A Distributed Particle Filter Applied in Single Object Tracking

Visual object-tracking is a fundamental task applied in many applications of computer vision. Particle filter is one of the techniques which has been widely used in object tracking. Due to the virtue of extendability and flexibility on both linear and non-linear environments, various particle filter-based trackers have been proposed in the literature. However, the conventional approach cannot handle very large videos efficiently in the current data intensive information age. In this work, a parallelized particle filter is provided in a distributed framework provided by the Hadoop/Map-Reduce infrastructure to tackle object-tracking tasks. The experiments indicate that the proposed algorithm has a better convergence and accuracy as compared to the traditional particle filter. The computational power and the scalability of the proposed particle filter in single object tracking have been enhanced as well.

Dynamic tracking and mobility analysis of single GLUT4 storage vesicle in live 3T3-L1 cells

Glucose transporter 4 (GLUT4) is responsible for insulin-stimulated glucose transporting into the insulin-sensitive fat and muscle cells. The dynamics of GLUT4 storage vesicles (GSVs) remains to be explored and it is unclear how GSVs are arranged based on their mobility. We examined this issue in 3T3-L1 cells via investigating the three-dimensional mobility of single GSV labeled with EGFP-fused GLUT4. A thin layer of cytosol right adjacent to the plasma membrane was illuminated and successively imaged at 5 Hz under a total internal reflection fluorescence microscope with a penetration depth of 136 nm. Employing single particle tracking, the three-dimensional subpixel displacement of single GSV was tracked at a spatial precision of 22 nm. Both the mean square displacement and the diffusion coefficient were calculated for each vesicle. Tracking results revealed that vesicles moved as if restricted within a cage that has a mean radius of 160 nm, suggesting the presence of some intracellular tethering matrix. By constructing the histogram of the diffusion coefficients of GSVs, we observed a smooth distribution instead of the existence of distinct groups. The result indicates that GSVs are dynamically retained in a continuous and wide range of mobility rather than into separate classes.

Hierarchical clustering based on single-pass for breaking topic detection and tracking

Single-pass is commonly used in topic detection and tracking( TDT) due to its simplicity,high efficiency and low cost. When dealing with large-scale data,time cost will increase sharply and clustering performance will be affected greatly. Aiming at this problem,hierarchical clustering algorithm based on single-pass is proposed,which is inspired by hierarchical and concurrent ideas to divide clustering process into three stages. News reports are classified into different categories firstly.Then there are twice single-pass clustering processes in the same category,and one agglomerative clustering among different categories. In addition,for semantic similarity in news reports,topic model is improved based on named entities. Experimental results show that the proposed method can effectively accelerate the process as well as improve the performance.

Structure defect prediction of single crystal turbine blade by dendrite envelope tracking model

The structure defects such as stray grains during unidirectional solidification can severely reduce the performance of single crystal turbine blades. A dendrite envelope tracking model is developed for predicting the structure defects of unidirectional solidification turbine blade. The normal vector of dendrite envelope is estimated by the gradient of dendrite volume fraction, and the growth velocity of the dendrite envelope (dendrite tips) is calculated with considering the anisotropy of grain growth. The solute redistribution at dendrite envelope is calculated by introducing an effective solute partition coefficient. Simulation tests show that the solute-build-up due to the rejection at envelope greatly affects grain competition and consequently solidification structure. The model is applied to predict the structure defects (e.g. stray grain) of single crystal turbine blade during unidirectional solidification. The results show that the developed model is reliable and has the following abilities: reproduce the growth competition among the different-preferential-direction grains; predict the stray grain formation; simulate the structure evolution (single crystal or dendrite grains).

