变光照环境下皮带撕裂视觉检测方法研究

针对变光照环境下给皮带撕裂视觉检测过程造成的识别干扰,提出了一种基于洛伦兹信息测度(Lorentz Information Measure,LIM)分块的改进Ostu分割算法的结构光视觉检测方法。首先将红色线性激光投射到皮带表面,通过CCD相机捕获高对比度图像;然后利用LIM方法将ROI(Region of Interesting)区域根据其水平方向光照强度分块,并逐块通过线性加权Ostu算法进行图像分割,进而合并出整幅图像的激光条纹;最后提取激光条纹中心线,并对其信息特征进行分析比较,判断皮带是否发生撕裂。通过现场在皮带机上进行试验,在轻度和重度污染环境下,892张样本的正确检测率为95.72%,平均单张样本检测时间为89ms,检测精度和速度满足实用要求。     

核燃料棒反光表面条纹自适应中心提取方法

针对水下检测中激光条纹中心难以准确提取的问题,提出了一种适用于水下核燃料棒反光表面的条纹自适应中心提取方法.根据检测环境中存在水体散射及物体表面高反光的特点,去除水下噪点、反光噪点,实现激光条纹分割提取;充分利用BP神经网络曲线拟合及由光条几何信息生成的自适应卷积模板,实现反光区域的轮廓与灰度分布修正,使光条截面灰度分...     

行车环境下钢轨轮廓激光条纹中心的提取方法

研究了行车环境下激光条纹图像中心线快速、准确且可靠的提取方法。基于ENet深度学习模型实现了激光条纹的多区段快速分割;通过统计各区段内光条梯度方向的直方图来确定各分段光条的法线主方向,并构造了相应的方向模板;利用分区域多模板匹配的灰度重心法实现了光条中心的亚像素坐标提取。研究结果表明,该方法可以有效克服室外行车环境中各类干扰信息对光条中心提取的影响,单幅钢轨轮廓图像的光条提取时间仅为2.1 ms,误差均值约为0.082 pixel,标准差为0.047 pixel,兼顾了光条中心提取的时效性和准确率。     

基于法线引导的激光中心线提取方法

在线结构光三维测量系统中,高精度激光条纹中心线提取是提高测量精度的关键。针对现有激光中心线存在提取精度不高、保留细节差等问题,提出了一种基于法线引导的激光中心线提取算法。该算法具体实现步骤为:首先,对图像进行预处理,结合边缘检测和几何中心法对激光线初步提取;然后,用主成分分析法(principal component analysis,PCA)求取其法线,在激光中心点处划分角度八邻域,通过法线角度引导搜寻有效点集;最后,利用灰度重心法对点集进行亚像素提取。实验结果表明:该算法均方根误差与灰度重心法相比提高了0.233 9像素,比Steger算法、方向模板法更好地保留了光条细节,可以更精确地提取光条中心,达到亚像素级的精度。     

汽车焊缝密封胶条的光条中心提取算法研究

汽车焊缝涂胶能够提高车身密封水平, 需要对胶条进行检测并且获得三维信息。激光三角法是应用广泛的检测方法, 其关键技术之一是准确提取光条中心。但针对黑色被检胶体, 光条中心提取鲁棒性差, 因此提出了一种基于光条分割的Hessian矩阵定位和高斯曲线拟合的算法。首先, 将整体图像对比度提高, 运用去噪算法消除干扰的噪声; 然后用极值法初步定位光条, 结合Hessian矩阵对激光条的中心进行定位; 最后利用高斯曲线准确拟合出激光条的中心。实验研究表明, 在对光条图像添加噪声幅度为100的白噪声之后进行处理, 同样能够将精度控制在亚像素级别, 能够提高黑色被检测物体的光条的稳定性。     

金属表面非相干线结构光条纹中心提取方法

采用线结构光法测量金属表面形貌时,由于受到金属表面光学特性和散斑噪声的影响,条纹中心的提取误差往往较大。为此,提出了一种非相干线结构光形貌测量方法,避免了散斑噪声的影响。通过分析该方法测量金属表面形貌时的条纹图像特点,提出一种适合非相干线结构光条纹的中心提取方法。该方法首先采用结合积分图像原理的自适应阈值分割算法,对原条纹图像进行分割。采用灰度重心法粗提取原条纹中心坐标,以该坐标为基准向条纹宽度方向延伸,从而确定阈值分割后条纹图的感兴趣区和背景区,并去除背景区的噪声。经中值滤波后,采用几何中心法提取条纹中心。实验结果表明：采用该方法提取粗糙度样块表面非相干光条纹中心的平均误差为1.5μm,提取齿轮渐开线样板表面非相干光条纹中心的平均误差为0.9μm,均比其线激光条纹中心的提取误差小。所提方法能实现金属表面非相干线结构光条纹中心的精确提取。     

