基于机器视觉的柱状弹簧质量综合检测系统

针对传统弹簧测量精度差和稳定性低的问题,提出一种基于机器视觉的弹簧尺寸与表面质量综合检测系统。采用远心镜头配合平行光源的方式获取弹簧图像,利用直方图均衡化对图像进行增强,利用形态学开运算使弹簧表面突变更加明显,以获取表面信息。推导了弹簧参数测量与计算的公式,建立了弹簧圆度与轴向直线度的测量模型,并搭建硬件系统进行测量。测量结果表明:系统测量误差在0.2-1.2μm范围内,重复测量结果保持稳定,系统的表面检测效果较优,该系统可用于柱状弹簧的综合测量。     

基于深度学习的电梯平层精度检测系统研究

电梯是日常生活中必不可少的一种交通工具,电梯平层精度的测量可以有效减少电梯事故的发生,为了实现对电梯平层精度的快速准确检测,开发了一套基于深度学习的电梯平层精度检测系统。系统将刻度尺安装在电梯停留时电梯井道的相应位置,在每个楼层安装一个刻度尺并进行校准,在电梯轿厢底部安装激光传感器、工业相机和光源,通过激光传感器对电梯轿厢位置在刻度尺上进行精准定位,通过在电梯轿厢底部安装光源以获取高质量的图像,通过工业相机获取实时获取激光线所在位置的图像信息。实验过程中共采集500张图像数据,通过随机旋转、镜像、裁剪、添加噪声等数据增强方式将图像样本数据扩充至5000张,通过开源软件LabelImg辅助标注工具对每张图像中的激光线和激光线最临近的刻度值进行标注,对标注的图像进行转化为可用于训练的电梯平层精度检测数据集,再将数据集按照4∶1划分为训练集和测试集,最后利用YOLOv3目标检测算法对刻度线图像数据集进行训练并测试,实现对激光线和激光线最邻近位置数字的检测和识别。在测试集中随机选取10组图像进行测试,采用该检测方法对电梯平层精度进行检测,对实验结果进行分析准确率高达90%。该系统的研究实现了电...     

基于特征提取的车道线快速检测方法研究

车道线检测是无人驾驶车辆及车道偏离预警等系统的关键技术。针对自动驾驶车辆当前道路可行驶区域检测精度不高、环境适应性差及实时性差的问题，提出了一种基于特征提取的快速检测车道线的方法。在图像预处理阶段，首先读取视频图像，把每一帧RGB图像转为灰度图，通过Canny算子对图像的边缘轮廓进行提取，然后绘制车道线的掩码区域，并与边缘检测结果图结合，对感兴趣区域（ROI），也就是对车道线所划分的汽车当前可行驶区域进行提取。最后进行概率霍夫变换和最小二乘拟合，并将得到的直线绘制到原图像中，最终对每一帧处理后的图像进行输出。该算法在多种边缘检测方法中选用Canny算法，并针对人为选取双阈值的复杂性问题，提出了自适应调节、选取阈值的方法；运用改进的概率霍夫变换对车道线进行检测，并进行离群值的过滤，有效提高了检测精度和速度。结果表明：该算法处理每一帧图像的时长为13.9 ms，且平均准确率达到97.6%，具有较好的实时性和准确性。     

机器视觉在气门盘检测中的应用

介绍一种气门盘在线检测系统。该系统采用非接触式方法进行图像的采集,利用一种改进的SUSAN算法对图像进行边缘检测,很好的保留了气门盘的边角特征。最后采用尺度标定,在不增加硬件成本的前提下,达到检测参数的精度和速度设计要求。     

饮料液位自动检测系统设计

为了提高饮料液位检测的效率,介绍了一种新型的瓶装饮料液位自动检测系统。该系统采用基于机器视觉的检测方法,采用CCD作为检测元件。并基于LabVIEW和IMAQ Vision,开发出了对饮料瓶进行图像获取、图像检测识别的软件。按识别结果进行分拣,从而摆脱了人工分拣的缺点,提高了检验的效率,实现了塑料瓶装饮料罐装线的全自动、非接触和在线检测。研究结果表明,该系统能有效地检测出饮料的液位位置,并判断灌装是否合格,具有准确率高、运行稳定、检验速度快等优点。     

一种基于逆透视变换的车道线检测方法

该文基于逆透视变换技术提出了一种车道线检测方法。将道路图像映射为逆透视图,用二维高斯滤波器对图像进行滤波,然后进行阈值处理;用Hough变换提取车道线的线条曲线,利用三阶Bezier样条曲线拟合算法进行直线拟合,获得车道线的检测结果。实验表明,该方法在图像较清晰的情况下,检测近距离的车道线获得了较好的效果,适用于平坦路面下的白色和黄色等车道线检测。     

