图像处理在医用纱布表面缺陷检测中的应用

针对医用纱布表面缺陷的特点,研究缺陷检测算法。运用迭代算法对图像进行预处理,有效地去除噪声对图像的影响。采用了一种特殊的图像分割方法即基于数学形态学的分割技术对图像进行边缘检测及分割,保证了边缘检测的连续性,能够分割出完整的缺陷目标图像。利用提取的纱布表面缺陷图像中的形态学特征对缺陷进行分类,实现纱布表面破洞、缺经、断纬等缺陷的识别目的。通过MATLAB软件对一些医用纱布图像进行测试,验证了该方法的可行性和准确率。     

基于图像角点匹配的机械加工零件表面缺陷检测

当前在对机械加工零件表面缺陷检测时,存在表面缺陷检测精度差的问题,导致检测结果不理想。提出一种基于图像角点匹配的机械加工零件表面缺陷检测方法,利用曲率空间检测零件图像的角点,采用泰勒级数删除伪角点。对特征点邻域梯度方向进行角度限度和就近投影,同时借助双向匹配方法进行机械加工零件图像角点匹配。在上述操作的基础上,利用一维直方图的阈值分割对零件图像进行分割处理,最终实现机械加工零件表面缺陷检测。实验结果表明,该方法能够获取高精度的零件表面缺陷检测结果,且对加工零件缺陷厚度、孔洞缺陷及缺陷最大边界距离的测量均较为准确。     

基于机器视觉的笔芯球珠表面缺陷检测系统研究

针对笔芯球珠表面缺陷检测识别问题,设计并实现了基于机器视觉的笔芯球珠表面缺陷检测系统。笔芯球珠在球面展开机构作用下,通过图像采集模块获取5张可以完全覆盖整个球面的图像。通过对每幅图像进行缺陷图像提取后,采用基于轮廓角点匹配的方法实现对每幅图像中缺陷图像的拼接;基于提取的有效特征组合通过KNN分类算法对完整的缺陷图像进行缺陷识别。试验结果表明,该方法能够对笔芯球珠表面缺陷进行精确有效的检测与识别。     

基于机器视觉的手机壳表面划痕缺陷检测

为解决手机壳表面缺陷检测采用人工目测法,检测效率低且漏检率高的问题,采用基于机器视觉的手机壳表面缺陷检测方法,实现产品缺陷的自动化检测。该检测算法采用八方向的各向异性高斯方向导数滤波器对图像进行卷积滤波,并做归一化处理;利用滤波结果图的直方图确定自适应阈值,并进行阈值分割;对图像进行细化后通过划痕缺陷长度特征进行缺陷的提取。实验结果表明,该划痕缺陷检测算法能够实现长度0.5 mm以上的划痕缺陷的准确检测,检测效率高,满足企业的实际需求。     

基于图像处理的电子元器件表面缺陷检测技术

针对电子元器件的表面缺陷对其性能和品质有直接影响的问题,基于图像处理技术,以晶振外壳缺陷的检测为例,研究了电子元器件表面缺陷的检测技术。首先建立了检测系统平台,然后对所获取的图像进行分析和检测目标分割,并由生成的灰度共生矩阵计算纹理特性值。基于惯性值对纹理的深浅具有敏感性,比较研究之后,采用了惯性值作为检测表面缺陷的指标。试验结果表明该方法能够有效检测平面缺陷。     

基于单幅图像的三维视觉测量与缺陷检测技术研究

研究了一种基于明暗恢复形状的三维视觉测量方法,可以通过单幅灰度图像获取表面三维形貌信息,进而对工件表面加工质艇或缺陷进行检测.本文采用了轴承表面质量进行检测分析,实验证明该三维测量方法能够精确获取工件表面的深度信息,并可以检测出轴承工件的表面缺陷区域的深度信息.同时,该方法具有操作简单,硬件成本低,处理速度快,精度较高的特点,能够用于工业在线检测.     

