Particle Image Velocimetry Algorithm Research Articles

SUAS-Based Remote Sensing of River Discharge Using Thermal Particle Image Velocimetry and Bathymetric Lidar

This paper describes a non-contact methodology for computing river discharge based on data collected from small Unmanned Aerial Systems (sUAS). The approach is complete in that both surface velocity and channel geometry are measured directly under field conditions. The technique does not require introducing artificial tracer particles for computing surface velocity, nor does it rely upon the presence of naturally occurring floating material. Moreover, no prior knowledge of river bathymetry is necessary. Due to the weight of the sensors and limited payload capacities of the commercially available sUAS used in the study, two sUAS were required. The first sUAS included mid-wave thermal infrared and visible cameras. For the field evaluation described herein, a thermal image time series was acquired and a particle image velocimetry (PIV) algorithm used to track the motion of structures expressed at the water surface as small differences in temperature. The ability to detect these thermal features was significant because the water surface lacked floating material (e.g., foam, debris) that could have been detected with a visible camera and used to perform conventional Large-Scale Particle Image Velocimetry (LSPIV). The second sUAS was devoted to measuring bathymetry with a novel scanning polarizing lidar. We collected field measurements along two channel transects to assess the accuracy of the remotely sensed velocities, depths, and discharges. Thermal PIV provided velocities that agreed closely ( R 2 = 0.82 and 0.64) with in situ velocity measurements from an acoustic Doppler current profiler (ADCP). Depths inferred from the lidar closely matched those surveyed by wading in the shallower of the two cross sections ( R 2 = 0.95), but the agreement was not as strong for the transect with greater depths ( R 2 = 0.61). Incremental discharges computed with the remotely sensed velocities and depths were greater than corresponding ADCP measurements by 22% at the first cross section and <1% at the second.

Particle Image Velocimetry of Oil-Water Two-Phase Flow with High Water Cut and Low Flow Velocity in a Horizontal Small-Diameter Pipe.

Velocity and flow field are both parameters to measure flow characteristics, which can help determine the logging location and response time of logging instruments. Particle image velocimetry (PIV) is an intuitive velocity measurement method. However, due to the limitations of image acquisition equipment and the flow pipe environment, the velocity of a horizontal small-diameter pipe with high water cut and low flow velocity based on PIV has measurement errors in excess of 20%. To solve this problem, this paper expands one-dimensional displacement sub-pixel fitting to two dimensions and improves the PIV algorithm by Kriging interpolation. The improved algorithm is used to correct the blank and error vectors. The simulation shows that the number of blank and error vectors is reduced, and the flow field curves are smooth and closer to the actual flow field. The experiment shows that the improved algorithm has a maximum measurement error of 5.9%, which is much lower than that of PIV, and that it also has high stability and a repeatability of 3.14%. The improved algorithm can compensate for the local missing flow field and reduce the requirements related to the measurement equipment and environment. The findings of this study can be helpful for the interpretation of well logging data and the design of well logging instruments.

Graphical Processing Unit-Accelerated Open-Source Particle Image Velocimetry Software for High Performance Computing Systems

Particle image velocimetry (PIV) data processing time can constrain data set size and limit the types of statistical analyses performed. General purpose graphics processing unit (GPGPU) computing can accelerate PIV data processing allowing for larger datasets and accompanying higher order statistical analyses. However, this has not been widespread likely due to limited accessibility to the GPU-PIV hardware and software. Most GPU-PIV software is platform dependent and proprietary, which restricts the computing systems that can be used and makes the details of the algorithm unknown. This work highlights the development of an open-source, cross-platform, GPU-accelerated, PIV algorithm. Validation of the algorithm is done using both synthetic and experimental images. The algorithm was found to accurately resolve the time-averaged flow, instantaneous velocity fluctuations, and vortices. All data processing was done on a GPU supercomputing cluster and notably outperformed the central processing unit version of the software by a factor of 175. The algorithm is freely available and included in the OpenPIV distribution.

TracTrac: A fast multi-object tracking algorithm for motion estimation

TracTrac is an open-source Matlab/Python implementation of a robust and efficient object tracking algorithm capable of simultaneously tracking several thousands of objects in very short time. Originally developed as an alternative to particle image velocimetry algorithms for estimating fluid flow velocities, its versatility and robustness makes it relevant to many other dynamic sceneries encountered in geophysics such as granular flows and drone videography. In this article, the structure of the algorithm is detailed and its capacity to resolve strongly variable and intermittent object motions is tested against three examples of geophysical interest.