基于数学形态学和MATLAB的车道线边缘检测方法

针对道路图像中的车道线边缘检测问题,提出了利用数学形态学和MATLAB进行车道线边缘检测,讨论了对其中出现的灰度图像转化、图像预处理、二值化图像、边缘检测等问题进行处理的关键方法,给出了基于数学形态学和MATLAB的车道线边缘检测方法。实验结果有效地显示和验证了此方法的高效性。     

振动环境下显示设备显示性能分析系统原型

研究了在振动环境下,基于计算机视觉技术实现显示设备显示性能客观分析的方法。给出了系统的硬件组成与软件原型的总体结构,说明了硬件系统设计和设备选型的方法;阐述了视频运动补偿、图像中线检测与定位等关键技术的实现方法。软件模型由C++编程实现,实验结果表明该系统能够采集清晰的图像,能够实现亚像素级的目标图像检测分析,并输出检测结果。     

激光刻写零件瑕疵图像在线检测算法

针对复杂背景条件下激光刻写零件瑕疵的机器视觉检测需求,提出一种基于导向线的刀路轮廓质量检测方法。先根据矢量文件生成导向线,对导向线中的线路与面域进行区分,再与待测图像配准,在导向线上生成采样点,最后对已滤波图像上轮廓的线宽、连通性和毛刺情况进行检测。有别于传统灰度或轮廓的差分检测方法,所提方法以导向线为路径,利用检测模板对采样点所在的局部图像进行检测。实验结果表明,该方法在复杂背景条件下对轮廓检测既具有良好的鲁棒性,又能够保证高检测精度与较好的实时性,满足激光刻写零件过程中的质量检测要求。     

基于机器视觉的扁弹簧分类方法研究

为实现扁弹簧的自动化分类,提出一种基于机器视觉的扁弹簧分类方法。首先对采集到的不同型号的扁弹簧图像进行预处理,对预处理图像提取黑色像素分布向量,作为特征参数向量。然后采用DDAG SVMs方法将不同型号的扁弹簧进行分类,最终对采集到的不同型号扁弹簧测试样本进行分类测试实验。结果表明,该分类方法效率高、准确率高,可以满足扁弹簧自动化分类的各项要求。     

囊式空气弹簧刚度特性研究

弹簧刚度作为空气弹簧最重要的参数,是隔振系统设计的基础。以囊式空气弹簧为研究对象,在深入分析空气弹簧受力及囊体材料特性的基础上推导了刚度计算模型,并分别对影响空气弹簧刚度的各个因素进行了仿真分析。仿真结果显示,空气弹簧的刚度主要由工作高度、工作压力决定,但囊体材料特性、加载方式等因素也会对空气弹簧的动态特性产生一定的影响。该模型可为囊式空气弹簧提供一种准确性高、实用性强的计算方法,对空气弹簧隔振系统分析计算具有一定的指导意义。     

A procedure for evaluating the positioning accuracy of reciprocating friction drive systems

The dynamic characteristics of stick-slip motion in reciprocating friction drive systems are investigated under dry contact using a 0.45% carbon steel pair. Based on this dynamic analysis, the stick-slip motion can be eliminated under certain experimental conditions depending upon driver speed, normal load and spring constant. The effects of normal load, driver speed, and spring constant on the positioning accuracy of the reciprocating friction drive system are examined under harmonic oscillation without stick-slip motion. Results show that at very low spring constants, the slip increases with increasing driver speed such that higher normal load has better positioning accuracy or smaller slip than does lower normal load. However, at high spring constants and high normal loads, there exhibits severe wear or peak at two limit positions on the sliding region. As a result, smaller normal loads have a better positioning accuracy than larger normal loads due to severe wear at high normal loads.     

预应力圆柱螺旋弹簧弯曲弹性特性研究

研究了预应力圆柱螺旋弹簧的弯曲弹性变形特性.通过对弹簧圈进行受力分析,利用材料力学的相关理论,推导了弹簧在受到弯矩作用时弯矩与弯曲角度之间的关系,得出了弹簧弯曲角度的计算公式,并对参数进行了优化.通过试验验证,表明所提出的公式精度较高,能满足工程设计的需要,为后续弹簧系统的动力学性能研究奠定了基础,同时也为新弹簧的设计提供了一定的理论依据.     

基于AMESim的微型精密比例阀迟滞性优化分析

针对微型比例阀流量控制精度低的问题,在结构设计上采用阻尼弹簧组件,分析比例阀系统的工作原理,利用AMESim建立了系统仿真模型,分析了线圈匝数、弹簧刚度、弹簧预紧力等参数对微型比例阀流量迟滞特性的影响.利用序列二次规划(NLPQL)优化算法,以线圈匝数为主变量,目标输出流量及最低迟滞性为优化目标,确定3组较优理论参数,...     

