基于DSP的主动视觉运动目标跟踪策略及实现

设计了一种基于TMS320DM642的单目主动视觉运动目标自动跟踪系统,该系统采用了一种期望极大化(Expectation Maximization,EM)自适应窗口的运动目标跟踪方法.本方法以考虑了像素空间位置信息的混合高斯模型建立目标的灰度特征模板,然后通过EM算法迭代估计出使每帧似然函数最大化的分布参数值,这些参数不但可确定出跟踪目标在图像序列中的位置和形状尺寸,而且为单目摄像机的自动变焦和基于分区逻辑的摇摆运动提供了控制信息.实验表明:系统可以自动而稳定地跟踪具有复杂运动状态的目标,对320pixelsx240pixels的图像可实现平均约20 frame/s的跟踪速度.     

机载光电跟踪系统连续帧图像快速识别定位

针对机载光电跟踪系统连续帧图像快速识别、准确定位的要求,本文在深入分析连续图像序列特性和机载光电跟踪系统特点的基础上,结合精密转台的运动变化信息,提出了基于转角信息的连续帧图像处理算法.该方法根据精密转台的转角变化信息,采用连续的图像序列进行多帧间差值运算以获得目标残差图,残差图之间相乘.除去大部分随机噪声点,结合中值滤波和自适应渡门跟踪算法,大大的减小了运算量,实现了运动目标的快速跟踪定位.通过对目标实测图像序列的试验,结果表明,该算法具有快速、稳定等优点,能满足机载光电跟踪系统实时图像跟踪的要求.     

轴对称回转体空间位置与姿态定位的立体视觉方法

针对虚拟现实技术中对空间刚体位置和姿态的定位与跟踪的需要,提出了一种确定轴对称回转体空间位置与姿态的立体视觉方法,开发了一个基于面阵CCD双目视觉传感的被动式光学立体定位与跟踪系统。试验结果表明,在执行机构运动速度为100mm/s的情况下,跟踪到的空间目标运动路径与真实值的最大偏差只有0.3mm,姿态角与真实值的最大偏差只有1.4°。实验结果证实,该系统满足虚拟现实技术中对空间目标定位与跟踪的精度要求。     

一种快速运动目标检测与跟踪算法

运动目标的检测与跟踪是计算机视觉和图象编码研究的主要内容,有着广泛的应用领域。基于光流场的检测与跟踪是其常用的方法之一,但其计算结果和效率是不能令人满意的。提出了基于帧间差阈值法和光流场相结合的快速运动目标检测与跟踪算法,诸结果表明该算法简单实用、运算速度快,克服了单纯光流场方法的不足。     

基于在线学习的Siamese网络视觉跟踪算法

基于Siamese网络的视觉跟踪算法是近年来视觉跟踪领域的一类重要方法,其在跟踪速度和精度上都具有良好的性能.但是大多数基于Siamese网络的跟踪算法依赖离线训练模型,缺乏对跟踪器的在线更新.针对这一问题,本文提出了一种基于在线学习的Siamese网络视觉跟踪算法.该算法采用双模板思想,将第一帧中的目标当作静态模板,...     

基于帧间差分和运动估计的Camshift目标跟踪算法

针对Camshift算法无法适应目标的高速运动、复杂背景和遮挡的局限性,本文提出结合帧间差分法和运动估计对Camshift算法进行改进.首先在颜色概率分布图计算中通过帧间差分法加入目标运动信息,实现自动初始化跟踪并排除与目标相似背景颜色的干扰.之后在搜索窗传递过程中预测目标的位置,根据跟踪状态对搜索窗进行调整,以实现对高速运动目标的跟踪.实验表明新算法在目标高速运动、遮挡、和同色干扰情况下,仍能进行有效跟踪.     

快速运动目标的Mean shift跟踪算法

针对Mean shift本身的理论缺陷,提出Mean shift和卡尔曼滤波器相结合的快速目标跟踪算法。利用卡尔曼滤波器来获得每帧Mean shift算法的起始位置,然后再利用Mean shift算法得到跟踪位置。在目标出现大比例阻挡情况时,利用卡尔曼残差的计算来关闭和打开卡尔曼滤波器,此时,目标位置的线性预测替代了卡尔曼的作用。试验证明,本算法可以实现对快速运动目标的跟踪,对阻挡也有很好的鲁棒性。     

无人机地面目标跟踪系统的建模与控制

对无人机地面目标跟踪系统进行了分析,并建立了各个子系统的数学模型,针对跟踪过程中可能导致目标丢失的两个关键问题--无人机的飞行轨迹和云台摄像机的控制,给出了一种轨迹规划与跟踪算法,该算法既可以使无人机与目标保持一定距离,又保证了无人机在飞行过程中机头始终朝向目标.此外,设计了一种云台摄像机控制方法,该方法利用无人机与目标的相对位姿和目标偏离图像中心的偏差作为反馈信息对云台摄像机进行控制.仿真结果表明本文所设计的无人机地面目标跟踪系统具有良好的性能.     