Determination of diffusion coefficient by image-based fluorescence recovery after photobleaching and single particle tracking system implemented in a single platform

Fluorescence recovery after photobleaching(FRAP)and single particle tracking(SPT)techni-ques determine the diffusion coefficient from average diffusive motion of high-concentration molecules and from trajectories of low-concentration single molecules,respectively.Lateral dif-fusion coefficients measured by FRAP and SPT techniques for the same biomolecule on cell membrane have exhibited inconsistent values across laboratories and platforms with larger dif-fusion coefficient determined by FRAP,but the sources of the inconsistency have not been investigated thoroughly.Here,we designed an image-based FRAP-SPT system and made a direct comparison between FRAP and SPT for diffusion coefficient of submicron particles with known theoretical values derived from Stokes-Einstein equation in aqueous solution.The combined iFRAP-SPT technique allowed us to measure the diffusion coefficient of the same fluorescent particle by utilizing both techniques in a single platform and to scrutinize inherent errors and artifacts of FRAP.Our results reveal that diffusion coefficient overestimated by FRAP is caused by inaccurate estimation of the bleaching spot size and can be corrected by simple image analysis.Our iFRAP-SPT technique can be potentially used for not only cellular membrane dynamics but also for quantitative analysis of the spatiotemporal distribution of the solutes in small scale analytical devices.

Single link tracking scheme based on UWB localization

To investigate the low-complex and high-precise tracking method, a novel single link tracking scheme based on UWB localization is proposed. Two antenna arrays are settled at the fixed station （FS） to receive the UWB signal from mobile terminal （MT）, one FS is enough for the proposed scheme to track the MT. The UWB multipath detection algorithm for time difference of arrival （TDOA） estimation is presented and TDOA is thus adopted to estimate angle of arrival （AOA）, avoiding the synchronization and complicated beamforming for AOA. The impacts of localization errors, concluding multipath and non-line-of-sight （NLOS） errors are simulated for the proposed track scheme. It is demonstrated that the simulation curve can match the real target moving, and the feasibility of the proposed scheme is proved.

SiamDLA: Dynamic Label Assignment for Siamese Visual Tracking

Label assignment refers to determining positive/negative labels foreach sample to supervise the training process. Existing Siamese-based trackersprimarily use fixed label assignment strategies according to human priorknowledge;thus, they can be sensitive to predefined hyperparameters and failto fit the spatial and scale variations of samples. In this study, we first developa novel dynamic label assignment (DLA) module to handle the diverse datadistributions and adaptively distinguish the foreground from the backgroundbased on the statistical characteristics of the target in visual object tracking.The core of DLA module is a two-step selection mechanism. The first stepselects candidate samples according to the Euclidean distance between trainingsamples and ground truth, and the second step selects positive/negativesamples based on the mean and standard deviation of candidate samples.The proposed approach is general-purpose and can be easily integrated intoanchor-based and anchor-free trackers for optimal sample-label matching.According to extensive experimental findings, Siamese-based trackers withDLA modules can refine target locations and outperformbaseline trackers onOTB100, VOT2019, UAV123 and LaSOT. Particularly, DLA-SiamRPN++improves SiamRPN++ by 1% AUC and DLA-SiamCAR improves Siam-CAR by 2.5% AUC on OTB100. Furthermore, hyper-parameters analysisexperiments show that DLA module hardly increases spatio-temporal complexity,the proposed approach maintains the same speed as the originaltracker without additional overhead.

Determination of optimal period of absolute encoders with single track cyclic gray code

Low cost and miniaturized rotary encoders are important in automatic and precise production. Presented here is a code called Single Track Cyclic Gray Code (STCGC) that is an image etched on a single circular track of a rotary encoder disk read by a group of even spread reading heads to provide a unique codeword for every angular position and features such that every two adjacent words differ in exactly one component, thus avoiding coarse error. The existing construction or combination methods are helpful but not sufficient in determining the period of the STCGC of large word length and the theoretical approach needs further development to extend the word length. Three principles, such as the seed combination, short code removal and ergodicity examination were put forward that suffice determination of the optimal period for such absolute rotary encoders using STCGC with even spread heads. The optimal periods of STCGC in 3 through 29 bit length were determined and listed.

