基于线结构光-机器视觉的CHN钢轨轨距动态检测方法

针对有轨电车传统轨距检测方法存在检测成本高、标定抗干扰能力弱等缺点,提出一种基于线结构光-机器视觉的轨距动态检测方法,通过线结构光照射左右铁路轨道,形成钢轨轮廓信息。首先通过标定,建立并融合左右轨道的空间姿态坐标系,获取轨距的计算公式。接着通过机器视觉中的灰度滤波、高斯滤波、开运算、骨架提取算法对轨道轮廓图像进行处理。然后将处理后的轮廓图像通过标定的空间姿态坐标系计算出轮廓的实际空间坐标,即可得到轨距。该方法不仅只用2组线结构激光摄像式传感器降低检测成本,且通过机器视觉算法能够大大提升环境干扰下轨距动态测量的鲁棒性,保证测量的精度。最后通过实验室平台以及轨检车试验场验证该方法的可行性和有效性。     

多台阶零件的视觉测量

为了快速准确地对复杂多台阶零件进行非接触检测，文中基于计算机视觉技术，提出了应用由十字线激光器和CCD相机构成的视觉测量系统来完成检测任务的设想。为此，文章重点介绍了该系统的各项必要参数及其标定方法，建立了测量复杂零件多台阶高度的数学模型，并用其实现了对复杂零件的非接触测量实验，从而验证了该方法的正确性和有效性。     

铣刀机器视觉检测系统设计

铣刀检测是铣刀生产的关键环节之一,针对传统铣刀检测过程中暴露的自动化程度低、效率低等问题,提出了基于机器视觉检测系统的检测方法.论文对系统的总体方案进行了设计,研究了系统的工作原理,并对系统硬件软件进行设计,最后通过实验分析得出结论.实验结果表明,通过视觉检测系统可以检测出铣刀是否合格,并且此法比传统方法有明显的优越性.     

基于神经网络的视觉检测系统的研究

随着电子学和计算机技术的日趋成熟,视觉检测系统取得了突飞猛进的发展并在许多领域得到了广泛地应用。本文提出了一种基于神经网络的结构光中心线提取的方法,实验结果表明神经网络方法可达到亚像素级精度,从而能有效地提高视觉检测的精度。     

荫罩板孔形尺寸的视觉检测算法研究

荫罩板是彩色显像管的重要元件,其质量的好坏直接决定显像管的质量,本文论述了采用计算机视觉检测技术实现荫罩板高精度,全自动检测的方法,发展了一个具有亚像素级精度的边缘提取及几何尺寸检测的算法,经试验证明检测精度优于1μm。     

基于线结构光的轮廓传感器的研究

自动化生产程度和产品质量的提高，对计量检测手段提出了更高的要求，视觉检测技术的发展适应了这一需求。结构光视觉系统为许多空间三维物体的检测问题提供了可行的解决办法。     

基于RBF神经网络的结构光三维视觉检测方法研究

研究了基于RBF神经网络的结构光三维视觉检测方法。该方法利用了RBF网络良好的非线性映射能力以及学习、泛化能力，通过采用高精度样本数据训练RBF网络，最终建立起了用于结构光三维视觉检测的RBF网络模型。与常规方法相比，它不需要考虑视觉模型误差、光学调整误差等对视觉检测系统测量精度的影响，因而能够有效地克服常规建模方法的不足，保证了检测系统具有较高的精度。     

机器视觉系统在皮带撕裂检测中的应用

文章对裂纹图像进行采集提取,然后在Visual Studio 2017、开发语言C++中借助功能模块,并通过边缘检测算法实现对图像的处理以便进行检测,之后对经处理的数字图像构建网络模型,以皮带裂纹图像的长宽几何特征量作为输入量,结合皮带裂纹图像样本,从测试出的裂纹特征值可得出整段皮带不同位置撕裂程度有所不同。     

基于机器视觉的工件检测系统研究

为检测工件质量,识别型号。采用机器视觉的方法,应用CCD相机在线拍摄工件图像,以图像采集卡作为与计算机的接口,进行数据的分析及处理,实现工件分拣。     

机械(模具)零件的敏捷制造--工艺基准定位系统的应用

通过分析传统制造业现状及存在的问题,提出采用工艺基准定位系统及引入"中断加工"概念可实现"一分钟换装",大大提高制造柔性,实现敏捷、快速、经济生产.     

