基于超像素的点追踪方法

目的由于当前大多数的追踪算法都是使用目标外观模型和特征进行目标的匹配,在长时间的目标追踪过程中,目标的尺度和形状均会发生变化,再加上计算机视觉误差,都会导致追踪的失误。提出一种高效的目标模型用于提高追踪的效率和成功率。方法采用分割后提取的目标特征来进行建模表示外观结构,利用图像分割的方法,将被追踪的目标区域分割成多个超像素块,结合SIFT特征,形成词汇本,并计算每个词在词汇本中的权值,作为目标的外观模型。利用外观模型确定目标对象的关键点位置后,通过使用金字塔Lucas-Kanade追踪器预测关键点在下一帧图像中的位置,并移动追踪窗口位置。结果结合点位移的加权计算有效地克服目标尺度和形状变化产生的问题。结论实验结果表明在目标发生形变或光照变化的情况下,算法也能准确地、实时地追踪到目标。     

自适应紧致特征的超像素目标跟踪

目的 针对现有的超像素目标跟踪算法（RST）对同一类中分别属于目标和相似干扰物体的超像素块赋予相同特征置信度,导致难以区分目标和相似干扰物的问题,为此提出自适应紧致特征的超像素目标跟踪算法（ACFST）。方法 该方法在每帧的目标搜索区域内构建适合目标大小的自适应紧致搜索区域,并将该区域内外的特征置信度分别保持不变和降低。处于背景中的相似干扰物体会被该方法划分到紧致搜索区域外,其特征置信度被降低。当依据贝叶斯推理框架求出对应最大后验概率的目标时,紧致搜索区域外的特征置信度低,干扰物体归属目标的程度也低,不会被误判为目标。结果 在具有与目标相似干扰物体的两个视频集进行测试,本文ACFST跟踪算法与RST跟踪算法相比,平均中心误差分别缩减到5.4像素和7.5像素,成功率均提高了11%,精确率分别提高了10.6%和21.6%,使得跟踪结果更精确。结论 本文提出构建自适应紧致搜索区域,并通过设置自适应的参数控制紧致搜索区域变化,减少因干扰物体与目标之间相似而带来的误判。在具有相似物体干扰物的视频集上验证了本文算法的有效性,实验结果表明,本文算法在相似干扰物体靠近或与目标部分重叠时,能够保证算法精确地跟踪到目标,提高算法的跟踪精度,具有较强的鲁棒性,使得算法更能适应背景杂乱、目标遮挡、形变等复杂环境。     

Unsupervised learning-based long-term superpixel tracking

Finding correspondences between structural entities decomposing images is of high interest for computer vision applications. In particular, we analyze how to accurately track superpixels - visual primitives generated by aggregating adjacent pixels sharing similar characteristics - over extended time periods relying on unsupervised learning and temporal integration. A two-step video processing pipeline dedicated to long-term superpixel tracking is proposed. First, unsupervised learning-based superpixel matching provides correspondences between consecutive and distant frames using new context-rich features extended from greyscale to multi-channel and forward-backward consistency constraints. Resulting elementary matches are then combined along multi-step paths running through the whole sequence with various inter-frame distances. This produces a large set of candidate long-term superpixel pairings upon which majority voting is performed. Video object tracking experiments demonstrate the accuracy of our elementary estimator against state-of-the-art methods and proves the ability of multi-step integration to provide accurate long-term superpixel matches compared to usual direct and sequential integration.     

Robust object tracking via multi-scale patch based sparse coding histogram

There are many visual tracking algorithms that are based on sparse representation appearance model. Most of them are modeled by local patches with fixed patch scale, which make trackers less effective when objects undergone appearance changes such as illumination variation, pose change or partial occlusion. To solve the problem, a novel appearance representation model is proposed via multi-scale patch based sparse coding histogram for robust visual tracking. In this paper, the appearance of an object is modeled by different scale patches, which are represented by sparse coding histogram with different scale dictionaries. Then a similarity measure is applied to the calculation of the distance between the sparse coding histograms of target candidate and target template. Finally, the similarity score of the target candidate is passed to a particle filter to estimate the target state sequentially in the tracking process. Additionally, in order to decrease the visual drift caused by partial occlusion, an occlusion handling strategy is adopted, which takes the spatial information of multi-scale patches and occlusion into account. Based on the experimental results on some benchmarks of video sequences, our tracker outperforms state-of-the-art tracking methods.     

