视觉目标跟踪方法研究综述

目的 随着军事侦察任务设备的发展,红外与可见光侦察技术成为军事装备中的主要侦察手段。研究视觉目标跟踪技术对提高任务设备的全天候目标侦察、目标跟踪、目标定位等战场情报获取能力具有重要意义。目前,对视觉目标跟踪技术的研究越来越深入,目标跟踪的方法和种类也越来越丰富。本文对目前应用较为广泛的4种视觉目标跟踪方法进行研究综述,为后续国内外研究者对目标跟踪相关理论及发展研究工作提供基础。方法 通过对视觉目标跟踪技术难点问题进行分析,根据目标跟踪方法建模方式的不同,将视觉目标跟踪方法分为生成式模型方法与判别式模型方法。分别对生成式模型跟踪算法中的均值漂移目标跟踪方法和粒子滤波目标跟踪方法,判别式模型跟踪算法中的相关滤波目标跟踪方法和深度学习目标跟踪方法进行研究。首先分别对4种跟踪算法的基本原理进行介绍,然后针对4种跟踪算法基本原理的不足和对应目标跟踪中的难点问题进行分析,最后针对目标跟踪的难点问题,给出对应算法的主流改进方案。结果 针对视觉目标跟踪相关技术研究进展,结合无人机侦察任务需求,对跟踪算法实际应用中存在的重点解决问题与相关目标跟踪的难点问题进行分析,给出目前的解决方案与不足,探讨研究未来无人机目标侦察跟踪技术的发展方向。结论 视觉目标跟踪技术已经取得了显著的进展,在侦察任务中的应用越来越广泛。但目标跟踪技术仍然是非常具有挑战性的问题,目标跟踪中的相关理论有待进一步完善和改进,由于实际应用中的场景复杂,目标跟踪的难点问题的挑战性更大,因此容易导致跟踪效果不佳。针对不同的应用环境,结合具体不同军事装备的特点,研究相对精确和鲁棒并且满足实时性要求的视觉目标跟踪算法,对提升装备的全天候侦察目标信息获取能力具有重要意义。     

基于人类视觉智能和粒子滤波的鲁棒目标跟踪算法

与计算机视觉目标跟踪算法遇到的各种困难和瓶颈不同的是,鲁棒视觉目标跟踪仅仅是人类视觉系统的一项基本功能.为此,基于自顶向下的视觉注意机制,构建了一种相应的计算模型,提出一种利用检测视觉注意力焦点区域的方法来模拟人类视觉系统的全局搜索过程,并用粒子滤波跟踪来模拟人类视觉系统的局部跟踪过程的目标跟踪算法.多组对比实验结果表明了所提出的算法在视觉目标跟踪中的优越性.     

Tracking a moving object via a sensor network with a partial information broadcasting scheme

Tracking a moving object is one of the key applications of wireless sensor networks (WSNs) for facility management, logistics, healthcare, etc. The quantitative relationship between tracking accuracy and resource consumption is crucial for designing a WSN-based tracking system. A partial information broadcasting scheme (PIBS) is developed as a decentralized tracking strategy in this paper, where only a part of the nodes broadcast their tracking estimation results to their neighbors. The relationship between tracking accuracy and resource consumption is quantified with this scheme. Two broadcasting policies are proposed by balancing the residual energy among sensors and reducing time delay, where the “Randomly Broadcast” (RB) policy takes energy consumption as a first priority and the “Good Estimates Broadcast First” (GEBF) policy consider time delay first. Both of them can be implemented as a decentralized tracking strategy. It is shown that the PIBS with RB or PIBS with GEBF can reduce required resources with minor degradation in tracking accuracy in comparison with the centralized tracking strategy. The finite sensing range of a sensor node is considered and a node activation scheme with variable activation radius is introduced for energy saving. The activation radius is adjusted to guarantee tracking coverage with trade-off between tracking accuracy and resource consumption. The analytical conditions on tracking coverage and the bounds on tracking accuracy are obtained. The numerical experiments demonstrate that this activation scheme outperforms the existing schemes.     