序列图像中目标的分割与跟踪

本文主要研究序列图像中目标自适应跟踪方法.文章首先分析比较了目前国内外常用的分割方法的优缺点,然后提出了一种可以实现快速跟踪的算法.该方法利用序列连续图像的帧间相关性和差异性检测目标,把目标历史运动信息经过预测滤波后的结果和差值的序列图像信息进行比较,以达到快速跟踪的目的.     

基于STM32的目标提取处理系统

本文探索了一种监控图像跟踪系统,该系统以STM32微处理器为控制核心,将广角摄像头采集的图像送至LCD显示,并分析获取图像的数据,针对常用的直接帧间差分法存在的误差大,运算量大,导致内存问题,影响运算速度等问题,提出了一种图像目标提取处理算法,准确快速计算出目标物体的位置坐标,从而控制云台移动,保证目标处于摄像头视野的中央,解决了在图像处理中误差大、运算速度慢、系统内存不足等问题,以低成本低功耗实现了视频目标的提取跟踪处理.     

一种新的Mean-Shift对象跟踪方法

由于存在外点干扰,传统的Mean-Shift对象跟踪方法不能有效的跟踪尺寸逐步变大的目标.猜想利用对象初始模型能有效排除外点对跟踪结果的干扰,在此基础上本文提出了一种利用对象初始模型的候选目标模型和相似度测量方法;为了使内核窗口中心及大小与对象形心及大小一致,根据内核窗口外一定宽度子带内像素分布情况,动态调整内核窗口大小和位置.通过对尺寸逐渐变大的汽车进行跟踪,验证了算法的有效性,同时试验结果也证明新算法具有更低的计算复杂度.     

基于两种轨道线路建模方法的地铁车轮磨耗预测对比分析

轨道线路对车轮磨耗具有重要影响。建立地铁车轮磨耗预测模型，模型包括车辆动力学模型、基于Hertz理论和FASTSIM算法的轮轨局部接触模型、基于Tγ/A-磨损率函数的磨耗模型、平滑与更新策略以及轨道线路模型。对比分析等效轨道线路法与真实轨道线路法两种轨道线路建模方法在建模复杂度、车轮磨耗预测结果与计算效率方面的差异。利用实测的车轮磨耗对Tγ/A-磨损率函数进行修正，并对磨耗预测模型进行验证。结果表明:与等效轨道线路法相比，真实轨道线路法建模较为复杂，但其可以更好地模拟实际轨道线路；二者磨耗预测结果差异较小，且均与实测结果较为吻合，但前者计算效率约是后者的3.7倍。采用等效轨道线路法建模可以较好地兼顾计算精度与效率。     

Piezoelectric actuation and sensing mechanisms––closed form solutions

The purpose of this study is to investigate the actuation and sensing mechanism of a shear PZT embedded in a composite laminated beam. The actuator/sensor is either continuous or in the form of patches. The effects of actuator location and number of patches on actuator's performance are studied for various configurations of patches and boundary conditions under mechanical and/or electric loads. A comparison between the shear PZT and the traditional (extension, surface mounted type) PZT was preformed in order to get a better understanding.

A relatively simple, yet accurate model was developed and implemented. The analytical model is based on a first order shear deformation theory. The three-coupled equations of motion of a general non-symmetric piezolaminated composite beam subjected to axial and lateral tractions are derived and solved for the static case.

It seems that in the actuation mode, the surface mounted type patch has a great efficiency compared with the shear type. However, by changing the dimensions of the beam to a short and ‘stubby' one, the efficiency of the shear piezoceramics becomes larger than the efficiency of the surface mounted piezoceramics. For the sensing mode, the shear type piezoelectric patch might provide a larger output as compared to the surface mounted type one, for certain loading cases.     

