基于模糊聚类的工业薄片件图像亚像素检测北大核心

针对机器视觉检测工业薄片件边缘存在精度不高的问题,基于模糊数学理论,引入模糊聚类思想,提出了一种图像亚像素综合定位算法。用OTSU法对源图像进行二值化处理;用5×5矩形域最大梯度模检测像素级轮廓边缘;用基于模糊聚类的亚像素边缘定位算法进行极片亚像素边缘定位,实现了边缘的高精度定位。实验结果表明,该方法算法简便,精度较高,易于编程实现,且兼具多种检测算法的优势,尤其适合工业生产现场的机器视觉检测系统。     

基于机器视觉的气门杆几何参数测量方法研究

文章给出了机器视觉系统的基本组成及高精度视觉测量的实施对策,并以气门杆几何及形位参数检测为例,讨论了图像的采集、处理及达到亚像素边缘检测精度的软件设计,通过LOG算子、最小二乘线性估计等手段,使直线边缘定位精度达到亚像素级,在不增加硬件成本的前提下,达到了检测参数的精度设计要求.     

基于机器视觉的柴油机缸套尺寸测量

为满足柴油机对缸套尺寸的高精度需求,提出了一种基于RANSAC(随机抽样一致性)拟合缸套亚像素轮廓的内外径尺寸检测方法。根据生产线上工业相机获取的CCD图像结构和灰度的信息特征,该测量方法先采用双边滤波和形态学梯度算法对图像预处理,去除噪声提高对比度;然后在用大律算法(QTSU)和Canny边缘检测算子获取图像边缘的像素级坐标;通过双线性插值算法对像素级坐标进行亚像素的重新定位,获取边缘亚像素坐标;最后采用RANSAC拟合算法对缸套内外径进行尺寸测量。通过仿真与实际尺寸对比,该方法具有更好的定位精度和抗噪性,相比最小二乘法RANSAC检测更可靠。     

基于机器视觉的柴油机曲轴轴线弯曲检测

针对传统的柴油机曲轴弯曲检测方法效率低下、检测精度低的特点,提出一种基于机器视觉的非接触测量方法。其主要方法是旋转曲轴,拍摄曲轴各轴颈在不同角度位置的图像,对每幅图像进行中值滤波、二值化、边缘检测、亚像素处理,获取亚像素边缘,然后利用最小二乘圆法对曲轴的每个像素列截面进行拟合,求出各截面中心坐标,最后,再利用最小二乘法拟合直线,计算轴线直线度大小,进而判断柴油机曲轴弯曲状况。实验表明该检测系统不仅检测精度高,检测速度也非常快。     

用于遥控焊接的焊缝特征检测算法

提出了一种焊缝特征检测算法,用来进行遥控焊接立体视觉中的立体匹配。首先使用灰度形态学中的底帽变换对图像进行预处理,突出了焊缝部分,并通过Canny算子边缘检测以及闭操作获得封闭的像素级焊缝边缘;根据焊缝边缘连续和宽度有限的特点对检测出的边缘点进行筛选,剔除其中的错误点;然后对边缘离散点进行三次平滑样条拟合获得连续、光滑的亚像素级精度的焊缝边缘特征点。最后利用检测出的特征进行焊缝特征立体匹配和三维重建试验,结果表明,文中提出的特征检测算法快捷、准确,且抗干扰能力强。     

基于图像像素级分割与标记映射的工件表面微小缺陷检测算法

为了解决当前工件表面微小缺陷在边缘模糊和特征不明显的情况下,导致其检测不准确的问题,文章提出了基于图像像素级分割与标记映射的工件表面微小缺陷检测算法。首先,利用工业显微镜采集工件图像,获取显微缺陷图;根据高效原则,以整数代替浮点运算,位移代替乘法运算,设计位移亮度滤波算子,进行平台数据转换,基于像素亮度阈值的标记映射,获取显微图像的像素亮度分布的二值图像。然后,耦合高斯平滑滤波与形状滤波,增强图像边缘轮廓特征,进行连通区域标记,准确识别与定位微小缺陷。最后,对缺陷目标进行特征计算和标注显示,完成微小缺陷的自动检测和定量计算。实验测试结果显示:与当前工件表面缺陷检测算法相比,文中检测技术拥有更高的识别精度。     

