Multi-scale patch-based sparse appearance model for robust object tracking

When objects undergo large pose change, illumination variation or partial occlusion, most existing visual tracking algorithms tend to drift away from targets and even fail to track them. To address the issue, in this paper we propose a multi-scale patch-based appearance model with sparse representation and provide an efficient scheme involving the collaboration between multi-scale patches encoded by sparse coefficients. The key idea of our method is to model the appearance of an object by different scale patches, which are represented by sparse coefficients with different scale dictionaries. The model exploits both partial and spatial information of targets based on multi-scale patches. Afterwards, a similarity score of one candidate target is input into a particle filter framework to estimate the target state sequentially over time in visual tracking. Additionally, to decrease the visual drift caused by frequently updating model, we present a novel two-step object tracking method which exploits both the ground truth information of the target labeled in the first frame and the target obtained online with the multi-scale patch information. Experiments on some publicly available benchmarks of video sequences showed that the similarity involving complementary information can locate targets more accurately and the proposed tracker is more robust and effective than others.     

正例投票下的L1目标跟踪算法

目的 传统的L1稀疏表示目标跟踪,是将所有候选目标表示为字典模板的线性组合,只考虑了字典模板的整体信息,没有分析目标的局部结构。针对该方法在背景杂乱时容易出现跟踪漂移的问题,提出一种基于正例投票的目标跟踪算法。方法 本文将目标表示成图像块粒子的组合,考虑目标的局部结构。在粒子滤波框架内,构建图像块粒子置信函数和相似性函数,提取正例图像块。最终通过正例权重投票估计跟踪目标的最佳位置。结果 在14组公测视频序列上进行跟踪实验,与多种优秀的目标跟踪算法相比,本文跟踪算法在目标受到背景杂乱、遮挡、光照变化等复杂环境干扰下最为稳定,重叠率达到了0.7,且取得了最低的平均跟踪误差5.90,反映了本文算法的可靠性和有效性。结论 本文正例投票下的L1目标跟踪算法,与经典方法相比,能够解决遮挡、光照变化和快速运动等问题的同时,稳定可靠地实现背景杂乱序列的鲁棒跟踪。     

Robust visual tracking via discriminative appearance model based on sparse coding

In this paper, we formulate visual tracking as a binary classification problem using a discriminative appearance model. To enhance the discriminative strength of the classifier in separating the object from the background, an over-complete dictionary containing structure information of both object and background is constructed which is used to encode the local patches inside the object region with sparsity constraint. These local sparse codes are then aggregated for object representation, and a classifier is learned to discriminate the target from the background. The candidate sample with largest classification score is considered as the tracking result. Different from recent sparsity-based tracking approaches that update the dictionary using a holistic template, we introduce a selective update strategy based on local image patches which alleviates the visual drift problem, especially when severe occlusion occurs. Experiments on challenging video sequences demonstrate that the proposed tracking algorithm performs favorably against several state-of-the-art methods.     

变分调整约束下的反向低秩稀疏学习目标跟踪

目的 低秩稀疏学习目标跟踪算法在目标快速运动和严重遮挡等情况下容易出现跟踪漂移现象,为此提出一种变分调整约束下的反向低秩稀疏学习目标跟踪算法。方法 采用核范数凸近似低秩约束描述候选粒子间的时域相关性,去除不相关粒子,适应目标外观变化。通过反向稀疏表示描述目标表观,用候选粒子稀疏表示目标模板,减少在线跟踪中L₁优化问题的数目,提高跟踪效率。在有界变差空间利用变分调整对稀疏系数差分建模,约束目标表观在相邻帧间具有较小变化,但允许连续帧间差异存在跳跃不连续性,以适应目标快速运动。结果 实验利用OTB（object tracking benchmark）数据集中的4组涵盖了严重遮挡、快速运动、光照和尺度变化等挑战因素的标准视频序列进行测试,定性和定量对比了本文算法与5种热点算法的跟踪效果。定性分析基于视频序列的主要挑战因素进行比较,定量分析通过中心点位置误差（central pixel error,CPE）比较跟踪算法的精度。与CNT（convolutional networks training）、SCM（sparse collaborative model）、IST（inverse sparse tracker）、DDL（discriminative dictionary learning）和LLR（locally low-rank representation）算法相比,平均CPE值分别提高了2.80、4.16、13.37、35.94和41.59。实验结果表明,本文算法达到了较高的跟踪精度,对上述挑战因素更具鲁棒性。结论 本文提出的跟踪算法,综合了低秩稀疏学习和变分优化调整的优势,在复杂场景下具有较高的跟踪精度,特别是对严重遮挡和快速运动情况的有效跟踪更具鲁棒性。     

Robust Visual Tracking Based on Convex Hull with EMD-L1

Factors such as drastic illumination variations, partial occlusion, rotation make robust visual tracking a difficult problem. Some tracking algorithms represent a target appearances based on obtained tracking results from previous frames with a linear combination of target templates. This kind of target representation is not robust to drastic appearance variations. In this paper, we propose a simple and effective tracking algorithm with a novel appearance model. A target candidate is represented by convex combinations of target templates. Measuring the similarity between a target candidate and the target templates is a key problem for a robust likelihood evaluation. The distance between a target candidate and the templates is measured using the earth mover’s distance with L1 ground distance. Comprehensive experiments demonstrate the robustness and effectiveness of the proposed tracking algorithm against state-of-the-art tracking algorithms.     