非金属工件电镀表面缺陷现场检测技术的研究

生产过程中对表面缺陷的检测非常需要。介绍了一种为现场设计的非金属电镀工件表面缺陷检测技术 ,系统在计算机信息技术、图像处理技术和自动测控技术相结合的基础上开发而成。该技术设计了具有均匀散射照明光源的 CCD光电检测系统 ,通过图像处理和计算机自动控制 ,由执行机构自动分选出带有表面缺陷的工件。该系统有效的解决了电镀工件表面的强反射光给图像处理和检测带来的困难 ,适于生产现场自动检测应用。     

基于RPCA的轴承表面缺陷视觉显著性检测

针对轴承缺陷图像存在的光照不均匀、整体对比度低、缺陷细节模糊等问题,结合鲁棒主成分分析和视觉显著性,提出一种新的轴承表面缺陷检测算法。首先,在鲁棒主成分的基础上,采用广朗日乘子算法计算稀疏矩阵,并根据稀疏矩阵计算缺陷区域的视觉显著值,生成凸显缺陷区域的显著图,然后,利用Otsu法对显著图进行缺陷分割,得到缺陷检测结果。并进行对比实验研究,结果表明,该检测方法能够显著突出轴承表面缺陷区域,实现对轴承表面各类缺陷的有效检测,与其他几种显著性检测方法相比,具有较好的查准率和召回率。     

基于机器视觉和改进形态学边缘检测算法的钢球缺陷检测技术研究

为提高钢球表面缺陷检测的效率和准确性,设计一种基于机器视觉的钢球表面缺陷分拣系统。对钢球表面图像进行图像分割、平滑去噪和二值化预处理,获取钢球表面图像的准确信息,并采用改进的中值滤波算法去除噪声;利用小波变换和多尺度形态学融合算法进行钢球表面缺陷的边缘检测;通过该融合算法和其他算法的检测结果对比和客观数据评价,验证了本文所提算法能够有效保留图像真实细节,并满足钢球分拣系统的需求。     

一种基于多目机器视觉的光学薄膜瑕疵检测系统

针对光学薄膜制备过程中的表面品质缺陷,给出一种基于机器视觉的多通道同步自动在线检测方案。瑕疵检测精度、薄膜幅宽决定系统须采用多相机同步采样,为此提出一种C/S系统架构处理和管理各通道数据,并采用高性能FPGA设计了多通道同步控制卡。在实验基础上分析了薄膜瑕疵成像特点,提出一种基于误差修正理论的快速瑕疵检测算法,即:无瑕疵的背景图像认为只存在呈正态分布的随机误差,瑕疵看作粗大误差,从而将瑕疵检测问题转化成粗大误差的判别问题。通过连续多次采样求平均的方法得到标准背景图像,算出每个像素坐标位置的标准差σ,然后对每幅待检图像应用3-σ准则进行瑕疵判别,可直接得到二值化的瑕疵分离结果。实验结果表明,该算法具有简单、快速、准确的优点,解决了普通差影算法阈值确定困难的问题。     

Morphologic and cathodoluminescence studies of diamond films by scanning electron microscopy

Using recent papers on scanning electron microscopy (SEM) of chemical vapor deposition (CVD) diamond films, two analytical applications of the SEM are discussed: the morphologic investigations (secondary electron emission mode) and the recognition of impurities and defects [cathodoluminescence (CL) mode]. Studies of CVD diamond films by SEM demonstrate that the morphologies of these films are affected by synthesis conditions, especially by substrate temperature, methane concentration, and total pressure in the reactor. CL spectra and images are useful tools for clarifying the relationship between emission centers and different types of defects generated during the process of diamond crystal growth. The paper shows that the investigations of the morphology, crystallinity, local CL emission, as well as the surface distribution of CL spectra on CVD diamond films by SEM led to the correlative information for quality estimation of films in comparison with natural diamond.     

Studies on ZDDP Thermal Film Formation by XANES Spectroscopy, Atomic Force Microscopy, FIB/SEM and 31P NMR

X-ray absorption near edge structure (XANES) spectroscopy has been used to characterize the chemistry of thermal films on steel samples, which were generated from a mineral base oil containing a zinc dialkyl dithiophosphate (ZDDP) additive. These films were formed at 150 °C by immersing steel coupons in ZDDP oil solutions. The phosphorus L-edge XANES spectra show that these films are composed of polyphosphates, unreacted ZDDP and other thiophosphate intermediates. Phosphorus K-edge FY XANES was used to monitor the thickness of these films, and the data are consistent with thickness derived by focussed ion beam (FIB) milling and SEM imaging. The sulphur K-edge TEY and FY XANES spectra show that these films are composed of different sulphur components, which depend upon the formation times. The surface morphology of these films was investigated using atomic force microscopy (AFM). These images show that the surface morphology of the thermal films changes with the formation time. 31P NMR spectra show that both primary and secondary ZDDP decomposes gradually at 150 °C.     

