Machine vision system for curved surface inspection

This application-oriented paper discusses a non-contact 3D range data measurement system to improve the performance of the existing 2D herring roe grading system. The existing system uses a single CCD camera with unstructured halogen lighting to acquire and analyze the shape of the 2D shape of the herring roe for size and deformity grading. Our system will act as an additional system module, which can be integrated into the existing 2D grading system, providing the additional third dimension to detect deformities in the herring roe, which were not detected in the 2D analysis. Furthermore, the additional surface depth data will increase the accuracy of the weight information used in the existing grading system. In the proposed system, multiple laser light stripes are projected into the herring roe and the single B/W CCD camera records the image of the scene. The distortion in the projected line pattern is due to the surface curvature and orientation. Utilizing the linear relation between the projected line distortion and surface depth, the range data was recovered from a single camera image. The measurement technique is described and the depth information is obtained through four steps: (1) image capture, (2) stripe extraction, (3) stripe coding, (4) triangulation, and system calibration. Then, this depth information can be converted into the curvature and orientation of the shape for deformity inspection, and also used for the weight estimation. Preliminary results are included to show the feasibility and performance of our measurement technique. The accuracy and reliability of the computerized herring roe grading system can be greatly improved by integrating this system into existing system in the future.     

基于圆结构光的复杂深孔内轮廓尺寸测量方法

涡流法、超声法测量深孔截面轮廓尺寸精度低,而光学单点扫描法测量效率低。介绍了一种基于圆结构光的复杂深孔内轮廓3维测量系统,提出了一种采用FFT分析和差分分析进行测试数据处理的方法,能够实现快速高精度测量。测试系统主要由圆结构光发生器、扩束锥镜、成像锥镜、镜头和CCD组成,对于获取的每一幅被测截面的结构光图像,首先提取光条中心,对光条中心上的点作最小二乘法拟合获取光条中心拟合圆心,并以此圆心点将光条中心线展开,展开波形中存在整体形状误差,主要由圆心偏心误差、椭圆形状误差两个周期性误差分量构成,由于二者振动频率与细节分量的振动频率不相同,借助于FFT分析,将两个误差分量分离出来,进而采取措施减小其对系统的影响;差分分析用于去除阴线和阳线之间的过渡线。该方法提高了阴线圆与阳线圆尺寸计算的精度和效率,实验结果表明,系统内径测量精度达到005mm。     

超限货物运输列车异物侵入数据智能检测仿真

对超限货物运输列车侵入异物进行智能检测,能够有效降低超限超重货物运输事故的概率,对保障铁路货运安全具有重要意义。针对当前超限货物运输列车图像异物侵入检测方法受到采集频率的影响,列车高速行驶时,异物侵入检测易存在漏检的情况,提出基于背景差分和形态学腐蚀操作的超限货物运输列车异物侵入数据智能检测方法。针对超限货物运输列车运行场景抖动特点,设计抖动补偿算法,消除超限货物运输列车图像抖动的像素部分。通过背景差分算法获取超限货物运输列车图像背景,再利用背景帧与当前帧的差分获得二值图像,最后对所得二值图像进行形态学腐蚀膨胀操作和阴影消除,获取异物侵入目标区域。实验结果表明,所提方法可准确检测出超限货物运输列车侵入异物,对保证铁路运输具有重要的现实意义。     

列车闸瓦钎及闸瓦故障图像检测算法研究

列车闸瓦钎及闸瓦状态正常与否对于货运列车安全运行极其重要,为此提出了一种改进的 SSD 目标检测算法,对货运列车闸瓦钎及闸瓦的缺失进行检测。首先将深度可分离卷积模块引入 ResNet50 网络模 型,使其参数数量减少约 50%,以提高检测效率。其次,将改进后的 ResNet50 网络模型替换 SSD 中的 VGG16 网络,以改善 SSD 网络模型的特征提取能力。然后利用高、低层特征融合的方法,将网络中 Conv5_3 和 Conv7_2 分别与 Conv4_6 和 Conv6_2 进行特征融合,提高检测精度。最后通过自建的货运列车制动部件缺失数据集对 网络进行训练获取比较准确的网络权重。实验结果表明,改进后的 SSD 算法在闸瓦钎丢失检测中准确率达到 96.85%,召回率达到 89.50%;在闸瓦丢失检测中准确率达到 97.01%,召回率达到 97.01%,可以满足列车闸瓦 钎及闸瓦缺失检测需求。     

