Image deconvolution techniques for motion blur compensation in DIC measurements

In the last two decades, the interest in digital image correlation (DIC) has grown steadily in various industrial and scientific fields, thanks to its contactless nature, capability to provide full field measurements, and ease of implementation. The possibility to carry out measurements in the presence of significant relative motion between the camera and the target has been thoroughly investigated. In dynamic conditions, the main additional source of uncertainty is represented by the motion blur effect, which is experienced whenever the relative displacement between the camera and the target during the exposure time is not negligible. Motion blur is a source of uncertainty, since it decreases the contrast in the image and the definition of the speckles in the acquired pattern. In this work, a complete procedure for motion blur estimation and compensation is considered. It was found that the uncertainty in the estimation of the blur intensity and orientation plays a crucial role in motion blur compensation. In this work, we propose a technique for the optimization of the blur estimation and compensation phases based on fitting the motion optical transfer function to a motion model. Compensation with a Wiener filter is then performed, studying different approaches to image noise estimation. Results show that the optimized procedure has a positive effect on DIC performances, being able to reduce the effect of motion blur on uncertainty, in particular for motion blur levels larger than about 1–2 px. The entire work has been validated by considering the procedure performances on DIC analysis performed on different kinds of speckle image with various amounts of motion blur applied.

The use of 2D ultrasound elastography for measuring tendon motion and strain

Analysis of mechanical behaviour and fracture interaction of multi-hole rock mass with DIC measurement

Robotic structures are typically fabricated using discrete components that serve only a single function, such as carbon fibers or polymer composites for structural reinforcement. The ability to integrate these discrete components into a single structure provide enormous opportunities to enhance the performance of robots by reducing weight and increasing the energy storage/harvesting. We have recently been pursuing the integration of sensors and solar cells into multifunctional “skin” structures. One issue that limits the integration of these components is the effect that that the interaction between them will have on the mechanical behavior, and its subsequent impact on multifunctional performance. To minimize these effects and better understand the mechanics of multifunctional skin structures, we first pursued compliant strain sensors integrated onto the compliant wings of a flapping wing MAV to sense deformations at the wing in real time. These measurements were correlated to the thrust force of the wing. The effects of these compliant strain sensors were also determined by comparing the new thrust measurements with the original wings and with 3D Digital Image Correlation (DIC) measurements. It was determined by SRS analysis that strain sensors tended to be more sensitive to lower integral modes of the flapping frequency than the thrust measurement data, while the DIC measurements correlated more directly with thrust measurements in the time domain. Thus, it would appear that the strain sensor and DIC measurements are sensitive to different aspects of the mechanics of the multifunctional skin structures that can be taken into account when using the strain sensors for real-time sensing and control. In addition to compliant strain sensor integration, flexible Solar Cells (SCs) were also integrated onto the compliant wings. Their ability to harvest energy during flight has the potential to prolong the time of flight for more autonomous operation. The effects of the integrated SCs on the 3D shapes of the compliant wings were also characterized during flapping. The solar cells had a significant effect on the wing shape, increasing the stiffness of the wings and reducing the volume of air that the wings need to capture to generate thrust, particularly at the apex and nadir of the flapping cycle where thrust is generated by blowback. Thus, a multifunctional analysis was developed to account for the tradeoff between the extra power require to generate the necessary thrust for the MAV to stay airborne versus what is gained from integrating the SCs in order to determine how to design the multifunctional skin structures to optimize performance.

