一种基于帧间差分和光流技术结合的运动车辆检测和跟踪新算法

为了改进智能交通中的运动车辆检测和跟踪方法,提出一种基于改进的帧间差分和光流技术结合的运动车辆检测和跟踪的新方法。先用帧间差分法检测出运动物体的运动区域,再计算差值图中不为零处的光流,然后利用其光流场来实现运动目标的跟踪。为了减少计算量,提出一种基于最优估计的点匹配技术和光流均匀采样策略的光流场计算方法,并通过对灰度化后的光流场进行自适应阈值分割、形态学滤波等处理,实现了实时的运动目标检测和跟踪。     

Fluid in Video: Augmenting Real Video with Simulated Fluids

We present a technique for coupling simulated fluid phenomena that interact with real dynamic scenes captured as a binocular video sequence. We first process the binocular video sequence to obtain a complete 3D reconstruction of the scene, including velocity information. We use stereo for the visible parts of 3D geometry and surface completion to fill the missing regions. We then perform fluid simulation within a 3D domain that contains the object, enabling one‐way coupling from the video to the fluid. In order to maintain temporal consistency of the reconstructed scene and the animated fluid across frames, we develop a geometry tracking algorithm that combines optic flow and depth information with a novel technique for “velocity completion”. The velocity completion technique uses local rigidity constraints to hypothesize a motion field for the entire 3D shape, which is then used to propagate and filter the reconstructed shape over time. This approach not only generates smoothly varying geometry across time, but also simultaneously provides the necessary boundary conditions for one‐way coupling between the dynamic geometry and the simulated fluid. Finally, we employ a GPU based scheme for rendering the synthetic fluid in the real video, taking refraction and scene texture into account.     

一种运动目标多特征点的鲁棒跟踪方法研究

提出了一种基于特征光流分割和卡尔曼滤波估计的鲁棒性的运动目标跟踪方法。该方法具有很多特点：首先在特征光流的计算中采用由粗到细的层级匹配算法，因而能够计算大的运动速度和具有更好的匹配精度；其次采用了有效的遮挡判决算法，该算法综合利用了先验的信息，对噪声的干扰不敏感；最后建立了线性卡尔曼滤波模型，当特征点被遮挡或丢失时，能够预测它们的位置，这使得跟踪更具有主动性。实验表明，该方法具有高精度、快速跟踪和很好的鲁棒性。     

An efficient and robust method for Lagrangian magnetic particle tracking in fluid flow simulations on unstructured grids

In this paper we report on a newly developed particle tracking scheme for fluid flow simulations on 3D unstructured grids, aiming to provide detailed insights in the particle behaviour in complex geometries. A possible field of applications is the magnetic drug targeting (MDT) technique, on which this paper will be focused. MDT is a promising medical technique that uses locally applied magnetic fields to capture magnetic drug carriers at the desired locations in the human body, strongly increasing the efficiency of medical drugs. The new particle tracking scheme combines the advantages of existing methods and is easy for implementation in a generic numerical code. The scheme is tested and validated for simple MDT cases that include effects of a non-homogeneous magnetic field on deposition of magnetic particles in laminar flow. The first test case is a validation study of the magnetic particle trajectories released in a horizontal circular pipe flow with a current-carrying wire parallel to the flow, for which analytical solutions are reported in literature. The second test case involves particle capture efficiencies in a 90° bent tube for different configurations of the imposed magnetic field. This configuration corresponds more closely to the conditions inside blood vessels, because of the presence of secondary motions. These results are compared with numerical studies from literature too. The obtained results demonstrate that the developed particle tracking scheme is a very robust, efficient and accurate method, which can give detailed insights in particle behaviour in complex geometries. As such it is a good candidate for future applications and optimisations of MDT technique for loco-regional cancer treatment or treatment of cardiovascular diseases.     

