基于频域滤波的电缆线芯图像纹理消除研究

电缆作为电力传输的重要部分,其线芯质量关系到用电安全及传输效率。 在利用机器视觉技术测量电缆线芯结构参数 时,刀痕纹理严重影响了图像分割效果,降低了测量精度。 针对这一问题,提出了一种基于改进频域滤波的线芯截面图像刀痕 纹理消除方法:首先采用相对总变差模型分解线芯截面图像,得到纹理图像描述;对线芯截面纹理图像进行快速傅里叶变换,分 析刀痕纹理在频域中的能量主方向;研究改进型巴特沃斯陷波滤波器,设计椭圆形的邻域窄阻带以降低刀痕纹理对应的带状区 域能量,从而消除刀痕纹理。 实验结果表明,该方法的刀痕纹理消除率可达 92% ,提升了电缆线芯质量检测的效率。     

Gabor滤波器在自动视觉表面检测中的应用

提出一种基于Gabor滤波器的纹理表面缺陷检测方法。该方法基于Gabor滤波方案，计算纹理图像与某一Gabor滤波器卷积能量输出响应。选择Gabor滤波器的最优参数，采用简单的二值化技术将纹理差别转化为可检测的滤波器输出。纺织品、铣磨面等结构纹理和皮革、砂纸等随机纹理的实验结果证明了该方法在识别不可定量测量缺陷方面的有效性。     

金属表面特征提取方法研究进展

粗糙表面几何特征的分析和提取是零件表面分类、缺陷识别、几何建模等研究工作中至关重要的环节。文中对灰度共生矩阵法、功率谱密度函数法、小波变换以及contourlet变换4种常用粗糙表面几何特征提取方法进行分析研究,认为contourlet变换在继承小波变换多尺度分析能力的基础上,其多方向分析能力有助于更加准确地捕捉表面的曲线纹理特征,因此是一种具有潜力、值得重视的表面几何特征提取新方法。     

基于纹理分布特征的虹膜识别算法

二维小波变换方向性差,不能从方向和频率同时描述虹膜纹理.基于此,分析了归一化虹膜纹理的分布特点,提出了基于纹理分布特征的虹膜识别算法.对原始人眼图像进行虹膜定位等预处理操作,得到归一化的虹膜纹理图像;对归一化虹膜图像进行了二维小波多尺度变换,结合虹膜纹理分布特点,选择小波分解水平通道;依据点能量贡献度,消除伪特征点并进行了点能量编码;计算了不同虹膜编码间海明距离,以其为依据进行分类.在给定距离阈值为0.25的前提下,可以达到99.91%的正确识别率.实验证明该算法是有效、可行的,并具有较高的识别率,识别速度也很快.     

基于环形对称Gabor变换的接触网销钉检测研究

针对接触网旋转双耳销钉,提出一种基于环形对称Gabor变换特征的销钉松脱与缺失状态检测方法。融合旋转不变LBP特征与HOG特征的方法训练SVM分类器,实现旋转双耳定位识别;采用圆弧检测完成销钉准确定位,实现销钉区域的分割;利用环形对称Gabor变换完成纹理边缘信息的特征提取,结合BP神经网络实现对旋转双耳销钉状态的判断识别。研究和实验结果表明,该方法实现了在销钉穿插方向任意和形状多样等复杂情况下的状态识别。     

随机减量法在工作模态参数识别中的应用研究

实际工程应用中施加于结构的激励及其响应信号难以测量,不能用传统的识别方法提取结构的模态参数。研究了运用随机减量法从易得的结构的随机响应中提取某种自由响应信号,然后利用小波变换识别模态参数。实验通过对悬臂梁的模态参数识别,结果表明随机减量法在模态参数识别过程中的适用性及用该方法获取的模态参数具有很高的精度。     

微分变换法在动力学响应分析及系统边界参数识别中的应用研究

引入求解非线性微分方程的微分变换法，将其推广为广义微分变换法。建立求解一般非线性振动微分方程的一般框架，将此方法用于求解著名的Vander pol方程。并且将微分变换法推广到结构边界参数识别，以一个典型的悬臂梁边界参数识别为例，对其进行数值仿真和实验研究，并将此方法的实验研究识别结果与用实测频率响应函数法的识别结果作比较。说明该方法具有良好的工程应用价值。     

基于小波变换纹理一致性测度的遥感图像融合算法

本文提出了一种基于小波变换的遥感图像融合算法，利用多分辨小波变换的系数，采用低频图像的小波系数最小值作为融合后的低频系数，高频图像根据纹理一致性测度的纹理检测确定融合规则，调整高频小波系数大小。利用小波变换对图像相对应的低频分量及各方向细节分量进行针对性融合处理，很好地将来自不同图像的特征与细节融合在一起，并对融合图像质量进行了对比评价。实验结果表明，这种方法能够在保留图像微小细节方面获得满意的结果，这种算法有效且优于传统的图像融合方法。     

