基于SOM神经网络聚类的车牌号码倾斜校正算法

经过摄像机摄入的图像会发生倾斜,给车牌的准确识别带来了困难.针对此问题,利用SOM神经网络良好的聚类性能,在水平倾斜校正时,把车牌号码图像中的像素坐标聚成两类,拟合成一条直线,计算出该直线倾斜角,完成水平校正;按照以上同样方法进行垂直倾斜校正.实验结果表明,该方法能准确获取车牌号码的倾斜角,算法结构简单,抗干扰能力较强,符合汽车牌照图像的特点,具有较好的处理效果.     

基于VSM的自适应信息过滤系统的研究与设计

为了从网络动态信息流中获得感兴趣的内容或过滤掉无关的垃圾信息,设计了一个基于向量空间模型的自适应信息过滤系统;描述了该系统的结构和工作流程;对该系统实现的关键技术,包括文本表示、用户模板与过滤阈值初始化、特征项选取、自适应过滤算法、模板更新和阈值调整等进行了深入的分析和研究。     

A Change-Point Approach for Phase-I Analysis in Multivariate Profile Monitoring and Diagnosis

Process monitoring and fault diagnosis using profile data remains an important and challenging problem in statistical process control (SPC). Although the analysis of profile data has been extensively studied in the SPC literature, the challenges associated with monitoring and diagnosis of multichannel (multiple) nonlinear profiles are yet to be addressed. Motivated by an application in multioperation forging processes, we propose a new modeling, monitoring, and diagnosis framework for phase-I analysis of multichannel profiles. The proposed framework is developed under the assumption that different profile channels have similar structure so that we can gain strength by borrowing information from all channels. The multidimensional functional principal component analysis is incorporated into change-point models to construct monitoring statistics. Simulation results show that the proposed approach has good performance in identifying change-points in various situations compared with some existing methods. The codes for implementing the proposed procedure are available in the supplementary material.     

Effects of T5 heat treatment on extrusion welds in AZ80 hollow profile

AZ80 hollow profile was produced by porthole-die extrusion. The mechanical behavior and microstructure characteristic of extrusion welds in AZ80 hollow profile during different T5 heat treatments were investigated. Five kinds of morphologies of β-Mg₁₇Al₁₂ were observed in turn in weld regions under various temperatures. Extrusion weld is not weakest region in AZ80 hollow profile in terms of hardness, the peak hardness of weld regions is roughly equal to the weld-free regions at different aging temperatures, which is attributed to similar volume fraction, morphology, and size of β-Mg₁₇Al₁₂ precipitate in weld regions and weld-free regions. The precipitate-free zones (PFZs) usually form near the welds interface in aging process, and the hardness of PFZs is usually lower than those of weld regions and weld-free regions. Aging at 200°C for 24?h is suitable heat treatment system for AZ80 hollow profile, which can achieve excellent mechanical property and simultaneously reduce the width of PFZs.     

Global optimum of microstructure parameters in the CMWP line-profile-analysis method by combining Marquardt-Levenberg and Monte-Carlo procedures

Line profile analysis of X-ray and neutron diffraction patterns is a powerful tool for determining the microstructure of crystalline materials. The Convolutional-Multiple-Whole-Profile (CMWP) procedure is based on physical profile functions for dislocations, domain size, stacking faults and twin boundaries. Order dependence, strain anisotropy, hkl dependent broadening of planar defects and peak shape are used to separate the effect of different lattice defect types. The Marquardt-Levenberg (ML) numerical optimization procedure has been used successfully to determine crystal defect types and densities. However, in more complex cases like hexagonal materials or multiple phases the ML procedure alone reveals uncertainties. In a new approach the ML and a Monte-Carlo statistical method are combined in an alternative manner. The new CMWP procedure eliminates uncertainties and provides globally optimized parameters of the microstructure.     

基于用户浏览行为的用户模型调整算法研究

个性化信息服务越来越成为信息检索领域中研究的热点。针对用户模型构造问题,文章结合手工定制建模与自动分析建模技术;并改进文档特征向量的权重计算方式,提高特征向量表示文档的能力;并结合用户浏览行为的页面反馈度计算,提出了一种用户模型智能调整算法。模拟实验表明,该结构和算法能够有效地提高检索结果的准确度,并且具有良好的可适应性。     

基于进程执行轮廓的缓冲区溢出攻击效果检测

缓冲区溢出攻击效果检测对缓冲区溢出安全防御工作具有重要意义,该文分析进程与Windows NativeAPI的关系,以Windows NativeAPI为数据源进行攻击效果检测。提出执行轮廓的概念及其建立方法,在分析缓冲区溢出攻击效果的基础上,提出基于进程执行轮廓的缓冲区溢出攻击效果检测方法,实验结果表明该方法的有效性。     

南水北调中线小南河串流片防洪水位计算

传统的多河流串流计算主要采用水库联调的方法,把串流的各条河流概化为静态水库,库水位水平且不考虑横向流动。为解决串流区水流的横向流动及水面线并不总像库水位一样水平的问题,采用二维非恒定流数学模型计算串流区洪水演进。通过概化地形、考虑入流、糙率、边界条件和水流的横向流动等因素,计算出了整个串流区域的水面线、水深、流速、出流等水力要素,并给出洪水演进模拟图,为确定河流与总干渠交叉建筑物的规模、总干渠左岸沿线水面线和对沿线村庄、农田的防洪影响提供了可靠的依据。     

A synergistic model-driven approach for persistence modeling with UML

The Model Driven Development (MDD) approach proposes that models (and model-to-model transformations) play the main role on system development. However, there is not a consensual notation to model persistence based upon object-relational mapping frameworks: while UML lacks specific resources for persistence modeling, the entity-relationship model does not make reference to the dynamic concepts existing in UML.This paper proposes MD-JPA, a UML profile for persistence modeling based on the well-known Java Persistence API 2 (JPA) standard for object-relational mapping, pursuing the modeling of transient and persistent elements in a more coherent and synergistic way. This paper describes the main characteristics of MD-JPA as well as the way that models that adopt such profile can them be used to generate a Java implementation by the application of the proposed model transformations on a MDD approach. Finally, an open source tool was developed to make the results of this work available to the community.     

基于支持向量机的中国地鼠分类特征基因选取

针对中国地鼠基因表达谱数据维数高和样本小的特点,提出一种基于支持向量机(SVM)的分类特征基因选取方法。该方法利用改进的Fisher判别(FDR)基因特征计分准则剔除分类无关基因,提出由空间距离和功能距离组成的新距离作为相似性度量的标准进行冗余基因的剔除,采用SVM作为分类器检验特征基因的分类性能。实验结果表明,该方法有效地剔除了分类无关基因和冗余基因,选取的特征基因满足对中国地鼠正确分类的最小基因数。