A novel particle filter implementation for a multiple-vehicle detection and tracking system using tail light segmentation

This paper proposes a vision-based multiple vehicle automatic detection and tracking system which can be applied in different environments. To detect vehicles, tail light position is utilized for fast vehicle candidate localization. A back propagation neural network (BPNN) trained by a Gabor feature set is used. BPNN verifies vehicle candidates and ensures detection system robustness. In the vehicle tracking step, to overcome multiple vehicle tracking challenges, partial vehicle occlusion and temporarily missing vehicle problems, this paper propose a novel method implementing a particle filter. The color probability distribution function (CPDF) of detected vehicles is used twice in the vehicle tracking sub-system. Firstly, CPDF is adopted to seek potential target vehicle locations; secondly, CPDF is used to measure the similarity of each particle for target vehicle position estimation. Because of various illuminations or target vehicle distances, the same vehicle will generate different CPDFs; the initial CPDF cannot guarantee long-term different scale vehicle tracking. To overcome these problems, an accurate tracking result, which is chosen by a trained BPNN, is used to update target vehicle CPDF. In our experiments, the proposed algorithm showed 84% accuracy in vehicle detection. Videos collected from highways, urban roads and campuses are tested in our system. The system performance makes it appropriate for real applications.     

基于时空结构张量的高速公路视频车辆检测

详细分析高速公路车辆运动模式与视频监控序列图像的特征,本文针对目标运动速度较大时,背景差分法运算复杂和连续帧差法容易产生虚影的情况,提出基于时空结构张量的光流分析法进行高速公路视频车辆检测。实验结果表明,该算法简洁并且能对高速公路视频序列中的运动车辆进行较准确的分割,为下一步的运动跟踪提供了可靠的依据。     

融合特征的卫星视频车辆单目标跟踪

目的 卫星视频作为新兴遥感数据,可以提供观测区域高分辨率的空间细节信息与丰富的时序变化信息,为交通监测与特定车辆目标跟踪等应用提供了不同于传统视频视角的信息。相较于传统视频数据,卫星视频中的车辆目标分辨率低、尺度小、包含的信息有限。因此,当目标边界不明、存在部分遮挡或者周边环境表观模糊时,现有的目标跟踪器往往存在严重的目标丢失问题。对此,本文提出一种基于特征融合的卫星视频车辆核相关跟踪方法。方法 对车辆目标使用原始像素和方向梯度直方图（histogram of oriented gradient,HOG）方法提取包含互补判别能力的特征,利用核相关目标跟踪器分别得到具备不变性和判别性的响应图;通过响应图融合的方式结合两种特征的互补信息,得到目标位置;使用响应分布指标（response distribution criterion,RDC）判断当前目标特征的稳定性,决定是否更新跟踪器的表征模型。本文使用的相关滤波方法具有计算量小且运算速度快的特点,具备跟踪多个车辆目标的拓展能力。结果 在8个卫星视频序列上与主流的6种相关滤波跟踪器进行比较,实验数据涵盖光照变化、快速转弯、部分遮挡、阴影干扰、道路颜色变化和相似目标临近等情况,使用准确率曲线和成功率曲线的曲线下面积（area under curve,AUC）对车辆跟踪的精度进行评价。结果表明,本文方法较好地均衡了使用不同特征的基础跟踪器（性能排名第2）的判别能力,准确率曲线AUC提高了2.9%,成功率曲线AUC下降了4.1%,成功跟踪车辆目标,不发生丢失,证明了本文方法的先进性和有效性。结论 本文提出的特征融合的卫星视频车辆核相关跟踪方法,均衡了不同特征提取器的互补信息,较好解决了卫星视频中车辆目标信息不足导致的目标丢失问题,提升了精度。     

基于概率算法自适应更新背景的运动车辆检测

交通流量检测是智能交通系统中的一个重要研究方向和热点问题,基于视频的车辆检测是交通流量采集分析的核心技术,它为交通流量参数的实时获取提供了可能。为实现在复杂交通视频场景中实时准确检测各类的运动车辆,在研究传统背景差分算法的缺点的工作基础上,提出一个自适应的贝叶斯概率背景检测算法,进而完成了较准确的运动车辆分类检测。实验结果表明该方法具有高效实时的特点,能够较准确地实现复杂交通路面的背景提取和运动车辆的检测,具有良好的鲁棒性。     