Welding is one of the most widely used joining techniques for manufacturing complex engineering components. However, due to transient temperature cycles in the adjacent materials being joined, the local microstructure is altered and plastic strain in the surrounding materials is introduced resulting in a variation of mechanical properties across the weldment. When an extracted cross-weld test specimen is subjected to an external uniaxial load, there is interaction between the various constituents of the weld (such as the weld fused zone, heat affected zone and parent metal) resulting in development of constraint in the weaker material due to the adjacent stronger material. This affects the distribution of stress along and within the test specimen. The level of constraint developed is dependent on the strength mismatch between adjacent materials which may be graded across the heat affected zone and the specimen cross-sectional geometry. With development of digital image correlation (DIC) and its application to measuring deformation of solids, it has become feasible to measure local stress-strain properties of cross-weld samples and also specimens of variable geometry at both room and high temperature. Local deformation measured by DIC is correlated with nominal stress based on an iso-stress assumption, where it is assumed that the various regions of the weld are arranged in series and that the stress is uniform over the cross-section area at any given location in the specimen. Thus, local mechanical properties are inferred by mapping the measured local strain against the global stress. In this context the effect of constraint is important because it introduces a local stress distribution that will differ from the global net section stress and can therefore introduce errors in determination of local stress-strain properties. The influence of these factors raises questions concerning the interpretation of apparent local mechanical properties of cross-welds measured using DIC. Specifically, there is uncertainty as to whether the measured variation in local strain versus average cross-sectional stress (average global stress) represents the local mechanical properties of the material with negligible error. The severity of potential constraint errors will be influenced by the specimen geometry (particularly the tensile specimen thickness to width ratio for 2D DIC strain measurements) and the gradient of material property inhomogeneity along the specimen gauge length. In the current study, measurement of deformation in varying stress fields using DIC was carried out. Two distinct methods of introducing varying stress fields were employed: in the first case, the stress field was varied by designing an hour-glass shaped specimen with the specimen width varying as a function of the gauge length. In the second case, stress fields were varied due to a change in the specimen thickness in mismatched dissimilar metal joints. To reduce the complexity in the developed stress fields in the region near the interface, two materials were joined using solid state diffusion bonding giving a step change in the material properties across a planar interface. Test specimens spanning the interface were studied by finite element modelling and experimental tests with DIC deformation monitoring. Using DIC and a single hour-glass shaped specimen, multiple creep curves at different stress levels along the specimen gauge length have been successfully extracted for a test temperature of 525oC. The creep data obtained at 525oC from the hour-glass specimen test have been fitted to RCC-MR, Garofalo and Graham-Walles creep deformation models. A better correlation is achieved between the experimental data and the RCC-MR model predicted values when the local deformations from the single specimen are plotted as a function of time and true stress at the end of the specimen’s initial loading. From both elastic-plastic FE models and experimental local stress-strain curves extracted from DIC measurements, it can be concluded that a thin (1mm thick or less) 6mm wide test specimen should be used to extract the local mechanical properties from tensile specimens that have varying material properties along the gauge length, such as found in cross-weld test specimens. This is because the thinner specimen reduces the size of the strain concentration region adjacent to the interface where errors in measured local mechanical properties can occur. As a ‘rule of thumb’, the intrinsic stress-strain properties of the material should not be inferred from regions within 1mm of the interface for these 1mm thick samples because of concentration errors. The results obtained demonstrate the capability of DIC as a measurement technique for characterising the local deformation of specimens with varying stress fields. The findings also highlight some aspects of the technique which need to be considered when interpreting data from DIC measurements of local deformation from specimens with varying stress fields.

Zinc alloys are used in a wide range of application such as electronics, automotive and building construction. Their various shapes are generally obtained by metal forming operation such as stamping. Therefore, it is important to characterize the material with adequate characterization tests. Sheet Bulging Test (SBT) is well recognized in the metal forming community. Different theoretical models of the literature for the evaluation of thickness and radius of the deformed sheet in SBT have been studied in order to get the hardening curve of different materials. These theoretical models present the advantage that the experimental procedure is very simple. But Koç et al. showed their limitation, since the combination of thickness and radius evaluations depend on the material. As Zinc alloys are strongly anisotropic with a special crystalline structure, a procedure is adopted for characterizing the hardening curve of a Zinc alloy. The anisotropy is first studied with tensile test, and SBT with elliptical dies is also investigated. Parallel to this, Digital Image Correlation (DIC) measures are carried out. The results obtained from theoretical models and DIC measures are compared. Measures done on post-mortem specimens complete the comparisons. Finally, DIC measures give better results and the resulting hardening curve of the studied zinc alloy is provided.

