Sensors

Research

Jump to: Other

21 pages, 1165 KiB

Open AccessArticle

δ-Generalized Labeled Multi-Bernoulli Simultaneous Localization and Mapping with an Optimal Kernel-Based Particle Filtering Approach

by Diluka Moratuwage, Martin Adams and Felipe Inostroza

Sensors 2019, 19(10), 2290; https://doi.org/10.3390/s19102290 - 17 May 2019

Cited by 20 | Viewed by 3118

Abstract

Under realistic environmental conditions, heuristic-based data association and map management routines often result in divergent map and trajectory estimates in robotic Simultaneous Localization And Mapping (SLAM). To address these issues, SLAM solutions have been proposed based on the Random Finite Set (RFS) framework, [...] Read more.

Under realistic environmental conditions, heuristic-based data association and map management routines often result in divergent map and trajectory estimates in robotic Simultaneous Localization And Mapping (SLAM). To address these issues, SLAM solutions have been proposed based on the Random Finite Set (RFS) framework, which models the map and measurements such that the usual requirements of external data association routines and map management heuristics can be circumvented and realistic sensor detection uncertainty can be taken into account. Rao–Blackwellized particle filter (RBPF)-based RFS SLAM solutions have been demonstrated using the Probability Hypothesis Density (PHD) filter and subsequently the Labeled Multi-Bernoulli (LMB) filter. In multi-target tracking, the LMB filter, which was introduced as an efficient approximation to the computationally expensive

δ

-Generalized LMB (

δ

-GLMB) filter, converts its representation of an LMB distribution to

δ

-GLMB form during the measurement update step. This not only results in a loss of information yielding inferior results (compared to the

δ

-GLMB filter) but also fails to take computational advantages in parallelized implementations possible with RBPF-based SLAM algorithms. Similar to state-of-the-art random vector-valued RBPF solutions such as FastSLAM and MH-FastSLAM, the performances of all RBPF-based SLAM algorithms based on the RFS framework also diverge from ground truth over time due to random sampling approaches, which only rely on control noise variance. Further, the methods lose particle diversity and diverge over time as a result of particle degeneracy. To alleviate this problem and further improve the quality of map estimates, a SLAM solution using an optimal kernel-based particle filter combined with an efficient variant of the

δ

-GLMB filter (

δ

-GLMB-SLAM) is presented. The performance of the proposed

δ

-GLMB-SLAM algorithm, referred to as

δ

-GLMB-SLAM2.0, was demonstrated using simulated datasets and a section of the publicly available KITTI dataset. The results suggest that even with a limited number of particles,

δ

-GLMB-SLAM2.0 outperforms state-of-the-art RBPF-based RFS SLAM algorithms. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

20 pages, 4945 KiB

Open AccessArticle

Information Fusion for Industrial Mobile Platform Safety via Track-Before-Detect Labeled Multi-Bernoulli Filter

by Tharindu Rathnayake, Ruwan Tennakoon, Amirali Khodadadian Gostar, Alireza Bab-Hadiashar and Reza Hoseinnezhad

Sensors 2019, 19(9), 2016; https://doi.org/10.3390/s19092016 - 29 Apr 2019

Cited by 6 | Viewed by 3036

Abstract

This paper presents a novel Track-Before-Detect (TBD) Labeled Multi-Bernoulli (LMB) filter tailored for industrial mobile platform safety applications. At the core of the developed solution is two techniques for fusion of color and edge information in visual tracking. We derive an application specific [...] Read more.

This paper presents a novel Track-Before-Detect (TBD) Labeled Multi-Bernoulli (LMB) filter tailored for industrial mobile platform safety applications. At the core of the developed solution is two techniques for fusion of color and edge information in visual tracking. We derive an application specific separable likelihood function that captures the geometric shape of the human targets wearing safety vests. We use a novel geometric shape likelihood along with a color likelihood to devise two Bayesian updates steps which fuse shape and color related information. One approach is sequential and the other is based on weighted Kullback–Leibler average (KLA). Experimental results show that the KLA based fusion variant of the proposed algorithm outperforms both the sequential update based variant and a state-of-art method in terms of the performance metrics commonly used in computer vision literature. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

19 pages, 903 KiB

Open AccessArticle

Joint Design of Transmit Waveforms for Object Tracking in Coexisting Multimodal Sensing Systems

by John S. Kota and Antonia Papandreou-Suppappola

Sensors 2019, 19(8), 1753; https://doi.org/10.3390/s19081753 - 12 Apr 2019

Cited by 2 | Viewed by 3066

Abstract

We examine a multiple object tracking problem by jointly optimizing the transmit waveforms used in a multimodal system. Coexisting sensors in this system were assumed to share the same spectrum. Depending on the application, a system can include radars tracking multiple targets or [...] Read more.