判别稀疏表示鲁棒快速视觉跟踪

目的 L1跟踪对局部遮挡具有良好的鲁棒性,但存在易产生模型漂移和计算速度慢的问题。针对这两个问题,该文提出了一种基于判别稀疏表示的视觉跟踪方法。方法 考虑到背景和遮挡信息的干扰,提出了一种判别稀疏表示模型,并基于块坐标优化原理,采用学习迭代收缩阈值算法和软阈值操作设计出了表示模型的快速求解算法。结果 在8组图像序列中,该文方法与现有的4种经典跟踪方法分别在鲁棒性和稀疏表示的计算时间方面进行了比较。在鲁棒性的定性和定量比较实验中,该文方法不仅表现出了对跟踪过程中的多种干扰因素具有良好的适应能力,而且在位置误差阈值从0~50像素的变化过程中,其精度曲线均优于实验中的其他方法;在稀疏表示的计算时间方面,在采用大小为16×16和32×32的模板进行跟踪时,该文算法的时间消耗分别为0.152 s和0.257 s,其时效性明显优于实验中的其他方法。结论 与经典的跟踪方法相比,该文方法能够在克服遮挡、背景干扰和外观改变等诸多不良因素的同时,实现快速目标跟踪。由于该文方法不仅具有较优的稀疏表示计算速度,而且能够克服多种影响跟踪鲁棒性的干扰因素,因此可以将其应用于视频监控和体育竞技等实际场合。     

Visual object tracking via sample-based Adaptive Sparse Representation (AdaSR)

When appearance variation of object and its background, partial occlusion or deterioration in object images occurs, most existing visual tracking methods tend to fail in tracking the target. To address this problem, this paper proposes a new approach for visual object tracking based on Sample-Based Adaptive Sparse Representation (AdaSR), which ensures that the tracked object is adaptively and compactly expressed with predefined samples. First, the Sample-Based Sparse Representation, which selects a subset of samples as a basis for object representation by exploiting L1-norm minimization, improves the representation adaptation to partial occlusion for tracking. Second, to keep the temporal consistency and adaptation to appearance variation and deterioration in object images during the tracking process, the object's Sample-Based Sparse Representation is adaptively evaluated based on a Kalman filter, obtaining the AdaSR. Finally, the candidate holding the most similar Sample-Based Sparse Representation to the AdaSR of the tracked object will be regarded as the instantaneous tracking result. In addition, we can easily extend the AdaSR for multi-object tracking by integrating the sample set of each tracked object (named Common Sample-Based Adaptive Sparse Representation Analysis (AdaSRA)). AdaSRA fully analyses Adaptive Sparse Representation similarity for object classification. Our experiments on public datasets show state-of-the-art results, which are better than those of several representative tracking methods.     

协同结构稀疏重构的判别性视觉跟踪

基于稀疏表示的表观似然模型在目标跟踪领域具有广泛的应用,但是这种单一产生式目标表观模型并未考虑完整的判别性结构信息,容易受复杂背景的干扰。为了缓解由该问题造成的目标跟踪漂移,提出了一种目标表观字典和背景字典协同结构稀疏重构优化的视觉跟踪方法。通过构建一个有判别力的基于稀疏表示的表观似然模型,实现了对目标表观模型更为准确的描述。通过合理选择约束候选目标区域和候选背景区域的稀疏系数,在表观似然模型中引入判别式信息,以进一步揭示候选目标区域的潜在相关性和候选背景区域的结构关系,从而更加准确地学习候选目标区域的表观模型。大量有挑战性的视频序列上的实验结果验证了算法在复杂背景下跟踪的鲁棒性,与其他相关算法的对比实验也体现了该算法的优越性。     

结合稀疏表示和均值偏移的运动目标跟踪算法

为了当出现目标尺度变化、方向变化、环境光照变化、目标部分遮挡等问题时,使得视觉跟踪算法具有更好的鲁棒性,提出一种结合稀疏编码和空间金字塔模型以及均值漂移的算法。首先扩展经典Meanshift算法使它不仅估计位置空间变化,还估计方向和尺度空间的变化。然后加入像素密度块采样技术和琐碎模板设计方案使直方图匹配更加准确,有效克服光照变化。最后取代原有算法中要么使用整体表示要么使用局部表示目标特征的方法,使得空间金字塔模型与两种表示方法相结合,有效解决目标遮挡等问题。实验表明,该算法实验结果明显优于同类算法,能很好地解决目标尺度变化、环境光照变化、目标部分遮挡等问题。     

