Multi-target tracking with PHD filter using Doppler-only measurements

In this paper, we address the problem of multi-target detection and tracking over a network of separately located Doppler-shift measuring sensors. For this challenging problem, we propose to use the probability hypothesis density (PHD) filter and present two implementations of the PHD filter, namely the sequential Monte Carlo PHD (SMC-PHD) and the Gaussian mixture PHD (GM-PHD) filters. Performances of both filters are carefully studied and compared for the considered challenging tracking problem. Simulation results show that both PHD filter implementations successfully track multiple targets using only Doppler shift measurements. Moreover, as a proof-of-concept, an experimental setup consisting of a network of microphones and a loudspeaker was prepared. Experimental study results reveal that it is possible to track multiple ground targets using acoustic Doppler shift measurements in a passive multi-static scenario. We observed that the GM-PHD is more effective, efficient and easy to implement than the SMC-PHD filter.     

HELOP: Multi-target tracking based on heuristic empirical learning algorithm and occlusion processing

Multi-target tracking is one of the important fields in computer vision, which aims to solve the problem of matching and correlating targets between adjacent frames. In this paper, we propose a fine-grained track recoverable (FGTR) matching strategy and a heuristic empirical learning (HEL) algorithm. The FGTR matching strategy divides the detected targets into two different sets according to the distance between them, and adopts different matching strategies respectively, in order to reduce false matching, we evaluated the trust degree of the target’s appearance feature information and location feature information, adjusted the proportion of the two reasonably, and improved the accuracy of target matching. In order to solve the problem of trajectory drift caused by the cumulative increase of Kalman filter error during the occlusion process, the HEL algorithm predicts the position information of the target in the next few frames based on the effective information of other previous target trajectories and the motion characteristics of related targets. Make the predicted trajectory closer to the real trajectory. Our proposed method is tested on MOT16 and MOT17, and the experimental results verify the effectiveness of each module, which can effectively solve the occlusion problem and make the tracking more accurate and stable.     

基于Markov随机场理论的鼠脑切片显微图像的分割研究

鼠脑中的神经细胞是生物学家的一个重要研究对象.随着计算机视觉技术的飞速进步,研究者们利用图像分割技术从鼠脑切片显微图像自动提取细胞,为进一步分析提供便利.文中提出一种基于马尔可夫随机场理论的鼠脑切片细胞分割算法.相对于传统的算法,文中创新是利用已有的专家标记图和原始图像的灰度特征,结合期望最大化算法,初步估计高斯混合模型的参数,作为条件迭代模式算法的初始值,不仅提高分割精度,且减少迭代次数;并将像素的灰度特征和像素间的距离加入到传统的Potts随机场模型中,更加合理地描述像素间的定量关系.实验结果表明,与传统方法相比,此方法具有较高的计算效率和分割精度.     

基于边缘约束局部区域MRF的图像分割方法

针对常规马尔科夫随机场(MRF)模型对复杂自然图像分割时,存在对噪声敏感且边缘模糊的问题,构建一种基于边缘约束局部区域MRF(ECLRMRF)的图像分割模型.利用欧氏距离度量局部区域内邻接像素的相似度,依据其相似度构建局部空间来约束高斯混合模型,有效描述丰富的局部区域统计特征,并建立MRF模型的局部区域一致性约束项.利...     

混合高斯模型与CamShift算法结合的台球跟踪方法

传统的CamShift跟踪算法是基于颜色信息的算法,这种算法在背景与目标颜色相近时会出现跟踪不准确的情况,而且不能自动跟踪.针对这两个缺点,本文结合混合高斯模型来对这一跟踪算法进行改进,通过混合高斯模型来实现对运动目标的检测得到目标的轮廓,将图像处理中的形态学变换应用到目标轮廓检测过程中,可以得到更好的检测结果和目标轮廓.对检测的目标再利用CamShift算法进行跟踪,在程序中引入Kalman预测算法,对目标的位置进行估计,提高跟踪的准确性,实验结果表明了这种方法的有效性,实现了对多目标的自动跟踪.     

一种基于改进粒子滤波的多目标跟踪算法

针对复杂背景环境下的多目标跟踪问题,论述了主要的数据关联技术,将目标检测算法与粒子滤波相结合,利用颜色直方图作为观测模型,并利用全领域(GNN)算法进行数据关联.提出一种改进的基于粒子滤波的多目标跟踪算法,实现了视频场景中的多个目标跟踪.该算法对于目标在场景中的频繁出现和消失、相似外表、交叉运动和短暂遮挡等均有较好的处理效果.     

A Gaussian mixture Bernoulli filter for extended target tracking with application to an ultra-wideband localization system

This paper presents a Gaussian mixture (GM) implementation of the Bernoulli filter for extended target tracking, which we call the extended target GM Bernoulli (ET-GM-Ber) filter. Closed form expressions for the ET-GM-Ber filter recursions are obtained. A clustering step is integrated into the measurement update stage in order to have a computationally tractable filter. Performance of the proposed filter is tested both on the simulated data and experimental data collected using an ultra-wideband (UWB) localization system. Simulations and experimental results demonstrate the accurate and effective performance of the proposed filter.     