We examine a multiple object tracking problem by jointly optimizing the transmit waveforms used in a multimodal system. Coexisting sensors in this system were assumed to share the same spectrum. Depending on the application, a system can include radars tracking multiple targets or multiuser wireless communications and a radar tracking both multiple messages and a target. The proposed spectral coexistence approach was based on designing all transmit waveforms to have the same time-varying phase function while optimizing desirable performance metrics. Considering the scenario of tracking a target with a pulse–Doppler radar and multiple user messages, two signaling schemes were proposed after selecting the waveform parameters to first minimize multiple access interference. The first scheme is based on system interference minimization, whereas the second scheme explores the multiobjective optimization tradeoff between system interference and object parameter estimation error. Simulations are provided to demonstrate the performance tradeoffs due to different system requirements. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

17 pages, 1860 KiB

Open AccessArticle

Tracking Multiple Targets from Multistatic Doppler Radar with Unknown Probability of Detection

by Cong-Thanh Do and Hoa Van Nguyen

Sensors 2019, 19(7), 1672; https://doi.org/10.3390/s19071672 - 8 Apr 2019

Cited by 19 | Viewed by 4688

Abstract

The measurements from multistatic radar systems are typically subjected to complicated data association, noise corruption, missed detection, and false alarms. Moreover, most of the current multistatic Doppler radar-based approaches in multitarget tracking are based on the assumption of known detection probability. This assumption [...] Read more.

The measurements from multistatic radar systems are typically subjected to complicated data association, noise corruption, missed detection, and false alarms. Moreover, most of the current multistatic Doppler radar-based approaches in multitarget tracking are based on the assumption of known detection probability. This assumption can lead to biased or even complete corruption of estimation results. This paper proposes a method for tracking multiple targets from multistatic Doppler radar with unknown detection probability. A closed form labeled multitarget Bayes filter was used to track unknown and time-varying targets with unknown probability of detection in the presence of clutter, misdetection, and association uncertainty. The efficiency of the proposed algorithm was illustrated via numerical simulation examples. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

17 pages, 1232 KiB

Open AccessArticle

State Transition for Statistical SLAM Using Planar Features in 3D Point Clouds

by Amirali Khodadadian Gostar, Chunyun Fu, Weiqin Chuah, Mohammed Imran Hossain, Ruwan Tennakoon, Alireza Bab-Hadiashar and Reza Hoseinnezhad

Sensors 2019, 19(7), 1614; https://doi.org/10.3390/s19071614 - 3 Apr 2019

Cited by 10 | Viewed by 3352

Abstract

There is a large body of literature on solving the SLAM problem for various autonomous vehicle applications. A substantial part of the solutions is formulated based on using statistical (mainly Bayesian) filters such as Kalman filter and its extended version. In such solutions, [...] Read more.

There is a large body of literature on solving the SLAM problem for various autonomous vehicle applications. A substantial part of the solutions is formulated based on using statistical (mainly Bayesian) filters such as Kalman filter and its extended version. In such solutions, the measurements are commonly some point features or detections collected by the sensor(s) on board the autonomous vehicle. With the increasing utilization of scanners with common autonomous cars, and availability of 3D point clouds in real-time and at fast rates, it is now possible to use more sophisticated features extracted from the point clouds for filtering. This paper presents the idea of using planar features with multi-object Bayesian filters for SLAM. With Bayesian filters, the first step is prediction, where the object states are propagated to the next time based on a stochastic transition model. We first present how such a transition model can be developed, and then propose a solution for state prediction. In the simulation studies, using a dataset of measurements acquired from real vehicle sensors, we apply the proposed model to predict the next planar features and vehicle states. The results show reasonable accuracy and efficiency for statistical filtering-based SLAM applications. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

21 pages, 15628 KiB

Open AccessArticle

A Grey Wolf Optimization-Based Track-Before-Detect Method for Maneuvering Extended Target Detection and Tracking

by Bo Yan, Xu Yang Zhao, Na Xu, Yu Chen and Wen Bo Zhao

Sensors 2019, 19(7), 1577; https://doi.org/10.3390/s19071577 - 1 Apr 2019

Cited by 9 | Viewed by 3093

Abstract

A grey wolf optimization-based track-before-detect (GWO-TBD) method is developed for extended target detection and tracking. The aim of the GWO-TBD is tracking weak and maneuvering extended targets in a cluttered environment using the measurement points of an air surveillance radar. The optimal solution [...] Read more.