An integrated head pose and eye gaze tracking approach to non-intrusive visual attention measurement for wide FOV simulators

Eye gaze tracking is very useful for quantitatively measuring visual attention in virtual environments. However, most eye trackers have a limited tracking range, e.g., ±35° in the horizontal direction. This paper proposed a method to combine head pose tracking and eye gaze tracking together to achieve a large range of tracking in virtual driving simulation environments. Multiple parallel multilayer perceptrons were used to reconstruct the relationship between head images and head poses. Head images were represented with the coefficients extracted from Principal Component Analysis. Eye gaze tracking provides precise results on the front view, while head pose tracking is more suitable for tracking areas of interest than for tracking points of interest on the side view.     

具有动态弹性稀疏表示的鲁棒目标跟踪算法

目标跟踪问题中目标所在环境的变化对跟踪效果有较大影响.鉴于此,提出一种基于弹性网结构的稀疏表示模型,并在粒子滤波框架下设计一种应用稀疏表示模型的抗干扰动态弹性网目标跟踪算法.同时,设计一种根据环境变化程度动态更新稀疏表示模型参数的方法,以克服光照变化等干扰对算法跟踪质量的影响.此外,所提出算法通过使用各向异性核函数计算各候选区域为跟踪目标所在位置的概率,能够提高跟踪算法的准确性,并改进字典模板更新方法,确保模板更新的准确性与及时性,保证跟踪质量.经实验验证,所提出的动态弹性网跟踪算法与其他跟踪算法相比,在光照等扰动下具有更好的跟踪效果,在遮挡及快速运动等情况下也能够有效保证跟踪精度.     

结合目标预测位置的压缩跟踪

目的：提出结合目标预测位置的压缩跟踪算法用于提高目标跟踪的准确度。方法：选择随机间距稀疏Toeplitz矩阵作为投影矩阵,对原始多尺度Haar-like特征进行压缩;然后,将样本与Mean Shift算法框架下的预测位置的距离权重输入Bayes分类器,形成分类背景与目标的判别函数;最后对参数的更新方式进行优化,提出了参数自适应的学习模式。结果：与目前较流行的6种目标跟踪算法在20个具有挑战性的序列中进行比较,实验结果表明本文提出的算法平均跟踪成功率比压缩跟踪算法将近高27%,平均运行时间为0.15秒/帧。结论：本文采用了结合预测位置的压缩跟踪算法,在参数更新阶段采用了非线性参数学习模式,实验表明结合目标预测位置的跟踪算法比一般的跟踪算法更具有鲁棒性,更能适应遮挡等情况,跟踪的效果也更加平滑。     

Moving people tracking with detection by latent semantic analysis for visual surveillance applications

The latent semantic analysis (LSA) has been widely used in the fields of computer vision and pattern recognition. Most of the existing works based on LSA focus on behavior recognition and motion classification. In the applications of visual surveillance, accurate tracking of the moving people in surveillance scenes, is regarded as one of the preliminary requirement for other tasks such as object recognition or segmentation. However, accurate tracking is extremely hard under challenging surveillance scenes where similarity among multiple objects or occlusion among multiple objects occurs. Usual temporal Markov chain based tracking algorithms suffer from the ‘tracking error accumulation problem’. The accumulated errors can finally make the tracking to drift from the target. To handle the problem of tracking drift, some authors have proposed the idea of using detection along with tracking as an effective solution. However, many of the critical issues still remain unsettled in these detection based tracking algorithms. In this paper, we propose a novel moving people tracking with detection based on (probabilistic) LSA. By employing a novel ‘twin-pipeline’ training framework to find the latent semantic topics of ‘moving people’, the proposed detection can effectively detect the interest points on moving people in different indoor and outdoor environments with camera motion. Since the detected interest points on different body parts can be used to locate the position of moving people more accurately, by combining the detection with incremental subspace learning based tracking, the proposed algorithms resolves the problem of tracking drift during each target appearance update process. In addition, due to the time independent processing mechanism of detection, the proposed method is also able to handle the error accumulation problem. The detection can calibrate the tracking errors during updating of each state of the tracking algorithm. Extensive, experiments on various surveillance environments using different benchmark datasets have proved the accuracy and robustness of the proposed tracking algorithm. Further, the experimental comparison results clearly show that the proposed tracking algorithm outperforms the well known tracking algorithms such as ISL, AMS and WSL algorithms. Furthermore, the speed performance of the proposed method is also satisfactory for realistic surveillance applications.     