As a representative method of optical non-interference measurement, digital image correlation (DIC) technology is a non-contact optical mechanics method that can measure the displacement and deformation of the whole field. However, when the measurement range of the field is too large, the existing DIC method cannot measure the full-field strain accurately, which limits the application of the DIC measurement in the case of a large size and wide-field view. To address this issue, a DIC measurement method for large-scale structures based on adaptive warping image stitching is proposed in this paper. First, multiple adjacent high-resolution images are collected at different locations of large-scale structures. Secondly, the collected images are stitched by applying the adaptive warping image stitching algorithm to obtain a panoramic image. Finally, the DIC algorithm is applied to solve the whole deformation field. In the experiments, we first verify the feasibility of the proposed method for image matching and fusion through the numerical simulation of a rigid body translation experiment. Then the accuracy and robustness of the proposed method in practical application are verified by rigid body translation and a three-point bending experiment. The experimental results demonstrate that the measurement range of DIC is improved significantly with the adaptive warping image stitching algorithm.

Railroad bridges in North America are an integral but aging part of the railroad network and are typically only monitored using visual inspections. When quantitative information is required for assessment, railroads often monitor bridges using accelerometers. However without a sensor to directly measure displacements, it is difficult to interpret these results as they relate to bridge performance. Digital Image Correlation (DIC) is a non-contact sensor technology capable of directly measuring the displacement of any visible bridge component. In this research, a railroad bridge was monitored under load using DIC and accelerometers. DIC measurements are directly compared to serviceability limits and it is observed that the bridge is compliant. The accelerometer data is also used to calculate displacements which are compared to the DIC measurements to assess the accuracy of the accelerometer measurements. These measurements compared well for zero-mean lateral data, providing measurement redundancy and validation. The lateral displacements from both the accelerometers and DIC at the supports were then used to determine the source of lateral displacements within the support system.

To study the influence of glass spacing on digital image correlation measurement in the optical path, an experimental method of displacement stepping for digital image correlation measurement was designed under nine working conditions (without glass, single glass, and double glass with different spacings of 20 mm, 70 mm, 120 mm, 170 mm, 220 mm, 270 mm, 320 mm respectively). The results of digital image correlation method and dial gauge method were compared, and the measurement error and standard deviation of digital image correlation measurement under different working conditions were analyzed. The experimental results show that the variation trend of displacement measured by DIC under nine working conditions is consistent with the step change of the specimen displacement, but there exist obvious errors between the nine conditions. Based on the results of dial gauge measurement, DIC measurement results are larger, and the maximum deviation is less than 3 μm. Under the condition of without glass, the relative error of displacement measured by DIC was the smallest, with an average error of 2.92%. Under the condition of double glass spacing of 170 mm, the relative error was the largest, with an average error of 11.37%. The standard deviation of double glass condition with a certain spacing was less than that of single glass condition. In general, when the double glass spacing is within a certain range, the measurement results are more accurate and less discrete than those of single glass, and the thermal protection performance is also better.

Applying a homogeneous pressure distribution to the upper vertebral endplate: Validation of a new loading system, pilot application to human vertebral bodies, and finite element predictions of DIC measured displacements and strains

Elucidating failure mechanisms in human femurs during a fall to the side using bilateral digital image correlation

An experimental investigation was conducted to evaluate the permanent strain characteristics of a granular pavement foundation layer using Repeated Load Triaxial (RLT) tests, assisted by a non-contact strain measurement system based on a Digital Image Correlation (DIC) technique. Permanent strains are generally determined from laboratory triaxial tests using global measurement of axial strains. Under cyclic loading there may be formations of localized strains and instabilities, which may not be captured by conventional measuring methods, but can have significant influence on the performance of the constructed granular layer. Results are reported from a series of 40 consolidated drained triaxial tests, to develop a better understanding of the permanent strain response of a granular material under simulated traffic loadings. The DIC technology was employed to record deformation measurements, which included acquiring high-resolution images of the specimen at every 100 cycles of repeated deviatoric stresses, using a Q-Imaging / QICAM 1394 camera, and VIC-2D image correlation software. Results indicate that there is a reasonable match between the LVDT and DIC measurements when the entire specimen length was used as the gage length. However, localized strains at the top, middle, and bottom of the specimens computed from DIC measurements deviated from the LVDT measurements. This factor should be taken into account during the pavement design phase, or for the development of pavement rutting performance models.

