A modified cross-correlation algorithm for PIV image processing of particle-fluid two-phase flow

PIV (Particle Image Velocimetry) technique for flow field measurement has achieved popular self-identify through over ten years development, and its application range is becoming wider and wider. PIV post-processing techniques have a great influence on the success of particle-fluid two-phase flow field measurement and thus become a hot and difficult topic. In the present study, a Phase Respective Identification Algorithm (PRIA) is introduced to separate low-density solid particles or bubbles and high-density tracer particles from the PIV image of particle-fluid two-phase flow. PTV (Particle Tracking Velocimetry) technique is employed to calculate the velocity fields of low-density solid particles or bubbles. For the velocity fields of high-density solid particles or bubble phase and continuous phase traced by high-density smaller particles, based on the thought of wavelet transform and multi-resolution analysis and the theory of cross-correlation of image, a delaminated processing algorithm (MCCWM) is presented to conquer the limitation of conventional Fourier transform. The algorithm is firstly testified on synthetic two-phase flows, such as uniform steady flow, shearing flow and rotating flow, and the computational results from the simulated particle images are in reasonable agreement with the given simulated data. The algorithm is then applied to images of actual bubble-liquid two-phase flow and jet flow, and the results also confirmed that the algorithm proposed in the present study has good performance and reliability for post-processing PIV images of particle-fluid two-phase flow.     

Phase discrimination and a high accuracy algorithm for PIV image processing of particle–fluid two-phase flow inside high-speed rotating centrifugal slurry pump

PIV technology is an efficient and powerful measurement method to investigate the characteristics of fluid flow field. But for PIV particle image post-processing, some problems still exit in two-phase particles discrimination and velocity field algorithm, especially for high-speed rotating centrifugal slurry pump. In this study, through summarization and comparison of the various phase discrimination methods, we proposed a two-phase identification method based on statistics of gray-scale level and particle size. The assessment of performance through experimental PIV images shows that a satisfying effect for particle identification. For high speed rotation of the impeller, a combination of adaptive cross-correlation window deformation algorithm and multistage grid subdivision is presented. The algorithm is applied to experimental PIV images of solid–liquid two-phase flow in a centrifugal slurry pump, the results show that the algorithm in the present study has less pseudo vector number and more matching particle pairs than those of fixed window and window translation methods, having the ability to remove pseudo vector efficiently. It confirmed that the algorithm proposed in the present study has good performance and reliability for PIV image processing of particle–fluid two-phase flow inside high-speed rotating centrifugal slurry pump.     

An improved particle image velocimetry method for efficient flow analyses

This study proposes a spatiotemporally Adaptive Search Area (SA) size selecting algorithm for Particle image velocimetry (PIV), ASAPiv. The presented method releases the constraint of conventionally used static SAs, offering a significant computational performance increase through optimizing the dimensions of the SA according to the local flow conditions in a transient manner. The algorithm is implemented as a part of a new PIV framework, developed within the MATLAB environment. The most relevant steps of PIV and the related methods are reviewed, starting from image pre-processing up to the post-processing of raw PIV results. The performance of the proposed algorithm and the PIV tool in general is demonstrated through three examples of different nature, including a synthetic image sequence, a conventional, laser illuminated PIV case and a large-scale, field application. The dynamic alteration of the SAs is found to be consistent with the prevailing flow conditions, while the accuracy of PIV in general is maintained. Total calculation times with static and dynamic SAs are compared. The benchmark cases highlight the relevance of adaptive SAs in cases of spatiotemporally varied flow conditions, where significant (up to 900%) computational performance increase is achieved. In case of unidirectional, steady flow conditions the method offers moderate speed-up compared to the employment of static SA sizes.     

Deep particle image velocimetry supervised learning under light conditions

Particle image velocimetry (PIV) is an important fluid visualization technology which extracts the velocity field from two successive particle images. Recently, some researchers have begun to use convolutional neural network (CNN) to tackle the PIV problem successfully. Some supervised learning methods make use of the PIV dataset with ground truth for network training. However, the existing dataset is composed of pairs of particle images under ideal light conditions and does not take into account the changes in actual experimental conditions. In this paper, we firstly generated a new and more challenging dataset called Light-PIV which fully simulates the change of the brightness of particle images in the real PIV experiment. Secondly, we present here a novel approach for fluid motion estimation which is based on an optical flow network LiteFlowNet. The proposed approach is verified by the application to a diversity of synthetic and experimental PIV images. We not only improve the structure, but also combine the traditional prior assumptions knowledge with the loss function to better guide the network training. The proposed approach is verified by the application to a diversity of synthetic and experimental PIV images. The experimental results show that our proposed method has advantages of high accuracy, obtaining detailed information and strong robustness in our PIV dataset compared with classical PIV methods such as HS optical flow and WIDIM, and even outperforms these existing approaches in some flow cases.     