面向PCB圆形基准点的权重式椭圆拟合定位方法

针对贴片机视觉系统对印刷电路板(PCB)圆形基准点快速精密定位的要求,提出基于形态学的基准点区域提取方法,对约化后的基准点进行像素级边缘提取,进而采用一维曲线拟合法提取出亚像素级边缘;在此基础上,提出一种圆形基准点边缘轮廓的权重式椭圆拟合定位算法,减少边缘轮廓点大离群值对拟合精度的影响,获到高精度的基准点中心位置。该算法通过设置椭圆约束条件,可避免零解和成比例解的出现,提高计算的速度和精度。仿真实验结果表明,所提的方法耗时25ms左右,基准点中心的位置误差小于0.03pixel,相比于其它定位方法,该方法在耗时和定位精度方面具有明显优势。在实验室搭建的视觉检测平台上开展进一步的实验验证,对实际采集到的PCB基准点图像可实现快速精确的基准点中心坐标提取,基准点拟合椭圆的形状误差均值为0.1pixel,能满足贴片机视觉系统快速精密的定位要求。     

基于数字图像的刀具磨损状态检测技术

采用数字图像分割和亚像素边缘检测等图像处理技术自动测量刀具磨损边界的磨损量并与刀具磨钝标准比较,判断刀具的磨损状态,实现刀具磨损状态的间歇式在线自动检测。针对刀具的边缘特征,运用了中值滤波、对比度拉伸、迭代自适应二值化、基于分水岭分割、Roberts边缘算子、图像匹配、Zernike正交矩边缘检测和曲线拟合等数字图像处理技术,完成了刀具磨损量微米级的高精度检测。     

胶辊微小齿形几何参数在机视觉检测方法研究

针对胶辊微小齿形几何参数无法在机检测的难题,提出一种胶辊在机视觉检测方法,通过合理的测量装置设计,基于机器视觉检测原理搭建在机检测系统。通过直线往复运动机构、摆臂机构及视觉检测机构相互配合的在机视觉检测装置实现胶辊检测。利用该装置获取胶辊图像,通过视觉处理软件依次对胶辊图像进行图像滤波、形态学分析等图像预处理;采用亚像素边缘检测方法进行螺纹轮廓的精准定位,然后利用Harris角点检测算法提取特征点;最后采用最小二乘法分段拟合方法得到胶辊螺纹轮廓,并计算胶辊螺纹齿形的主要参数,判断胶辊的合格性。实验结果证明:所提胶辊齿形几何参数在机视觉检测方法可以实现胶辊齿形的高效高精度检测。     

基于亚像素的埋弧焊焊缝图像处理

埋弧焊是普遍使用的熔焊方法之一,具有生产效率高、焊缝成形好、没有弧光辐射等诸多优点,应用前景广阔.为了提高埋弧焊焊接过程的自动化和智能化,对埋弧焊精密跟踪技术的要求不断提高,需对焊缝间隙和坡口边缘进行精密检测.在此提出了一种基于B样条变换的亚像素定位算法,该算法是在对图像进行一次Marr-Hildreth边缘检测后再进...     

Edge location in the automatic track system of NGW with small slope angle

Based on the analysis of the groove images in gas metal are welding with small slope angle, a new algorithm of the groove edge location is presented. The groove edge was effectively detected by combining Roberts Detector with the general nonlinear gradient operator. In addition, using Norton Quadratic Polynomial Interpolation, the edge location precision reached sub-pixd level. The experimental results show that the edge detection system works well under the condition of short transfer are welding.     

基于小波变换的大型锻件图像边缘检测

大型锻件热态在线尺寸测量一直是在线生产的难题,近来,提出了用CCD测量尺寸的原理,其中CCD图像的边缘检测是其中的关键环节。文章根据Mallat小波理论中图像边缘对应于小波变换模的局部极大值点的理论,采用B样条小波对大型锻件图像进行了多分辨率小波变换,并根据不同尺度下噪声的小波系数变化较大的特点,采用固定阈值,有效抑制了噪声,成功地检测出图像的单像素边缘。试验结果表明,用小波变换处理大型锻件热态CCD图像的效果较好。     

远距离采集图像损伤区域三维重建仿真研究

远距离采集图像的损伤区域形状基属性形变和噪音难以消除,三维重建难度大,传统重建方法不能有效修复损伤区域.提出基于体像素插值的图像损伤区域三维重建方法,在远距离图像采集中形变不规则的前提下,利用产生奇异点的规律进行奇异点插值补偿,提取到图像仿真投射映像属性点的非齐次矩阵后,利用体像素补偿损伤区域奇异点,并通过边缘冲突原理进一步消除奇异点,获取到更加真实的远距离采集三维重建图像.实验结果证明:所提方法增强了远距离采集图像损伤区域三维重建的信息量.     

