使用单彩色相机的单相机三维数字图像相关方法

介绍了一种基于单个彩色相机的新型全靶面、单相机三维数字图像相关(3D-DIC)方法。借助于设计巧妙的颜色分光光路,被测物体表面图像可以通过两条不同的光路达到相机靶面,采集的标定靶和实验件表面的彩色图像可以分离得到蓝色和红色子图像。通过使用3D-DIC分析标定靶和实验件表面分离后的蓝色和红色子图像,可以获得物体表面的三维形貌和变形。形貌测量、面内和离面平移、以及静动态三维变形实验验证了该单彩色相机3D-DIC方法的有效性和测量精准度。由于可避免双相机同步,且能实现无分辨率损失的全靶面三维形貌和变形测量,本文方法在需要实现瞬态位移和变形测量的爆炸、冲击、振动等领域中具有广阔重要的应用前景。     

Full-Field Surface 3D Shape and Displacement Measurements Using an Unfocused Plenoptic Camera

Full-field surface 3D shape and displacement measurements using a single commercial unfocused plenoptic camera (Lytro Illum) are reported in this work. Before measurements, the unfocused plenoptic camera is calibrated with two consecutive steps, including lateral calibration and depth calibration. Each raw image of a checkerboard pattern recorded by Lytro Illum is first extracted to an array of sub-aperture images (SAIs), and the center sub-aperture images (CSAIs) at diverse poses are used for lateral calibration to determine intrinsic and extrinsic parameters. The parallax maps between the CSAI and the remaining SAIs at each pose are then determined for depth parameters estimation using depth calibration. Furthermore, a newly developed physical-based depth distortion model is established to correct the serious distortion of the depth field. To realize shape and deformation measurements, the raw images of a test sample with speckle patterns premade on its surface are captured by Lytro Illum and extracted to arrays of SAIs. The parallax maps between the CSAI and the target SAIs are obtained using subset-based digital image correlation. Based on the pre-computed intrinsic and depth parameters and the disparity map, the full-field surface 3D shape and displacement of a test object are finally determined. The effectiveness and accuracy of the proposed approach are evaluated by a set of experiments involving the shape reconstruction of a cylinder, in-plane and out-of-plane displacement measurements of a flat plate and 3D full-field displacement measurements of a cantilever beam. The preliminary results indicate that the proposed method is expected to become a novel approach for full-field surface 3D shape and displacement measurements.     

Diffraction Assisted Image Correlation: A Novel Method for Measuring Three-Dimensional Deformation using Two-Dimensional Digital Image Correlation

Digital Image Correlation (DIC) provides a full-field non-contact optical method for accurate deformation measurement of materials, devices and structures. The measurement of three-dimensional (3D) deformation using DIC in general requires imaging with two cameras and a 3D-DIC code. In the present work, a new experimental technique, namely, Diffraction Assisted Image Correlation (DAIC) for 3D displacement measurement using a single camera and 2D-DIC algorithm is presented. A transmission diffraction grating is placed between the specimen and the camera, resulting in multiple images which are then used to obtain apparent in-plane displacements using 2D-DIC. The true in-plane and out-of-plane displacements of the specimen are obtained from the apparent in-plane displacements and the diffraction angle of the grating. The validity and accuracy of the DAIC method are demonstrated through 3D displacement measurement of a small thin membrane. This technique provides new avenues for performing 3D deformation measurements at small length scales and/or dynamic loading conditions.     

