Single Object Tracking Research Articles

Hierarchical fuzzy logic based approach for object tracking

In this paper a novel tracking approach based on fuzzy concepts is introduced. A methodology for both single and multiple object tracking is presented. The aim of this methodology is to use these concepts as a tool to, while maintaining the needed accuracy, reduce the complexity usually involved in object tracking problems. Several dynamic fuzzy sets are constructed according to both kinematic and non-kinematic properties that distinguish the object to be tracked. Meanwhile kinematic related fuzzy sets model the object’s motion pattern, the non-kinematic fuzzy sets model the object’s appearance. The tracking task is performed through the fusion of these fuzzy models by means of an inference engine. This way, object detection and matching steps are performed exclusively using inference rules on fuzzy sets. In the multiple object methodology, each object is associated with a confidence degree and a hierarchical implementation is performed based on that confidence degree.

Dynamic learning, retrieval, and tracking to augment hundreds of photographs

Tracking is a major issue of virtual and augmented reality applications. Single object tracking on monocular video streams is fairly well understood. However, when it comes to multiple objects, existing methods lack scalability and can recognize only a limited number of objects. Thanks to recent progress in feature matching, state-of-the-art image retrieval techniques can deal with millions of images. However, these methods do not focus on real-time video processing and cannot track retrieved objects. In this paper, we present a method that combines the speed and accuracy of tracking with the scalability of image retrieval. At the heart of our approach is a bi-layer clustering process that allows our system to index and retrieve objects based on tracks of features, thereby effectively summarizing the information available on multiple video frames. Dynamic learning of new viewpoints as the camera moves naturally yields the kind of robustness and reliability expected from an augmented reality engine. As a result, our system is able to track in real-time multiple objects, recognized with low delay from a database of more than 300 entries. We released the source code of our system in a package called Polyora.

Sparsity based Single Object Tracking

Object tracking has importance in various video processing applications like video surveillance, perceptual user interface driver assistance, tracking etc. This paper deals with a new tracking technique that combines the dictionary based background subtraction along with sparsity based tracking. The speed and performance challenges faced during the sparsity based tracking alone are addressed, as it is based on a background subtraction preprocessing and local compressive tracking. It also overcomes the challenges faced by the traditional techniques due to illumination variation, pose and shape change of the object. Output of the proposed technique is compared with that of compressive tracking technique.

On the Hardware/Software Design and Implementation of a High Definition Multiview Video Surveillance System

This paper proposes a distributed architecture for high definition multiview video surveillance system. It adopts a modular design where single view/stereo intelligent internet protocol (IP)-based video surveillance cameras are connected to a front-end field programmable gate array (FPGA) board(s) which are connected to a back-end local video server through the IP network. The data intensive video analytics (VA) algorithms such as background modeling, connected component labeling and single view object tracking are implemented in the FPGA using an efficient fix-point based architecture. Each back-end video server is equipped with a storage and graphics processing units for supporting high-level VA and other processing algorithms such as video decompression/display, mean depth estimation and consistent labeling. A real-time prototype system was constructed to illustrate the architecture and VA algorithms involved. Satisfactory results were obtained for both publicly available data set and real surveillance video data.

Sub-sample swapping for sequential Monte Carlo approximation of high-dimensional densities in the context of complex object tracking

In this paper, we address the problem of complex object tracking using the particle filter framework, which essentially amounts to estimate high-dimensional distributions by a sequential Monte Carlo algorithm. For this purpose, we first exploit Dynamic Bayesian Networks to determine conditionally independent subspaces of the object’s state space, which allows us to independently perform the particle filter’s propagations and corrections over small spaces. Second, we propose a swapping process to transform the weighted particle set provided by the update step of the particle filter into a “new particle set” better focusing on high peaks of the posterior distribution. This new methodology, called Swapping-Based Partitioned Sampling, is proved to be mathematically sound and is successfully tested and validated on synthetic video sequences for single or multiple articulated object tracking.

Object tracking based on particle filter with discriminative features

This paper presents a particle filter-based visual tracking method with online feature selection mechanism. In color-based particle filter algorithm the weights of particles do not always represent the importance correctly, this may cause that the object tracking based on particle filter converge to a local region of the object. In our proposed visual tracking method, the Bhattacharyya distance and the local discrimination between the object and background are used to define the weights of the particles, which can solve the existing local convergence problem. Experiments demonstrates that the proposed method can work well not only in single object tracking processes but also in multiple similar objects tracking processes.

Single and Multiple Object Tracking Using Log-Euclidean Riemannian Subspace and Block-Division Appearance Model

Object appearance modeling is crucial for tracking objects, especially in videos captured by nonstationary cameras and for reasoning about occlusions between multiple moving objects. Based on the log-euclidean Riemannian metric on symmetric positive definite matrices, we propose an incremental log-euclidean Riemannian subspace learning algorithm in which covariance matrices of image features are mapped into a vector space with the log-euclidean Riemannian metric. Based on the subspace learning algorithm, we develop a log-euclidean block-division appearance model which captures both the global and local spatial layout information about object appearances. Single object tracking and multi-object tracking with occlusion reasoning are then achieved by particle filtering-based Bayesian state inference. During tracking, incremental updating of the log-euclidean block-division appearance model captures changes in object appearance. For multi-object tracking, the appearance models of the objects can be updated even in the presence of occlusions. Experimental results demonstrate that the proposed tracking algorithm obtains more accurate results than six state-of-the-art tracking algorithms.