基于多幅X射线数字图像的缺陷自动识别技术

在现有的X射线数字图像自动识别方法中,多采用对单幅数字图像进行孤立评判的方法。由于此类方法中阈值选取难以最优化,因而存在一定的误判率。为了解决这一问题,提出了一种X射线数字图像自动识别新方法。该方法将识别过程分为两步:缺陷提取和缺陷跟踪。第一步利用传统方法在每幅图像中分离出潜在缺陷。这一步保证真缺陷能全部提取出来,而不考虑伪缺陷的数量。第二步力图找出同一试件不同图像中分离出的缺陷之间的相互关系。如果第一步某一图像中分离出的某一缺陷在其他图像中都找不到相对应的缺陷区域,就定义该缺陷为伪缺陷,也就是说,真缺陷在不同图像中必须满足一定的几何关系。多幅图像中的缺陷跟踪综合利用了极线约束、三维重建和三线性约束等立体视觉算法。该方法的检测效果已经利用航空发动机叶片X射线数字图像得到验证。试验结果表明:利用该方法可以提高真缺陷的识别率,降低误判率。     

Polymers on graphite and gold: molecular images and substrate defects

Sub-monomolecular layers of a derivatized Phthalocyaninato-poly-siloxane (PCPS) and a Polydiacetylene (PDA) have been prepared on highly orientated pyrolytic graphite and gold by Langmuir-Blodgett and Langmuir-Schäfer techniques. Raman scattering and grazing-incidence-reflection infrared spectroscopy were used to characterize packing and molecular orientations within these films. It was found that PCPS forms similarly well ordered monolayers on both graphite and gold while PDA does not. Scanning tunnelling microscopy (STM) was performed in air using both the constant height and constant current mode. On highly orientated pyrolytic graphite substrates various defect structures based on a √3X√3 R 30° superstructure were found near localized defects and small graphite steps. They were carefully distinguished from molecular images of densely packed hydrocarbon chains. Defect rich graphite was characterized with both STM and Raman spectroscopy. Similarly, disordered and graphite-like regions were found on carbon fibres. The sensitivity of STM to surface defects in graphitic material turned out to be large compared to that of the Raman spectroscopic method. STM images of PCPS monolayers on graphite and gold exhibited parallel polymer rods, 2 nm apart from each other.     

Simple experimental method for alpha particle range determination in lead iodide films

An experimental method for determining the range of alpha particles in films based on I-V(s) analysis has been suggested. The range of 5.5 MeV alpha particles in PbI(2) films determined by this technique is 30+/-5 microm, and this value is in agreement with the value calculated by SRIM (the stopping and range of ions in matter), r=24 microm in PbI(2). More than 100 I-V(s) of PbI(2) films with different thicknesses and quality have been analyzed, and the influence of alpha particle radiation on PbI(2) I-V(s) curves has been studied. Developed analytical methods (dependence of current density on electric field and conception of surface defects) were used, and the method limitations are discussed. It was shown that I-V(s) demonstrate the tendency to obey Ohm's law under alpha radiation. On the other hand, dark conductivity of the lead iodide films shows a typical impure character that can lead to an overestimation of the alpha particles' range in PbI(2) films. After films were exposed to alpha radiation, the dark resistivity and I-V shape of some films improved. Also, a weak decrease of the charge carrier concentration, due to a decrease of the "surface defect" concentration ("surface refining"), was registered after successive measurements of I-V(s).     