A grey wolf optimization-based track-before-detect (GWO-TBD) method is developed for extended target detection and tracking. The aim of the GWO-TBD is tracking weak and maneuvering extended targets in a cluttered environment using the measurement points of an air surveillance radar. The optimal solution is the trajectory constituted by the points of an extended target. At the beginning of the GWO-TBD, the measurements of each scan are clustered into alternative sets. Secondly, closely sets are associated for tracklets. Each tracklet equals a candidate solution. Thirdly, the tracklets are further associated iteratively to find a better solution. An improved GWO algorithm is developed in the iteration for removal of unappreciated solution and acceleration of convergence. After the iteration of several generations, the optimal solution can be achieved, i.e. trajectory of an extended target. Both the real data and synthetic data are performed with the GWO-TBD and several existing algorithms in this work. Result infers that the GWO-TBD is superior to the others in detecting and tracking maneuvering targets. Meanwhile, much less prior information is necessary in the GWO-TBD. It makes the approach is engineering friendly. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

20 pages, 5741 KiB

Open AccessArticle

A Performance Evaluation Scheme for Multiple Object Tracking with HFSWR

by Kun Wang, Pengju Zhang, Jiong Niu, Weifeng Sun, Lun Zhao and Yonggang Ji

Sensors 2019, 19(6), 1393; https://doi.org/10.3390/s19061393 - 21 Mar 2019

Cited by 5 | Viewed by 2851

Abstract

High-frequency surface wave radar (HFSWR) can detect and continuously track ship objects in real time and beyond the horizon. When ships navigate in a sea area, their motions in a time period form a scenario. The diversity and complexity of the motion scenarios [...] Read more.

High-frequency surface wave radar (HFSWR) can detect and continuously track ship objects in real time and beyond the horizon. When ships navigate in a sea area, their motions in a time period form a scenario. The diversity and complexity of the motion scenarios make it difficult to accurately track ships, in which failures such as track fragmentation (TF) are frequently observed. However, it is still unclear how and to what degrees the motions of ships affect the tracking performance, especially which motion patterns can cause tracking failures. This paper addresses this problem and attempts to undertake a first step towards providing an intensive quantitative performance assessment and vulnerability detection scheme for ship-tracking algorithms by proposing an evolutionary and data-mining-based approach. Low-dimensional scenarios in terms of multiple maneuvering ship objects are generated using a grammar-based model. Closed-loop feedback is introduced using evolutionary computation to efficiently collect scenarios that cause more and more tracking performance loss, which provides diversified cases for analysing using data-mining technique to discover indicators of tracking vulnerability. Results on different tracking algorithms show that more cluster and convergence patterns and longer duration of our convoy and cluster patterns in the scenarios can cause severer TF to HFSWR ship tracking. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Graphical abstract

21 pages, 917 KiB

Open AccessArticle

Target Tracking in Clutter with Extremum Seeking Control for Adaptive Detection Thresholding

by Seung Hyo Park and Taek Lyul Song

Sensors 2019, 19(6), 1386; https://doi.org/10.3390/s19061386 - 20 Mar 2019

Cited by 2 | Viewed by 2771

Abstract

If the signal strength obtained from sonar is higher than the predefined detection threshold, it is considered as a candidate for target tracking. This detection threshold is a parameter that affects the detection probability of targets and the distribution of clutter measurements, so [...] Read more.