Energy-based multi-view piecewise planar stereo

The piecewise planar model (PPM) is an effective means of approximating a complex scene by using planar patches to give a complete interpretation of the spatial points reconstructed from projected 2D images. The traditional piecewise planar stereo methods suffer from either a very restricted number of directions for plane detection or heavy reliance on the segmentation accuracy of superpixels. To address these issues, we propose a new multi-view piecewise planar stereo method in this paper. Our method formulates the problem of complete scene reconstruction as a multi-level energy minimization problem. To detect planes along principal directions, a novel energy formulation with pair-wise potentials is used to assign an optimal plane for each superpixel in an iterative manner, where reliable scene priors and geometric constraints are incorporated to enhance the modeling efficacy and inference efficiency. To detect non-principal-direction planes, we adopt a multi-direction plane sweeping with a restricted search space method to generate reliable candidate planes. To handle the multi-surface straddling problem of a single superpixel, a superpixel sub-segmenting scheme is proposed and a robust Pⁿ Potts model-like higher-order potential is introduced to refine the resulting depth map. Our method is a natural integration of pixel- and superpixel-level multi-view stereos under a unified energy minimization framework. Experimental results for standard data sets and our own data sets show that our proposed method can satisfactorily handle many challenging factors (e.g., slanted surfaces and poorly textured regions) and can obtain accurate piecewise planar depth maps.     

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。     

GLSC: LSC superpixels at over 130 FPS

Superpixel has been successfully applied in various computer vision tasks, and many algorithms have been proposed to generate superpixel map. Recently, a superpixel algorithm called “superpixel segmentation using linear spectral clustering” (LSC) has been proposed, and it performs equally well or better than state-of-the art superpixel segmentation algorithms in terms of several commonly used evaluation metrics in superpixel segmentation. Although LSC is of linear complexity, its original implementation runs in few hundreds of milliseconds for images with resolution of 481 × 321 stated by the authors, which is a limitation for some real-time applications such as visual tracking which may needs, for instance, 30 FPS for standard image resolution (e.g., 480 × 320, 640 × 480, 1280 × 720 and 1920 × 1080). Instead of inventing new algorithms with lower complexity than LSC, we will explore LSC to modify its structure and make it suitable to be implemented by parallel technique. The modified LSC algorithm is implemented in CUDA and tested on several NVIDIA graphics processing unit. Our implementation of the proposed modified LSC algorithm achieves speedups of up to 80× from the original sequential implementation, and the quality, measured by two commonly used evaluation metrics, of our implementation keeps being similar to the original one. The source code will be made publicly available.     

结合深度学习和超像元的高分遥感影像变化检测

目的 随着遥感影像空间分辨率的提升,相同地物的空间纹理表现形式差异变大,地物特征更加复杂多样,传统的变化检测方法已很难满足需求。为提高高分辨率遥感影像的变化检测精度,尤其对相同地物中纹理差异较大的区域做出有效判别,提出结合深度学习和超像元分割的高分辨率遥感影像变化检测方法。方法 将有限带标签数据分割成切片作训练样本,按照样本形式设计一个多切片尺度特征融合网络并对其训练,获得测试图像的初步变化检测结果;利用超像元分割算法将测试图像分割成许多无重叠的同质性区域,并将分割结果与前述检测结果叠合,得到带分割标记的变化检测结果;用举手表决算法统计带分割标记的变化检测结果中超像元的变化状况,得到最终变化检测结果。结果 在变化检测实验结果中,本文提出的多切片尺度特征融合卷积网络模型在广东数据集和香港数据集上,优于单一切片尺度下卷积神经网络模型,并且结合超像元的多切片尺度特征融合卷积网络模型得到的Kappa系数分别达到80%和82%,比相应的非超像元算法分别提高了6%和8%,在两个测试集上表现均优于长短时记忆网络、深度置信网络等对比算法。结论 本文提出的卷积神经网络变化检测方法可以充分学习切片的空间信息和其他有效特征,避免过拟合现象;多层尺度切片特征融合的方法优于单一切片尺度训练神经网络的方法;结合深度学习和超像元分割算法,检测单元实现了由切片到超像元的转变,能对同物异谱的区域做出有效判决,有利于提升变化检测精度。     

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.     