高噪声背景下激光条纹亚像素中心的提取

针对线结构光三维形貌测量中大量噪声易干扰激光条纹中心提取准确度的问题,提出了一种条纹亚像素中心提取方法。分析条纹图像中的噪声,采用平均法和中值滤波预处理图像;利用迭代阈值分割及形态学方法,获取条纹目标,引入距离变换提取条纹的像素级中心;根据像素级中心、二值信息及光强灰度,结合曲线拟合及重心法精确提取条纹的亚像素中心。仿真分析和实验验证下,相邻行条纹中心列坐标最大偏差值像素小于2 ,平均偏差像素值约为0.3,与传统方法相比,2项指标值更小。实验结果表明,算法有效利用条纹灰度分布规律,可降低噪声对中心定位精度的影响,更逼近条纹真实中心位置,抗噪能力极强。     

平顶线结构光的中心检测算法及光刀平面标定

针对传统线结构光光刀平面标定方法测量精度不高,应用范围小的问题,提出基于平面标靶的线结构光系统光刀平面标定,对无激光的标靶图片进行迭代摄像机标定,有激光的标靶图片进行光刀平面标定.提出光强符合均匀分布的平顶激光检测中心算法,将平顶激光建模为矩形的台阶函数,估计背景亮度和前景亮度,确定亮条纹宽度,再将窗口内的有效像素参与重心计算,得到光条纹中心.用该算法对不同噪声及不同量块的图片进行处理,结果表明,处理后图像的均方根误差分别在0.149pixel和0.176pixel内,表明该算法抗噪声能力强、精度高.用该算法提取光条中心,计算光条在标靶上的位置,根据至少两个姿态下的光条中心三维点,基于最小二乘法拟合光刀平面.通过迭代摄像机标定和光刀平面标定,利用三角测量法,在立体视觉模型下获取物体的三维点云数据.实验测量两个距离为100.5mm的标准球,相机与标准球距离为500mm,比较两球心距离与标准距离,测得平均误差为0.236mm.表明平顶激光检测中心算法切实可行,光刀平面标定方法基本满足要求.     

基于多尺度分析的激光光条中心点坐标提取方法

在线结构光测量系统中,激光光条图像中心准确提取是影响整个测量系统精度的关键因素之一。针对高反光等复杂环境下光学测量中激光光条图像中心提取问题,提出了一种基于多尺度分析的激光光条图像中心高精度定位方法。该方法利用骨骼化图像处理方法确定激光光条图像中心的初值,根据每一个光条中心初值处的光条宽度确定该位置处对应的高斯核均方差σ的初值。对图像进行多尺度卷积,确定最优的σ值,并求得光条亚像素级中心点,最后完成光条链接。结果表明该算法抗噪声能力强,可实现光条宽度变化较大的激光光条图像中心的高精度提取。     

恒星光干涉仪主动镜偏转角的检测

采用激光干涉检测的方法,综合应用计算机控制、数字图像处理等技术对高精密仪器主动镜的偏转角进行了检测.设计了主动镜的光干涉CCD数字图像检测系统,推导出了干涉条纹与主动镜偏转角之间的对应数学关系式,对条纹图像进行了噪音分析和相应的预处理,在干涉条纹细化的基础上对每条干涉条纹采用最小二乘法进行直线拟合,得到条纹间的间距以及条纹的方向角,最终检测出主动镜X轴方向的偏转分辨率为0″.113,最大偏转角为56″.872,Y轴方向的偏转分辨率为0″.112,最大偏转角为49″.835,完成了主动镜偏转角的检测.     

On-line conveyor belts inspection based on machine vision

Under the background that mining conveyor belts are prone to failure in operation, the on-line fault detection technique based on machine vision for conveyor belts is investigated. High-brightness linear light sources arranged to a vaulted shape provide light for a line-array CCD camera to capture high-quality belt images. A fast image segmentation algorithm is proposed to deal belt images on-line. The algorithm for detecting longitudinal rip and belt deviation which are serious threat to the mine safety production from binary belt images is presented. Then, an on-line visual belt inspection system is developed. The laboratory testing results testify the validity of the visual inspection system.     