A SINGLE-OBJECT TRACKING METHOD FOR ROBOTS USING OBJECT-BASED VISUAL ATTENTION

It is a quite challenging problem for robots to track the target in complex environment due to appearance changes of the target and background, large variation of motion, partial and full occlusion, motion of the camera and so on. However, humans are capable to cope with these difficulties by using their cognitive capability, mainly including the visual attention and learning mechanisms. This paper therefore presents a single-object tracking method for robots based on the object-based attention mechanism. This tracking method consists of four modules: pre-attentive segmentation, top-down attentional selection, post-attentive processing and online learning of the target model. The pre-attentive segmentation module first divides the scene into uniform proto-objects. Then the top-down attention module selects one proto-object over the predicted region by using a discriminative feature of the target. The post-attentive processing module then validates the attended proto-object. If it is confirmed to be the target, it is used to obtain the complete target region. Otherwise, the recovery mechanism is automatically triggered to globally search for the target. Given the complete target region, the online learning algorithm autonomously updates the target model, which consists of appearance and saliency components. The saliency component is used to automatically select a discriminative feature for top-down attention, while the appearance component is used for bias estimation in the top-down attention module and validation in the post-attentive processing module. Experiments have shown that this proposed method outperforms other algorithms without using attention for tracking a single target in cluttered and dynamically changing environment.

What Are We Tracking: A Unified Approach of Tracking and Recognition

Tracking is essentially a matching problem. While traditional tracking methods mostly focus on low-level image correspondences between frames, we argue that high-level semantic correspondences are indispensable to make tracking more reliable. Based on that, a unified approach of low-level object tracking and high-level recognition is proposed for single object tracking, in which the target category is actively recognized during tracking. High-level offline models corresponding to the recognized category are then adaptively selected and combined with low-level online tracking models so as to achieve better tracking performance. Extensive experimental results show that our approach outperforms state-of-the-art online models in many challenging tracking scenarios such as drastic view change, scale change, background clutter, and morphable objects.

Effect of occlusion and landmarks on single object tracking during disrupted viewing

(blue) and motion path of tracked object (red). This participants uses discrete eye-movements to accurately track the square up until sound probe (orange) when eyes stabilize on same position until tap response is made (about 1 second later). This sustained gaze-response may serve as a “place-holder” to help keep track of the position of the tracked object. Effect of Occlusion and Landmarks on Single Object Tracking During Disrupted Viewing Meriam Naqvi1 ,2, Kevin Zish1, Ronald Planer1,3, Deborah Aks1, Zenon W. Pylyshyn1 Rutgers Center for Cognitive Science1, Rutgers Department of Biological Sciences2, Rutgers Department of Philosophy3

Effect of Occlusion and Landmarks on Single Object Tracking During Disrupted Viewing

Effect of Occlusion and Landmarks on Single Object Tracking During Disrupted Viewing

High performance memetic algorithm particle filter for multiple object tracking on modern GPUs

This work presents an effective approach to visual tracking using a graphics processing unit (GPU) for computation purposes. In order to get a performance improvement against other platforms it is convenient to select proper algorithms such as population-based ones. They expose a parallel-friendly nature needing from many independent evaluations that map well to the parallel architecture of the GPU. To this end we propose a particle filter (PF) hybridized with a memetic algorithm (MA) to produce a MAPF tracking algorithm for single and multiple object tracking problems. Previous experimental results demonstrated that the MAPF algorithm showed more accurate tracking results than the standard PF, and now we extend those results with the first complete adaptation of the PF and the MAPF for visual tracking to the NVIDIA CUDA architecture. Results show a GPU speedup between 5×–16× for different configurations.

SiMOR: Single Moving Object Recognition

Automatic moving object detection and tracking is very important task in video surveillance applications. In the present work the well known background subtraction model and use of Gaussian Mixture Models (GMM) have been used to implement a robust automated single object tracking system. In this implementation, background subtraction on subtracting consecutive frame-by-frame basis for moving object detection is done. Once the object has been detected it is tracked by employing an efficient GMM technique. After successful completion of tracking, moving object recognition of those objects using well known Principal Component Analysis (PCA), which is used for extracting features and Manhattan based distance metric is used for subsequent classification purpose. The system is capable of handling entry and exit of an object. Such a tracking system is cost effective and can be used as an automated video conferencing system and also has applications like human tracking, vehicles monitoring, and event recognition for video surveillance. The proposed algorithm was tested on standard database on complex environments and the results were satisfactory.