Application of weight functions to the magnetic localization of anobject

A method to localize a magnetized object from the readings of simple array of magnetometers that is selected to the advantage of magnetic field measurement is applied. In the method the dipole field curve is fit to the readings by weighted least squares, where the weight function that is proportional is applied to the sixth power of the distance between the assumed location of the object and the point for measurement to ensure convergence of the iteration. By an experiment using a ship's magnetic model, it was ascertained that the convergence could be ensured if the initially approximated location is taken farther than the real one     

基于ABAQUS的钢-混凝土组合结构纤维梁模型的开发及应用

在大量已有研究成果的基础上,基于大型通用有限元软件ABAQUS提供的用户子程序UMAT开发了一组适用于组合结构构件的单轴材料滞回本构模型,并编制了可用于组合截面纤维自动离散的前处理程序,由此集成了钢-混凝土组合结构纤维梁模型,可用于完全剪力连接的组合框架体系在竖向荷载及水平往复荷载作用下的整体弹塑性分析。与国内外大量试验的对比结果表明,该文开发的钢-混凝土组合结构纤维梁模型具有较高的精度和较为广泛的适用性。纤维梁模型虽然不能考虑钢梁与混凝土板之间的滑移效应,但与试验结果的对比表明是否考虑滑移效应对结构体系整体计算结果的影响不大,但建模工作量大大减少,计算效率显著提高,在分析组合框架体系在地震往复荷载作用下的受力性能时具有较大的优势。     

Piezoelectric Active-Sensor Diagnostics and Validation Using Instantaneous Baseline Data

This paper presents a signal processing tool that efficiently performs piezoelectric (PZT) sensor diagnostic and validation. Validation of the sensor/actuator functionality during structural health monitoring (SHM) operation is a critical component to successfully implement a complete and robust SHM system, especially with an array of PZT active-sensors involved. The basis of this method is to track the capacitive value of PZT transducers, which manifests in the imaginary part of the measured electrical admittance. Both degradation of the mechanical/electrical properties of a PZT transducer and the bonding defects between a PZT patch and a host structure can be identified by the proposed process. However, it is found that the temperature variations in sensor boundary conditions manifest themselves in similar ways in the measured electrical admittances. Therefore, we examine the effects of temperature variation on the sensor diagnostic process and develop an efficient signal processing tool that enables the identification of a sensor validation feature that can be obtained instantaneously without relying on prestored baselines. This paper concludes with experimental results to demonstrate the effectiveness of the proposed technique.     

Tucker visual search-based hybrid tracking model and Fractional Kohonen Self-Organizing Map for anomaly localization and detection in surveillance videos

Anomaly detection (AD) in video is a challenging task employed in the intelligent video surveillance applications. This paper presents a technique for localizing and detecting anomalies in surveillance videos by proposing hybrid tracking model and Fractional Kohonen Self-Organizing Map (FKSOM). At first, the objects in the initial frames are detected by extracting the background and comparing with the succeeding frames. Then, a tracking model is developed to track the objects in the frame. Further, the features, such as object shape, speed, energy, correlation, and homogeneity, are extracted in the feature extraction process. Finally, the proposed FKSOM algorithm performs AD by identifying anomalous and normal events in the frame. The performance of the proposed technique is evaluated using the metrics, such as Multiple Object Tracking Precision (MOTP), accuracy, sensitivity, and specificity, where it obtains MOTP of 0.9895 with an average accuracy of 0.9339, the sensitivity of 0.9288 and specificity of 1.     

Variation of natural frequencies of beams using the active stiffening effect

The present study investigates the stiffening effects of a simply supported and clamped–free symmetric piezolaminated composite type beams. The structure consists of PZT layers or two sets of patches bonded to the surface of the beam. The analysis considers the linear piezoelectric constitutive relations and a first order shear deformation theory (FSDT). The influence of the actuators is evaluated by means of the pin-force model and their size and location along the beam are taken into account. Two coupled equations of motion for the lateral displacement and bending rotation and one uncoupled equation for the axial displacement of symmetric piezolaminated composite beam are solved numerically to obtain the natural frequencies and mode shapes. Numerical experiments demonstrate that the natural frequencies and mode shapes of the beam can be actively altered using the piezoelectric bonded actuators. The results of the present degenerated model are compared to results presented in the literature. The comparisons yielded excellent matches.     

基于传递矩阵法的车辆-轨道垂向耦合系统动力分析方法

为准确高效求解车辆-轨道耦合系统动力响应,利用轨道结构的周期性特征,基于传递矩阵法(transfer matrix method,TMM)提出了一种便捷的轨道子系统建模和求解方法.该方法根据轨道系统结构特点,分别将有砟轨道和CRTSⅡ型无砟轨道系统的周期性重复部分划分为不同轨道元胞结构,在元胞结合面引入刚度方程假定,基...     

Micromachined piezoelectric force sensors based on PZT thin films

A micromachined lead zirconate titanate (PZT) force sensor for scanning force microscope (SFM) is conceptualized by its piezoelectricity. The fabrication procedure is interpreted, and mechanical characteristics of the micromachined PZT force sensors with various lengths are studied in this paper. A compact SFM is constructed by using the piezoelectric PZT sensor. A very clear image is taken by this SFM. The current study of the micromachined PZT force sensor can be considered as a breakthrough of design of SFM as well as a good example of integrated piezoelectric microdevices