Three-Dimensional Model Reconstruction of Nonwovens from Multi-Focus Images

The three-dimensional(3D)model is of great significance to analyze the performance of nonwovens.However,the existing modelling methods could not reconstruct the 3D structure of nonwovens at low cost.A new method based on deep learning was proposed to reconstruct 3D models of nonwovens from multi-focus images.A convolutional neural network was trained to extract clear fibers from sequence images.Image processing algorithms were used to obtain the radius,the central axis,and depth information of fibers from the extraction results.Based on this information,3D models were built in 3D space.Furthermore,self-developed algorithms optimized the central axis and depth of fibers,which made fibers more realistic and continuous.The method with lower cost could reconstruct 3D models of nonwovens conveniently.

Three-Dimensional Measurement and Reconstruction of Fabric Drape Shape

This paper introduces a new method of measuring the three-dimensional drape shape of fabrics with structural light. First, we apply parallel annular structural light to form light and shade alternating contour stripes on the surface of fabrics. We then collect the images of contour stripes using Charge Coupled Device (CCD). Subsequently, we process the images to identify the contour stripes and edges of fabrics, and obtain the fabric contour lines of curved surfaces. Finally, we apply three-dimensional curved surface modeling method based on a network of polar coordinates, and reconstruct the three-dimensional drape shape of fabrics. Experiments show that our method is effective in testing and reconstructing three-dimensional drape shape of fabrics.

Multi-lens stereo reconstruction of wave surface in a wave basin

A new method is developed to measure the wave surface elevation in a wave basin.A thin film of tiny granules floating on the free surface ensures the formation of the dot grid pattern on the wave surface.The dot grid pattern is generated by a high-brightness projector.The granule film has little influence on the wave propagation in a wave basin.The measuring system is validated by the experiments of the propagation of solitary wave in water of uniform depth.It turns out that the time series of the surface elevation measured by the multi-lens stereo reconstruction method agree well with that measured by wave gauges.The errors of the measuring system are discussed.This work sheds a light on measuring the local wave field with high precision and efficiency in laboratory.

Three-dimensional modeling and reconstructive change of residual stress during machining process of milling, polishing, heat treatment, vibratory finishing, and shot peening of fan blade

Residual stress during the machining process has always been a research hotspot,especially for aero-engine blades.The three-dimensional modeling and reconstructive laws of residual stress among various processes in the machining process of the fan blade is studied in this paper.The fan blades of Ti-6Al-4V are targeted for milling,polishing,heat treatment,vibratory finishing,and shot peening.The surface and subsurface residual stress after each process is measured by the X-ray diffraction method.The distribution of the surface and subsurface residual stress is analyzed.The Rational Taylor surface function and cosine decay function are used to fit the characteristic function of the residual stress distribution,and the empirical formula with high fitting accuracy is obtained.The value and distribution of surface and subsurface residual stress vary greatly due to different processing techniques.The reconstructive change of the surface and subsurface residual stress of the blade in each process intuitively shows the change of the residual stress between the processes,which has a high reference significance for the research on the residual stress of the blade processing and the optimization of the entire blade process.

复杂型面结构超声成像检测研究进展

工程结构普遍具有复杂型面,而且此类区域是制造和服役阶段损伤甚至失效的高发区。超声无损检测是评估结构制造质量和服役安全性的重要手段,但复杂型面给超声检测带来了声耦合困难、随形扫描控制难、超声入射/接收难、超声传播行为复杂等挑战。本文分别从超声耦合策略、超声换能器、随形扫描控制、复杂型面轮廓重建、成像算法、复杂型面与材料复杂耦合的影响几大方面,概述了现有研究已取得的重要进展。最后探讨了复杂型面超声成像检测研究仍面临的挑战以及未来发展趋势,包括超复杂型面三维异形结构检测、高柔性二维阵列超声换能器、以及同时考虑宏/微观结构和材料自身声学特性的影响改进超声成像算法。