The development of an automated PIV image processing software—SmartPIV

Since the popularity of digital particle image velocimetry technique (DPIV), many PIV image processing algorithms have been proposed. Amongst them, fast Fourier transform (FFT) Cross Correlation, Discrete Window Offset Cross Correlation, Iterative Multigrid Cross Correlation, Iterative Image Deformation Cross Correlation and cross correlation based particle tracking methods are widely used algorithms and have been extensively studied by researchers. All of these algorithms have their advantages and disadvantages in terms of computational load and measurement accuracy. To choose a suitable algorithm, researchers not only need to understand the complex principles of these algorithms, but also need to find out their applicable flow conditions. This could greatly increase work load for PIV users who focus more on flow structure itself instead of PIV algorithms. It is therefore necessary to develop a method which can choose PIV algorithms wisely according to the input PIV images. This paper firstly reviews the development of PIV algorithm with mainly focus on analysing advantages and disadvantages of six widely used algorithms. By using both synthetic and real PIV images, comparative studies are then carried out among these algorithms. The tests give a rate for the performance of the algorithms and provide a parameter to automatically separate pattern match and particle tracking algorithms. Based on qualitative and quantitative analysis, an automated PIV image processing method—SmartPIV is proposed and tested by both synthetic and real PIV images. For all the three test cases, the SmartPIV successfully picks the most suitable algorithm and gives very promising results.     

基于几何光学的图像矫正法在圆管PIV实验中的应用研究

粒子图像速度技术被广泛用于流体流动测量,介质折射率差异使光在圆管壁面发生偏折,导致图像失真,直接影响速度测量精度。本文建立了光学折射的物理模型,得到圆形管道中物点和图像点之间的函数关系进而得到矫正后图像的像素坐标,使用双线性插值算法得到像素灰度值重建出矫正后的粒子图像,最后根据多重网格迭代算法计算管内速度场。分别对流体进行管内静态流体与管内层流速度场测量实验,对比了光学矫正箱法、线性矫正以及基于光学模型的畸变矫正方法误差。结果表明,本文提出的基于几何光学的图像矫正方法精度优于光学矫正箱法和线性矫正方法,并通过静态与流动实验充分验证了所建立几何光学模型的准确性和有效性。     

Improvements to time-series TR-PIV algorithms using historical displacement and displacement variation information

Improvements to two widely used particle-image velocimetry (PIV) algorithms, e.g., multi-grid and iterative image deformation cross-correlations, are proposed here to reduce the computational costs associated with time-resolved PIV (TR-PIV) data-processing. TR-PIV typically involves capturing significant time-series particle-image datasets across to allow statistically meaningful temporal and spectral analyses; hence considerable computational cost-savings can be realised. The improvements involve using the historical particle displacement field and its variation to determine the required window offsets and image deformations in the above-mentioned algorithms, respectively. In this case, cross-correlation based on the smallest interrogation window size can be used directly instead of multi-pass cross-correlations based on decreasing interrogation window sizes. To evaluate their efficacy, the proposed improvements were implemented and evaluated using synthetic PIV images of a Rankine vortex flow, numerical solutions for a square cylinder wake flow, as well as actual experimental time-series TR-PIV measurements. Comparisons show that the proposed improvements save up to 50% computational time while maintaining relatively similar measurement accuracy levels as conventional algorithms. In particular, the new algorithms successfully resolve unsteady flow fields where particle displacements vary by more than 20% between successive particle-images, where error propagations associated with large displacement variations are mitigated by employing suitable recalculation thresholds.     

基于MATLAB的PIV软件(MPIV)的开发与应用

研究一种基于MATLAB的PIV软件MPIV，可处理各种粒子密度的PIV图像。选用两种有代表性的算法，分别是对应于高粒子密度图像的基于快速傅立叶变换的互相关法，以及对应于低粒子密度图像的二值化图像相关法，以旋转流场为例，通过计算机模拟粒子图像，检验软件的可行性。     

Particle image velocimetry of a flow at a vaulted wall

The assessment of flow along a vaulted wall (with two main finite radii of curvature) is of general interest; in biofluid mechanics, it is of special interest. Unlike the geometry of flows in engineering, flow geometry in nature is often determined by vaulted walls. Specifically the flow adjacent to the wall of blood vessels is particularly interesting since this is where either thrombi are formed or atherosclerosis develops. Current measurement methods have problems assessing the flow along vaulted walls. In contrast with conventional particle image velocimetry (PIV), this new method, called wall PIV, allows the investigation of a flow adjacent to transparent flexible surfaces with two finite radii of curvature. Using an optical method which allows the observation of particles up to a predefined depth enables the visualization solely of the boundary layer flow. This is accomplished by adding a specific dye to the fluid which absorbs the monochromatic light used to illuminate the region of observation. The obtained images can be analysed with the methods of conventional PIV and result in a vector field of the velocities along the wall. With wall PIV, the steady flow adjacent to the vaulted wall of a blood pump was investigated and the resulting velocity field as well as the velocity fluctuations were assessed.     