Automated recognition system to classify subcellular protein localizations in images of different cell lines acquired by different imaging systems

Systemic analysis of subcellular protein localization (location proteomics) provides clues for understanding gene functions and physiological condition of the cells. However, recognition of cell images of subcellular structures highly depends on experience and becomes the rate-limiting step when classifying subcellular protein localization. Several research groups have extracted specific numerical features for the recognition of subcellular protein localization, but these recognition systems are restricted to images of single particular cell line acquired by one specific imaging system and not applied to recognize a range of cell image sources. In this study, we establish a single system for automated subcellular structure recognition to identify cell images from various sources. Two different sources of cell images, 317 Vero (http://gfp-cdna.embl.de) and 875 CHO cell images of subcellular structures, were used to train and test the system. When the system was trained by a single source of images, the recognition rate is high and specific to the trained source. The system trained by the CHO cell images gave high average recognition accuracy for CHO cells of 96%, but this was reduced to 46% with Vero images. When we trained the system using a mixture of CHO and Vero cell images, an average accuracy of recognition reached 86.6% for both CHO and Vero cell images. The system can reject images with low confidence and identify the cell images correctly recognized to avoid manual reconfirmation. In summary, we have established a single system that can recognize subcellular protein localizations from two different sources for location-proteomic studies. studies.     

Novel torsional spring with corrugated flexible units for series elastic actuators for cooperative robots

Reasonable robot joint stiffness is necessary to guarantee the safety and control accuracy of cooperative robots. In this study, a novel flat torsional spring with corrugated flexible units for a series elastic actuator (SEA) was developed to meet the requirements of cooperative robots. The torsional spring can absorb impact energy. The spring was optimized through the design of experiment method, and its theoretical stiffness was verified through numerical calculation and finite element analysis. Compared with other existing flat torsional springs in the simulation, the proposed torsional spring showed reasonable torsional stiffness and high radial and axial stiffness, which could guarantee safety and control accuracy. The proposed torsional spring was also evaluated through real experiments. Simulation and experimental results revealed that the proposed torsional spring has a linear torque versus angle characteristic.

     

IC封装焊头机构接触力分析

针对IC封装焊头机构(bonder)高速运动、精密定位、微接触力的要求,运用虚拟样机技术,建立焊头机构虚拟样机模型.通过虚拟仿真,揭示了焊头部件的运动和动力学规律.为了保证取晶和固晶时的接触力要求,对弹簧参数和焊头与晶圆、引线框架的间隙进行优化.结果表明,焊头的位移控制精度则直接影响到接触力的大小,弹簧刚度则不仅影响接触力的大小,还影响它的稳定几性.     

Measurement of the surrounding liquid motion of a single rising bubble using a Dual-Camera PIV system

A new type of Particle Image Velocimetry technique, called “Dual-Camera PIV System”, was developed in order to achieve higher-accuracy measurement at a high time resolution. It is very difficult to measure precisely a complex flow field such as a gas–liquid two phase flow using PIV, because of the existence of a strong turbulence. In the conventional dynamic PIV, a time interval of two images required for analysis depends basically on a camera frame rate. A time interval of a set of PIV images affects the measurement accuracy significantly. Therefore, it is necessary to shorten the time interval of a set of PIV images as well as to achieve high frame rates. For this specific purpose, we developed a measurement system composed of two high speed cameras. The interval of two images obtained from each camera was controlled arbitrarily. Furthermore, a recursive cross-correlation method was adopted as PIV algorithm in order to achieve high spatial resolution. The interrogation areas were determined from the number density of PIV particles. The developed system was evaluated by cross-correlation coefficient and signal–noise (S/N) ratio. As the demonstration, the surrounding liquid motion in the vicinity of a single rising bubble was measured via this measurement system.     

低成本柔顺板式精密定位平台的理论与试验

基于柔顺薄板的弹性变形原理,提出一种低成本、整体式柔顺精密定位平台的设计方法。设计一种平板弹簧机构,使平台能够在较低精度驱动器的驱动下获得高精度的位移输出,从而大幅度降低精密定位平台的成本。建立平台的刚度模型,分析驱动器的精度与平台的定位精度之间的相互关系。最后,分析加工误差对平台定位精度的影响,得出影响定位精度的关键结构参数。通过试验测试,得出驱动器的输入位移和精度分别为4 mm和10m,平台的输出位移和精度分别为9 m和40 nm,线性度为0.33,从而验证了该设计方法的正确性。该平台结构紧凑,成本低,具有很好的性价比,应用前景广阔。     

光纤共聚焦显微内窥镜活体内实时成像系统的设计和研究

介绍了一种光纤共聚焦显微内窥镜实时活体内成像系统。该系统采用高速扫描振镜、超细成像光纤束、大尺度几何形变理论图像拼接技术,并结合自主研发的综合软件,使该系统具有实时检测、探头物理尺寸小、图像分辨力高、用户界面友好、操作方便等多方面优势。实验表明:该系统可为医生提供丰富的组织学和病理学影像信息,提高诊断准确率。该系统为临床医学提供了一种能在活体内进行实时细胞尺寸检测的医疗仪器,是癌症早期诊断的重要工具。