基于多线索融合的目标跟踪算法研究

目标跟踪是计算机视觉领域一项核心技术,通用的目标跟踪算法以单线索或简单融合多个线索为主,当背影突然改变时,单一线索或多个线索的简单融合便难以跟踪成功.针对这个问题,在粒子滤波的框架下提出了自适应融合颜色线索和角点线索的跟踪方法,通过判断各个线索的可信程度,自适应给不同线索分配不同的权重,很好地解决了在复杂背景、互相遮挡情况下的跟踪问题.实验证明,采用的自适应多线索融合方法在实际应用中有更强的鲁棒性.     

随机场中运动一致性的多线索目标跟踪

目的 通过建立各线索间的关联,提高多线索目标跟踪方法的鲁棒性,利用简单而有效的模型使多线索目标跟踪方法的表达和实现变得容易.方法 在不同线索描述下的目标对象间引入运动一致性约束,利用链状结构随机场模型表达不同线索描述下的目标对象及其约束关系,将多线索目标跟踪问题转化为随机场目标函数的简单优化求解.实验中结合亮度直方图、方向梯度直方图和局部二进制模式描述目标对象.结果 15组公测视频序列上的实验结果表明,所提方法相对于多种优秀的目标跟踪方法,在目标受到遮挡、运动模糊、光照变化、背景杂乱等因素干扰时,获得了较低中心位置误差和较高的精度值,反映了所提方法的有效性.结论 运动一致性约束能够较好地增强各线索间的关联,通过链状结构的随机场模型表达该约束关系和各线索描述下的目标对象,在提高跟踪鲁棒性的同时,使跟踪方法的实现变得简单.     

基于子空间联合模型的视觉跟踪

目标跟踪是计算机视觉的重要组成部分,其鲁棒性一直受到目标遮挡,光照变化,目标姿态变化等因素的制约。针对这个问题,提出了基于子空间联合模型的视觉跟踪算法。算法为了克服遮挡对目标跟踪的影响,采用局部动态稀疏表示进行遮挡检测,根据遮挡检测结果来修正增量子空间误差。此外,在稀疏子空间基础上计算目标模板和候选模板的相似性。在粒子滤波框架下,联合候选目标增量误差和相似性实现目标跟踪。通过在多个具有挑战性的视频序列上进行实验,表明该算法具有较好的鲁棒性。     

改进权值计算的均值移动目标跟踪

针对基于Bhattacharyya相似度的均值移动跟踪算法精度较差的问题,提出一种基于直方图交集思想的新型颜色分量加权方法,该方法利用参考模板与候选模板归一化颜色概率密度对应颜色分量的比值作为均值移动算法的加权系数。新权值计算方法在目标快速运动,有场景相似颜色干扰等情况下具有很好的适用能力,从而提高目标的跟踪精度。另外处理跟踪过程中,因摄像机抖动、光照变化等因素导致跟踪线索变化的情况,利用基于辅助模板的目标更新机制,有效地解决了目标短暂遮挡以及更新过程中的累积误差问题。通过多组对比实验结果可以看出,算法具有更强地抑制背景干扰以及特征自适应的能力,从而提高了均值移动跟踪算法的鲁棒性。     

基于稀疏子空间选择的在线目标跟踪

本文在粒子滤波框架下提出一种基于稀疏子空间选择的两步在线跟踪方法.在跟踪的第一步,利用稀疏子空间选择算法筛选出与目标状态相似性较高的候选区域,并将目标与背景间的过渡区域定义为单独的类别以降低目标发生漂移的可能;第二步则通过构建有效的观测模型计算候选区域与目标状态间的相似性,其中相似性函数综合考虑二者在整体和局部特征上的相似性,且将目标的原始状态和当前状态都作为参考,因此增强了观测模型的可靠性;最后利用最大后验概率估计目标状态.此外,该算法通过对目标数据的更新来适应目标的表观变化.实验结果表明该算法能有效处理目标跟踪中的遮挡、运动模糊、光流与尺度变化等问题,与当前流行的9种跟踪方法在多个测试视频上的对比结果验证了该算法的有效性.     

基于梯度方向直方图特征的多核跟踪

提出了基于梯度方向直方图特征的多核跟踪算法, 对跟踪过程中的光线变化和部分遮挡具有较强的鲁棒性. 该算法将目标分块, 分别提取出每块的核函数加权的梯度方向直方图特征. 目标模型和候选目标模型的相似度用所有块直方图间的Bhattacharyya系数之和进行度量, 目标的跟踪通过Mean shift算法最大化两者的相似度实现. 对车辆、人体等多个目标的跟踪验证了本文提出算法的有效性.     