Stereo-DIC Calibration and Speckle Image Generator Based on FE Formulations

Stereo digital image correlation (stereo-DIC) is being accepted by the industry as a valid full-field measurement technique for measuring shape, motion and deformation, and it is therefore of utmost importance to provide uncertainties on the obtained measurements. However, the influences on a stereo-DIC measurement are not fully understood; indeed, stereo-DIC is a complex optical-numerical process and it is not always clear how errors are propagating throughout the measurement chain. In order to investigate the magnitude of the different error-sources a simulator for stereo-DIC is proposed. This simulator is able to generate realistic synthetic images as if they were made during a real set-up, so the error sources can be investigated separately and an optimal set-up can be chosen before any physical test is performed. We present in this paper the mathematical approach to the DIC simulator including details on how to convert FE displacement field results to stereo-DIC images. The simulator includes the ability to control the lighting and to create synthetic calibration images. The synthetic images are compared to simulations for a bulge test as a validation of the simulator. Synthetic calibration images are compared to experimental calibration studies to verify those. Finally a brief look at how the simulator could be used for looking at calibration quality is conducted.     

Instantaneous phase-stepping interferometry using polarization imaging with a micro-retarder array

An instantaneous phase-stepping and subsequent phase analysis method, using a CCD camera with a form-birefringent micro-retarder array, is proposed for interferometry. An optical setup of a polarization interferometry using a Twyman-Green interferometer with two polarizers is constructed to analyze the distribution of out-of-plane displacement. Light emerging from the interferometer is recorded using a CCD camera that has micro-retarder array on the CCD plane. This micro-retarder array has four different principal directions. That is, an image obtained by the CCD camera contains four types of data corresponding to four different optical axes of the retarder. The four images separated from the image recorded by the CCD camera are reconstructed using gray-level interpolation. Then, the distributions of the Stokes parameters that represent the state of polarization are calculated from the four images. The phase distribution of the interference fringe pattern produced by the Twyman-Green interferometer is then obtained from these Stokes parameters. This method is applicable to time-dependent phenomena because multiple exposures are unnecessary for sufficient data acquisition in the completion of phase analysis.     

彩色光弹性干涉影像分析系统

自行开发的“彩色光弹性干涉影像分析系统”首先利用CCD成像和图像采集设备,将光弹图像以数字图像的形式存储到计算机,然后通过对存储的光弹图像进行处理,得到物体边界、等差线、等倾线等数据。最后根据这些数据,绘制出主应力迹线,并进行二维的和三维的应力分析。本文着重介绍了系统整体设计以及系统研制的难点问题(彩色光弹图像处理、主应力迹线的绘制等)。系统可以通过对彩色图像进行分解,应用目前已经比较成熟的灰度光弹图像处理技术,来完成彩色图像的处理;也可以直接应用彩色信息来确定条纹级数,进行相关处理。彩色图像能够比灰度图像提供更精确的图像信息,以满足高精度测量的要求。     

Bubble deformation and flow structure measured by double shadow images and PIV/LIF

An experiment on bubble motion in a simple shear layer was performed in order to obtain fundamental knowledge of the force on the bubble and its lateral motion induced by the surrounding flow field. We explored the flow structure in the vicinity of the bubble in one plane and its deformation in two planes by particle image velocimetry (PIV)–laser-induced fluorescence (LIF) and a projection technique for two perpendicular planes, respectively. For our experiment, we chose a single air bubble with an equivalent bubble diameter D _eq of 2~6 mm in a vertical shear flow. Velocity measurements were made using a digital high-speed CCD camera for PIV with fluorescent tracer particles. The second and third CCD cameras were used to detect the bubbles shape and motion via backlighting from an array of infrared LEDs. We quantitatively studied the three-dimensional wake structure from measurements of the two-dimensional vortex structure and approximated three-dimensional shape deformation arranged from two perpendicular bubble images.     

Image correlation velocimetry applied to discrete smoke-wire streaklines in turbulent pipe flow

Two-colour image correlation velocimetry is applied to distinct smoke-wire streaklines in turbulent pipe flow. A single frame from a 35-mm film camera is exposed to two consecutive flashes, one from a red and one from a blue stroboscope. The resulting composite image is digitized, and the color components are extracted to obtain two images. Image correlation is then applied to estimate the velocity which is then compared with hot-wire velocity measurements.     