Two-dimensional dam-break wave analysis: particle image velocimetry versus optical flow

Dam-break waves have been the focus of a number of research papers; next to analytical investigations, scaled experimental models and numerical simulations are used to determine dam-break wave propagation. Experimental models and data analysis tools are necessary to generate calibration data for numerical models. Often, particle image velocimetry (PIV) algorithms are used to calculate two-dimensional velocity fields of propagating dam-break waves. This paper deals with a visual analysis method called optical flow (OF). An experimental model was built to identify the initial stage of dam-break waves. A comparison between results using the PIV method and the OF method is given to identify flow velocities for two-dimensional dam-break waves. It can be shown that the OF method is able to predict resulting velocity profiles accurately. The mean relative error for the comparison of manually observed data and OF results is less than 6%, while for the PIV comparison an error of 11% can be calculated. Since velocity vectors are produced for all image pixels, the resulting information density is high compared to results produced by the PIV method.

Measuring the velocity fields of granular flows – Employment of a multi-pass two-dimensional particle image velocimetry (2D-PIV) approach

Measuring velocity fields plays a crucial role in investigating the dynamics of granular flows, which can improve the modeling of hazardous geophysical flows (e.g. avalanches and debris flows) and the control of powder flows in industrial applications. Non-invasive optical methods are invaluable tools for estimating this physical quantity at the laboratory scale. Despite the recent improvements of particle image velocimetry (PIV) algorithms, the employment of PIV to granular flows is still a non-trivial application, where there are several specific aspects to be carefully addressed. Here, we address the main challenges of granular PIV applications and systematically test the open-source window deformation multi-pass code, PIVlab [Thielicke and Stamhuis, J. Open Res. Soft., 2014], for dry granular flows in rotating drum and chute flow experiments. Three granular media (glass spheres, Ottawa sand and acetalic resin beads) with different optical properties are used as a broad test bench for validating the PIV approach. As well, comparisons between the estimations by PIVlab and those obtained by the commercial code, IDT ProVision-XS, are reported, where the advantages of the multi-pass approach are highlighted. This extensive experimental investigation allowed the evaluation of the accuracy of PIVlab in granular flow applications and also helped to assess the reliability of measurements of second-order statistics, such as the granular temperature. Finally, a guideline for setting a reliable PIV arrangement is suggested.

Optimized Time-Resolved Echo Particle Image Velocimetry- Particle Tracking Velocimetry Measurements Elucidate Blood Flow in Patients With Left Ventricular Thrombus.

Contrast ultrasound is a widely used clinical tool to obtain real-time qualitative blood flow assessments in the heart, liver, etc. Echocardiographic particle image velocimetry (echo-PIV) is a technique for obtaining quantitative velocity maps from contrast ultrasound images. However, unlike optical particle image velocimetry (PIV), routine echo images are prone to nonuniform spatiotemporal variations in tracer distribution, making analysis difficult for standard PIV algorithms. This study introduces optimized procedures that integrate image enhancement, PIV, and particle tracking velocimetry (PTV) to obtain reliable time-resolved two-dimensional (2D) velocity distributions. During initial PIV analysis, multiple results are obtained by varying processing parameters. Optimization involving outlier removal and smoothing is used to select the correct vector. These results are used in a multiparameter PTV procedure. To demonstrate their clinical value, the procedures are implemented to obtain velocity and vorticity distributions over multiple cardiac cycles using images acquired from four left ventricular thrombus (LVT) patients. Phase-averaged data elucidate flow structure evolution over the cycle and are used to calculate penetration depth and strength of left ventricular (LV) vortices, as well as apical velocity induced by them. The present data are consistent with previous time-averaged results for the minimum vortex penetration depth associated with LVT occurrence. However, due to decay and fragmentation of LV vortices, as they migrate away from the mitral annulus, in two cases with high penetration, there is still poor washing near the resolved clot throughout the cycle. Hence, direct examination of entire flow evolution may be useful for assessing risk of LVT relapse before prescribing anticoagulants.