短时记忆与CenterTrack的车辆多目标跟踪

目的 车辆多目标跟踪是智能交通领域关键技术,其性能对车辆轨迹分析和异常行为鉴别有显著影响。然而,车辆多目标跟踪常受外部光照、道路环境因素影响,车辆远近尺度变化以及相互遮挡等干扰,导致远处车辆漏检或车辆身份切换（ID switch,IDs）问题。本文提出短时记忆与CenterTrack的车辆多目标跟踪,提升车辆多目标跟踪准确度（multiple object tracking accuracy,MOTA）,改善算法的适应性。方法 利用小样本扩增增加远处小目标车辆训练样本数;通过增加的样本重新训练CenterTrack确定车辆位置及车辆在相邻帧之间的中心位移量;当待关联轨迹与检测目标匹配失败时通过轨迹运动信息预测将来的位置;利用短时记忆将待关联轨迹按丢失时间长短分级与待匹配检测关联以减少跟踪车辆IDs。结果 在交通监控车辆多目标跟踪数据集UA-DETRAC （University at Albany detection and tracking）构建的5个测试序列数据中,本文方法在维持CenterTrack优势的同时,对其表现不佳的场景获得近30%的提升,与YOLOv4-DeepSort（you only look once—simple online and realtime tracking with deep association metric）相比,4种场景均获得近10%的提升,效果显著。Sherbrooke数据集的测试结果,本文方法同样获得了性能提升。结论 本文扩增了远处小目标车辆训练样本,缓解了远处小目标与近处大目标存在的样本不均衡,提高了算法对远处小目标车辆的检测能力,同时短时记忆维持关联失败的轨迹运动信息并分级匹配检测目标,降低了算法对跟踪车辆的IDs,综合提高了MOTA。     

一种基于灰度直方图的交通检测系统

在中国由于行人和机动车混行的情况非常普遍,因此行人的检测和机动车一样也是交通检测的重要方面。然而国内外在行人与机动车混合交通流检测方面的研究却不多见。文章提出并阐述了一套基于视频的综合交通检测系统。该系统通过视频技术实时检测获得路段或路口上机动车、行人的综合交通数据。系统运用了实用性很强自适应背景差减法进行前景图像分割,并提出了基于灰度直方图的目标跟踪方法,在一定程度上解决了在人车混行中比较严重的遮挡问题。实验结果表明,系统在前景图像提取以及目标跟踪分类等环节都取得了令人满意的成果。     

Moving vehicles detection based on adaptive motion histogram

As one of the important topics in computer vision, moving vehicle segmentation has attracted considerable attention of researchers. However, robust detection is hampered by the interferential moving objects in dynamic scenes. In this paper, we address the problem of the moving vehicles segmentation in the dynamic scenes. Based on the distinct motion property of the dynamic background and that of the moving vehicles, we present an adaptive motion histogram for moving vehicles segmentation. The presented algorithm consists of two procedures: adaptive background update and motion histogram-based vehicles segmentation. In the adaptive background update procedure, we make use of the lighting change of the scene and present a novel method for background evolving. In the motion histogram-based vehicles segmentation procedure, an adaptive motion histogram is maintained and updated according to the motion information in the scenes, and the moving vehicles are then detected according to the motion histogram maintained. Experimental results of typical scenes demonstrate robustness of the proposed method. Quantitative evaluation and comparison with the existing methods show that the proposed method provides much improved results.     

视频中运动目标的实时检测和跟踪

基于视频的自动目标检测和跟踪是计算机视觉中一个重要的研究领域，特别是基于视频的智能车辆监控系统中的运动车辆的检测和跟踪。提出了一种自适应的背景相减法来分割运动物体，为了准确地定位运动车辆的区域，采用差分图像投影和边缘投影相结合的方法来定位车体，同时利用双向加权联合图匹配方法对运动车辆区域进行跟踪，即将对运动车辆区域跟踪问题转化为搜索具有最大权的联合图的问题。该算法不仅能实时地定位和跟踪直道上运动的车辆，同时也能实时地定位和跟踪弯道上运动的车辆，从实验结果看，提出的背景更新算法简单，并且运动车辆区域的定位具有很好的鲁棒性，从统计的检测率和运行时间来看，该算法具有很好的检测效果，同时也能满足基于视频的智能交通监控系统的需要。     