Object tracking via appearance modeling and sparse representation

This paper proposes a robust tracking method by the combination of appearance modeling and sparse representation. In this method, the appearance of an object is modeled by multiple linear subspaces. Then within the sparse representation framework, we construct a similarity measure to evaluate the distance between a target candidate and the learned appearance model. Finally, tracking is achieved by Bayesian inference, in which a particle filter is used to estimate the target state sequentially over time. With the tracking result, the learned appearance model will be updated adaptively. The combination of appearance modeling and sparse representation makes our tracking algorithm robust to most of possible target variations due to illumination changes, pose changes, deformations and occlusions. Theoretic analysis and experiments compared with state-of-the-art methods demonstrate the effectivity of the proposed algorithm.     

利用增广拉格朗日乘子的鲁棒跟踪算子

针对视频目标鲁棒跟踪问题,提出了一种基于稀疏表示的生成式算法。首先提取特征构建目标和背景模板,并利用随机抽样获得足够多的候选目标状态;然后利用多任务反向稀疏表示算法得到稀疏系数矢量构造相似度测量图,这里引入了增广拉格朗日乘子(ALM)算法解决L₁-min难题;最后从相似度图中使用加性池运算提取判别信息选择与目标模板相似度最高并与背景模板相似度最小的候选目标状态作为跟踪结果,该算法是在贝叶斯滤波框架下实现的。为了适应跟踪过程中目标外观由于光照变化、遮挡、复杂背景以及运动模糊等场景引起的变化,制定了简单却有效的更新机制,对目标和背景模板进行更新。对仿真结果的定性和定量评估均表明与其他跟踪算法相比,所提算法的跟踪准确性和稳定性有了一定的提高,能有效地解决光照和尺度变化、遮挡、复杂背景等场景的跟踪难题。     

局部模板更新逆向联合稀疏表示目标跟踪算法

逆向联合稀疏表示算法可充分利用跟踪过程中的时间相似性和空间连续性,但由于遮挡、光照变化等的影响,易出现跟踪漂移.为解决上述问题,提出一种基于局部模板更新逆向联合稀疏表示目标跟踪算法,其通过逆向局部重构目标模板集完成逆向联合稀疏表示.首先,在首帧初始化目标模板集,利用粒子滤波获取候选图像,并对其分块处理,构建逆向联合稀疏...     

Accurate target tracking via Gaussian sparsity and locality-constrained coding in heavy occlusion

This paper presents a Gaussian sparse representation cooperative model for tracking a target in heavy occlusion video sequences by combining sparse coding and locality-constrained linear coding algorithms. Different from the usual method of using ?₁-norm regularization term in the framework of particle filters to form the sparse collaborative appearance model (SCM), we employed the ?₁-norm and ?₂-norm to calculate feature selection, and then encoded the candidate samples to generate the sparse coefficients. Consequently, our method not only easily obtained sparse solutions but also reduced reconstruction error. Compared to state-of-the-art algorithms, our scheme achieved better performance in heavy occlusion video sequences for tracking a target. Extensive experiments on target tracking were carried out to show the results of our proposed algorithm compared with various other target tracking methods.     

基于子空间和直方图的多记忆自适应相关滤波目标跟踪算法

为了增强相关滤波算法(CF)在目标遮挡或背景干扰情况下跟踪的鲁棒性,提出基于子空间和直方图的多记忆自适应相关滤波目标跟踪算法.首先,针对CF使用的模板单一无法应对不同时期相邻帧目标表现的差异,提出利用随机更新策略学习多个目标模板,应对不同时期的目标变化.然后,针对不同的更新模板得到多个候选目标,利用子空间学习上一帧的表示系数,综合判断候选目标的准确性.同时,因为CF与子空间表示均利用模板判断跟踪结果,对背景杂乱等情况判断容易造成偏差,所以引入颜色直方图,利用统计特征作为独立的判断依据,增强算法对候选目标判断结果的准确性.在标准视频集上的实验表明,文中算法具备一定的抗遮挡及抗背景干扰能力.     

