一种在MPEG-4编码过程中提高实时性的算法

本文提出了一种新的在MPEG-4编码过程中提高实时性的算法。算法采用双重筛选机制,充分利用了运动估计的结果提取运动对象。首先在运动估计过程中加入背景优先准则进行第一重筛选;然后根据运动估计后宏块的最小平均绝对差值分利用区域自适应门限进行第二重筛选,从而获得运动区域掩模;最后对掩模进行形态滤波和填充消除噪声点,最终得到运动区域。算法分别在硬件和软件平台上进行测试,结果表明提出的算法得到了理想的分割效果,较传统的在编码前进行运动对象提取的方法实时性大大提高了。     

解读大学人事管理的以人为本原则

大学人事管理的对象是教学科研的主体——教师。大学教师职业的独特属性要求在大学人事管理中对以人为本原则给予充分解读。本分析了大学教师职业独特的内在属性和作用于大学教师队伍的几对主要矛盾。阐述了以人为本原则对大学人事管理的启示。     

虫媒登革病毒的研究——Ⅱ.登革病毒非结构蛋白NS1的纯化、鉴定及其在临床检测应用

本文从纯化Den2 NS1,Mab入手,自制了Mab亲和层析系统,从感染Den2的细胞膜中,分离提取了天然的Den 2 NS1双体,比较其NGC株与我国地方株(H-87)NS1的理化性质及组成,并对某些重要的生物学性质进行了综合性研究。同时,利用上述研究结果建立了检测DF病人血清特异NS1抗体的新方法,探索其协助临床诊断及研究的可能性。     

智能型入侵检测模型的设计

针对目前大多数入侵检测系统存在的局限性,提出了一种较为完善的入侵检测模型,即将专家系统和神经网络技术同时应用于入侵检测系统中,重点分析了规则的生成.为进一步提高检测性能,提出了遗传算法优化BP(Back Propagation)神经网络,经论证本设计降低了漏报率和误报率.     

The interplay of representations and patterns of classroom discourse in science teaching sequences

The purpose of this study is to examines the relationship between the communicative approach of classroom talk and the modes of representations used by science teachers. Based on video data from two physics classrooms in Singapore, a recurring pattern in the relationship was observed as the teaching sequence of a lesson unfolded. It was found that as the mode of representation shifted from enactive (action based) to iconic (image based) to symbolic (language based), there was a concurrent and coordinated shift in the classroom communicative approach from interactive–dialogic to interactive–authoritative to non-interactive–authoritative. Specifically, the shift from enactive to iconic to symbolic representations occurred mainly within the interactive–dialogic approach while the shift towards the interactive–authoritative and non-interactive–authoritative approaches occurred when symbolic modes of representation were used. This concurrent and coordinated shift has implications on how we conceive the use of representations in conjunction with the co-occurring classroom discourse, both theoretically and pedagogically.     

基于Python语言的模式识别应用

以模式识别中英文文字识别和人脸检测为例,利用PIL库、pytesser库和OpenCV库实现了图像灰度转换、降噪处理、文字识别和目标检测。应用表明,Python处理实验数据的方法简便灵活,表征数据图形直观明了,适合在数字图像处理课程的模式识别章节中应用。     

Clinical detection and movement recognition of neuro signals

INTRODUCTION Human limbs are complex, delicate systems.Their reconstruction, both for appearance and forfunction compensations, proves to be complicatedprojects. With the development of science and tech-nology, both researchers and disabilities hope to haveartificial limb much more perky, comfortable, dex-terous and effective. Although thousands of pros-thetic limbs are commercially available worldwide,most of them, from mechanical styles to electricstyles, from motion controlled t…     

基于AT89C51单片机的苯类检测仪的设计

基于AT89C51单片机的苯类检测仪由AT89C51单片机、苯类传感器、12位A／D转换器TLC2543组成,以液晶屏显示苯类含量．软件系统包括初始化程序、数据采集程序和显示程序三部分．     

基于视频的城市快速路交通流检测及车辆行为监控

提出一种针对城市快速路的视频监控系统,完成交通流及道路异常行为检测.流量检测模块对车辆进行实时跟踪获得各类交通流统计数据,实现快速路车辆运行状况的全面掌控.异常监控模块运用隐马尔科夫模型,对带有时间和空间信息的车辆轨迹进行训练,获得路径划分后对道路车辆轨迹进行参数匹配,提取诸如超低(高)速行驶、违章停车、违规掉头等异常...