基于多传感器数据融合的智能小车避障的研究

针对智能小车避障问题,提出了一种将模糊逻辑和神经网络相结合的融合方法—Takagi-Sugeno(T-S)模糊神经网络方法。基于此方法的数据融合算法应用在智能小车避障运动中,采用多只超声波传感器和红外线传感器探测障碍物的距离和方向,采集的各种数据利用T-S模糊神经网络进行融合。通过实验仿真表明:此方法能够使智能小车对障碍物的灵活避障和导航行进。     

基于混合模型的CCD镜头畸变精校正算法

针对机器视觉检测和高精度图像测量中使用的CCD镜头都存在不同程度光学畸变的问题,提出基于混合模型的CCD镜头畸变校正算法。用经典模型对畸变图像进行初次校正,用多面函数拟合法进行二次精校正,用三次B样条函数进行灰度重建。实验结果表明,该方法在不依赖摄像机内部参数条件下,相比单一的镜头畸变校正模型,精度提高,鲁棒性增强,校正后径向均方根误差为0.3个像素。     

面向社会Agent的网络监测系统框架

在社会Agent理论的基础上,提出面向社会Agent的网络监测系统框架。结合软件复用的思想,根据网络监测数据,将该框架分为自治功能层、个体意识层和社会意识层的3层监测系统抽象模型。基于该框架设计的网页内容访问监测软件开发速度快、资源复用率高和鲁棒性强。     

基于第三方的安全移动支付方案

在现有移动支付方案研究的基础上,提出一种新的基于第三方的安全移动支付方案。该方案以第三方支付平台为基础,在交易过程中采用“一次一密”的密钥分配机制,改善了现有移动支付方案的缺陷,在安全性上实现交易信息的保密性、不可伪造性及不可否认性等特性,并利用串空间模型的形式化分析方法对相关协议进行安全性证明。     

数据挖掘在银行CRM中的应用研究

该文阐述了数据挖掘和CRM的概念,并对在银行CRM中如何使用数据挖掘技术进行了研究,并给出了银行CRM中B1分析的一个简单的应用方式。     

ATmega16与PC机的移动机器人定位系统研究

为提高移动机器人定位系统的可靠性,设计了组合使用光纤陀螺仪、光电码盘和超声波传感器的定位系统,系统采用CAN总线的数据传输方式。ATmega16采集各传感器数据,再以CAN总线方式传输给PC机;PC机平台综合处理光纤陀螺、光电码盘与超声波返回的数据,实现移动机器人定位。定位算法以航迹推算为主,超声波传感器起辅助定位作用。实验表明定位系统可靠有效。     

应用改进的C-V模型分割牛肉眼肌图像

牛肉眼肌图像的分割是牛肉自动分级的重要步骤。基于水平集的C-V模型采用两个具有不同均值的分块连续区域分割图像,在迭代过程中,为了约束水平集保持为符号距离函数,必须对其进行重新初始化,大大增加了计算量,降低了曲线演化速度。为此,通过加入边缘检测函数和惩罚项因子,提出了改进的C-V水平集模型,并用于牛肉图像分割,同时与模糊C均值聚类、阈值分割法进行了对比。结果表明:采用改进的C-V模型对牛肉图像进行分割,准确提取了眼肌边缘,分割出了脂肪和肌肉区域,准确率高;曲线演化过程中无需对水平集函数符号初始化,加快了曲线演化速度。     

基于XML的电子文档智能化编辑方法研究

分析了信息系统提供的电子文档编辑功能和实际应用中在该方面的需求,针对电子文档编辑效率低问题,提出一种文档智能编辑方法。该方法采用XML技术,将文档内容归纳为既有文本、输入文本、单选文本和多选组合文本4种类型,分别采用编辑框、单选框和复选框控件界面,实现文档内容的智能交互编辑。实验结果表明,该方法缩短了电子文档的编辑时间,为信息化系统电子文档快速编辑提供了新的技术。     

Preliminary study evaluating manual dexterity tests assessing gloves protecting against cuts and stabs by hand knives

The paper presents preliminary results on the research methodology of evaluating the ergonomic properties of gloves protecting against cuts and stabs by hand knives. Four manual dexterity tests were selected for the study: TEST 1: evaluation of fine finger movements pursuant to PN-EN 420; TEST 2: cylinder grip and pull test according to PN-EN 1082 evaluating gross movements of the arms and hands; TEST 3: Purdue Pegboard Test evaluating fine finger movements; TEST 4: evaluation of gross movements of the arms and hands while performing simulated occupational tasks. The tested gloves differed in terms of construction and material (metal mesh or knitwear incorporating a metal yarn) and were selected depending on the scope of work activities performed during meat cutting and boning. The tests were conducted on a homogeneous group of subjects. During the performance of the four dexterity tests, we monitored the loading of four groups of muscles of the upper limb (adductor pollicis, extensor carpi radialis, flexor carpi ulnaris, and biceps brachii) using surface electromyography (EMG). Additionally, subjective sensations concerning the strain on the upper limb were evaluated using a questionnaire survey.The objective of the study was to identify a group of tests that would be characterized by high sensitivity and ensure reliable assessment of gloves protecting against cuts and stabs by hand knives. The results showed that not all of the dexterity captured differences in the way the gloves affected dexterity. This was corroborated by electromyographic measurements, which revealed considerable differences in load of upper limb muscles when using various gloves and further supported by the subjective sensations of the participants, as reported in the questionnaire study. It was concluded that in order to reliably evaluate the ergonomic properties of gloves protecting against cuts and stabs by hand knives it is necessary to design new dexterity tests that would reflect actual workplace conditions, and could simulate occupational activities. Workplace observations helped to identify specific aspects of manual dexterity (fine, medium, gross), and types of hand movement associated with professional activities including different force configurations (finger flexion, wrist abduction, and pressing with the fingers and wrist). Based on those observations three new manual dexterity tests dedicated exclusively to evaluation of the ergonomic properties of gloves protecting against cuts and stabs by hand knives have been proposed.     

基于图像层级的机器素描研究

针对运用机械臂绘制的机器素描图像存在缺失真实感、艺术感,硬件设备复杂,难以普及的问题,提出基于图像层级的机器素描方法.该方法在运用灰度变换、滤波、二值化等图像预处理方法处理原始人物图像的基础上,分别提取出图像轮廓和图像内部填充区域两部分二值目标图像,通过简易路径规划遍历目标图像,根据遍历轨迹控制XY型简易机械臂绘制图像轮廓及按层级绘制模式填充图像内部区域,实现人物素描图像的绘制.实验结果表明,该系统可根据图像轮廓实现任意曲线的绘制,通过层级绘制模式包含环境因素达到了衬托素描图像使其更具层次感的目的.