Robust Stabilization and Tracking Control for a Class of Switched Nonlinear Systems

This paper addresses the problem of robust stabilization and tracking control for a class of switched nonlinear systems via the multiple Lyapunov functions (MLFs) approach. First, a state feedback controller and a state dependent switching law are designed to globally asymptotically stabilize the switched system via linear matrix inequalities (LMIs). The main objective of this paper is to develop a tracking control approach that assures global asymptotic output and state tracking with zero tracking error in the steady state. Then, the tracking control is formulated such that the robust H_∞ tracking performance is achieved. Finally, a simulation example is provided to demonstrate the effectiveness of the main method.     

不确定移动机器人编队间接自适应模糊动力学控制

研究不确定非完整移动机器人Leader/Follower编队动力学控制问题,对Follower提出了自适应队形跟踪控制方法.首先,基于队形方程设计了运动学队形跟踪控制器,使用Backstepping技术扩展出队形跟踪误差动力学方程;然后,利用自适应模糊系统逼近其不确定项,构造了间接自适应模糊动力学队形跟踪控制器;最后,证明了队形跟踪误差可收敛到原点的小邻域内.仿真结果验证了所提出控制策略的有效性.     

粒子群优化的压缩跟踪算法

目的 针对基于压缩感知理论的跟踪算法跟踪效率不高和难以抗遮挡的问题,提出一种结合压缩感知和粒子群优化的跟踪算法。方法 将粒子群优化算法结合到压缩跟踪算法中,提出了采用粒子群优化的搜索方法替代在确定候选目标时,采用每隔一个像素选取一个候选目标的搜索策略;在目标发生遮挡时,采用粒子群优化的方法进行整幅图全局搜索。结果 20个视频序列数据库的目标跟踪结果表明,本文算法极大地提高了跟踪效率,并有很强的抗目标遮挡和形变的能力从而提高了跟踪的成功率。20个视频数据库进行了定量的分析,平均成功率达到了65.2%,平均中心位置偏差为33.4,平均每秒运行155.5帧。结论 提出的跟踪算法优化了搜索目标的计算次数,提高了算法的运行效率,当在目标发生遮挡时,采用粒子群优化进行全局搜索直到目标重新出现,从而提高了跟踪算法的跟踪成功率,本文算法能适用于不同场景,能够提高智能视频监控系统的智能监控性能。     

基于Tiny-YOLOV3的无人机地面目标跟踪算法设计

为了提高四旋翼无人机对地面目标跟踪的稳定性和跟踪精度,提出了一种结合Tiny-YOLOV3和卡尔曼滤波的跟踪算法;首先分析了Tiny-YOLOV3的原理和网络结构,并基于Tiny-YOLOV3的目标检测结果,结合无人机状态和目标的几何关系建立了目标跟踪系统的数学模型;接着对目标相对运动关系进行分析,建立目标的运动学模型,考虑到目标检测结果受干扰影响较大,应用卡尔曼滤波器实现对目标轨迹的滤波和预测,进而提升目标跟踪的精度;最后根据经过卡尔曼滤波后的目标轨迹信息设计无人机控制律,在轨迹控制的同时引入对无人机偏航角的控制,从而实现无人机对目标的稳定跟踪;仿真结果表明无人机对目标的位置跟踪精度在0.5 m以内,速度跟踪误差在0.2 m/s以内,偏航角跟踪误差在3°以内,跟踪效果良好,从而论证了所提算法的有效性。     