A PTV method based on ultrasound imaging and feature tracking in a low-concentration sediment-laden flow

This study aims to provide a particle tracking velocimetry (PTV) method based on ultrasound imaging and feature-tracking in a low-concentration sediment-laden flow. A phased array probe is used to generate a 2D ultrasound image at different times. Then, the feature points are extracted to be tracked instead of the centroids of the particle image. In order to better identify the corresponding feature point, each feature is described by an oriented angle and its location. Then, a statistical interpolation procedure is used to yield the displacement vector on the desired grid point. Finally a correction procedure is adopted because the ultrasound image is sequentially acquired line by line through the field of view. A simple test experiment was carried out to evaluate the performance. The ultrasound PTV system was applied to a sediment-laden flow with a low concentration of 1‰, and the speed was up to 10 cm s−1. In comparison to optical particle image velocimetry (PIV), ultrasound imaging does not have a limitation in optical access. The feature-tracking method does not have a binarisation and segmentation procedure, which can result in overlapping particles or a serious loss of particle data. The feature-tracking algorithm improves the peak locking effect and measurement accuracy. Thus, the ultrasound PTV algorithm is a feasible alternative and is significantly more robust against gradients than the correlation-based PIV algorithms in a low-concentration sediment-laden fluid.

A New Feature Based Algorithm for Image Analysis of Deformable Materials for Laboratory Investigations of Slope Stability

Physical modelling of slope stability scenarios can provide new insights into failure mechanisms as well as assistance with interpretation of numerical modelling investigations. To increase the value of such experiments, algorithms that support rapid analysis and quantification of the slope deformation occurring in the experiment are needed. Feature based image analysis has advantages in this respect over area or patch based approaches but suffers from robustness issues. To this end, a new image processing algorithm for measurement of deformation of granular media in laboratory experiments is presented. Our novel algorithm combines a feature detector with model based constraints and outlier detection to achieve fast and robust particle tracking. Comparison with a high precision particle image velocimetry algorithm shows excellent results with much improved processing times. Application of the algorithm for analysis of a laboratory simulation of slope stability is demonstrated and comparison with numerical modelling confirms the algorithm’s flexibility and robustness.

An optical flow algorithm based on gradient constancy assumption for PIV image processing

Particle image velocimetry (PIV) has matured as a flow measurement technique. It enables the description of the instantaneous velocity field of the flow by analyzing the particle motion obtained from digitally recorded images. Correlation based PIV evaluation technique is widely used because of its good accuracy and robustness. Although very successful, correlation PIV technique has some weakness which can be avoided by optical flow based PIV algorithms. At present, most of the optical flow methods applied to PIV are based on brightness constancy assumption. However, some factors of flow imaging technology and the nature property of the fluids make the brightness constancy assumption less appropriate in real PIV cases. In this paper, an implementation of a 2D optical flow algorithm (GCOF) based on gradient constancy assumption is introduced. The proposed GCOF assumes the edges of the illuminated PIV particles are constant during motion. It comprises two terms: a combined local-global gradient data term and a first-order divergence and vorticity smooth term. The approach can provide accurate dense motion fields. The approach are tested on synthetic images and on two experimental flows. The comparison of GCOF with other optical flow algorithms indicates the proposed method is more accurate especially in conditions of illumination variation. The comparison of GCOF with correlation PIV technique shows that the proposed GCOF has advantages on preserving small divergence and vorticity structures of the motion field and getting less outliers. As a consequence, the GCOF acquire a more accurate and better topological description of the turbulent flow.

Finding the location of the instantaneous center of rotation using a particle image velocimetry algorithm

This work is about planar rigid-body kinematics and, in particular, the principle of the instantaneous center of rotation (IC). Using a computer simulated approach, a workflow is presented that results in a visual representation of the locus of the IC, based on particle image velocimetry (PIV). Here, a small number of digital animations of textured objects are created with multibody dynamics software, and later imported in PIV software to extract the velocity field (magnitude and direction) of objects moving within a plane. We believe the workflow presented may help learners improve their understanding of the concept of the IC, thus enhancing their knowledge of rigid body kinematics.