运动香烟滤棒在线图像检测系统

柔软滤棒是容易变形的工件,不能用接触法测量.作者对滤棒的生产过程进行了研究,并提出了一种采用亚像素图像测量技术对香烟滤棒的圆周长、圆度误差进行实时检测的方法.本文对香烟滤棒在线检测系统的图像测量软件的开发过程进行阐述,并通过滤棒试验结果验证了系统的有效性、可行性.     

GTAW填丝增材制造熔池尾部固液分界点跟踪算法

针对GTAW填丝增材制造熔池尾部固液分界点检测难题,基于熔池液态边缘波动大,固态成形金属边缘波动小的特性,提出一种连续图像采集处理的熔池尾部固液分界点跟踪算法. 采用去噪、高斯滤波、边缘检测和阈值分割等算法预处理熔池尾部图像,根据连续图像熔池尾部边缘像素点坐标变化,计算其差异系数,通过多项式函数拟合像素点差异系数曲线,并求取其一阶导数,据此判定差异系数曲线折点,确定熔池尾部固液分界点. 结果表明,该算法有效地实现了GTAW填丝增材制造熔池尾部固液分界点的实时跟踪,算法简易、高效,具有良好的抗干扰能力.     

Robust welding seam tracking using image seam extraction

Abstract

Extracting single pixel wide paths is the core image processing problem in many vision based welding seam tracking applications. Traditional techniques such as edge detection and image thinning may not be effective for solving this problem on low quality welding images polluted by process noises. Based on the understanding of the prior knowledge about the welding images, we present in this paper a welding seam tracking method using the image seam extraction technique. Owing to its optimisation nature, the proposed method is robust to strong image noises caused by environmental factors such as uneven illumination, light interferences and welding spatters. In addition, the proposed method is of low time complexity and is highly adaptive to different types of welding images. Experimental results on a variety of welding images validate the effectiveness and efficiency of the proposed method.     

Nano-precision measurement of diamond tool edge radius for wafer fabrication

In this paper, a non-destructive nano-precision measurement method for diamond tool cutting edge radius is presented. The basis of the method is that the profile of a tool cutting edge can be copied by indenting the tool cutting edge into the surface of a selected material, and that the copy of the profile can be measured at nano-precision level using AFM. The selected material elastic error compensation coefficient has to be determined to cancel out the effect of elastic spring-back. Copper was selected as the indentation piece material due to its (1) high rigidity and high density, (2) large Young’s modulus and (3) low yield strength. The elastic error compensation coefficient for the copper material is determined through the indentation of a tungsten carbide tool edge on the copper surface. By comparing the actual tool edge radius measured using scanning electron microscope (SEM) on the sectional view of the tungsten carbide tool with the one measured from the copied profile of the tool edge on the copper surface, the coefficient is obtained. Three diamond tool edge radii were obtained using the proposed method. Analysis is given for the accuracy of the proposed method, showing that as far as the error elastic compensation coefficient is consistent with the copper material used, the only source of errors with the measurement will come from the device for measuring the indented profile on the surface.     

NDT data fusion at pixel level

No single non-destructive testing (NDT) technique or method is satisfactory to fully assess the structural integrity of a material. Indeed, each method presents some limitations in terms of defect detection and characterisation. Additionally, poor signal-to-noise ratio may make signal interpretation complex or unreliable. For these reasons, the concept of NDT data fusion — based on the synergistic use of information from multiple sources in order to facilitate signal interpretation and increase defect detection and characterisation — is expanding rapidly. A carbon fibre reinforced composite material previously subject to low energy impact damage has been non-destructively inspected using eddy current and infrared thermography. Following these non-destructive examinations, the information gathered from these two NDT methods has been fused at pixel level. Several well established and recognised data fusion processes based on statistical and probabilistic algorithms were used to combine eddy current and infrared thermographic data. The results obtained with these processes are presented in this paper and their efficiency discussed. It is shown that, in certain cases, NDT data fusion at pixel level may be adequate to increase knowledge about defect location and characterisation and to reduce ambiguity.