A color-coded backlighted defocusing digital particle image velocimetry system

Defocusing digital particle image velocimetry (DDPIV), as a true three-dimensional (3D) measurement technique, allows for the measurement of 3D velocities within a volume. Initially designed using a single CCD and 3-pinhole mask (Willert and Gharib in Exp Fluids 12:353–358, 1992), it has evolved into a multi-camera system in order to overcome the limitations of image saturation due to multiple exposures of each particle. In order to still use a single camera and overcome this limitation, we have modified the original single CCD implementation by placing different color filters over each pinhole, thus color-coding each pinhole exposure, and using a 3-CCD color camera for image acquisition. Due to the pinhole mask, there exists the problem of a significant lack of illumination in a conventional lighting setup, which we have solved by backlighting the field-of-view and seeding the flow with black particles. This produces images with a white background superimposed with colored triple exposures of each particle. A color space linear transformation is used to allow for accurate identification of each pinhole exposure when the color filters’ spectrum does not match those of the 3-CCD color camera. Because the imaging is performed with a multi-element lens instead of a single-element lens, an effective pinhole separation, d _e, is defined when using a pinhole mask within a multi-element lens. Calibration results of the system with and without fluid are performed and compared, and a correction of the effective pinhole separation, d _e, due to refraction through multiple surfaces is proposed. Uncertainty analyses are also performed, and the technique is successfully applied to a buoyancy-driven flow, where a 3D velocity field is extracted.     

High-Accuracy 2D Digital Image Correlation Measurements with Bilateral Telecentric Lenses: Error Analysis and Experimental Verification

By comparing two digital images of a test planar specimen surface recorded in different configurations, two-dimensional digital image correlation (2D-DIC) provides full-field displacements to sub-pixel accuracy and full-field strains in the recorded images. For the 2D-DIC systems using an optical lens, a simple pinhole imaging model is commonly used to describe the linear relationship between the measured sensor plane displacements and the actual displacements in the object surface. However, in a practical measurement, various unavoidable disadvantageous factors, such as small out-of-plane motion of the test object surface occurred after loading, small out-of-plane motion of the sensor target due to the self-heating or temperature variation of a camera, and geometric distortion of the imaging lens, may seriously impair or slightly change the originally assumed linear correspondence. In certain cases, these disadvantages may lead to significant errors in displacements and strains measured by 2D-DIC. In this work, the measurement errors of 2D-DIC due to the above three disadvantageous factors are first described in detail. Then, to minimize the errors associated with these disadvantages, a high-accuracy 2D-DIC system using a bilateral telecentric lens is established. The performance of the established 2D-DIC system and other two 2D-DIC systems using a conventional lens and an object-side telecentric lens are investigated experimentally using easy-to-implement stationary, out-of-plane and in-plane rigid body translation tests. A detailed examination reveals that a high-quality bilateral telecentric lens is not only insensitive to out-of-plane motion of the test object and the self-heating of a camera, but also demonstrates negligible lens distortion. Uniaxial tensile tests of an aluminum specimen were also performed to quantitatively compare the axial and transversal strains measured by the proposed 2D-DIC system and those measured by strain gage rosettes. The perfect agreement between the two measurements further verifies the accuracy of the established 2D-DIC system.     

Towards RGB photoelasticity: Full-field automated photoelasticity in white light

In this paper a new full-field method for the automatic analysis of isochromatic fringes in white light is presented. The method, named RGB photoelasticity, eliminates the typical drawbacks of the classical approach to photoelasticity in white light which requires a subjective analysis of colors and an experienced analyst to acquire and interpret the results. The proposed method makes it possible to determine retardations uniquely in the range of 0–3 fringe orders. For this purpose the isochromatics are acquired by means of a color video camera and the colors are decomposed in the three primary colors (red, green and blue) and compared to those stored in a calibration array in the system. Furthermore, the influence of various spurious effects on the accuracy of the proposed method is experimentally evaluated.     