Dynamic Analysis of the Seafloor Pilot Miner Based on Single-Body Vehicle Model and Discretized Track-Terrain Interaction Model

In order to achieve the complex dynamic analysis of the self-propelled seafloor pilot miner moving on the seafloor of extremely cohesive soft soil and further to make it possible to integrate the miner system with some subsystems to form the complete integrated deep ocean mining pilot system and perform dynamic analysis, a new method for the dynamic modeling and analysis of the miner is proposed and developed in this paper, resulting in a simplified 3D single-body vehicle model with three translational and three rotational degrees of freedom, while the track-terrain interaction model is built by partitioning the track-terrain interface into discrete elements with parameterized force dements built on the theory of terramechanics acting on each discrete dement. To evaluate and verify the correctness and effectiveness of this new modeling and analysis method, typical comparative studies with regard to computational efficiency and solution accuracy are carried out between the traditional modeling method of building the tracked vehicle as a multi-body model and the new modeling method. In full consideration of the particMar structure design of the pilot miner, the special characteristics of the seafioor soil and the hydrodynamic force of near-seafloor currnt, the dynamic simulation analysis of the miner is performed and discussed, which can provide useful guidance and reference for the practical miner system in design and operation. This new method can not only realize the rapid dynamic simulation analysis of the miner but also make possible the integration and rapid dynamic analysis of the complete integrated deep ocean mining pilot system in further researches.

A Real-Time Multi-Vehicle Tracking Framework in Intelligent Vehicular Networks

In this paper,we provide a new approach for intelligent traffic transportation in the intelligent vehicular networks,which aims at collecting the vehicles’locations,trajectories and other key driving parameters for the time-critical autonomous driving’s requirement.The key of our method is a multi-vehicle tracking framework in the traffic monitoring scenario..Our proposed framework is composed of three modules:multi-vehicle detection,multi-vehicle association and miss-detected vehicle tracking.For the first module,we integrate self-attention mechanism into detector of using key point estimation for better detection effect.For the second module,we apply the multi-dimensional information for robustness promotion,including vehicle re-identification(Re-ID)features,historical trajectory information,and spatial position information For the third module,we re-track the miss-detected vehicles with occlusions in the first detection module.Besides,we utilize the asymmetric convolution and depth-wise separable convolution to reduce the model’s parameters for speed-up.Extensive experimental results show the effectiveness of our proposed multi-vehicle tracking framework.

An Adaptive Single Neural Control for Variable Step-Size P&O MPPT of Marine Current Turbine System

Marine current energy has been increasingly used because of its predictable higher power potential.Owing to the external disturbances of various flow velocity and the high nonlinear effects on the marine current turbine(MCT)system,the nonlinear controllers which rely on precise mathematical models show poor performance under a high level of parameters’uncertainties.This paper proposes an adaptive single neural control(ASNC)strategy for variable step-size perturb and observe(P&O)maximum power point tracking(MPPT)control.Firstly,to automatically update the neuron weights of SNC for the nonlinear systems,an adaptive mechanism is proposed to adaptively adjust the weighting and learning coefficients.Secondly,aiming to generate the exact reference speed for ASNC to extract the maximum power,a variable step-size law based on speed increment is designed to strike a balance between tracking speed and accuracy of P&O MPPT.The robust stability of the MCT control system is guaranteed by the Lyapunov theorem.Comparative simulation results show that this strategy has favorable adaptive performance under variable velocity conditions,and the MCT system operates at maximum power point steadily.

基于改进Single-pass 算法的新闻话题演化跟踪算法

随着信息技术的发展,每天都有大量的新闻文本在互联网上发布、转发,在这样的海量信息环境下,如何快速定位自己感兴趣的话题、追踪其发展趋势已成了近年来的研究热点。面向互联网上新闻文本,提出聚类阈值的估计方法对已有的Single-pass算法进行优化,进而基于时间片设计一个新闻文本演化算法。在新华网等四个网站上采集新闻数据并进行实验,实验表明所提算法可有效跟踪新闻话题的演化过程。

RECALLABLE TRACKING TECHNIQUE FOR TOP WELD BEADS OF NARROW GAP SUBMERGED ARC WELDING

According to the wire and nozzle movement track in groove, the movement parameters of wire were memorized and recalled for the following top welds by using a single chip computer. In this paper, it was also discussed that the design problems of correcting deviation of wire movement track in narrow gap submerged arc welding process must be noticed in order to obtain the sound welding result.