自适应分块多线索融合粒子滤波跟踪方法

多线索融合是解决复杂情况下跟踪问题的有效手段, 为此提出一种基于自适应分块目标模型的多线索融合 粒子滤波跟踪方法. 根据目标颜色分布自适应分块建立目标描述模型, 可提高对目标初始描述的适应性; 采用多线索融合粒子滤波跟踪, 在跟踪过程中能根据子块可靠程度动态调整权重, 提高对剧烈光照变化、目标姿态变化、遮挡等复杂情况的适应性. 实验结果表明, 所提出的跟踪方法在多种复杂情况下能准确有效地跟踪目标.

     

基于稀疏表示的目标跟踪方法

提出了一个新的基于稀疏表示的目标跟踪方法.在粒子滤波框架下,将目标模板线性表示为所有目标候选的线性组合.当假设目标候选中存在与目标模板相似的候选时,线性表示的系数满足稀疏性约束,可以通过(L)1范式最小化求解.每一个目标候选在线性表示中的系数反映了该候选与目标模板的相似程度,因此可以将系数作为目标候选的权重.目标跟踪的结果为权重最大的候选.实验结果表明本文提出的算法比文献中现有的基于(L)1范式最小化的跟踪方法性能更稳定、计算效率更高.     

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。     

基于Mean-shift的改进目标跟踪算法

传统的Mean-shift目标跟踪算法对背景因素比较敏感,采用核加权直方图的方法计算目标模板与候选区域目标特征往往无法实现对运动目标的准确定位。在研究传统算法的基础上,改进了Mean-shift算法中目标特征选取机制,即目标模板采用背景加权,候选目标区域采用核加权。仿真结果表明,该方法实现了在复杂环境背景下对运动目标更加准确的跟踪。     

A Particle Filtering and DSmT Based Approach for Conflict Resolving in case of Target Tracking with Multiple Cues

In this paper, we propose an efficient and robust method for multiple targets tracking in cluttered scenes using multiple cues. Our approach combines the use of Monte Carlo sequential filtering for tracking and Dezert-Smarandache theory (DSmT) to integrate the information provided by the different cues. The use of DSmT provides the necessary framework to quantify and overcome the conflict that might appear between the cues due to the occlusion. Our tracking approach is tested with color and location cues on a cluttered scene where multiple targets are involved in partial or total occlusion.     

基于DST-PCR5多目标自适应视觉跟踪方法

复杂场景下实现快速稳定地自适应跟踪是视觉领域亟需解决的课题之一, 利用目标的多特征信息进行高效融合是提升跟踪算法鲁棒性能的重要途径。本文首先基于DST(Dempster-Shafer Theory)和PCR5(Proportional Conflict Redistribution No.5)设计一种新的合并策略融合运动目标的颜色和纹理特征,其次在粒子滤波框架下建立复杂场景下的多目标自适应跟踪模型,最终实现了复杂场景下多特征信息融合的自适应视觉跟踪。实验结果及性能分析表明,该方法在不良的跟踪条件下,高冲突证据的自适应处理能力得到明显改善,有效提高了粒子的使用效率和跟踪的鲁棒性,可以较好实现复杂场景下准确、稳定地多目标跟踪。     

Cooperative fusion particle filter tracker

Robust visual tracking has become an important topic in the field of computer vision. Integrating multiple cues has proved to be a promising approach to visual tracking in situations where no single cue is suitable. In this work, a new particle filter based visual tracking algorithm is proposed. By introducing a new cooperative fusion strategy, the proposed tracker has better fault tolerance ability than the traditional methods. Experiments are performed in various tracking scenes to evaluate the proposed algorithm, and the results show improved tracking accuracy.     

Visual tracking by fusing multiple cues with context-sensitive reliabilities

Many researchers argue that fusing multiple cues increases the reliability and robustness of visual tracking. However, how the multi-cue integration is realized during tracking is still an open issue. In this work, we present a novel data fusion approach for multi-cue tracking using particle filter. Our method differs from previous approaches in a number of ways. First, we carry out the integration of cues both in making predictions about the target object and in verifying them through observations. Our second and more significant contribution is that both stages of integration directly depend on the dynamically changing reliabilities of visual cues. These two aspects of our method allow the tracker to easily adapt itself to the changes in the context, and accordingly improve the tracking accuracy by resolving the ambiguities.     

Markerless tracking of complex human motions from multiple views

We present a method for markerless tracking of complex human motions from multiple camera views. In the absence of markers, the task of recovering the pose of a person during such motions is challenging and requires strong image features and robust tracking. We propose a solution which integrates multiple image cues such as edges, color information and volumetric reconstruction. We show that a combination of multiple image cues helps the tracker to overcome ambiguous situations such as limbs touching or strong occlusions of body parts. Following a model-based approach, we match an articulated body model built from superellipsoids against these image cues. Stochastic Meta Descent (SMD) optimization is used to find the pose which best matches the images. Stochastic sampling makes SMD robust against local minima and lowers the computational costs as a small set of predicted image features is sufficient for optimization. The power of SMD is demonstrated by comparing it to the commonly used Levenberg–Marquardt method. Results are shown for several challenging sequences showing complex motions and full articulation, with tracking of 24 degrees of freedom in ≈1 frame per second.     