光条亮度对线结构光测量精度的影响

光条中心的提取精度直接影响线结构光传感器最后的测量精度,而光条亮度是影响提取精度的主要因素。提出了光条中心提取精度的评价方法,通过实验获得光条亮度与光条中心提取精度之间的关系曲线;为了在实际测量中能获得最佳光条亮度(对应最高的提取精度)设计了基于D/A转换的光条亮度自适应调节电路。实验结果表明:光条过暗或过亮都会使线结构光传感器的测量精度降低,在测量不同颜色的表面时通过光条亮度的自适应调节均可获得最高的提取精度。     

New optical measuring system for solder joint inspection

Due to specularity and complex 3D geometry, the visual inspection of solder joints has been regarded as one of the most difficult tasks and thus has not guaranteed accurate inspection results. This paper deals with an optical sensing system designed to inspect solder joints automatically. Taking into account a specular characteristic, a new optical solder joint inspection system is proposed to obtain the external profile of solder joints accurately. A laser scanning unit scans the area of solder joint and observes the angle of the reflected beam. To realize this measuring principle, the system is so composed that a galvanometer steers the direction of laser beam to a point to be measured, a mirror unit (a parabolic and a conic mirror) gathers the specular components of the reflected beam toward its center and a beam-receiving unit positioned along the mirror unit's center line detects the reflected beam. To classify the defects of solder joint, a statistical pattern recognition method is utilized. To verify the validity of the developed system, a series of experiments was performed for SOPs and QFPs in insufficient, normal and excess soldering condition. Based upon observation of the experimental results, the proposed system is found to show good performance for inspection of solder joint defects.     

A machine vision technique for subpixel estimation of three-dimensional profilometry

A quick and efficient technique for a charge-coupled device (CCD)-based profilometric measurement system is proposed and verified by experiment. When a laser stripe is projected onto an object, the deformed laser line image is detected by a CCD camera with subpixel resolution and is automatically transferred to its space position using a least-squares mapping algorithm. Measurements are executed on each separate surface of a three-dimensional object and the multi-surface data are integrated through a process of coordinate transformation. Large arrays of data points are acquired on each surface by means of a scanning mechanism. Because the measuring data from each laser stripe position can provide two-dimensional information concerning the surface contour, the laser stripe and the CCD camera are integrated into a measurement system for obtaining the entire three-dimensional information. Theoretical analysis shows that this technique can determine the three-dimensional profile of an object with an accuracy below 0.2 mm.     

基于B样条迭代法的激光光条噪声去除技术研究

在基于线结构光的视觉测量系统中,激光光条中心的提取精度是影响系统最后精度的关键因素。目前常用的光条中心提取方法都没有去除光条上的干扰噪声,在分析现有方法的基础上提出了一种利用B样条迭代去除光条噪声的方法。以光条上的点作为控制顶点拟合B样条曲线,在同一位置下用B样条曲线上的点取代噪声点,这样反复修改噪声点的位置使其逐渐逼近实际曲线,从而达到去除噪声的目的。利用重心法提取去噪后光条的中心,着重分析了参与运算的像素的取值范围。通过对比实验证明该方法的提取精度远高于其它方法,从而验证了去除噪声的有效性。     

Measured anisotropic air flow resistivity and sound attenuation of glass wool

The air flow resistivity of glass wool has been measured in different directions. The glass wool was delivered from the manufacturer as slabs measuring 100x600x900 mm3, where the surface 600x900 mm2 was parallel with the conveyor belt used in the manufacturing. Directions in the glass wool are described by a coordinate system with the X axis perpendicular to the conveyor belt, the Z axis in the direction the conveyor belt moves, and the Y axis perpendicular to the two other axes. It was found that the resistivities in the Y and Z directions were equal in all cases. For density 14 kg/m3 the mean resistivity in the X direction was 5.88 kPa s m(-2) and in the Y direction 2.94 kPa s m(-2). For density 30 kg/m3 the mean resistivity in the X direction was 15.5 kPa s m(-2) and in the Y direction 7.75 kPa s m(-2). A formula for prediction of resistivity for other densities is given. By comparing measured values of sound attenuation with results calculated from resistivity data, it is demonstrated that the measured attenuation can be predicted in a simple manner.