An automatic statistical segmentation algorithm for extraction of fire and smoke regions

Estimation of the extent and spread of wildland fires is an important application of high spatial resolution multispectral images. This work addresses a fuzzy segmentation algorithm to map fire extent, active fire front, hot burn scar, and smoke regions based on a statistical model. The fuzzy results are useful data sources for integrated fire behavior and propagation models built using Dynamic Data Driven Applications Systems (DDDAS) concepts that use data assimilation techniques which require error estimates or probabilities for the data parameters. The Hidden Markov Random Field (HMRF) model has been used widely in image segmentation, but it is assumed that each pixel has a particular class label belonging to a prescribed finite set. The mixed pixel problem can be addressed by modeling the fuzzy membership process as a continuous Multivariate Gaussian Markov Random Field. Techniques for estimating the class membership and model parameters are discussed. Experimental results obtained by applying this technique to two Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) images show that the proposed methodology is robust with regard to noise and variation in fire characteristics as well as background. The segmentation results of our algorithm are compared with the results of a K-means algorithm, an Expectation Maximization (EM) algorithm (which is very similar to the Fuzzy C-Means Clustering algorithm with entropy regularization), and an MRF-MAP algorithm. Our fuzzy algorithm achieves more consistent segmentation results than the comparison algorithms for these test images with the added advantage of simultaneously providing a proportion or error map needed for the data assimilation problem.     

基于层次Gamma混合模型的高分辨率SAR影像分割方法

为了准确地对高分辨率合成孔径雷达(SAR)影像内像素强度统计分布建模并得到高精度的分割结果,提出基于层次Gamma混合模型(HGaMM)的高分辨率SAR影像分割方法.HGaMM由多个Gamma混合模型构成,用于对非对称、重尾和多峰等复杂的像素强度统计分布建模.为了减少影像噪声对分割的影响,采用马尔科夫随机场建模像素标号场,将像素邻域位置关系引入HGaMM.根据贝叶斯理论,利用后验分布构建影像分割模型.马尔科夫链蒙特卡罗算法用于模拟影像分割模型.在模拟和真实SAR影像上的分割实验表明,文中方法可得到较高精度的分割结果.     

基于得分域多维特征分类器的声纹密码系统

针对不同类型数据对目标发音人区分能力不同的现象,在传统系统基础上提出利用UBM模型对测试数据进行分类,使用分类后的似然比得分形成多维特征,在此基础上利用SVM分类器进行声纹密码确认。该方法把传统的似然比检验策略转换成多维特征空间上的二类分类问题。测试与注册数据同信道情况时,在4种手机数据集上,文中系统相对文本相关GMM-UBM声纹密码系统等错误率分别下降41。25%、33。33%、37。49%和26。03%,在交叉信道上系统性能也获得改善。     

基于形状感应的运动目标跟踪算法

Mean-Shift算法无法自动跟踪目标,且对目标形状要求较苛刻。针对该问题,提出一种基于形状感应的运动目标跟踪算法,采用混合高斯分布对背景建模,协助Mean-Shift算法自动定位初始目标,增加描述形状的协方差参数,使跟踪能感受到目标形状的变化。实验结果表明,该算法基本解决了自动定位问题及形状变化问题,在保证实时性的前提下,跟踪准确度提高40%以上。     

高斯混合模型结合加权似然的目标跟踪算法

鉴于高斯混合模型对背景变化快时无法精确检测出目标和目标跟踪的适应性差等瑕疵，提出了基于加权似然跟踪器来改进高斯混合模型实现运动目标跟踪算法。主要引入了自适应高斯混合模型来实时检测运动目标，然后空间加权似然来进行视频中的目标定位，引入加权似然期望值来改进高斯混合模型处理视频中的多尺度、多角度变化的目标跟踪不精准问题。通过VOT 2014 dataset对比实验结果表明提出的基于加权似然跟踪（Weighted Likelihood Tracking，WLT）和改进高斯混合模型（Improved Gaussian Mixture Model，IGMM）的目标跟踪算法较传统高斯混合模型跟踪算法在跟踪的精度有较大提高。在应对多尺度、多角度变化的目标跟踪表现出了较大的优势。     

一种模糊算法及其在多假设多目标跟踪中的应用

研究了基于模糊交互式多模型算法解决多假设多目标跟踪中的新目标数据相关问题。使用模糊交互式多模型算法处理新目标的数据相关问题,能够提高跟踪精度、时间效率。用隶属度表示每个模型与所跟踪新目标的隶属关系,根据隶属度选择最佳的模型对新目标进行跟踪。仿真结果表明,该方法不仅能够处理新目标的数据相关,同时又能够提高算法的时间效率。     