We present a sensor-fusion method that exploits a depth camera and a gyroscope to track the articulation of a hand in the presence of excessive motion blur. In case of slow and smooth hand motions, the existing methods estimate the hand pose fairly accurately and robustly, despite challenges due to the high dimensionality of the problem, self-occlusions, uniform appearance of hand parts, etc. However, the accuracy of hand pose estimation drops considerably for fast-moving hands because the depth image is severely distorted due to motion blur. Moreover, when hands move fast, the actual hand pose is far from the one estimated in the previous frame, therefore the assumption of temporal continuity on which tracking methods rely, is not valid. In this paper, we track fast-moving hands with the combination of a gyroscope and a depth camera. As a first step, we calibrate a depth camera and a gyroscope attached to a hand so as to identify their time and pose offsets. Following that, we fuse the rotation information of the calibrated gyroscope with model-based hierarchical particle filter tracking. A series of quantitative and qualitative experiments demonstrate that the proposed method performs more accurately and robustly in the presence of motion blur, when compared to state of the art algorithms, especially in the case of very fast hand rotations.

We present a method that can automatically determine the optimal camera exposure time for high-quality deformation measurement with digital image correlation (DIC) techniques. The proposed method needs to capture a series of surface images of a test sample at its reference state with different camera exposure times. The relationship between the mean intensity gradients (MIGs) and average grayscales of these images reveals that the best quality (i.e. maximum MIG) of a speckled sample surface always corresponds to a certain average grayscale. Thus, the proposed method can serve two purposes in DIC practice. First, at the initial state, the camera exposure time can be adjusted automatically to obtain a reference image with the best speckle pattern quality. Second, by adjusting the camera exposure time to make the average grayscale of an image close to the predetermined optimal value, the proposed method can adaptively output high-quality deformed images with an almost constant speckle pattern quality, regardless of serious ambient light variations. Experimental results demonstrated that the proposed method can automatically obtain stable and high-quality speckle pattern images, thus delivering better DIC measurement compared with regular DIC techniques using a fixed camera exposure time. Because the present automatic camera exposure time control method allows a nonprofessional operator to consistently obtain high-quality speckle pattern images that warrant high-accuracy DIC measurements, it is therefore suggested that the present method should be used as a routine practice in practical DIC applications.

Digital Image Correlation (DIC) provides measurements without disturbing the specimen, which is a major advantage over contact methods. Additionally, DIC techniques provide full-field maps of response quantities like strains and displacements, unlike traditional methods that are limited to a local investigation. In this work, an experimental application of DIC is presented to investigate a problem of relevant interest in the civil engineering field, namely the interface behavior between externally bonded fabric reinforced cementitious mortar (FRCM) sheets and concrete substrate. This represents a widespread strengthening technique of existing reinforced concrete structures, but its effectiveness is strongly related to the bond behavior between composite fabric and underlying concrete. To investigate this phenomenon, a set of notched concrete beams are realized, reinforced with FRCM sheets on the bottom face, subsequently cured in different environmental conditions (humidity and temperature) and finally tested up to failure under three-point bending. Mechanical tests are carried out vis-à-vis DIC measurements using two distinct cameras simultaneously, one focused on the concrete front face and another focused on the FRCM-concrete interface. This experimental setup makes it possible to interpret the mechanical behavior and failure mode of the specimens not only from a traditional macroscopic viewpoint but also under a local perspective concerning the evolution of the strain distribution at the FRCM-concrete interface obtained by DIC in the pre- and postcracking phase.