Single Molecule Applications of Quantum Dots

Fluorescent nanocrystals composed of semiconductor materials were first introduced for biological applications in the late 1990s. The focus of this review is to give a brief survey of biological applications of quantum dots (QDs) at the single QD sensitivity level. These are described as follows: 1) QD blinking and bleaching statistics, 2) the use of QDs in high speed single particle tracking with a special focus on how to design the biofunctional coatings of QDs which enable specific targeting to single proteins or lipids of interest, 3) a hybrid lipid-DNA analogue binding QDs which allows for tracking single lipids in lipid bilayers, 4) two-photon fluorescence correlation spectroscopy of QDs and 5) optical trapping and excitation of single QDs. In all of these applications, the focus is on the single particle sensitivity level of QDs. The high applicability of QDs in live cell imaging experiments held together with the prospects in localization microscopy and single molecule manipulation experiments gave QDs a promising future in single molecule research.

Parasitic Effects on the Performance of DC-DC SEPIC in Photovoltaic Maximum Power Point Tracking Applications

This paper presents an analysis of the effect of parasitic resistances on the performance of DC-DC Single Ended Pri- mary Inductor Converter (SEPIC) in photovoltaic maximum power point tracking (MPPT) applications. The energy storage elements incorporated in the SEPIC converter possess parasitic resistances. Although ideal components significantly simplifies model development, but neglecting the parasitic effects in models may sometimes lead to failure in predicting first scale stability and actual performance. Therefore, the effects of parasitics have been taken into consideration for improving the model accuracy, stability, robustness and dynamic performance analysis of the converter. Detail mathematical model of SEPIC converter including inductive parasitic has been developed. The performance of the converter in tracking MPP at different irradiance levels has been analyzed for variation in parasitic resistance. The converter efficiency has been found above 83% for insolation level of 600 W/m2 when the parasitic resistance in the energy storage element has been ignored. However, as the parasitic resistance of both of the inductor has increased to 1 ohm, a fraction of the power managed by the converter has dissipated;as a result the efficiency of the converter has reduced to 78% for the same insolation profile. Although the increasing value of the parasitic has assisted the converter to converge quickly to reach the maximum power point. Furthermore it has also been observed that the peak to peak load current ripple is reduced. The obtained simulation results have validated the competent of the MPPT converter model.

Diffusion tensor imaging fiber tracking with reliable tracking orientation and flexible step size

We propose a method of reliable tracking orientation and flexible step size fiber tracking. A new directional strategy was defined to select one optimal tracking orientation from each directional set, which was based on the single-tensor model and the two-tensor model. The directional set of planar voxels contained three tracking directions： two from the two-tensor model and one from the single- tensor model. The directional set of linear voxels contained only one principal vector. In addition, a flexible step size, rather than fixable step sizes, was implemented to improve the accuracy of fiber tracking. We used two sets of human data to assess the performance of our method; one was from a healthy volunteer and the other from a patient with low-grade glioma. Results verified that our method was superior to the single-tensor Fiber Assignment by Continuous Tracking and the two-tensor eXtended Streamline Tractography for showing detailed images of fiber bundles.

Initial refinement of data from video‐based single‐cell tracking

Background: Video recording of cells offers a straightforward way to gainvaluable information from their response to treatments. An indispensable stepin obtaining such information involves tracking individual cells from therecorded data. A subsequent step is reducing such data to represent essentialbiological information. This can help to compare various single‐cell trackingdata yielding a novel source of information. The vast array of potential datasources highlights the significance of methodologies prioritizing simplicity,robustness, transparency, affordability, sensor independence, and freedomfrom reliance on specific software or online services.Methods: The provided data presents single‐cell tracking of clonal (A549)cells as they grow in two‐dimensional (2D) monolayers over 94 hours,spanning several cell cycles. The cells are exposed to three differentconcentrations of yessotoxin (YTX). The data treatments showcase theparametrization of population growth curves, as well as other statisticaldescriptions. These include the temporal development of cell speed in familytrees with and without cell death, correlations between sister cells, single‐cellaverage displacements, and the study of clustering tendencies.Results: Various statistics obtained from single‐cell tracking reveal patternssuitable for data compression and parametrization. These statistics encompassessential aspects such as cell division, movements, and mutual informationbetween sister cells.Conclusion: This work presents practical examples that highlight theabundant potential information within large sets of single‐cell tracking data.Data reduction is crucial in the process of acquiring such information whichcan be relevant for phenotypic drug discovery and therapeutics, extendingbeyond standardized procedures. Conducting meaningful big data analysistypically necessitates a substantial amount of data, which can stem fromstandalone case studies as an initial foundation.