一种非重叠多摄像头的实时监控系统

大范围场景的监控需要使用多个摄像头。论文利用运动目标的颜色信息和路径特征,设计了一种非重叠多摄像头的实时监控系统。系统采用分布式多层次结构,在进行单摄像头层的处理时,根据像素点亮度变化检测和跟踪运动目标,同时获取运动目标的外形信息和路径特征;在进行多摄像头层的处理时,使用估计目标外形变化和建立路径模型方法融合多个摄像头信息,实现目标在非重叠多摄像头的跟踪。该系统不要求校准摄像头,也不要求建立完整的场景模型,即便在有亮度变化的环境中,仍能立即准确跟踪目标。实验证明提出的方法有好的跟踪效果。     

Robust Visual Tracking by Integrating Multiple Cues Based on Co-Inference Learning

Visual tracking can be treated as a parameter estimation problem that infers target states based on image observations from video sequences. A richer target representation may incur better chances of successful tracking in cluttered and dynamic environments, and thus enhance the robustness. Richer representations can be constructed by either specifying a detailed model of a single cue or combining a set of rough models of multiple cues. Both approaches increase the dimensionality of the state space, which results in a dramatic increase of computation. To investigate the integration of rough models from multiple cues and to explore computationally efficient algorithms, this paper formulates the problem of multiple cue integration and tracking in a probabilistic framework based on a factorized graphical model. Structured variational analysis of such a graphical model factorizes different modalities and suggests a co-inference process among these modalities. Based on the importance sampling technique, a sequential Monte Carlo algorithm is proposed to provide an efficient simulation and approximation of the co-inferencing of multiple cues. This algorithm runs in real-time at around 30 Hz. Our extensive experiments show that the proposed algorithm performs robustly in a large variety of tracking scenarios. The approach presented in this paper has the potential to solve other problems including sensor fusion problems.     

Adaptive texture and color segmentation for tracking moving objects

Color segmentation is a very popular technique for real-time object tracking. However, even with adaptive color segmentation schemes, under varying environmental conditions in video sequences, the tracking tends to be unreliable. To overcome this problem, many multiple cue fusion techniques have been suggested. One of the cues that complements color nicely is texture. However, texture segmentation has not been used for object tracking mainly because of the computational complexity of texture segmentation. This paper presents a formulation for fusing texture and color in a manner that makes the segmentation reliable while keeping the computational cost low, with the goal of real-time target tracking. An autobinomial Gibbs Markov random field is used for modeling the texture and a 2D Gaussian distribution is used for modeling the color. This allows a probabilistic fusion of the texture and color cues and for adapting both the texture and color over time for target tracking. Experiments with both static images and dynamic image sequences establish the feasibility of the proposed approach.     

Pedestrian detection and tracking in infrared imagery using shape and appearance

In this paper, we present an approach toward pedestrian detection and tracking from infrared imagery using joint shape and appearance cues. A layered representation is first introduced and a generalized expectation-maximization (EM) algorithm is developed to separate infrared images into background (still) and foreground (moving) layers regardless of camera panning. In the two-pass scheme of detecting pedestrians from the foreground layer: shape cue is first used to eliminate non-pedestrian moving objects and then appearance cue helps to locate the exact position of pedestrians. Templates with varying sizes are sequentially applied to detect pedestrians at multiple scales to accommodate different camera distances. To facilitate the task of pedestrian tracking, we formulate the problem of shot segmentation and present a graph matching-based tracking algorithm that jointly exploits the shape, appearance and distance information. Experimental results with both OSU Infrared Image Database and WVU Infrared Video Database are reported to demonstrate the accuracy and robustness of our algorithm.     

Improvement of mean shift tracking performance using a convex kernel function and extracting motion information

Color feature is now taken into real consideration as one of the important cues in the area of objects tracking, in image sequences. This feature has attracted considerable attention, in recent years. One of the well-known tools in color feature extraction is to use mean shift (MS) tracking algorithm. The probability of finding the object location in line with this tracking algorithm is somehow desirable, in image sequences, by maximizing the Bhattacharyya coefficient between both objects and corresponding candidate models. Even though the MS tracking algorithm is just known as a popular tool in the field of object tracking, it does not have sufficient merit to be realized in complex environments, i.e., background with object’s similar color, sudden light changes, occlusion types and so on. In such a case, the amount of the present coefficient could truly be decreased, during the tracking process, because of the mentioned environmental problems. A convex kernel function in association with the motion information of video sequences is used in this investigation to improve the MS tracking algorithm for the purpose of overcoming the existing problems. The proposed approach is employed to present the MS kernel function, directly. Thus, by using the investigation in its present form, the capability of the MS kernel is increased. Moreover, by using both color feature and motion information, simultaneously, in comparison with single color feature, noises and also uninterested regions can actually be eliminated. Experimental results on data set illustrate that the proposed approach has an optimum performance in real-time object tracking under the severe conditions.