High-quality indoor scene 3D reconstruction with RGB-D cameras:A brief review

High-quality 3D reconstruction is an important topic in computer graphics and computer vision with many applications,such as robotics and augmented reality.The advent of consumer RGB-D cameras has made a profound advance in indoor scene reconstruction.For the past few years,researchers have spent significant effort to develop algorithms to capture 3D models with RGB-D cameras.As depth images produced by consumer RGB-D cameras are noisy and incomplete when surfaces are shiny,bright,transparent,or far from the camera,obtaining highquality 3D scene models is still a challenge for existing systems.We here review high-quality 3D indoor scene reconstruction methods using consumer RGB-D cameras.In this paper,we make comparisons and analyses from the following aspects:(i)depth processing methods in 3D reconstruction are reviewed in terms of enhancement and completion,(ii)ICP-based,feature-based,and hybrid methods of camera pose estimation methods are reviewed,and(iii)surface reconstruction methods are reviewed in terms of surface fusion,optimization,and completion.The performance of state-of-the-art methods is also compared and analyzed.This survey will be useful for researchers who want to follow best practices in designing new high-quality 3D reconstruction methods.

Error evaluation of binocular vision method for reconstructing breaking wave surface

Quantization error and its source of the binocular stereo vision technique are evaluated by measuring breaking wave surface in wave flume.The verification test focuses on the run-up process of a breaking solitary wave on a slope.The three-dimensional topography of the free surface is reconstructed at three representative stages of wave propagation.The two-dimensional profile from the side view and statistical data are used to evaluate the quantization error.The breaking degree of wave surface has a mild influence on the performance of stereo reconstruction in the current configuration.The mean error of detecting the surface elevation is about 2 mm,around 0.1%of the measurement distance.Sixty percent of the error comes from the deficient match in image interrogation,and the remaining forty percent arises from the limited system configuration.The precise stereo measurement of the breaking surface shows that the present binocular vision system provides a new technique in experiments in hydrodynamics with the breaking surface.

基于暗场线扫技术的球体表面缺陷检测技术

钢球的球体几何特性令表面上的缺陷难以连贯成像引起漏检或缺陷畸变被误判,且钢球的镜面反射效应可能造成缺陷信号被周围环境的像和高反亮斑淹没,导致缺陷漏检。为此,提出了一种基于光学暗场线扫描技术的钢球表面缺陷检测方法,设计搭建了一套适用于高反射率钢球的暗场线扫表面缺陷检测系统,开发了图像预处理、畸变校正、缺陷提取、点云重建、分割和过滤等算法,建立球面图像点间的空间几何关系,成功将钢球表面缺陷在三维球面上重构,实现了轴承钢球表面缺陷的三维连贯检测。实验结果表明,对表面完好的无损球测量重复性为0.14%,对表面带有缺陷的磨损球测量重复性为0.11%。

海量散乱点的曲面重建算法研究

基于海量散乱点的曲面重建在机械产品测量造型、计算机视觉、根据切片数据的医学图像重建等领域中有重要应用 .给出了一种以物体表面上不附加任何几何和拓扑信息 (包括测点法矢、曲面边界信息 )的散乱点集为处理对象 ,自动生成物体表面的三角网格模型的算法 .该算法首先根据测点的邻近测点估算曲面在该测点处的法矢 ,并采用优化的顺序对法矢方向进行调整以使各测点处的法矢都指向曲面外侧 ,最后用步进立方体算法输出三角网格模型 .采用新的方法计算切平面 ,不但进一步提高了效率 ,而且改善了曲面边界及尖锐棱边区域的重建效果 .还提出并解决了法矢方向传播中可能出现的局部“孤岛”问题 .同时 ,提出了一种对海量数据进行空间划分的算法 ,从而大大提高了海量数据的处理效率 .应用实例表明 ,算法效果良好 .