A Mass Conservative Scheme for Fluid–Structure Interaction Problems by the Staggered Discontinuous Galerkin Method

In this paper, we develop a new mass conservative numerical scheme for the simulations of a class of fluid–structure interaction problems. We will use the immersed boundary method to model the fluid–structure interaction, while the fluid flow is governed by the incompressible Navier–Stokes equations. The immersed boundary method is proven to be a successful scheme to model fluid–structure interactions. To ensure mass conservation, we will use the staggered discontinuous Galerkin method to discretize the incompressible Navier–Stokes equations. The staggered discontinuous Galerkin method is able to preserve the skew-symmetry of the convection term. In addition, by using a local postprocessing technique, the weakly divergence free velocity can be used to compute a new postprocessed velocity, which is exactly divergence free and has a superconvergence property. This strongly divergence free velocity field is the key to the mass conservation. Furthermore, energy stability is improved by the skew-symmetric discretization of the convection term. We will present several numerical results to show the performance of the method.     

2D Articulated Pose Tracking Using Particle Filter with Partitioned Sampling and Model Constraints

In this paper, we develop a two-dimensional articulated body tracking algorithm based on the particle filtering method using partitioned sampling and model constraints. Particle filtering has been proven to be an effective approach in the object tracking field, especially when dealing with single-object tracking. However, when applying it to human body tracking, we have to face a “particle-explosion” problem. We then introduce partitioned sampling, applied to a new articulated human body model, to solve this problem. Furthermore, we develop a propagating method originated from belief propagation (BP), which enables a set of particles to carry several constraints. The proposed algorithm is then applied to tracking articulated body motion in several testing scenarios. The experimental results indicate that the proposed algorithm is effective and reliable for 2D articulated pose tracking.     

Using ART2 networks to deduce flow velocities

A novel algorithm for obtaining flow velocity vectors using ART2 networks (based on adaptive resonance theory) is presented. The method involves tracking the movement of groups of seeding particles in a fluid space through the analysis of two successive images. Simulated flows, created artificially by shifting the particles through known distances or rotating through known angles, were used to establish the accuracy of the technique in predicting displacements. Accuracies were quantified by comparison with known displacements and were found to improve with increasing displacement, angle of rotation and size of the sampling window. In addition, the technique has been extended to derive qualitative and quantitative information for a practical case of natural convective flow.     

基于微分光流和粒子滤波的视频运动目标跟踪

针对传统跟踪方法中存在的背景遮挡问题,提出一种基于微分光流的粒子滤波跟踪方法。该方法建立一个用于描述目标的微分光流模型,该模型在每一帧的跟踪处理之前获得更新。使用光流相似函数计算目标与粒子所表示区域的相似度,通过粒子     

Viscous flow and colloid transport near air–water interface in a microchannel

In order to understand the transport behavior of colloids near an air–water interface (AWI), two computational methods are applied to simulate the local water flow field near a moving AWI in a 2D microfluidic channel. The first method is a mesoscopic multicomponent and multiphase lattice Boltzmann (LBM) model and the second is the macroscopic, Navier–Stokes based, volume-of-fluid interface tracking method. In the LBM, it is possible to predict the dynamic contact angles after the static contact angle is correctly set, and the predicted dynamic contact angles are in good agreement with previous observations. It is demonstrated that the two methods can yield a similar flow velocity field if they are applied properly. The flow field relative to AWI depends on the direction of the flow, and exhibits curved streamlines that transport fluid between the center of the channel and the wall region. Using the obtained flow, the motion of sub-micron colloids in a de-ionized water solution is then studied by a Lagrangian approach. The observed colloid trajectories are in qualitative agreement with our visualizations using a confocal microscope.     

Physical‐based spatio‐temporal resolution enhancement of scalar data for fluid visualization

We present a method that improves the spatio‐temporal resolution of original data with fluid scalar data. The basis of the method, velocity estimation, is constructed to be inverse and optimized problem. We reduce the calculation cost through convex optimization and make the velocity field more accurate by coupling with Navier–Stokes equations. The spatial resolution receives significant enhancement by applying advection on original data with higher‐resolution velocity field data generated by our method. The temporal resolution is improved by generating intermediate velocity fields through the solution of Navier–Stokes equations. In this paper, we demonstrate that the accuracy of our velocity estimation method is clearly better than that of optical flow methods and the enhanced data show an attractive performance in fluid visualization. Copyright © 2016 John Wiley & Sons, Ltd.     