踝关节功能评价三维测量系统设计

在临床上医师通过对踝关节的X线拍片和多年经验来判断踝关节功能性不稳,这将增加临床诊断成本和诊断门槛.对于此问题,介绍一种基于三维测量与重建模型的踝关节功能评价系统.该系统采用结构光正-反投影技术,通过结构光正-反条纹图像来检测结构光边缘并实现亚像素定位;在此基础上,利用立体视觉原理来实现踝关节姿势的三维测量与重建,最后通过三维重建模型来分析踝关节功能性不稳.实验结果表明,该系统在无任何机械接触的情况下,有效地分析评价踝关节功能性不稳,为踝关节功能性不稳的早期发现和康复医疗提供有利的诊断与决策依据.     

阔大货物装载方案中超限自动识别方法

为了改善阔大货物运输安全,提高铁路货物运输效率,开发了一种铁路阔大货物运输超限自动识别系统该自动识别系统利用阔大货物三视图,结合阔大货物运输车辆,在综合分析货物三维空间结构与铁路限界几何关系的基础上,实现阔大货物超限自动识别技术．该系统可以帮助货物运输前选车,并对货物超限等级做出预先判断,保证货物运输的安全性．     

3D coordinates measurement based on structured light sensor

This paper mainly discusses several challenges in exploiting a structured light sensor to realize 3D coordinates measurement: a practical and flexible sensor structure is proposed with accuracy analysis, a very robust and accurate line extraction method is adopted to localize the center of the light stripe under considerable disturbances, the invariance of double cross-ratio based method is used to obtain enough accurate calibration points and a novel planar target based calibration method is presented. An application system of the structured light sensor for tyre inspection is set up, and the real online measurement results show that the application system has good repeatability and accuracy.     

The meta-defect-detection system for gear pitting based on digital twin

Gear pitting is a typical surface defect, and its accurate three-dimensional (3D) detection is significant for the operation and maintenance of equipment. Therefore, this paper develops a 3D gear pitting detection method based on digital twin. Firstly, a geometric model of gear pitting is built to expand the samples which are helpful improving the detection accuracy. Secondly, based on the gear pitting model and Unity, a virtual fringe projection profilometry (FPP) system is established in the metaverse for generating the deformation fringe patterns and retrieving their phases. To improve the accuracy of pitting detection, a gear pitting detection network (GPD-Net) is proposed. It can not only extract local and global characteristics from different spaces and scales but also fuse these features via the proposed attention-based conditional random field modules. Thus GPD-Net can retrieve the high-precision wrapped phase by a single fringe pattern. Meanwhile, an FPP system is built in the physical world, and the reconstructed actual point cloud images are imported into the virtual FPP system for the precise measurement of gear pitting. The developed measurement system can be regarded as a meta-defect-detection system. The experimental results show the superiority of GPD-Net over the state-of-the-art phase retrieval algorithms and the effectiveness of the proposed three-dimensional gear pitting detection method. The proposed 3D detection method based on FPP is not limited by the fault type, and it can be effectively applied to the 3D measurement of various surface defects.     

Real-Time Range Measurement Device for Three-Dimensional Object Recognition

This correspondence describes a real-time range measurement device for detecting the range, shape, and position of a three-dimensional object by a light stripe projection method. The device consists of a light stripe projection unit, a light stripe scanning control unit, a TV camera, a signal processing unit, and a microcomputer. While a laser light stripe is projected and scanned on an object at a high speed, range measurements and their calculations are done with special hardware. Range data are obtained by triangulation. This device can detect the range and shape of the object placed in a 60 cm square area at a distance of 100 cm in 490 ms by 48 × 50 points. The measurement accuracy of the device is ±2.0 cm. A sorting system for machine parts using this device has also been developed. By use of a unique range gating method, the system can discriminate three types of machine parts on a conveyor belt in 4.0 s.     