一种基于压缩域的实时跟踪改进算法

在基于压缩域的实时跟踪算法中,判别函数对目标外观考虑不足易造成跟踪精度较低。为此,提出一种改进的基于压缩域的实时跟踪算法。利用稀疏测量矩阵提取候选目标的低维多尺度特征,并根据在线更新的特征概率分布,采用朴素贝叶斯分类器判别目标与背景,实现粗跟踪。通过视频帧间候选目标内部区域所具有的相似性,在粗跟踪的基础上实施基于动态目标外观模型的二次跟踪,在线寻找目标的最佳跟踪位置。对多种跟踪视频库的测试结果表明,该算法在不过量增加计算负荷的情况下能有效提高跟踪精度。     

余弦权重逆稀疏框架视频目标跟踪算法

针对目标跟踪的遮挡与局部形变,提出局部余弦相似度训练权重的逆稀疏视觉目标跟踪策略。借鉴参数空间的粒子滤波的核心思想,以逆稀疏表示为理论框架,用候选目标重构模板获得候选目标的稀疏表示系数,依据表示系数分布特征筛选出最佳候选目标。为克服遮挡影响,引入新的目标函数构建模板的局部块判别能力：计算正负样本与模板之间的局部余弦相似度差值,利用二次优化,更新具有判别能力的权重。依据权重信息综合进行有选择的模板更新,避免模板更新的无效性。多组实验结果表明,该算法在部分遮挡等复杂环境下,仍然可以准确地跟踪目标,相比已有算法具有自己的优势。     

基于特征选择与时间一致性稀疏外观模型的目标追踪算法

为了更有效利用追踪目标的判别特征信息,提高目标追踪的精度和鲁棒性,在粒子滤波追踪框架下提出基于特征选择与时间一致性稀疏外观模型的目标追踪算法.首先,采集目标的正负模板和候选目标,根据特征选择模型对正负模板和候选目标进行特征选择,去除多余的干扰信息,得到关键的特征信息.然后,利用正负模板和候选目标的特征建立多任务稀疏表示模型,引入时间一致性正则项,促进更多的候选目标与先前帧的追踪结果具有稀疏表示的相似性.最后,求解多任务稀疏表示模型,得到判别稀疏相似图,获取每个候选目标的判别分,根据目标追踪结果更新正负模板.实验表明,即使在复杂的环境下,文中算法仍然比其它一些追踪算法具有更高的准确性.     

Robust visual tracking using information theoretical learning

This paper presents a novel online object tracking algorithm with sparse representation for learning effective appearance models under a particle filtering framework. Compared with the state-of-the-art ? ₁ sparse tracker, which simply assumes that the image pixels are corrupted by independent Gaussian noise, our proposed method is based on information theoretical Learning and is much less sensitive to corruptions; it achieves this by assigning small weights to occluded pixels and outliers. The most appealing aspect of this approach is that it can yield robust estimations without using the trivial templates adopted by the previous sparse tracker. By using a weighted linear least squares with non-negativity constraints at each iteration, a sparse representation of the target candidate is learned; to further improve the tracking performance, target templates are dynamically updated to capture appearance changes. In our template update mechanism, the similarity between the templates and the target candidates is measured by the earth movers’ distance(EMD). Using the largest open benchmark for visual tracking, we empirically compare two ensemble methods constructed from six state-of-the-art trackers, against the individual trackers. The proposed tracking algorithm runs in real-time, and using challenging sequences performs favorably in terms of efficiency, accuracy and robustness against state-of-the-art algorithms.     

分块多特征目标描述子的移动机器人目标跟踪

为解决机器人目标跟踪过程中的遮挡和外观改变等问题, 提出一种分块多特征描述子的方法. 该方法将候选样本分块, 提取图像片的深度、颜色、纹理特征来表示目标构造检测器. 结合目标与机器人的运动构造运动卡尔曼滤波器(MEKF) 作为跟踪器. 跟踪过程中根据目标深度信息调整其尺寸, 结合深度特征及图像片外观相似度进行检测并处理遮挡. 实验结果表明, 该算法对目标的尺度变化、光照改变和遮挡现象具有较强的鲁棒性.

     

Robust visual tracking with structured sparse representation appearance model

In this paper, we present a structured sparse representation appearance model for tracking an object in a video system. The mechanism behind our method is to model the appearance of an object as a sparse linear combination of structured union of subspaces in a basis library, which consists of a learned Eigen template set and a partitioned occlusion template set. We address this structured sparse representation framework that preferably matches the practical visual tracking problem by taking the contiguous spatial distribution of occlusion into account. To achieve a sparse solution and reduce the computational cost, Block Orthogonal Matching Pursuit (BOMP) is adopted to solve the structured sparse representation problem. Furthermore, aiming to update the Eigen templates over time, the incremental Principal Component Analysis (PCA) based learning scheme is applied to adapt the varying appearance of the target online. Then we build a probabilistic observation model based on the approximation error between the recovered image and the observed sample. Finally, this observation model is integrated with a stochastic affine motion model to form a particle filter framework for visual tracking. Experiments on some publicly available benchmark video sequences demonstrate the advantages of the proposed algorithm over other state-of-the-art approaches.