Optimization in PIV algorithm for visualizing vortices in bubble wake

The PIV (Particle image velocimetry) is the most commonly used method for flow field observation because of its high efficiency and non-interference to the flow field. This study aims to clarify the optimal parameter conditions used in the cross-correlation algorithm of PIV for flow fields with vortices. The influence factors on the error of the cross-correlation algorithm are analyzed and discussed using a synthetic flow field, including the seeding conditions, the velocity gradient and vortex size. It is confirmed that the minimum particle number density per interrogation window is about 10, which generally limits the minimum size of the window. For a vortex, when the vortex size is fixed, the velocity gradient corresponding to the characteristic velocity both controls the lower and upper limitation of window size. For a relatively small vortex, generally a window not larger than the vortex size is asked. Then, a strategy to improve the observation based on the existing equipment is discussed and applied to visualize a rising bubble wake based on LIF (Laser-induced fluorescence) images.     

光流法运动估计在FPGA上的实现与性能分析

图像序列的光流估计理论在机器视觉领域已被提出多年,但算法的高计算复杂度限制了其在工业领域的应用。为了满足应用的实时性要求,阐述了一种光流实时估计的实现方法。为了提高算法精度及减少FPGA片内资源消耗,对L&K光流计算方法进行改进。首先,通过设计两层光流计算架构来提高精度。针对在此过程中出现的外部存储器读写速率不够的问题,提出一次读取同时分层缓存、分时计算的方法。考虑到两层光流在计算过程中的迭代关联性,设计了满足要求的外部存储器数据读出顺序表;然后,针对卷积运算资源消耗大的问题,设计了新的卷积权重函数,能够将卷积计算量降低73%,从而节省了大量逻辑资源;最后通过实验验证,所提出的FPGA光流计算方法的精度高于运行在PC平台的L&K方法,卷积计算资源消耗明显降低。设计的系统可以完成1280×1024 pixel、60 frame/s输入视频的计算,满足光流计算的实时性要求。     

基于自适应窗口选择PIV技术的序列星图运动估计方法

对星空观测CCD相机获得的序列图象,由于相机姿态的变化或者观测平台位置的改变,观测的序列图象存在全局运动,需要对星图的运动参数进行准确的估计。粒子图象测速技术（Particle Image Velocimetry, PIV）的发明和发展标志着现代实验流体测量技术的重大突破。PIV技术在流体力学及空气动力学研究领域具有较高的学术意义和实用价值,本文提出一种基于自适应窗口选择PIV技术的序列星图运动参数估计算法。算法通过自适应选取窗口使得分析窗口内的恒星成像密度达到最大,且窗口边缘区域不存在高亮恒星成像,然后对分析窗口进行互相关计算,求得序列星图的运动参数。结果证明本文算法在很大程度上提高了互相关运动参数估计的精度,将运动参数估计误差降低到0.01pixel以内。     

Increasing microscopy resolution with photobleaching and intensity cumulant analysis

Super‐resolution fluorescence microscopy and its applications for analysis of biological structures are evolving rapidly field. A number of approaches aimed at overcoming the fundamental limit imposed by diffraction have been proposed in recent years. Here we present a modification of super‐resolution optical fluctuation imaging (SOFI), a technique based on spatio‐temporal evaluation of the optical signal from independently fluctuating emitters. Instead of rapid, reversible photoswitching, photobleaching is used to produce irreversible transitions between emitting and nonemitting states of the fluorochrome molecules. Simulated images are used to demonstrate that, in the absence of noise, the proposed SOFI modification increases the efficiency of transfer of high spatial frequencies in a fluorescence microscope. Correspondingly, a decrease of the point spread function (PSF) width is obtained. Moreover, the modified SOFI algorithm is capable of resolving point emitters in the presence of simulated noise. Using real biological images we demonstrate that an increase of resolution is obtained in 2D optical sections through densely packed chromatin in cell nuclei and lamin layer at the nuclear envelope. Finally, the approach is extended to 3D wide‐field microscopy, allowing reduction of out‐of‐focus image blurring. Microsc. Res. Tech. 78:958–968, 2015. © 2015 Wiley Periodicals, Inc.     

