance system for forensic handwriting experts and has been in use by the Bundeskriminalamt, the federal police bureau in Germany, for two years. In the following, the layered framework will be presented, fundamental document-independent filters for textured, homogeneous background removal and for foreground removal will be described, as well as aspects of the implementation. Results of the framework-application will also be given. Received July 12, 2000 / Revised October 13, 2000     

三级大豆油酸价的近红外光谱检测

为改进现有的油脂检测方法,以三级大豆油脂酸价检测为研究对象,利用近红外透射光谱技术,结合化学计量学方法,选择4500～6000cm-1为特征波段建立油脂酸价校正模型,首先研究基于小波变换的光谱预处理方法,通过详细比较不同小波分解层数对建模的影响。结果确定db4小波的四层分解去噪效果最佳。结合滤波后重构的光谱信号建立油脂酸价BP神经网络校正模型,利用预测集样本对模型进行验证,决定系数R2和预测均方根误差分别为0.9743和0.1036。证明利用近红外光谱分析技术快速检测油脂酸价是完全可行的。     

Development of foundation models for Internet of Things

With the advent of the Internet of Things (IoT) that offers capabilities to identify and connect worldwide physical objects into a unified system, the importance of modeling and processing IoT data has become significantly accentuated. IoT data is substantial in quantity, noisy, heterogeneous, inconsistent, and arrives at the system in a streaming fashion. Due to the unique characteristics of IoT data, the manipulation of IoT data for practical applications has encountered many fundamental challenging problems, such as data modeling and processing. This paper proposes the infrastructure for an IoT prototype system that aims to develop foundation models for IoT data. We illustrate major modules in the IoT prototype, as well as their functionalities, and provide our vision of the key techniques used for tacking the critical problems in each module.