BGA焊点空洞缺陷的数字形态学检测

空洞缺陷是BGA焊接缺陷中比较常见的一种, 主要由回流焊时产生的气体没有及时排出而导致. X射线无损检测技术可以使空洞缺陷在焊球图像上显示为白色区域供技术人员查看, 但在噪声、不均匀照射、存在与缺陷类似目标的干扰下如何准确地自动提取缺陷一直是个难题. 提出使用Otsu算法分割焊球, 使用数学形态学中的开闭运算、 顶帽变换提取空洞缺陷的技术方案, 实现每个焊点空洞缺陷的准确提取和自动分析. 实验结果表明, 提出的技术方案鲁棒性强, 可应用于BGA焊点的空洞缺陷检测.     

Indirect monitoring of the dynamics at the contact interfaces of a pair of abrading titanium specimens

The novel method of using continuous tribo-electrification variations to monitor the dynamic tribological properties between metal films has been applied successfully [Y.P. Chang, A novel method of using continuous tribo-electrification variations for monitoring the tribological properties between the pure metal films, Wear, 262 (2007) 411–423]. The method was shown to produce clear and strong signals, superior to monitoring continuous friction coefficient variations. However, the above method was only shown to be suitable for the tested material pairs that were studied. In this paper, the method was improved and applied to monitoring the dynamic tribological properties between titanium oxide (TiO₂) films in the friction process. The experiment was conducted on a purposed-designed friction tester with a suitable measuring system. In order to investigate the tribological property of titanium (Ti) sliding against Ti with TiO₂ films in detail, the continuous variations of electrical contact resistance and friction coefficient were measured for monitoring the onset of film rupture between the TiO₂ films and the chemical reactions between the interfaces. Wear loss was measured by an accuracy balance and scan electron microscopy was used to observe the microstructures and material transfer. The experiments demonstrated that the novel method of using electrical contact resistance variations has great potential for monitoring the dynamic tribological properties and the chemical reactions of titanium specimens.     

Quantification of defects depth in glass fiber reinforced plastic plate by infrared lock-in thermography

The increasing use of composite materials in various industries has evidenced the need for development of more effective nondestructive evaluation methodologies in order to reduce rejected parts and to optimize production cost. Infrared thermography is a noncontact, fast and reliable non-destructive evaluation technique that has received vast and growing attention for diagnostic and monitoring in the recent years. This paper describes the quantitative analysis of artificial defects in Glass fiber reinforced plastic plate by using Lockin infrared thermography. The experimental analysis was performed at several excitation frequencies to investigate the sample ranging from 2.946 Hz down to 0.019 Hz and the effects of each excitation frequency on defect detachability. The four point method was used in post processing of every pixel of thermal images using the MATLAB programming language. The relationship between the phase contrast with defects depth and area was examined. Finally, phase contrast method was used to calculate the defects depth considering the thermal diffusivity of the material being inspected and the excitation frequency for which the defect becomes visible. The obtained results demonstrated the effectiveness of Lock-in infrared thermography as a powerful measurement technique for the inspection of Glass fiber reinforced plastic structures.     

X-Ray Absorption Spectroscopy and Morphology Study on Antiwear Films Derived from ZDDP Under Different Sliding Frequencies

X-ray absorption near-edge structure (XANES) spectroscopy has been used to characterize the chemistry of antiwear (AW) films generated from mineral base oil containing a zinc dialkyl dithiophosphate additive. These films were formed on rubbed steel surfaces with a reciprocating boundary contact using different sliding frequencies. The phosphorus L-edge XANES spectra show that these films have slightly different chemical natures. Longer chain polyphosphates were present on the steel surface prepared at the higher sliding frequencies. The surface morphology of these films was investigated using atomic force microscopy. These images show that the surface morphology of the AW films changes with the sliding frequency. Round and bigger antiwear pads were formed at a lower frequency while higher frequencies resulted in thinner films and flattened surfaces. Nanomechanical properties of these antiwear films were investigated by nanoindentation measurement and the elastic moduli extracted from force–displacement (f–d) curves are similar for all antiwear films, 100 ± 10 GPa.     

Extraction of weld defects dimension from radiographic images using the level set segmentation without re-initialization

Radiographic images segmentation is the major interest for the weld defect diagnosis and monitoring in the field of industry. In this work we present a method that takes ownership of local segmentation geodesic active contours. The goal of the method presented in this paper is to automate the process of extracting dimension of weld defects from radiographic images using level set segmentation. The information would be used in the area of nondestructive testing (NDT). This method is found to be effective and robust.