3D Single Object Tracking with Multi-View Unsupervised Center Uncertainty Learning

Center point localization is a major factor affecting the performance of 3D single object tracking.Point clouds themselves are a set of discrete points on the local surface of an object,and there is also a lot of noise in the labeling.Therefore,directly regressing the center coordinates is not very reasonable.Existing methods usually use volumetric-based,point-based,and view-based methods,with a relatively single modality.In addition,the sampling strategies commonly used usually result in the loss of object information,and holistic and detailed information is beneficial for object localization.To address these challenges,we propose a novel Multi-view unsupervised center Uncertainty 3D single object Tracker(MUT).MUT models the potential uncertainty of center coordinates localization using an unsupervised manner,allowing the model to learn the true distribution.By projecting point clouds,MUT can obtain multi-view depth map features,realize efficient knowledge transfer from 2D to 3D,and provide another modality information for the tracker.We also propose a former attraction probability sampling strategy that preserves object information.By using both holistic and detailed descriptors of point clouds,the tracker can have a more comprehensive understanding of the tracking environment.Experimental results show that the proposed MUT network outperforms the baseline models on the KITTI dataset by 0.8%and 0.6%in precision and success rate,respectively,and on the NuScenes dataset by 1.4%,and 6.1%in precision and success rate,respectively.The code is made available at https://github.com/abchears/MUT.git.

基于单天线航向测量的股道占用判别方法研究

随着北斗卫星系统的建设完成,以卫星导航为核心的车载化列车定位成为列车定位系统的发展方向。在这一背景下,精准判别列车股道占用成为确保列车高精度位置感知的必要前提,对避免冲突、支撑列车调度决策和保障列车安全具有重要的意义。针对传统的股道占用判别方法通常依赖于地面设备,易受到成本、复杂性和维护困难等因素制约的问题,研究利用低成本的单GNSS天线进行高精度的列车航向角测量。特别是在道岔咽喉区,计算单天线测量结果与数据库中列车航向的偏差,评估对不同股道的占用概率。为此,首先构建了载波相位星间差分与历元差分的双差分模型,消除了电离层、对流层、钟差、模糊度等一系列卫星定位误差源的影响;其次,通过扩展卡尔曼滤波估计相邻历元间列车的位移变化量,同时计算伪姿态坐标系下的列车航向;最后,基于航向角偏差进行列车的股道占用判别,并利用列车进出青藏线联通河站的实测数据对所提算法进行了验证,并结合轨道电子地图,判别列车占用正线/侧线不同股道的置信概率。实验结果表明:通过单天线解算列车航向角的误差均方根为0.111°,根据航向角偏差进行股道判别的模糊区距离减少了约50m,模糊区时间减少了约5s。仅通过单天线航向解算完成高精度的航向测量,并根据航向角偏差实现股道占用判别,可作为轨道占用的一种判别方式,具有一定的工程价值。

不锈钢激光选区熔化单道成形表面介观模拟与实验验证

目的为了获得不锈钢激光选区熔化单道成形质量好的加工参数,提高成形质量。方法对于316L不锈钢材料的激光选区熔化(SLM)成形,采用离散元方法(DEM)构建颗粒随机分布的三维介观模型,采用流体体积法(VOF)动态追踪SLM成形过程熔池传热、流动和凝固等行为,该模拟考虑相变潜热、热物性参数随温度非线性变化、Marangoni效应,研究了温度梯度引起的表面张力和蒸汽反冲力等现象对熔池液态金属流动和凝固的影响,并对不同激光参数和扫描速度对熔池温度场和内部流动的影响规律进行了分析。结果当激光功率为150 W、扫描速度为400 mm/s时,激光能使不锈钢粉末充分熔化,凝固后的轨迹形貌连续光滑。通过提高激光扫描速度和降低激光功率使得热输入减小,熔道表面成形出现球化等缺陷。结论通过单道成形实验观察和分析熔池与熔道的三维尺寸与形貌,有效验证了数值模拟的正确性,较好地预测出SLM过程中的缺陷种类,为深入理解SLM过程中的复杂物理现象和优化工艺参数提供了参考。