Extended Keyframe Detection with Stable Tracking for Multiple 3D Object Tracking

We present a method that is able to track several 3D objects simultaneously, robustly, and accurately in real time. While many applications need to consider more than one object in practice, the existing methods for single object tracking do not scale well with the number of objects, and a proper way to deal with several objects is required. Our method combines object detection and tracking: frame-to-frame tracking is less computationally demanding but is prone to fail, while detection is more robust but slower. We show how to combine them to take the advantages of the two approaches and demonstrate our method on several real sequences.

A color-based particle filter for multiple object tracking in an outdoor environment

Tracking multiple objects is more challenging than tracking a single object. Some problems arise in multiple-object tracking that do not exist in single-object tracking, such as object occlusion, the appearance of a new object and the disappearance of an existing object, updating the occluded object, etc. In this article, we present an approach to handling multiple-object tracking in the presence of occlusions, background clutter, and changing appearance. The occlusion is handled by considering the predicted trajectories of the objects based on a dynamic model and likelihood measures. We also propose target-model-update conditions, ensuring the proper tracking of multiple objects. The proposed method is implemented in a probabilistic framework such as a particle filter in conjunction with a color feature. The particle filter has proven very successful for nonlinear and non-Gaussian estimation problems. It approximates a posterior probability density of the state, such as the object's position, by using samples or particles, where each state is denoted as the hypothetical state of the tracked object and its weight. The observation likelihood of the objects is modeled based on a color histogram. The sample weight is measured based on the Bhattacharya coefficient, which measures the similarity between each sample's histogram and a specified target model. The algorithm can successfully track multiple objects in the presence of occlusion and noise. Experimental results show the effectiveness of our method in tracking multiple objects.

Real Time Color Object Tracking on Cell Broadband Engine Using Particle Filters

Real time visual tracking by particle filter has been implemented on Cell Broadband Engine in parallel. Major problem for the implementation is small size of Local Store (LS) in SPEs (Synergistic PEs), which are computational cores, to deal with image of large size. As a first step for the implementation, we focus on color single object tracking, which is one of the most simple case of visual tracking. By elaborating to compress the color extracted image into bit-wise representation of binary image, all information of the color extracted image can be stored in LS for 640×480 size of original image. By applying our previous implementation of general particle filter algorithm on Cell/B.E. to this specific case, we have achieved real time performance of visual tracking on PlayStation®3 about 7 fps with a camera of maximum 15 fps.

Object tracking using AM-FM image features

AM-FM models analyse an image in terms of amplitude (AM) and frequency modulated (FM) sinusoids. In this study, the authors present detection and tracking of single and multiple objects in video sequences using AM-FM features. The authors use the particle filtering framework for estimating the motion parameters. The single object tracking algorithm uses an affine motion model and a subspace-based appearance model. The multiple object tracking algorithm, which is a logical extension of single object tracking, can handle varying number of interacting objects. The performance of both single and multiple object tracking is illustrated on real-world videos.

2D Articulated Pose Tracking Using Particle Filter with Partitioned Sampling and Model Constraints

In this paper, we develop a two-dimensional articulated body tracking algorithm based on the particle filtering method using partitioned sampling and model constraints. Particle filtering has been proven to be an effective approach in the object tracking field, especially when dealing with single-object tracking. However, when applying it to human body tracking, we have to face a "particle-explosion" problem. We then introduce partitioned sampling, applied to a new articulated human body model, to solve this problem. Furthermore, we develop a propagating method originated from belief propagation (BP), which enables a set of particles to carry several constraints. The proposed algorithm is then applied to tracking articulated body motion in several testing scenarios. The experimental results indicate that the proposed algorithm is effective and reliable for 2D articulated pose tracking.

Multiple object detection and tracking with pseudo-particle filter

To tackle the divergence of the classical particle filter method for multiple object tracking in image sequences, a new particle filter, called pseudoparticle filter (PPF), is proposed. The PPF invokes subset particles of generic particle filters to form a continuous estimate of the posterior density function of the objects. After sampling-importance resampling (SIR), the subset particles converge to the observations. It is proved that, using an appropriate kernel function of the mean shift algorithm, we can get the subset particles of the observations and the fixed points of clustering results as the state of the objects. A multiple object data association and state estimation technique is proposed to resolve the subset particles correspondence ambiguities that arise when multiple objects are present. Experimental results demonstrate the efficiency and effectiveness of the algorithm for single and multiple object tracking.

Contextual Mixture Tracking

Multiple object tracking (MOT) poses three challenges to conventional well-studied single object tracking (SOT) algorithms: 1) multiple targets lead the configuration space to be exponential to the number of targets; 2) multiple motion conditions due to multiple targets' entering, exiting and intersection make the prediction process degrade in precision; 3) visual ambiguities among nearby targets make the trackers error prone. In this paper, we address the MOT problem by embedding contextual proposal distributions and contextual observation models into a mixture tracker which is implemented in a Particle Filter framework. The proposal distributions are adaptively selected by motion conditions of targets which are determined by context information, and the multiple features are combined according to their discriminative power between ambiguity prone objects. The induction of contextual proposal distribution and observation model can help to surmount the incapability of conventional mixture tracker in handling object occlusions, meanwhile retain its merits of flexibility and high efficiency. The final experiments show significant improvement in variable number objects tracking scenarios compared with other methods.