多用户眼动跟踪数据的可视化共享与协同交互

随着数字图像处理技术的发展,以及计算机支持的协同工作研究的深入,眼动跟踪开始应用于多用户协同交互.但是已有的眼动跟踪技术主要针对单个用户,多用户眼动跟踪计算架构不成熟、标定过程复杂,眼动跟踪数据的记录、传输以及可视化共享机制都有待深入研究.为此,建立了基于梯度优化的协同标定模型,简化多用户的眼动跟踪标定过程;然后提出面向多用户的眼动跟踪计算架构,优化眼动跟踪数据的传输和管理.进一步地,探索眼动跟踪数据的可视化形式在协同交互环境下对用户视觉注意行为的影响,具体设计了圆点、散点、轨迹这3种可视化形式,并验证了圆点形式能够有效地提高多用户协同搜索任务的完成效率.在此基础上,设计与开发了基于眼动跟踪的代码协同审查系统,实现了代码审查过程中多用户眼动跟踪数据的同步记录、分发,以及基于实时注视点、代码行边框和背景灰度、代码行之间连线的可视化共享.用户实验结果表明,代码错误的平均搜索时间比没有眼动跟踪数据可视化分享时减少了20.1%,显著提高了协同工作效率,验证了该方法的有效性.     

非连续轨迹下的公路车辆智能跟踪技术研究

针对运动车辆轨迹跟踪控制中的遮挡问题,研究创新地提出以SVM分类检测器优化STAPLE跟踪算法,以保证在目标退出遮挡时可以重新搜索并定位目标;同时对颜色特征直方图、HOG算法进行了改进,以提高算法特征提取效率;最后选择VOT2016国际标准序列集对优化后的STAPLE跟踪算法进行验证;研究结果显示,改进STAPLE算法能在目标退出遮挡后更为快速地重新捕捉目标位置;改进算法对目标中心的跟踪精度达到了0.81,对目标框的跟踪精度达到了0.89;在目标进入遮挡状态时,改进算法的跟踪精度最高;这次研究提出的STAPLE优化算法表现出较好的跟踪效果,其面对长时遮挡的跟踪能力具有较高的应用价值。     

Object tracking method based on data computing

The data computing process is utilized in various areas such as autonomous driving. Autonomous vehicles are intended to detect and track nearby moving objects avoiding collisions and to navigate in complex situations, such as heavy traffic and dense pedestrian areas. Therefore, object tracking is the core technology in the environment perception systems of autonomous vehicles and requires the monitoring of surrounding objects and the prediction of the moving states of objects in real time. In this paper, a multiple object tracking method based on light detection and ranging (LiDAR) data is proposed by using a Kalman filter and data computing process. We suppose that the movements of the tracking objects are captured consecutively as frames; thus, model-based detection and tracking of dynamic objects are possible. A Kalman filter is applied for predicting posterior state of tracking object based on anterior state of the tracking object. State denotes the positions, shapes, and sizes of objects. By computing the likelihood probability between predicted tracking objects and clusters which registered from tracking objects, the data association process of the tracking objects can be generated. Experimental results showed enhanced object tracking performance in a dynamic environment. The average matching probability of the tracking object was greater than 92.9%.

     

深度学习辅助的多行人跟踪算法

目的 目标的长距离跟踪一直是视频监控中最具挑战性的任务之一。现有的目标跟踪方法在存在遮挡、目标消失再出现等情况下往往会丢失目标,无法进行持续有效的跟踪。一方面目标消失后再次出现时,将其作为新的目标进行跟踪的做法显然不符合实际需求;另一方面,在跟踪过程中当相似的目标出现时,也很容易误导跟踪器把该相似对象当成跟踪目标,从而导致跟踪失败。为此,提出一种基于目标识别辅助的跟踪算法来解决这个问题。方法 将跟踪问题转化为寻找帧间检测到的目标之间对应关系问题,从而在目标消失再现后,采用深度学习网络实现有效的轨迹恢复,改善长距离跟踪效果,并在一定程度上避免相似目标的干扰。结果 通过在标准数据集上与同类算法进行对比实验,本文算法在目标受到遮挡、交叉运动、消失再现的情况下能够有效地恢复其跟踪轨迹,改善跟踪效果,从而可以对多个目标进行持续有效的跟踪。结论 本文创新性地提出了一种结合基于深度学习的目标识别辅助的跟踪算法,实验结果证明了该方法对遮挡重现后的目标能够有效的恢复跟踪轨迹,适用在监控视频中对多个目标进行持续跟踪。     

Dynamic control of a robot manipulator for high-performance conveyor tracking

A new robot control scheme for the specific application to conveyor tracking is presented. To improve the performance of conveyor tracking, the robot arm dynamics is incorporated into the control scheme. The tracking problem for the workpiece on a variable-speed conveyor is formulated as a stochastic optimal tracking problem with specified criteria. Dividing the conveyor speed into the nominal term and the perturbed term, a two-stage control strategy is employed to cope with the nonlinearity and uncertainty of the robot-conveyor system. Simulation results are given to verify the good tracking performance with fast cycle time and high accuracy obtained in a robotic workcell. © 1995 John Wiley & Sons, Inc.     