The use of phase-stepping for the measurement of characteristic parameters in integrated photoelasticity

A three-dimensional photoelastic body can be represented by an optically equivalent model, which consists of a linear retarder, δ, at a certain angle, θ, and a pure rotator, χ. These have been described as the characteristic retardation, δ, and the primary and secondary characteristic directions, θ and θ+χ. Until now these characteristic parameters have only been determined using manual, point-by-point collection methods which are involved and time consuming. Therefore an automated phase-stepping method has been developed to enable the determination of the three characteristic parameters for three-dimensional or integrated photoelasticity. Expressions have been derived to obtain δ, θ and θ+χ from six phase-stepped images. These images are collected using a CCD camera and the full-field data is processed using a standard personal computer. This novel method allows accurate, full-field maps of all three characteristic parameters to be obtained in a relatively short time, which makes full-field tomographic reconstruction of photoelastic data a real possibility. Both are SEM Members.     

Simultaneous Observation of Phase-Stepped Images for Photoelasticity Using Diffraction Gratings

Phase-stepped photoelasticity is a powerful method for full-field stress analysis, but sequential collection of the multiple required images limits the technique to static loading applications. We have developed a system that utilizes diffraction gratings to collect four phase-stepped images simultaneously with a single camera for transient loading applications. Two adjacent, perpendicularly oriented, 1D Ronchi rulings are placed after a transparent sample to split the light into equal intensity beams for each diffraction order. The four beams that are diffracted once in the x direction and once in the y direction transmit through arrays of analyzing polariscope elements, with different combinations of fast-axis orientations for four phase-stepped images. The mirrors and imaging lenses in the system work in concert to focus each beam onto separate quadrants of the same CCD. We demonstrate the system for stress analysis of compressive loading of a Homalite-100 disk and of a Homalite-100 plate with a central hole. This system has the potential for photoelastic analysis of time-dependent materials and of dynamic events, when equipped with a high-speed camera.     

Two-color digital PIV employing a single CCD camera

 An extension of two color particle image velocimetry (PIV) is described where the color images are recorded onto a single high-resolution (3060×2036 pixel) color CCD sensor. Unlike mono-color CCD sensors, this system not only eliminates the processing time and the subsequent digitization time of film-based PIV but also resolves the directional ambiguity of the velocity vector without using conventional image-shifting techniques. For comparing the spatial resolutions of film and CCD data, a calibration experiment is conducted by recording the speckle pattern onto 35 mm color film and using a CCD sensor under identical conditions. This technique has been successfully implemented for simulated turbine film-cooling flows in order to obtain a more detailed characterization of the coolant-injection phenomenon and its interaction with freestream disturbances. Received: 20 November 1996/Accepted: 29 January 1998     

Influence of Camera Rotation on Stereo-DIC and Compensation Methods

Camera rotation during stereo digital image correlation (DIC) measurements is always present in some amount and is caused by environmental vibrations during testing (e.g. blast testing, testing in industrial environments, vibration testing Helffrick et al., Mech Syst Signal Process 25(3):917–927, 2011, etc.). The impact of changing extrinsic parameters on stereo-DIC measurements was investigated in this paper by simulating several DIC experiments with a shaking camera system. To obtain an appropriate order of magnitude of the camera motion, the actual camera rotation during a test was observed and reported. The impact hereof was investigated by analyzing static images and by simulating camera rotation with a stereo-DIC simulator presented in Balcaen et al. (Exp Mech 57(5):703–718, 2017). Insight on this underestimated error source is explained by investigating which rotations are more critical to the resulting displacements and strains. Two possible compensation methods were investigated and their capability of compensating camera motion was evaluated.     

Real-time analysis of isochromatics and isoclinics using the phase-shifting method

This paper studies a new real-time phase-shifting method for the analysis of isochromatic and isoclinic parameters in photoelasticity. By rotating an analyzer at a constant rate and an output quarter-wave plate at a double rate of the analyzer and recording images by a CCD camera continuously, sequential images which brightness is integrated by sensors in a CCD camera during phase-shifting are obtained. Then, the distributions of the isochromatic and isoclinic parameters are obtained immediately and quantitatively using the proposed phase-shifting algorithm. The proposed method can be applied to high-speed inspection of optical elements or glass products. Also, it is expected that slowly varying time-dependent problems can be analyzed by the proposed method.     