基于改进高斯混合模型的前景检测

针对自适应混合高斯背景模型执行速度慢、检测前景时容易产生“鬼影”等问题,提出一种改进的混合高斯背景建模方法。该方法通过对高斯分布权值和生存时间的限制,建立高斯分布退出机制,使模型能根据场景自适应选择每个像素的高斯分布个数,从而去除多余高斯分布,加快算法执行速度。在模型更新过程中,通过融入帧间差分,将每帧图像分成运动像素、背景像素及非真实运动像素,并通过对非真实运动像素赋予较大学习率来加速移出背景的恢复,从而避免“鬼影”和拖影现象。实验结果表明,与传统检测方法相比,该方法可以获得更好的目标检测效果。     

基于视频监控的人脸检测跟踪识别系统研究

设计了一种基于视频监控的人脸检测跟踪识别系统,该系统的功能是检测并实时跟踪视频中的人脸图像,同时进行身份识别。针对Gentle AdaBoost算法构造的级联分类器检测效率偏低的问题,提出了一种递进复杂度的级联分类器。针对传统粒子滤波器最高权重粒子不准确的问题,提出了均值权重粒子滤波器。针对传统粒子滤波器样本衰退的问题,提出了一种同时结合人脸检测和人脸跟踪算法的跟踪校正策略。对于检测和跟踪到的人脸,利用基于Gabor变换和HMM的方法进行身份识别。实验结果表明,系统能够准确地检测并实时跟踪视频中的人脸,可以实现人脸的快速识别,是一种能够应用到视频监控系统中的有效方法。     

基于信息分享机制的粒子滤波算法及其在视觉跟踪中的应用

针对基本粒子滤波方法存在的权值退化和计算效率低问题,提出了一种基于信息分享机制的粒子滤波算法.该方法将粒子群优化算法和蚁群优化算法的优化思想共同作用到粒子更新中,实现粒子之间信息共享,从而增强粒子的多样性和最优估计能力.同时分析了该算法的收敛性.视觉跟踪实验表明,该算法能用较少的粒子实现单目标和多目标跟踪,综合跟踪性能优于基本粒子滤波和基于粒子群优化的粒子滤波方法,验证了本算法的有效性.     

基于OpenCV的摄像头动态手势轨迹识别及其应用

利用OpenCV计算机视觉库在vs2008平台上设计了一个基于实时摄像头的集动态手势检测、动态手势跟踪、动态手势轨迹识别的应用.首先,该应用基于静止的背景更新,利用背景差分检测运动手势,再结合颜色直方图的粒子滤波进行动态手势跟踪,最后利用隐马尔可夫模型(HMM)进行运动轨迹识别.在运动检测部分结合了背景差分图与通过颜色直方图获得的反投影图,达到比较满意的实时运动检测效果;在运动手势跟踪部分,改进的颜色直方图的粒子跟踪能够在经过类肤色人脸的干扰后迅速地找回运动手势,基本达到了跟踪的要求,但是同时对于HMM识别轨迹时需要的运动轨迹序列采集造成了影响;在识别轨迹部分,HMM的训练达到了识别的要求,但是识别的效果主要取决于实时运动轨迹序列的采集工作与采集方法的优化.     

基于双层背景的遗弃物检测方法

提出一种基于双层背景的遗弃物检测方法。分别采用滑动平均算法和改进的高斯混合模型,对参考背景和动态背景进行建模,通过2个背景得到前景间的差异,以此提取静止前景,对检测到的静止物体进行直方图匹配以消除鬼影,在前景检测的基础上引入均值漂移算法和粒子滤波算法,处理物体间遮挡问题。实验结果证明了该方法的有效性。     

一种基于分裂高斯混合模型的背景建模算法

针对运动检测算法中目标空洞和环境噪声难以消除等问题,提出一种背景建模算法。将待处理图像通过三层高斯模糊滤镜,分层抽取图像信息,通过分层建立分裂高斯混合模型,进行建模并计算运动区域。将上层提取的低频信息以及建模后提取的粗块化运动目标,加入到下层的背景判定计算流程中,根据综合判定结果纠正下层错误的模型参数。在公共数据库中的实验结果表明,该算法在高效地去除了环境噪声的情况下,可保证所提取运动目标的准确性,并且对光照突变不敏感,有较好的检测率和较低的误检率。     

Joint latent class model of survival and longitudinal data: An application to CPCRA study

There has been an increasing interest in the joint analysis of repeated measures and time to event data. In many studies, there could also exist heterogeneous subgroups. Thus a new model is proposed for the joint analysis of longitudinal and survival data with underlying subpopulations identified by latent class model. Within each latent class, a joint model of longitudinal and survival data with shared random effects is adopted. The proposed model is applied to Terry Beirn Community Programs for Clinical Research on AIDS study (CPCRA) to characterize the underlying heterogeneity of the cohort and to study the relation between longitudinal CD4 measures and time to death. The proposed model is desirable when the heterogeneity among subjects cannot be ignored and both the longitudinal and survival outcomes are of interest.