基于光度立体法的中医舌体三维表面重建

在中医舌诊客观化研究进程中,以2维舌图像为基础进行分析所获得的舌体特征信息有限,制约了舌诊的客观化发展。在将光度立体法用于静态舌模型表面3维重建的可行性研究基础之上,为了对真实舌体表面进行3维重建,该文设计一种舌体3维动态信息采集系统,用于舌像采集;然后采用光度立体法依次求出舌体表面的法向量、纹理反射率和深度信息,并基于Direct X进行3维显示;最后针对舌体表面高光区域对重建结果的影响,采用高光剔除算法剔除高光。经高光剔除后的重建结果和原来相比,在重建精度和鲁棒性等方面都有更好的表现。经实验证实,该重建方法的平均相对误差约为7.24%,重建结果可形象地表达舌体的表面形态和齿痕等细节信息,可辅助中医进行诊断。

体素重建中的快速移动立方体方法

利用二维切片数据重建三维结构,在医学等领域有着重要的应用。在大数据量的情况下,等值面抽取的计算效率是富有挑战性的课题。本文提出了一种改进的等值面抽取算法。该算法采用区域增长和一种优化的搜索策略加速边界体素的寻找,避免了整个体素空间的搜索,从而加速等值面抽取。实验结果表明,本算法大大加速了传统的移动立方体(Marching Cubes)方法,特别是在大数据量时,加速的效果更加明显。

股骨上段的三维重建

人体髋关节在发生严重病变、骨折等情况时，进行人工髋关节的置换是一条可行的途径。股骨的三维重建为髋关节假体的优化设计提供了依据。计算机断层扫描技术的发展，使得以无损伤技术获得病人关节骨断层的几何形状成为可能。将不同层面上的二维图像数据按照断层的空间位置依次排列，就可以形象地显示病人关节的三维图像。本文以人体髋关节为例，从ＣＴ片的图像处理、轮廓数据筛选、三维网格面建立以及股骨实体化等几方面介绍了股骨上端的三维重建过程。

多视角几何三维重建法识别工程结构缺损与变形

裂缝等工程结构表面缺损与弯曲等构件变形是既有工程结构安全性检测鉴定与监测的重要指标,采用数字图像法可有效改善当前人工检查技术的缺点,但在工程实践中显现出定量检测中图像几何变形修正困难、局部损伤难以在整体结构中定位、无法测量钢结构空间变形等问题。结合多视角几何三维重建法的研究与应用可有效解决上述问题。该文对多视角几何三维重建的方法原理与算法实现进行讨论,列举方法实践的经典算法与高效算法;围绕数字图像法检测在工程实践中存在的问题,提出表面投影法解决成像几何变形与损伤定位、逆向工程建模与特征提取识别几何变形损伤;通过结构表面裂缝识别、大型构筑物表面损伤定位、钢结构构件变形识别等3个应用研究,阐述多视角几何三维重建法在结构表面损伤与变形识别中的具体操作方法与比较优势。多视角几何三维重建法具有设备要求低、操作方便快捷、模型色彩丰富、点云精度较高等特点,与数字图像法结合后在工程结构检测监测领域的研究与应用潜力巨大。

基于SFS原理的SMT焊点表面三维重构技术研究

基于明暗重构形状原理重构表面组装焊点的表面三维形状过程是:先通过图像采集设备,采集到SMT焊点图像,使用相关的图像处理技术,对SMT焊点图像进行处理;根据一个确定的反射模型建立物体表面形状与图像亮度之间的约束关系和物体表面形状的先验知识建立物体表面形状参数的约束关系,然后对这些约束关系联立求解,可得到物体表面的三维形状。同时针对不可接受SMT焊点图像重构出的三维图像不够理想的缺点进行了改进。在阐述其基本思想和原理的基础上,结合实例介绍了该技术的实现方法与步骤,对其中焊点图像的获取与处理、焊点三维重构技术算法等主要内容与关键技术进行了研究和探讨,并对结果进行了分析验证。