Visual Inspection of Highway Bridges

The Federal Highway Administration recently conducted an investigation to assess the reliability of visual inspection of highway bridges as implemented by state departments of transportation. The study focused on routine and in-depth inspection, the two most prevalent forms of visual inspection of bridges. Because visual inspection is a highly subjective nondestructive evaluation technique, the results of these bridge inspections can be highly variable and are dependent on many factors. This study included the completion of 10 bridge inspection tasks by 49 practicing state bridge inspectors. With regard to routine inspections, this study showed that condition ratings are assigned with significant variability. Specifically, approximately 68% of condition ratings are within 1 rating point of the average. The results of the in-depth superstructure inspections indicated that these inspections are not likely to detect and identify the specific types of defects for which this inspection procedure is sometimes prescribed. Specifically, widespread defects such as those normally noted during a routine inspection are likely to be noted; however, local deficiencies, such as crack indications, are unlikely to be detected. The results of the deck delamination survey indicate that the accuracy of this type of inspection is relatively poor, with few inspection teams providing results that could be considered to accurately portray the condition of the deck.     

视觉检测技术及应用

视觉检测技术,尤其是基于三角法的主动和被动视觉检测技术具有非接触、速度快、柔性好等特点,是一种先进的检测手段,适合现代制造业的需要。文章论述了视觉检测技术原理,讨论了已经研制的多个实际视觉检测系统,从不同角度展示了视觉检测技术在现代制造业中广阔的应用前景。     

Contemporary Approaches to Paced Visual Inspection

An approach to the study of inspection task design has been investigated. Twenty female inspectors did a simulated paced visual inspection task for 90 minutes each. The experimental variables were display size, product pacing, and fault information as a measure of defect probability. They were tested in a 15 point design with 3 replications of the center point. The criterion of performance was the fraction of total products correctly identified as being defect free. From the experimental data quantitative estimates of performance increments or decrements associated with changes in inspection task variables were determined. The implications of observations and their importance in the design of an inspection task have been discussed.     

在线视觉检测过程及图形化输出

在线检测是提高产品加工精度的有力保证,视觉检测技术是利用机器视觉原理,对物体的几何尺寸进行准确定量的非接触检测 形化输出直观地显示在线测量结果,得于对加工生产线的分析和调整,两者的结合可以尽量减少废品,实现“零废品”工程。     

Performance Effects of Variables in Dynamic Visual Inspection

Three task variables of dynamic visual inspection were examined in laboratory experiments: (V) conveyor-belt velocities, (T) target exposure time and (P) the viewing position of the inspector. Measurements were obtained on both error types in inspection of equally-spaced targets for 36 experimental condition. Task variables V and T provided significant main effects on inspection accuracy and variables T and P had an interacting effect. The probability of correctly identifying a target was found to be describable as l-2.58e^-a where a = 10.587T- 0.0.75/V-0.012/[T+ P²] and the coefficient of determination, R² = 0.88. Inspector's eye motions were recorded and gross eye-motion patterns were identified but none appeared to be directly associated with inspection accuracy. Shifts in the inspection error criterion were observed and described for changes in these task variables. Both inspection speed and accuracy are combined in a cost model. This study shows that different conveyor speeds have better economics with different error costs but that shorter exposure times were never economically better.     

An Elementary Model of Human Performance on Paced Visual Inspection Tasks

An analysis of type one and type two error committed by human inspectors on a paced visual inspection task was conducted. The analysis was accomplished by development of a two-phase quantitative model describing the inspector's behavior when performing such inspection tasks: (1) the observe-not observe decision model described the inspector's decision to attend or not attend his task at any particular time; (2) the accept-reject decision model, which was based on signal detection theory, described the inspector's decision to classify a product unit as good or defective, given that he has made the decision to observe. Experimental evaluation supported the model, but did not provide complete validation.     

基于图像配准技术的二维平台位置参数视觉测量

提出一种基于图像配准技术测量移动平台二维平移、旋转等参数的视觉测量方法．固定基准图和镜头,使CCD随平台在基准图的像平面内移动或旋转,再由图像配准技术对当前CCD接收到的图像在整个像平面内的位置进行定位,从而可实现微位移平台位置参数的测量．对基准图进行分块设计,并且对每一分块进行二进制编码．如果保证至少有一分块落到CCD成像区域内．实验结果表明系统的位移测量精度可达到亚微米级．     

轮轨接触状态可视化检测装置研究及试验

为了研究列车高速运行过程中的轮轨接触状态,利用图像测量原理研制了便携式的轮轨接触状态可视化检测装置.通过钢轨内外两侧安装的高速摄像机,同步采集列车高速通过特定测量区域时的轮轨接触图像,再利用图像处理算法获得每个采样位置的轮轨外形轮廓,识别出轮轨异常接触状态,计算测量区域的轮轨接触轨迹.本文介绍了系统测量原理,重点设计了高速摄像机系统的参数,并给出了静态、动态和高速正线试验条件下装置安装可行性和检测性能的验证结果.     

零件位置度的测量能力分析

在几何量测量中,需要对零件进行形位公差的检测,测量过程中,测量结果是否准确,是大家关注的问题,零件位置度的测量在机加检测中具有一定的代表性,通过不确定度的评定,考虑各种误差对测量结果的影响,对测量结果进行分析,评价测量能力,来验证测量水平是否满足生产的需要。