背景估计与运动目标检测跟踪

基于视频的自动目标检测和跟踪是计算机视觉中一个重要的研究领域,特别是基于视频的智能车辆监控系统中的运动车辆的检测和跟踪。提出一种自适应的背景估计方法来实时获得当前背景图像,从而分割出运动物体。为了准确地定位运动车辆的区域,采用差分图像投影和边缘投影相结合的方法来定位车体,同时利用双向加权联合图匹配方法对运动车辆区域进行跟踪,即将对运动车辆区域跟踪问题转化为搜索具有最大权的联合图的问题。该算法不仅能实时地定位和跟踪直道上运动的车辆,同时也能实时地定位和跟踪弯道上运动的车辆,从实验结果看,提出的背景更新算法简单,并且运动车辆区域的定位具有很好的鲁棒性,从统计的检测率和运行时间来看,该算法具有很好的检测效果,同时也能满足基于视频的智能交通监控系统的需要。     

基于仿射特征描述的鲁棒车辆遮挡检测方法

在交通场景下进行多目标跟踪时,往往会存在车辆之间的相互遮挡,使得基于视觉的车流量计算准确度受到严重影响。针对这一问题,提出一种基于仿射特征描述的鲁棒车辆遮挡检测方法,该方法利用车辆的角点特征,建立基于仿射变换的车辆特征描述模型,判断车辆之间是否存在遮挡现象,再对遮挡车辆进行特殊分割,保证了车辆跟踪以及车流量统计的准确性。通过50段5分钟连续视频,每个视频中出现车辆遮挡时段大于1分钟,遮挡频率25%,遮挡帧数大于500幅进行试验,算法的车流量统计的正确性达到95%,正确跟踪率达到98%,验证了该算法的有效性。     

An efficient and flexible approach for multiple vehicle tracking in the aerial video sequence

ABSTRACT

Multiple vehicle tracking (MVT) in the aerial video sequence can provide useful information for the applications such as traffic flow analysis. It is challenging due to the high requirement for the tracking efficiency and variable number of the vehicles. Furthermore, it is particularly challenging if the vehicles are occluded by the shadow of the trees, buildings, and large vehicles. In this article, an efficient and flexible MVT approach in the aerial video sequence is put forward. First, as the pre-step to approach the MVT problem, the superpixel segmentation-based multiple vehicle detection (MVD) is achieved by combining the two-frame difference and superpixel segmentation. The two-frame difference is utilized to reduce the search space. By scanning the search space via the centre of the superpixels, the moving vehicles are detected efficiently. Then, the deterministic data association is proposed to tackle the MVT problem. To improve the tracking accuracy, we fuse multiple types of features to establish the cost function. With respect to the variable number of the vehicles, the tracker management is designed by establishing or deleting the trackers. Furthermore, for the occlusion handling, we focus on the accurate state estimation, and it is realized by the driver behaviour-based Kalman filter (DBKF) method. In the DBKF method, we take seriously into account the driver behaviour, including the speed limit and rear-end collision avoidance with the front vehicle. Both tracker management and occlusion handling make the MVT approach flexibly cope with varieties of traffic scenes. Finally, comprehensive experiments on the DARPA VIVID data set and KIT AIS data set demonstrate that the proposed MVT algorithm can generate satisfactory and superior results.     

Multiple objects tracking in the presence of long-term occlusions

We present a robust object tracking algorithm that handles spatially extended and temporally long object occlusions. The proposed approach is based on the concept of “object permanence” which suggests that a totally occluded object will re-emerge near its occluder. The proposed method does not require prior training to account for differences in the shape, size, color or motion of the objects to be tracked. Instead, the method automatically and dynamically builds appropriate object representations that enable robust and effective tracking and occlusion reasoning. The proposed approach has been evaluated on several image sequences showing either complex object manipulation tasks or human activity in the context of surveillance applications. Experimental results demonstrate that the developed tracker is capable of handling several challenging situations, where the labels of objects are correctly identified and maintained over time, despite the complex interactions among the tracked objects that lead to several layers of occlusions.     