基于区域生长法的自适应图像分割的网眼织物瑕疵检测

为了解决网眼织物在织造过程中的实时检测问题,提出了一种基于频域滤波和区域生长法的瑕疵区域自适应分割方法。通过分析傅里叶变换后的功率谱获得网眼织物纱线分布特征后,通过设计滤波器自动滤除了与织物纹理背景相关的频率成分以削弱织物自身纱线分布结构对瑕疵检测精度的干扰。利用滤波后图像的灰度分布特点自适应地求取了分割参数,并使用区域生长法实现了对瑕疵区域的自动分割。利用对获得的二值图像进行形态学运算,得到了最终的瑕疵区域分割图像。实验结果表明,该算法能较准确地对瑕疵进行定位,瑕疵区域分割精度较高,自适应性和抗干扰能力较强,可以满足工业检测的要求。     

基于Chirp-Z变换的结构热模态试验方法

本文提出了一种基于调频Z变换(Chirp-Z变换)的时变系统模态参数识别算法,通过对响应数据进行Chirp-Z变换获得细化频率后的结构频响特性,进而通过模态参数识别算法获得模态参数,并利用三自由度系统对该方法进行了验证,最后对受热机翼结构进行了动态特性参数识别,为高超声速热模态测试提供了一种方法.     

一种编织物三维模拟的新方法

首先从编织物的结构中提取出三种基本编织结构：Knit、Tuck、Welt,并对这三种结构进行模型构造;然后分析纹理图案的bmp图片,从中得到编织物针珐数据;接着把这三种基本结构按照得到的针珐数据连接起来,生成一条条粗糙的纱线中心线,并由此得到粗略的编织物框架;再从纱线的每个结构中选取一定的点作为控制点,由这些控制点生成NURBS曲线,得到精确、平滑的纱线中心线;然后构造纱线的横截面,根据横截面生成纱线表面的三角面片;最后按照每条纱线的颜色显示这些面片,从而生成逼真的编织物模拟效果。同时,运用弹簧一质点模型,模拟了编织物受到拉力后变形的情况。     

Influence of weave structures on the tribological properties of hybrid Kevlar/PTFE fabric composites

The existing research of the woven fabric self-lubricating liner mainly focus on the tribological performance improvements and the service life raised by changing different fiber type combinations, adding additive modification, and performing fiber surface modification. As fabric composites, the weave structures play an important role in the mechanical and tribological performances of the liners. However, hardly any literature is available on the friction and wear behavior of such composites with different weave structures. In this paper, three weave structures (plain, twill 1/3 and satin 8/5) of hybrid Kevlar/PTFE fabric composites are selected and pin-on-flat linear reciprocating wear studies are done on a CETR tester under different pressures and different frequencies. The relationship between the tensile strength and the wear performance are studied. The morphologies of the worn surfaces under the typical test conditions are analyzed by means of scanning electron microscopy (SEM). The analysis results show that at 10 MPa, satin 8/5 performs the best in friction-reduction and antiwear performance, and plain is the worst. At 30 MPa, however, the antiwear performance is reversed and satin 8/5 does not even complete the 2 h wear test at 16 Hz. There is no clear evidence proving that the tensile strength has an influence on the wear performance. So the different tribological performance of the three weave structures of fabric composites may be attributed to the different PTFE proportions in the fabric surface and the different wear mechanisms. The fabric composites are divided into three regions: the lubrication region, the reinforced region and the bonding region. The major mechanisms are fatigue wear and the shear effects of the friction force in the lubrication region. In the reinforced region fiber-matrix de-bonding and fiber breakage are involved. The proposed research proposes a regional wear model and further indicates the wear process and the wear mechanism of fabric composites.     

Extraction of yarn positional information from three-dimensional CT image of textile fabric using a yarn model for its structure analysis

This paper proposes a novel method for analyzing a textile fabric structure to extract positional information regarding each yarn using three-dimensional X-ray computed tomography (3D CT) image. Positional relationship among the yarns can be reconstructed using the extracted yarn positional information. In this paper, a sequence of points on the center line of each yarn of the sample is defined as the yarn positional information, since the sequence can be regarded as the representative position of the yarn. The sequence is extracted by tracing the yarn. The yarn is traced by estimating the yarn center and direction and correlating the yarn part of the 3D CT image with a 3D yarn model, which is moved along the estimated yarn direction. The trajectory of the center of the yarn model corresponds to the positional information of the yarn. The application of the proposed method is shown by experimentally applying the proposed method to a 3D CT image of a double-layered woven fabric. Furthermore, the experimental results for a plain knitted fabric show that this method can be applied to even knitted fabrics.     