High-resolution traction force microscopy on small focal adhesions - improved accuracy through optimal marker distribution and optical flow tracking

The accurate determination of cellular forces using Traction Force Microscopy at increasingly small focal attachments to the extracellular environment presents an important yet substantial technical challenge. In these measurements, uncertainty regarding accuracy is prominent since experimental calibration frameworks at this size scale are fraught with errors – denying a gold standard against which accuracy of TFM methods can be judged. Therefore, we have developed a simulation platform for generating synthetic traction images that can be used as a benchmark to quantify the influence of critical experimental parameters and the associated errors. Using this approach, we show that TFM accuracy can be improved >35% compared to the standard approach by placing fluorescent beads as densely and closely as possible to the site of applied traction. Moreover, we use the platform to test tracking algorithms based on optical flow that measure deformation directly at the beads and show that these can dramatically outperform classical particle image velocimetry algorithms in terms of noise sensitivity and error. We then report how optimized experimental and numerical strategy can improve traction map accuracy, and further provide the best available benchmark to date for defining practical limits to TFM accuracy as a function of focal adhesion size.

Peak-locking reduction for particle image velocimetry

A parametric study of the factors contributing to peak-locking, a known bias error source in particle image velocimetry (PIV), is conducted using synthetic data that are processed with a state-of-the-art PIV algorithm. The investigated parameters include: particle image diameter, image interpolation techniques, the effect of asymmetric versus symmetric window deformation, number of passes and the interrogation window size. Some of these parameters are found to have a profound effect on the magnitude of the peak-locking error. The effects for specific PIV cameras are also studied experimentally using a precision turntable to generate a known rotating velocity field. Image time series recorded using this experiment show a linear range of pixel and sub-pixel shifts ranging from 0 to ±4 pixels. Deviations in the constant vorticity field (ωz) reveal how peak-locking can be affected systematically both by varying parameters of the detection system such as the focal distance and f-number, and also by varying the settings of the PIV analysis.A new a priori technique for reducing the bias errors associated with peak-locking in PIV is introduced using an optical diffuser to avoid undersampled particle images during the recording of the raw images. This technique is evaluated against other a priori approaches using experimental data and is shown to perform favorably. Finally, a new a posteriori anti peak-locking filter (APLF) is developed and investigated, which shows promising results for both synthetic data and real measurements for very small particle image sizes.

Adaptive PIV algorithm based on seeding density and velocity information

An adaptive particle image velocimetry (PIV) processing algorithm to increase local spatial resolution and obtain accurate results is presented. Adaptive sampling criterion is based on seeding density and the analysis of velocity information. The proposed methodology places more measurement points in the region where high seeding density exists or flow parameters vary drastically. This methodology is effective in PIV image processing, especially in non-optimal PIV experimental conditions. The working principle of the adaptive processing method in this paper is to generate spring force on the bias of the deviation between the ideal and actual distances of measurement points. Under artificial force, the sampling points are moved to desirable places to satisfy sampling density function. Sampling density function is important in processing and ingeniously conceived based on the combination of vorticity and velocity gradients. The viability of the method is elaborated through synthetic and experimental tests. The synthetic test verifies the effectiveness of the adaptive processing on the basis of flow information. The bias and random error can be reduced by 7.3% and 4.0%, respectively. The result of the experimental test shows that the sampling points and the interrogation window size can be arranged according to the spatial distribution of seeding density and the actual flow features. Compared with iterative window deformation method, spatial resolution is locally enhanced, and the robustness and reliability of the result are increased, particularly in poor seeding regions.

Porosity and geometry control ductile to brittle deformation in indentation of porous solids

We report a series of plane strain wedge indentation experiments on a model porous brittle solid (soft rock) and study the effect of porosity and the indenter geometry on the mechanics of indentation. A series of images of the region surrounding the indenter were captured during indentation and were analysed using a particle image velocimetry algorithm. From the image analyses, deformation parameters were analysed as functions of porosity, including velocity fields and localization features such as compaction bands and fractures. The initiation of fracture during indentation was observed in the transition from brittle to ductile behaviour as a function of indentation depth. Higher ductility of the rock during indentation was observed for samples of high porosity and for indentation utilizing sharper wedges. The implications of these results for controlling deformation and fracture in rock drilling by tool geometry and rock porosity control are briefly discussed.