基于改进TPDS长阵列的车辆超偏载检测研究

由于车辆超偏载检测过程中,受天气、路况等因素以及车辆行驶状态影响,导致信号组成的阵列过短,利用传统的TPDS(Trackside Performance Detection System,TPDS)车辆超偏载检测精度较低、耗时较长。提出基于改进TPDS长阵列的车辆超偏载检测研究。通过车辆运行品质轨边动态监测系统,利用测距传感器与数据采集仪将力学信号转换成电压信号,将输出信号并将其组合为长阵列输入到系统的测距传感器中;采用EDM算法对长阵列中的噪声点实施消除处理;将4个测距传感器安置到货运车辆底部,采用货运车辆动态称重方法,计算车辆上各弹簧组承载力的总和,并结合车辆有用信号,对车辆的载重情况展开测量,实现对车辆的超偏载检测。实验结果表明,利用长阵列改进TPDS系统后,测试具有较高的测量精度、检测效率以及检测稳定性。     

线结构光光条中心提取算法

目的 线结构光视觉测量是一种利用可控光源和数字图像的主动视觉测量方法,光条中心提取是线结构光视觉测量的关键技术,直接影响到线结构光视觉测量的精度。传统灰度重心法只在图像的横向或纵向上计算光条的灰度重心,没有考虑光条的法线方向,精度较低。本文提出一种改进的光条中心提取算法,以期实现光条中心的精确提取。方法 在分析线结构光的光条灰度特性基础上,基于传统的灰度重心法,提出一种改进的两步提取算法。基于图像差分法从原始图像中分离出有效的线结构光光条,采用传统灰度重心法对光条中心进行粗提取;在粗提取的光条中心点处通过自定义的方向模板确定光条的法线方向,以粗提取的光条中心点为中心,沿法线方向采用灰度重心法进行二次提取,获取线结构光光条的中心。结果 本文采用CCD相机、镜头、线激光器及辅助机构搭建线结构光视觉系统,采用提出的算法对线激光器投影产生的直线型光条、非连续光条和弯曲光条的中心进行提取。通过光条中心提取实验获取的光条中心线的走向与光条的走向大致相同,符合预期的光条中心线。本文将Steger法作为评价标准,分别计算本文算法、传统灰度重心法与Steger法提取的光条中心的偏差,通过对比实验可知,本文算法提取的光条中心的偏差更小,并且程序运行时间比Steger法减少了3 s以上。结论 本文研究线结构光的光条中心提取算法,对传统灰度重心法进行改进,能够实现直线型光条、非连续光条和弯曲光条等不同形状光条的亚像素级中心提取,并且在保证较少的程序运行时间的同时,能够提高传统灰度重心法的光条中心提取精度。     

结构光截面轮廓测量中激光条纹的处理

在结构光测量系统中需要对采集到的激光条纹图像进行有效处理。该文提出了一种基于差图像的阈值分割方法,应用该方法对采集到的激光条纹图像进行阈值分割从而得到二值激光条纹图像,利用神经网络对二值激光条纹图像进行细化处理,处理过的细化条纹图像满足0.5％的精度要求,图像二值化处理时间不到1s,细化过程的时间小于2s,相对传统的图像处理时间有所增加,已经成功地应用在物体三维重建时特征参数匹配的预处理中。     

Detection of Wheat Stripe Rust Using Solar-induced Chlorophyll Fluorescence and Reflectance Spectral Indices

Detection of wheat stripe rust is important for agriculture management and decision,this paper aims to improve detection accuracy of the disease severity of wheat stripe rust by integrating the advantages of reflectance spectroscopy in the detection of crop biochemical parameters and the advantages of chlorophyll fluorescence in photosynthetic physiology diagnosis.Firstly,the solar-induced chlorophyll fluorescence (SIF) at O₂-A band (760 nm) was calculated using the 3FLD algorithm,and seven spectral indices sensitive to wheat stripe rust were investigated for estimating the disease severity.Then,three classic statistical modelling methods,including Support Vector Machine (SVM),Stepwise Regression (SR) and BP neural network (BP),were used to quantitatively investigated the performance of the spectral indices and SIF for detection of winter wheat stripe rust severity.The results show that:(1) there is a significantly negative correlation between SIF and the severity of wheat stripe rust.The relationship between SIF and DI can be effectively applied to detect wheat stripe rust.(2) the spectral models based on SIF combined with spectral indices are more accurate than those based on spectral indices.SIF can significantly improve the detection accuracy of the disease severity of winter wheat stripe rust.(3) compared to the SVM and SR methods,the training model constructed by the BP neural network has the highest prediction accuracy whether using the spectral indices or SIF combined spectral indices.However,the verification results show that the disease severity prediction model constructed by SVM and SR method have a better prediction.     