Stability analysis of nonlinear multi-agent relay tracking systems over a finite time interval

ABSTRACT

This paper explores the stability of a class of multi-agent systems with proposed relay tracking strategy over a finite time interval. In this paper, agents are deployed in an area to monitor and track the intruded targets. According to proposed relay tracking strategy, the tracking agents and their communication topologies switch during the whole tracking process. This results in impulsive effects on the overall tracking errors. The relationship of finite time interval against desired overall tracking errors and control parameter is derived quantitatively for the multi-agent relay tracking system. Moreover, stability conditions for the system with disturbances and impulsive effects are obtained. Performance of proposed relay tracking strategy and correctness of derived results on stability over finite time intervals are demonstrated by a set of simulations.     

噪声方差未知条件下的视频目标跟踪

目的 基于卡尔曼滤波的视频目标跟踪算法需要事先获得过程噪声和观测噪声方差,但在实际应用中,无法得知这两种噪声方差的准确值。此外,由于目标运动的随机性和视频场景中背景的复杂性,噪声方差也会随时间发生动态变化。如果设定的噪声方差不准确,跟踪精度会受影响,严重时会导致目标跟踪失败。考虑到上述问题,提出一种新的解决方法。方法 将带遗忘因子的推广递推最小二乘法(EFRLS)运用到视频目标跟踪研究领域。在该算法中,无需使用噪声方差,首先利用Mean Shift算法获得目标位置的初步估计,再利用EFRLS算法估计下一帧目标的位置。结果 该算法明显好于传统Mean Shift算法,并且与Kalman结合Mean Shift算法的跟踪性能相当。此外,在目标发生严重遮挡时,该算法优于Kalman结合Mean Shift算法,具有较好的跟踪性能。结论 本文算法无需设置噪声参数,可以实现目标在发生严重遮挡和遮挡后目标重新出现的情况下的准确跟踪,提高了跟踪的鲁棒性,具有一定的工程使用价值。     

结合连续卷积算子的自适应加权目标跟踪算法

目的 在视觉跟踪领域中,特征的高效表达是鲁棒跟踪的关键,观察到在相关滤波跟踪中,不同卷积层表达了目标的不同方面特征,提出了一种结合连续卷积算子的自适应加权目标跟踪算法。方法 针对目标定位不准确的问题,提出连续卷积算子方法,将离散的位置估计转换成连续位置估计,使得位置定位更加准确;利用不同卷积层的特征表达,提高跟踪效果。首先利用深度卷积网络结构提取多层卷积特征,通过计算相关卷积响应大小,决定在下一帧特征融合时各层特征所占的权重,凸显优势特征,然后使用从不同层训练得到的相关滤波器与提取得到的特征进行相关运算,得到最终的响应图,响应图中最大值所在的位置便是目标所在的位置和尺度。结果 与目前较流行的3种目标跟踪算法在目标跟踪基准数据库（OTB-2013）中的50组视频序列进行测试,本文算法平均跟踪成功率达到85.4%。结论 本文算法在光照变化、尺度变化、背景杂波、目标旋转、遮挡和复杂环境下的跟踪具有较高的鲁棒性。     

State unilateral tracking control of positive switched systems via designing a switching law

This paper focuses on the model reference state unilateral tracking control problem of positive switched systems. The state unilateral tracking means that the state always keeps in the same direction of the reference objective, either above or below. This feature makes unilateral tracking play a role of avoiding the overshoot. Via the dual design of both the controller and the switching signal, the state unilateral tracking control problem with L₁-gain performance for positive switched systems is solved through adopting multiple linear co-positive Lyapunov functions technique. The design method is applicable to the case that the unilateral tracking control problem is unsolvable for any subsystem of a positive switched system. Moreover, two techniques are presented in order to deal with the state unilateral tracking problem. Finally, a practical example demonstrates the validity of the presented result.