An optimal constrained approach for divergence-free velocity interpolation and multilevel VOF method

The use of mesh refinement techniques is becoming more and more popular in computational fluid dynamics, from multilevel approaches to adaptive mesh refinement. In this paper we present a new method to interpolate the coarse velocity field which is based on an optimal approach and is characterized by a constrained minimization of an objective functional. The functional contains the sum of the square difference between the velocity components and their target average value subject to a number of divergence-free constraints. In this work we describe this approach in two- and three-dimensional geometries with different discrete velocity field configurations. This technique is applied to a multilevel Volume-of-Fluid (VOF) method where the volume fraction function is used to reconstruct and advect the interface between two immiscible phases. The coarse velocity field is interpolated to a fixed fine grid with the optimal approach over a given number of refinement levels. The results of several kinematic tests are presented, where the mass and geometrical errors are compared with those obtained with refined velocity fields interpolated with a simple midpoint rule.     

Three-dimensional flow velocity and wall shear stress distribution measurement on a micropillar-arrayed surface using astigmatism PTV to understand the influence of microstructures on the flow field

In this study, the influence of the microstructure in a microchannel on the three-dimensional (3D) flow field and shear stress distribution on the wall was investigated with 3D velocity measurement method. In a micro-total analysis system or a lab-on-a-chip application, the control of the flow is necessary. Thus, microstructures are often applied to the fluidic system for passive flow control. However, the flow field which interacts with microstructures becomes complicated three-dimensionally. The 3D measurement of such microfluidic flow would give insight on the interaction of the flow with the structures and be also useful for other applications. In this study, micropillar array was introduced in a microchannel and we investigated the influence of the micropillar on the 3D flow field by the astigmatism particle tracking velocimetry which enables to determine three-dimensional and three-component velocity by single-viewing. Furthermore, the wall shear stress distribution was also investigated. From measurement results, it was confirmed that the pillar changes the wall shear stress distribution and 3D velocity distribution. Compared to a flat channel (no-pillar array), the wall shear stress in our channel varied spatially in a range of approximately ??80 to +?20%. Moreover, we also conducted a numerical simulation to consolidate the measurement results.     

Virtual Rheoscopic Fluids

We present a visualization technique for simulated fluid dynamics data that visualizes the gradient of the velocity field in an intuitive way. Our work is inspired by rheoscopic particles, which are small, flat particles that, when suspended in fluid, align themselves with the shear of the flow. We adopt the physical principles of real rheoscopic particles and apply them, in model form, to 3D velocity fields. By simulating the behavior and reflectance of these particles, we are able to render 3D simulations in a way that gives insight into the dynamics of the system. The results can be rendered in real time, allowing the user to inspect the simulation from all perspectives. We achieve this by a combination of precomputations and fast ray tracing on the GPU. We demonstrate our method on several different simulations, showing their complex dynamics in the process.     

Markov-type velocity field for efficiently animating water stream

In computer graphics, one of the most challenging tasks is continuously varying phenomena such as waving, swaying, and flowing motions. In this paper, we present a novel hybrid model (physical-stochastic) to create an endless animation in which offline simulation is used to produce an infinitely varying real-time animated result. In this particular case, a water stream model is proposed. Most fully 3D physically based simulation methods for depicting fluid flows are very time and memory consuming. Thus, these methods are still reserved for offline simulations and small-domain real-time simulations, especially in the case of fluid flows with irregularly repeating patterns. The proposed model is based on the tracer particle technique, uses a non-static velocity field, and consists of two main phases. In the first phase, we construct the stochastic velocity field by using the physically based method. The second phase is the main part, in which we create real-time endless animation. Here, we introduce a new type of velocity field which we refer to as a Markov-type velocity field (MTVF). MTVF allows us to animate a water stream endlessly in real-time by avoiding the time-consuming process of solving the corresponding equations for every simulation step.     