Evaluation of recursive PIV algorithm with correlation based correction method using various flow images

The hierarchical recursive local-correlation PIV algorithm with CBC (correlation based correction) method was employed to increase the spatial resolution of PIV results and to reduce error vectors. The performance of this new PIV algorithm was tested using synthetic images, PIV standard images of Visualization Society of Japan, real flows including ventilation flow inside a vehicle passenger compartment and wake behind a circular cylinder with riblet surface. As a result, most spurious vectors were suppressed by employing the CBC method, the hierarchical recursive correlation algorithm improved the sub-pixel accuracy of PIV results by decreasing the interrogation window size and increased spatial resolution significantly. However, with recursively decreasing of interrogation window size, the SNR (signal-to-noise ratio) in the correlation plane was decreased and number of spurious vectors was increased. Therefore, compromised determination of optimal interrogation window size is required for given flow images, the performance of recursive algorithm is also discussed from a viewpoint of recovery ratio and error ratio in the paper.     

Mallat算法的光学实现方法

现有的光学小波变换方法均基于连续小波变换,基于离散信号的小波变换算法（Mallat算法）的光学小波变换还没有出现,这阻碍了光学小波变换应用的发展。针对这一问题,分析利用光学4f系统实现Mallat算法的基本原理,提出Mallat算法的光学实现方法。针对空间光调制器只能实现非负的实函数,且CCD只能记录光的强度,给出一种应用于光学4f系统的光学小波滤波器的设计方法。使用该种光学小波滤波器,利用光学4f系统实现Mallat算法的小波分解部分,并通过数值计算实现Mallat算法的小波重构部分。仿真分析和光学实验结果验证了方法的正确性。     

卫星解码数字图像块效应计算机后处理方法

分块余弦变换(BDCT)已经被广泛应用于国际图像压缩标准并取得了良好的效果。然而,在有损图像处理中都存在着块效应问题。为了消除卫星数字图像块效应,这里结合卫星图像的特点,在文献^［4］的DCT后处理方法基础上进行改进,提出一种改进的解码图像块效应后处理方法。此方法不仅可以减少卫星图像块效应和提高图像质量,而且减少了DCT块效应后处理时间,提高了解码图像后处理效率。具有广泛的应用性。     

基于激光扫描的密集泡状流三维重建与优化

为测量密集气液泡状流的流动形态及参数,建立了基于激光扫描的三维可视化测量系统。采用片状激光结合旋转正多边形棱镜实现对流场的光学扫描,高速摄像机采集扫描切片图像,首先对图像进行预处理。针对扫描成像中产生的切片重复曝光问题,提出二阶微分平均卷积优化算法,该方法不仅可以有效提取多切片图像中重复曝光的特征点,而且可以去除冗余及噪声信息。实验结果表明,针对分散相遮挡的密集泡状流,基于激光扫描可完整重建其三维结构,二阶微分优化算法可以有效降低重建畸变影响,重建后体积含气率的相对误差优于6%。激光扫描方法非侵入、重建精度高,具有传统方法不可比拟的优势。     

基于光流的物体移动速度检测技术

通过采集目标物体的运动图像,根据移动物体运动场和光流场的关系,由运动主方向的原理,确定出物体移动的方向;采用一种基于金字塔式的L-K算法对移动物体的光流速度进行计算;把图像中的光流速度转化为物体的实际移动速度。实验结果表明,该算法能有效地检测出物体的实际移动速度。     

应用自适应模拟退火和多分辨率搜索实现医学超声图像的拼接

为提高图像拼接的成功率,提出了一种基于自适应模拟退火和多分辨率搜索策略的图像自动拼接新方法。该新算法先自适应地选取配准区域,再以互信息为相似度评价标准,结合自适应模拟退火和多分辨率搜索策略的思想分别进行图像平移和旋转参数的全局优化和局部搜索,最后实现图像的拼接。通过对含噪声数字图像和医学超声图像进行的24次模拟拼接实验表明,该新算法较传统的多分辨率直接搜索法有精度高、速度快和抗噪声能力强的优点。由于结合了模拟退火算法的高精度和多分辨率搜索法的高效率,改进后的图像拼接算法将拼接成功率提高了12.5%,并将运算时间控制在可接受的范围内。     

微球测速聚类分析的流式液路稳定性评估

提出了一种基于90°Mie散射的高速图像采集微球测速方法,用于准确评估流式细胞仪流动室内的层流状态及单细胞流的稳定性。利用流动室内微球速度的稳定性对流动室内单细胞流的稳定性进行了评估。首先,利用高速显微成像系统采集90°Mie散射光的图像,选取90°侧向散射光以避免激发光源直射光的干扰,同时去除背景光源并提高图像对比度;然后,利用基于梯形白化权函数的灰色聚类分析方法对微球拖尾图像进行分类,实现对不足、正常、衍射和重叠4种情况的准确分类;最后,利用中点法确定正常图像上升沿及下降沿的边界,提高拖尾长度计算的准确性。搭建了高速微球测速实验系统,对本文方法进行验证。结果表明,该方法能够获得清晰的微球拖尾图像并对微球拖尾图像进行准确分类。对本文实验系统测得的微球拖尾长度平均值为116.9个像素点,标准差为1.7。