列车车轴温度无线监测系统设计

火车运输时因车轴温度升高而产生热切轴现象,直接导致火车发生故障翻车,造成巨大经济损失;针对货运列车的轴温故障问题,设计了对列车故障轴温度进行实时监测的车轴无线实时监测系统;本系统基于热电阻PT100、电桥式温度传感器、ATmega16L单片机、RF905无线射频模块等,采用功能强大的ATMEGA48芯片进行电源控制。本系统温度输出误差不超过±0.5℃,通信距离达到不少于2km的要求,具有效率高,实时性好,精度高等特点,已成功应用于某轴温型测试系统当中。     

BP神经网络和遗传算法在货车车锁检测中的应用

为实现货车自动检测记录系统，需要根据化的图像检测进站货车车锁是否存在，提出一种基于BP神经网络和遗传算法的货车车锁检测方法，首先提取图像的投影特征，边缘图像的矩向量特征以及灰度直方图特征，然后用BP神经网络进行检测和定位；同时引入遗传算法，利用遗传算法的高并行性和鲁棒性，可以较快地完成全局搜索，而不会陷入局部最优，实验结果表明，该方法能有效地克服货车车锁种类多，变形大以及光照变化的影响，具有较高的检测速度和检测成功率。     

基于亮度差和纹理的货车类型的识别

提出了基于亮度差和纹理的货车类型识别方法，能有效区分棚式货车与敞式货车，实验表明这种方法具有较高的速度和识别成功率，能够满足实际应用。     

CMOS传感器在可编码三维测量中的应用

设计了一种可纠错的光栅条纹编码三维测量方案。采用CMOS图像采集传感器构建了一种三维测量系统,通过采用汉明码作为可纠错编码方案,解决了光栅交界处像素点编码错误的问题,并验证了可行性。     

基于三线结构光的相机平面标定方法研究

相机标定技术是结构光三维视觉测量的关键技术之一,结构光测量系统的相机标定的精度对三维测量的精度有很大影响;首先对三线结构光系统图的相机标定方法进行了分析,简单介绍了工业相机成像的几何模型及标定的原理;其次利用Harris角点检测方法提取特征点坐标,并选用了BP神经网络来校正工业相机的畸变模型,以提高标定算法的优化速度和标定精度;最后采用张正友的平面标定法对校正后的摄像机模型进行标定实验,由实验结果知,该方法具有一定的准确性和有效性,在一定误差范围内,基于神经网络畸变校正的张正友相机标定能够有效提高视觉检测的精度。     

三维服装仿真中的"服装-人体"快速冲突检测及响应算法

维胆装仿真可以为三维人体动画生成逼真的服装动态效果，但其中的冲突检测与仿真计算的时间复杂度太高，其实用性一直受到很大限制．提出了一种快速的“服装-人体”冲突检测友响应算法，在人体运动状态下，快速检测服装与人体之间的位置冲突，其时间复杂度仅为O（n）（n为服装模型上的顶点数目）．在此基础上，提出一种合理有效的冲突响应机制，并实现快速稳定的三维服装仿真，取得了真实的仿真结果．     

列车测速测距设备故障检测系统研究

本文设计了一种列车测速测距设备故障检测系统,该系统对轮轴速度传感器及测速测距处理单元进行动态闭环离线检测,可再现列车运行状态及设备故障工况,同时在检测过程中分析待测传感器及测速测距处理单元的各项测试数据,从中选取速度、脉冲值和工作电流等信息进行故障模式辨别。测试结果显示,该系统故障诊断正确率高,实现了对测速测距设备服役状态参数的自动化检测。