基于髋关节CT图像三维实体髋骨重建及髋骨承受力数据有限元分析

背景:髋骨的解剖结构与受力复杂,传统的X 射线片及CT 扫描等二维影像检查在髋骨诊断中具有一定的局限性。目的:在CT 图像的基础上,找寻一种生物骨骼实体模型准确且快捷的重建方法,研究人体内髋骨的受力情况,建立出术前优化髋关节模型,以便3D 打印。方法:采集1 例患者的髋关节CT 图像(实验已获包头医学院第一附属医院伦理批准,受试者对实验知情同意),使用核心医学Mimics 软件,从光学图像和噪声点中找出髋关节CT 图像,并且获得矢量断层图像,人工仔细分离后拟合多层图像,形成3D 图像结构。之后,在Geomagic 软件中完成对模型的曲面拟合,与此同时生成曲面和实体化。最后,在Abaqus 软件中完成髋关节的网格划分,模型赋值和有限元模拟分析。至此,完成从原始CT 到实体的所有数据转换,同时使用Anybody 软件用于模拟人体步态模拟,进行有限元分析。结果与结论:使用最简单方法完成了CT 图像的重建,经过Geomagic 完成曲面拟合并得到了良好的髋骨模型,通过Anybody 逆动力学分析得到了重要的髋骨承受力数据,得到体内承受力较大的地方为髋骨远端处,有限元分析验证了此结果。

基于图像分割的三维立体匹配改进算法

传统基于分割的立体匹配算法所采用的局部匹配方法误匹配率较高,同时平面拟合也具有一定的局限性。为了获得高准确率的视差图,本文将改进的局部匹配方法与全局匹配相结合,来获取初始视差图;同时利用曲面拟合的方法对各分割区域进行处理,并在区域合并中提出了新的判决准则。仿真实验结果表明,本文算法能够获得较高精度的视差图,实现了各纹理区域之间的平滑过渡,同时,在低纹理区域和被遮挡区域也取得了比较理想的效果。

基于特征的扫描电镜立体对三维重建

应用扫描电子显微镜（ＳＥＭ）立体对成像技术和图像处理、计算机视觉方法建立了一套微结构形貌三维自动重建系统。在Ｂａｒｎａｒｄ算法基础上根据ＳＥＭ立体对成像的特点提出了一种基于边界特征的对应点匹配概率松弛迭代方法。系统中融合了基于区域的立体匹配原理，检出的特征定位准确。匹配过程引入多种几何及边界强度和位向约束，具有较高的匹配效率和精度。应用本系统进行了典型金属断裂表面的三维重建。

基于相位法的三维面型测量及曲面重建技术

文中采用相位法对三维面型测量和曲面重建技术进行了探讨,实现了三维信息获取和曲面重建的方法。该方法是用非相干光源投射的虚拟光栅和 CCD 摄像机构成的视觉系统对物体进行一次扫描获取物体表面的三维测量数据,这些数据经过处理后转换成*.ibl 文件,借助于 Pro/Engineer 软件重建三维物体表面。该方法快速、简便,适合于实时测量、加工,同时可实现测量、处理、加工的一体化,从而为反求工程的发展提供了一个很好的新方法。最后给出了实验,实验结果证明了方法的可行性和有效性。

基于机器视觉的管道内表面三维重建研究

针对管道内表面腐蚀测量和三维可视化的特殊背景和实际需求,研究了基于双目视觉的管道内表面三维重建方法.图像获取部分采用双目立体视觉系统,在机器视觉技术的理论基础上给出了相应的三维重建流程.采用张正友平面标定法进行摄像机标定,采用BP神经网络对图像进行边缘检测,采用自适应SIFT匹配算法进行立体匹配,综合计算摄像机标定获取的参数与立体匹配获取的视差,最终实现了管道内表面局部图像的三维重建.

基于改进光照模型由SFS方法重构SMT焊点三维形状技术

在采用SFS方法重构SMT焊点三维形状技术中,由于焊点表面镜面反射而在图像中产生的亮点会影响重构形状,从而降低SMT焊点三维重构精度,为此提出一种改进光照模型,该模型不仅对物体表面漫反射分量进行了改进,同时考虑了物体表面镜面反射分量对表面重构所产生的影响,把漫反射分量和镜面反射分量线性叠加,在采用图像处理技术对焊点图像进行适当处理后,利用改进模型对SMT焊点进行三维重构.结果表明,采用改进光照模型与传统光照模型相比,改进光照模型能较好地减少镜面反射对重构结果的影响.