基于Gabor滤波器对织造前帘子布疵点检测研究

针对织造前帘子布纱线疵点自动检测问题,对织造前帘子布纱线疵点分布特点开展了研究分析,提出了一种基于Gabor滤波器的帘子布纱线疵点检测算法,利用Gabor滤波器的能量值来描述帘子布纱线疵点特征。通过对疵点Gabor图像阈值化处理,得到了疵点的二值图像。然后对疵点二值图像沿水平方向进行了均等分割,得到了疵点所在的条状图像并除去随机噪声。在1 792个疵点图像测试集上对不同尺度参数的Gabor滤波器进行疵点检测准确度分析,建立了最优的方向和尺度参数组合。并搭建了基于机器视觉技术的帘子布纱线疵点检测系统。研究结果表明,该检测算法对帘子布纱线疵点检测准确度达到99.2%。经过4个月生产实践表明,该系统对织造前帘子布纱线疵点检测效果比人工检测效果提高86%,实现了机器换人的目的。     

Fabric model for clothing design by wrinkle simulation

During the design process in clothing production, fashion designers must imagine the state of the fabric used. The state of wrinkles formed on fabric according to human posture and movements is an important design element. However, the selection process of the fabric for realizing the design envisioned by the designer takes a lot of time and effort. By using clothing simulation, it is possible to identify the parameters of weave and mechanical properties of the yarn fabric for realizing the design that the designer requires, and this can facilitate the selection process. Many fabric models taking into account the mechanical properties of fabric based on mass spring models have been proposed, but none of the models proposed take into account the state of friction and slipping at the contact point of the warp and weft. In this paper, a method for setting the weave and mechanical properties of the fabric as parameters in order to simulate a state in which the fabric covers an object or human, and a system based on this method were proposed. The proposed system can perform simple simulation using a mass spring model that takes into account slipping at the contact points of the warp and weft of the fabric worn.     

Underwater cable detection in the images using edge classification based on texture information

In this paper, a new approach is proposed for detection of an underwater cable, which makes an Autonomous Underwater Vehicle (AUV) capable for automatic tracking. In this approach instead of traditional image segmentation, first, edges of the images are extracted. Then they are classified using Multilayer Perceptron (MLP) neural network and Support Vector Machine (SVM) using texture information. Then the edge points belonged to the background information are removed and the remaining ones are used for the next processes. Finally, the filtered edges are repaired by morphological operators and are fed into the Hough transform for cable detection. Some texture information methods are used for feature extraction but the results confirm that the 2D Fourier transform in combination with MLP network is the best method for edge classification in this environment. Hough transform, is used in two strategies, which in the first one, the whole information of the edges in the image, are used for line detection, and in the second approach because of curve like shape of the cable, a center part of the image, is used for line detection. In the experiments, many different scenes was used for testing the cable detection algorithm, which first method, resulted to good accuracy but the second one, provided better recognition rate for the cable detection task.     

A Study of the Static/Kinetic Friction Behavior of PTFE-Based Fabric Composites

Three different polytetrafluoroethylene (PTFE)-based fabric composites were prepared. The static/kinetic friction behaviors of these composites under different loads and speeds were studied. A 3D laser microscope and profile measurement apparatus were used for analysis of the morphology and weave structure of the composites, and the contact temperature of these composites under different loads and speeds was monitored continuously using a high-precision thermal resistor. In addition, a dynamic mechanical analysis (DMA) apparatus was used to explore the thermal and mechanical properties of PTFE-based fabric composites. The results demonstrated that speed/load, weave structure, and fiber form have an important influence on static and kinetic friction behavior of the fabric composites. Generally, the static friction coefficient is greater than the kinetic friction coefficient, except when considering light load conditions. Under light load conditions, the static friction coefficient is equal to the kinetic friction coefficient. In addition, the kinetic friction coefficient first increased and then decreased with increased speed, but the static coefficient increased first and then remained at an almost constant value. At all sliding speeds, the static and kinetic friction coefficients of tape yarn composites are better than those of the multifilament yarn composites. Weave structure has no effect on the static friction coefficient, but it has a significant influence on the kinetic friction coefficient.     

织物密度的数字图像检测方法

针对织物的纱线密度检测,提出一种对织物图像,信息进行检测的方法。通过数字相机获得织物图像,对织物图像进行最优二值化处理。利用像素黑白颜色区分织物上的纱线和间隙,通过扫描经线和纬线方向的像素颜色值获得纱线处像素颜色特性,从而计算出经线方向和纬线方向的纱线数目。根据各方向最大像素数得织物的大小,即可得知织物的密度。     

基于傅里叶变换的数字散斑照相测量

提出了基于傅里叶变换的双曝光散斑场的全场数字化测量。对双曝光散斑图进行傅里叶变换得到条纹图以灰度归一法和相位构造法将其优化,获取低噪声、高对比度的散斑干涉条纹。平面位移的测量实验结果表明,数字散斑照相技术适合计算机进行自动条纹识别和位移场计算,能得到较为精确的水平位移值,可实现变形场的精准测量。