Numerical and experimental comparison of 3D Particle Tracking Velocimetry (PTV) and Particle Image Velocimetry (PIV) accuracy for indoor airflow study

Experimental measurement still plays an important role in indoor airflow study. To obtain three-dimensional and high-quality experimental data in building's indoor airflow study, Particle Tracking Velocimetry (PTV) and Particle Image Velocimetry (PIV) have been increasingly used. The accuracy and precision of measurement technologies is always a crucial issue. This paper first numerically compares the measuring performance of 3D PTV and typical 2D PIV algorithms on three laminar macro scale flows of known displacement, as a function of the particle tracking density defined as the ratio of mean particle spacing to mean particle displacement. Then, 3D PTV and 2D PIV results are compared using an experimental low-turbulence indoor airflow generated by a low-speed tailpipe. Results suggest that when the tracking density is smaller than two, 3D PTV generally cannot yield reliable measurement results. As the tracking density increases, 3D PTV has a better ability to measure larger displacement than PIV.

Estimation of flow discharge by an airborne velocimetry system

A low-cost airborne velocimetry system has been developed and tested, consisting of an action cam, a quadrocopter, ground reference points, and seeding material. The obtained video frames were ortho-rectified and geo- referenced by computer vision techniques, with the by-product that each image got scaled as well. Flow velocity vector fields were determined using a particle image velocimetry algorithm covering a total reach length of 310 m. The data generally confirm depth-averaged velocity profiles obtained by a 3D acoustic Doppler current profiler. Flow discharges are estimated from that. For areas with homogeneous flow conditions we conclude that a correction factor of 0.85-0.9 should be applied to reduce surface velocities to depth-averaged velocities.

Estimating Horizontal Displacement between DEMs by Means of Particle Image Velocimetry Techniques

To date, digital terrain model (DTM) accuracy has been studied almost exclusively by computing its height variable. However, the largely ignored horizontal component bears a great influence on the positional accuracy of certain linear features, e.g., in hydrological features. In an effort to fill this gap, we propose a means of measurement different from the geomatic approach, involving fluid mechanics (water and air flows) or aerodynamics. The particle image velocimetry (PIV) algorithm is proposed as an estimator of horizontal differences between digital elevation models (DEM) in grid format. After applying a scale factor to the displacement estimated by the PIV algorithm, the mean error predicted is around one-seventh of the cell size of the DEM with the greatest spatial resolution, and around one-nineteenth of the cell size of the DEM with the least spatial resolution. Our methodology allows all kinds of DTMs to be compared once they are transformed into DEM format, while also allowing comparison of data from diverse capture methods, i.e., LiDAR versus photogrammetric data sources.

An assessment of the potential of shadow sizing analysis and Particle Image Velocimetry (PIV) to characterise hot trub morphology

Methods for mixture particles imaging are increasingly popular in many fields of engineering research. Particularly interesting is application of imaging techniques in the case of mixtures which by their nature take the form of agglomerates inside the flow. There are cases in which introduction of sensors into the flow system is not possible. What is also very often impossible is to perform mixture initial homogenization, required by many measurement apparatuses, due to the fact that such a measurement does not provide reliable information about dispersed phase's morphology. An example of such a case is analysis of the so-called hot trub which is formed during beer wort boiling. Determination of the hot trub amount and knowledge of distribution of its particles' size are the key elements impacting the criteria of optimization of this type of separator's structure. While the sediment particles are traveling towards the bottom the sediment's volumetric share is considerable, which further impedes the analysis of the investigated system. Therefore, in the case of hot trub (which is a mixture with high concentration of dispersed phase), using a non-invasive method based on analysis of digital images is a justified choice. This paper presents an application of the Shadow Sizing method for identification of morphological parameters (diameter) of hot trub particles. Abilities and limitations in detection of hot trub particles traveling within an interrogation cell have been determined. An application of the shadow analysis as a method for determination of dispersed phase's velocity. A way of drawing conclusions from obtained images, later used for PIV analysis with LMS and Adaptive PIV algorithms, has been presented.

Correlation algorithm for computing the velocity fields in microchannel flows with high resolution

A cross-correlation algorithm, which enables the obtaining of the velocity field in the flow with a spatial resolution up to a single pixel per vector, has been realized in the work. It gives new information about the structure of microflows as well as increases considerably the accuracy of the measurement of the flow velocity field. In addition, the realized algorithm renders information about the velocity fluctuations in the flow structure. The algorithm was tested on synthetic data at a different number of test images the velocity distribution on which was specified by the Siemens star. The experimental validation was done on the data provided within the international project “4th International PIV Challenge”. Besides, a detailed comparison with the Particle Image Velocimetry algorithm, which was realized previously, was carried out.