Enhancement of separation efficiency on continuous magnetophoresis by utilizing L/T-shaped microchannels

In this article a novel design of on-chip continuous magnetophoretic separator was proposed by utilizing the magnetic field and L-turning/T-junction effect of the flow field for high throughput applications. The motion of the magnetic bead was simulated based on Lagrangian tracking method and the separation efficiency was calculated according to the trajectories. Impact parameters including geometrical configuration, fluid velocity, magnetic flux density, magnetic bead size, and temperature on separation efficiency were discussed. The results show that both the L- and T-microchannel separators have higher separation efficiency as compared with the conventional straight-microchannel separator because of the L-turning/T-junction effect of the flow field. The separation efficiencies for L- and T-microchannel separators are 63.4 and 100%, respectively, while it is only 43.7% for straight-microchannel separator at the same conditions. Above a critical flow rate the separation efficiency drops drastically from nearly 100% to zero while this decrease is much slower for T-shaped configurations. The separation efficiency increases initially with the increase of the external magnetic flux density and keeps nearly constant at high magnetic flux density owing to saturated magnetization of the beads. It is also found that both the magnetic bead diameter and fluid temperature have great effect on the separation efficiency. The L/T-microchannel separators presented in this article are simple and efficient for magnetophoretic separation at high flow rates and thus useful for the high-efficiency on-chip enrichment of analytes with very low concentrations.     

粒子滤波目标跟踪算法综述

随着人工智能科学的发展,目标跟踪成为中外学者研究的热点,近年来很多目标跟踪算法相继被提出,其中,经典的卡尔曼滤波算法常被用于目标跟踪领域。然而,在实际情况中,目标跟踪过程常涉及到非线性非高斯问题,由于粒子滤波算法在非线性非高斯系统中有较好的性能,因此将其引入目标跟踪研究领域。针对粒子滤波算法存在的跟踪精度差、实时性不高等问题,近年来国内外学者提出很多改进方法。从特征融合、算法融合和自适应粒子滤波三个方面介绍了相关改进方法的基本思想,展望了粒子滤波算法在目标跟踪领域的发展方向。     

Optical measurement of flow field and concentration field inside a moving nanoliter droplet

Droplets-based method has been employed to enhance mixing in microfluidic systems. This paper presents experimental studies of the recirculating flow field inside a moving droplet and the characterization of the mixing of two aqueous droplets. In the first part, the velocity field inside the moving water droplet was measured using the micro-particle image velocimetry (micro-PIV) technique. The PIV measurements showed that recirculation flow exists inside the droplet. However, the findings suggested that the outer layer of droplets move at a faster velocity than the central part. The result is different from what is reported by other researchers. In the second part, two water droplets, a de-ionized (DI) water droplet and another DI water droplet with fluorescent dye, were brought together by the carrier fluid to form a bigger droplet. The mixing between the two aqueous droplets was characterized by the fluorescent dye concentration distribution.     

形态滤波和空间滤波在两相流速度测量中的应用

为了快速、准确地对气固两相流速度进行测量,介绍一种利用形态滤波和空间滤波处理气固两相流信号的基本方法,首先研究电容传感器的空间滤波效应,并找出固体速度和电容传感器带宽之间的关系.然后通过对一维形态滤波算法理论进行分析,推导出可用于实时运算的形态滤波方法,此方法具有处理速度快,滤波效果好,适用性广的特点,可应用于多种信号的实时处理中.然后利用形态滤波确定传感器的带宽,进而求出固体速度.最后给出仿真实验结果,仿真实验结果表明:该方法可以满足气固两相流速度的测量要求.     

Image and Video Abstraction by Coherence‐Enhancing Filtering

In this work, we present a non‐photorealistic rendering technique to create stylized abstractions from color images and videos. Our approach is based on adaptive line integral convolution in combination with directional shock filtering. The smoothing process regularizes directional image features while the shock filter provides a sharpening effect. Both operations are guided by a flow field derived from the structure tensor. To obtain a high‐quality flow field, we present a novel smoothing scheme for the structure tensor based on Poisson's equation. Our approach effectively regularizes anisotropic image regions while preserving the overall image structure and achieving a consistent level of abstraction. Moreover, it is suitable for per‐frame filtering of video and can be efficiently implemented to process content in real‐time.     

Flow velocity measurement in microchannels using temperature-dependent fluorescent dye

The size of microfluidic channels prevents the use of conventional methods for flow velocity measurement. This paper presents and evaluates a method of flow velocity measurement using a temperature dependent fluorescent dye and a microheater. The microheater applied a heat pulse to the dye flowing in a glass capillary, resulting in a plug of fluid of lower fluorescent intensity. By tracking this low intensity region, the velocity of the heat pulse travelling with the flow was determined and used to calculate average flow velocity using correlations. The method was verified by measuring a range of flow velocities in two different sized glass capillaries.