基于堆栈式消噪自编码机的分块目标跟踪

在视觉目标跟踪系统中, 特征的表达和提取是重要的组成部分. 本文提出基于多个自编码机网络相联合的特征提取机, 通过对输入数据进行一定程度的重组, 采用深度学习的理论对其局部特征进行描述并对结果进行联合决策. 结合该网络结构, 本文提出一种融合局部特征的深度信息进行目标跟踪的算法. 将输入图像分块使得大量的乘法运算转化为加法和乘法的混合运算, 相对于全局的特征表达, 大幅降低了运算复杂度. 在跟踪过程中, 目标候选区的各分块权重能够根据相应网络的置信度进行自适应的调整, 提升了跟踪器对光照变化、目标姿态和遮挡的适应. 实验表明, 该跟踪算法在鲁棒性和跟踪速度上表现优秀.     

基于相关滤波的尺度自适应目标跟踪

针对视频目标跟踪中的尺度问题,提出了一种基于相关滤波的尺度自适应目标跟踪方法。首先利用核相关滤波获得目标的中心位置,然后将目标均分为四个子块,通过计算找出子块中心的最大响应位置,最后根据前后两帧目标子块中心位置的相对变化计算出尺度的伸缩系数,进而计算出目标尺度。在具有尺度变化的公开数据集上对本文方法进行测试,并和多种跟踪方法作对比,实验结果表明,该方法将尺度的计算问题转化为对子块中心的定位,其平均跟踪性能优于其它方法,验证了方法的有效性。     

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.     

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.     

多模型协作的分块目标跟踪

为解决复杂场景下,基于整体表观模型的目标跟踪算法容易丢失目标的问题,提出一种多模型协作的分块目标跟踪算法.融合基于局部敏感直方图的产生式模型和基于超像素分割的判别式模型构建目标表观模型,提取局部敏感直方图的亮度不变特征来抵制光照变化的影响;引入目标模型的自适应分块划分策略以解决局部敏感直方图算法缺少有效遮挡处理机制的问题,提高目标的抗遮挡性;通过相对熵和均值聚类度量子块的局部差异置信度和目标背景置信度,建立双权值约束机制和子块异步更新策略,在粒子滤波框架下,选择置信度高的子块定位目标.实验结果表明,本文方法在复杂场景下具有良好的跟踪精度和稳定性.     

一种健壮的超像素跟踪算法

在目标跟踪中,传统的超像素跟踪算法在发生遮挡等情况后,会将非目标超像素标记为目标加入到特征空间. 在对候选样本置信度计算中,利用特征空间中最近邻超像素来划定样本中超像素的簇归属会产生错误;而依据的近邻超像素数量过多时,又会造成分类误差的积累. 为解决上述问题,本文提出一种健壮的超像素跟踪算法. 本算法以贝叶斯算法为框架,首先,将前几帧进行超像素切割,提取特征并使用均值漂移聚类算法和基于超像素的外观表示模型进行分类和计算类置信度,放入特征空间中. 其次,根据接下来几帧的平均中心误差确定最佳近邻数目. 最后,在跟踪过程中,对获取帧的指定区域进行超像素切割,提取特征、进行软分类和计算置信度;根据上一帧目标位置进行高斯采样,累加样本内超像素置信度,获得样本置信度;在发生严重遮挡时,不进行滑动窗口更新和外观模型修改,使用当前模型继续跟踪. 与传统的最近邻超像素算法相比,本算法能够有效提升跟踪成功率和降低平均中心误差.     

Image clustering segmentation based on SLIC superpixel and transfer learning

Traditional fuzzy C-means clustering algorithm has poor noise immunity and clustering results in image segmentation. To overcome this problem, a novel image clustering algorithm based on SLIC superpixel and transfer learning is proposed in this paper. In the proposed algorithm, SLIC superpixel method is used to improve the edge matching degree of image segmentation and enhances the robustness to noise. Transfer learning is adopted to correct the image segmentation result and further improve the accuracy of image segmentation. In addition, the proposed algorithm improves the original SLIC superpixel algorithm and makes the edge of the superpixel more accurate. Experimental results show that the proposed algorithm can obtain better segmentation results.     

基于超像素和判别稀疏的运动目标跟踪算法

针对目标遮挡、非刚性变换、光照变换等因素干扰产生的漂移问题,提出基于超像素和判别稀疏的运动目标跟踪算法。算法首先利用SLIC方法对运动目标的观测区域进行超像素分割,然后通过K-Means算法构建包含目标和背景的超像素字典,再基于判别稀疏表示和[?1]范数最小化框架求解候选目标的稀疏系数,同时结合粒子滤波框架和在线字典更新策略完成目标跟踪。实验结果表明,该算法在多种因素干扰的环境中具有较强的鲁棒性,能够准确稳定地进行在线目标跟踪。