Research on synchronous measurement technique of temperature and deformation fields using multispectral camera with bilateral telecentric lens

The hot-section parts easily occur the creep-fatigued interaction under the condition of mechanical-thermal coupled load during the period of service, which may lead to the damage of the parts, and therefore, the measurement and characterization of thermal-deformed fields of the parts are important to understand its damage process. Aiming at relevant demand, the bilateral telecentric-multispectral imaging system was established, the research of synchronous measurement technique of the temperature and deformation fields was developed. On the one hand, the measurement technology for surface temperature of the object was developed using the two-color images captured by the multispectral camera with bilateral telecentric lens and combined with colorimetric method. On the other hand, the 2D-DIC measurement technique of the multispectral camera was developed by conducting digital image correlation analysis using the blue light images before and after deformation, which can measure the high temperature deformation field of the object (the blue light images were filtered by multispectral camera). Results showed that the bilateral telecentric lens is used to replace the ordinary optical lens for imaging, which can effectively eliminate the distortion of the multispectral imaging system. Since the temperature measurement process of this measurement system is little affected by the emissivity of the object, therefore, it has excellent robustness. The thermal expansion coefficients of the nickel alloys are evaluated at the temperature ranges of 700–1000 ℃, indicating this system can achieve the synchronous and precise measurement of the temperature and deformation fields of the object.     

Stereoscopic imaging and reconstruction of the 3D geometry of flame surfaces

Stereoscopic imaging and reconstruction of turbulent impinging diffusion flames have been demonstrated. The 3D geometry of the flame surface is reconstructed from any pair of stereo images through digital image processing and computer vision. Only one camera is required to capture the pair of stereo images due to the use of a stereo adapter. As a result the pair of flame images is captured on a single CCD chip. They are then post-processed to produce a 3D geometry of the flame. The 3D results and surface area calculations demonstrate the potential of stereo visualisation, data extraction and the validity of the proposed methodology.     

载荷作用下泡沫铝的波速与变形关系的实验研究

对单轴压缩下两种密度的泡沫铝进行了三种频率(50kHz、200kHz和300kHz)的声波测试,同时运用CCD对实验进行跟踪拍摄。分析结果表明:P波波速随载荷增大而增加,当载荷达到屈服力时,波速有所下降;S波有相同趋势,但变化比P波小。应用数字图像相关方法对CCD拍摄图片进行处理,得到不同载荷阶段下泡沫铝的全场应变;采用Weibull函数对全场应变分布进行拟合,研究了Weibull分布参数随载荷变化的规律。由此,初步建立了载荷作用下材料结构变化与波速的关系。此研究对于声波探测领域有很好的指导意义。     

PTV investigation of phase interaction in dispersed liquid–liquid two-phase turbulent swirling flow

An investigation of dispersed liquid–liquid two-phase turbulent swirling flow in a horizontal pipe is conducted using a particle tracking velocimetry (PTV) technique and a shadow image technique (SIT). Silicone oil with a low specific gravity is used as immiscible droplets. A swirling motion is given to the main flow by an impeller installed in the pipe. Fluorescent tracer particles are applied to flow visualization. Red/green/blue components extracted from color images taken with a digital color CCD camera are used to simultaneously estimate the liquid and droplet velocity vectors. Under a relatively low swirl motion, a large number of droplets with low specific gravity tend to accumulate in the central region of the pipe. With increasing droplet volume fraction, the liquid turbulence intensity in the axial direction increases while that in the wall-normal direction decreases in the central region of the pipe. In addition, the turbulence modification in the present flow is strongly dependent on the droplet Reynolds number; however, the interaction of droplet-induced turbulences is significant due to vortex shedding, particularly at high droplet Reynolds numbers and higher droplet volume fraction.