On-road vehicle detection: a review

Developing on-board automotive driver assistance systems aiming to alert drivers about driving environments, and possible collision with other vehicles has attracted a lot of attention lately. In these systems, robust and reliable vehicle detection is a critical step. This paper presents a review of recent vision-based on-road vehicle detection systems. Our focus is on systems where the camera is mounted on the vehicle rather than being fixed such as in traffic/driveway monitoring systems. First, we discuss the problem of on-road vehicle detection using optical sensors followed by a brief review of intelligent vehicle research worldwide. Then, we discuss active and passive sensors to set the stage for vision-based vehicle detection. Methods aiming to quickly hypothesize the location of vehicles in an image as well as to verify the hypothesized locations are reviewed next. Integrating detection with tracking is also reviewed to illustrate the benefits of exploiting temporal continuity for vehicle detection. Finally, we present a critical overview of the methods discussed, we assess their potential for future deployment, and we present directions for future research.     

Robust ride comfort control of vehicles without measurements of tire deflection

In this article, a robust ride comfort control scheme for vehicles is proposed in which measurements of the tire deflections are not required. The controller has the property that we can specify a location where the ride comfort will be best. To achieve this end, an estimator for the tire deflections and the road disturbances is proposed. Next, a combined ideal vehicle is designed. In the ideal vehicle, the location where ride comfort will be best can be moved by setting only one design parameter. Finally, a robust tracking controller is developed so that a real vehicle tracks the motion of the combined ideal vehicle.     

基于光流和压缩感知的目标跟踪

由于姿态变化,光照改变,遮挡以及运动模糊等因素,开发有效和实时的目标跟踪算法是一个有挑战性的任务。因此,在复杂的实际环境中使用单一的检测与跟踪算法已无法满足目标跟踪的实用性要求。本文基于光流法和压缩跟踪算法,提出一种融合跟踪、在线学习以及检测技术的目标跟踪方法。实验结果表明,该方法提高了跟踪精度,而且具有较强的鲁棒性。     

以颜色和形状直方图为线索的粒子滤波人脸跟踪

跟踪器的设计和跟踪线索的选择与表达是人脸跟踪中的两大关键因素,针对一般人脸跟踪算法中常用简单椭圆来描述人脸形状线索时易受背景干扰的缺点,以及视频目标跟踪中动态模型和观测模型的非线性非高斯特点,提出了一种以颜色和形状直方图为线索的粒子滤波人脸跟踪算法,该算法在粒子滤波基本框架之下,引入了一种新的用直方图来描述人脸形状的方法,并对其进行了改进,用来作为人脸跟踪的形状线索。同时,为了减轻背景干扰,提出了一种经验有效边缘的检测方法。实验表明,该跟踪方法不仅能有效地处理人脸旋转、背景中的肤色干扰和部分遮掩问题,并且能够在由于大面积遮掩等原因而丢失目标的情况下,及时有效地重新捕获已丢失的目标。     

运动像机下基于粒子滤波的多目标跟踪算法

针对摄像机运动情况下的多目标跟踪问题,提出了基于粒子滤波的跟踪算法.在粒子滤波算法基础上,将二阶自回归过程作为系统状态转移模型,HSV颜色直方图作为观测模型,对视频中多个目标的位置、大小进行跟踪.实验结果表明,该算法能实时正确地跟踪多个目标,并对局部遮挡有较好的鲁棒性,也能在目标短暂消失导致跟踪失败后,在目标重新出现后及时捕获并继续进行跟踪.     

基于透视点和区块特征的多车粘连分割方法

针对交通视频车辆检测与跟踪中经常出现的车辆前景粘连问题,提出了一种利用透视点在图像轮廓上搜索车辆分割点并通过区块特征识别车辆的粘连车辆分割方法.根据路面场景中车道线信息提取出透视点和车道区域,结合混合高斯模型与形态学梯度轮廓算法提取出车辆前景团块.利用透视点原理从前往后顺序分割粘连的车辆前景团块.对分割开的待检定区域,利用车辆区块特征进行检验识别,修正错误分割,将粘连的多辆车逐一分割.实验结果表明,该方法对直线道路上多辆车粘连的分割有较好的准确度和适应性.     

基于Radon变换的视频测速算法

为了从视频监测图像中自动提取车辆行驶速度,提出了一种基于Radon变换的视频测速算法。根据车辆行驶轨迹的时空特点,利用道路交通标线为图像与道路建立距离映射关系,简化了现场标定的条件和过程;利用时间堆栈成像方法,建立了车辆行驶轨迹的时空关系;基于Radon变换的图像处理,实现了行驶车辆的视频测速技术。算法具有较好的鲁棒性,与传统方法相比,计算复杂性也要低很多。