If the signal strength obtained from sonar is higher than the predefined detection threshold, it is considered as a candidate for target tracking. This detection threshold is a parameter that affects the detection probability of targets and the distribution of clutter measurements, so it is important to determine a proper threshold to improve target tracking performance. There are various techniques for adjusting the detection threshold. To apply these techniques, it is assumed that the probability density functions of the signal strength for clutter are known in advance. However, in a real environment, the probability density function of the signal strength is unknown in general. In this paper, we propose a detection threshold control method using extremum seeking control in realistic environments. The extremum seeking control is a method used in complex nonlinear systems. We propose a new structure for extremum seeking control that is applicable to detection processes with nonlinear characteristics. This structure is used to adjust the detection threshold of the received signal amplitude to make the estimated clutter measurement density converge to a designed clutter measurement density to achieve the best target tracking performance in the current environment. Simulation studies for the proposed extremum seeking control applied to target tracking in an unknown clutter signal distribution demonstrate the effectiveness and improved target tracking performance. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

25 pages, 424 KiB

Open AccessArticle

An Efficient Multi-Path Multitarget Tracking Algorithm for Over-The-Horizon Radar

by Yuan Huang, Yifang Shi and Taek Lyul Song

Sensors 2019, 19(6), 1384; https://doi.org/10.3390/s19061384 - 20 Mar 2019

Cited by 12 | Viewed by 3418

Abstract

In target tracking environments using over-the-horizon radar (OTHR), one target may generate multiple detections through different signal propagation paths. Trackers need to jointly handle the uncertainties stemming from both measurement origin and measurement path. Traditional multitarget tracking algorithms suffer from high computational loads [...] Read more.

In target tracking environments using over-the-horizon radar (OTHR), one target may generate multiple detections through different signal propagation paths. Trackers need to jointly handle the uncertainties stemming from both measurement origin and measurement path. Traditional multitarget tracking algorithms suffer from high computational loads in such environments since they need to enumerate all possible joint measurement-to-track assignments considering the measurements paths unless they employ some approximations regarding the measurements and their corresponding paths. In this paper, we propose a novel algorithm, named multi-path linear multitarget integrated probabilistic data association (MP-LM-IPDA), to efficiently track multitarget in multiple detection environments. Instead of generating all possible joint assignments, MP-LM-IPDA calculates the modulated clutter measurement density for each measurement cell of each track. The modulated clutter measurement density considers the possibility that the measurement cells originate from the clutter as well as from other potential targets. By incorporating the modulated clutter measurement density, the single target tracking structure can be applied for multitarget tracking, which significantly reduces the computational load. The simulation results demonstrate the effectiveness and efficiency of the proposed algorithm. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

19 pages, 3812 KiB

Open AccessArticle

Adaptive Unscented Kalman Filter for Target Tracking with Unknown Time-Varying Noise Covariance

by Baoshuang Ge, Hai Zhang, Liuyang Jiang, Zheng Li and Maaz Mohammed Butt

Sensors 2019, 19(6), 1371; https://doi.org/10.3390/s19061371 - 19 Mar 2019

Cited by 45 | Viewed by 6247

Abstract

The unscented Kalman filter (UKF) is widely used to address the nonlinear problems in target tracking. However, this standard UKF shows unstable performance whenever the noise covariance mismatches. Furthermore, in consideration of the deficiencies of the current adaptive UKF algorithm, this paper proposes [...] Read more.

The unscented Kalman filter (UKF) is widely used to address the nonlinear problems in target tracking. However, this standard UKF shows unstable performance whenever the noise covariance mismatches. Furthermore, in consideration of the deficiencies of the current adaptive UKF algorithm, this paper proposes a new adaptive UKF scheme for the time-varying noise covariance problems. First of all, the cross-correlation between the innovation and residual sequences is given and proven. On this basis, a linear matrix equation deduced from the innovation and residual sequences is applied to resolve the process noise covariance in real time. Using the redundant measurements, an improved measurement-based adaptive Kalman filtering algorithm is applied to estimate the measurement noise covariance, which is entirely immune to the state estimation. The results of the simulation indicate that under the condition of time-varying noise covariances, the proposed adaptive UKF outperforms the standard UKF and the current adaptive UKF algorithm, hence improving tracking accuracy and stability. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

23 pages, 1031 KiB

Open AccessArticle

Resolvable Group State Estimation with Maneuver Based on Labeled RFS and Graph Theory

by Weifeng Liu and Yudong Chi

Sensors 2019, 19(6), 1307; https://doi.org/10.3390/s19061307 - 15 Mar 2019

Cited by 12 | Viewed by 2455

Abstract

In this paper, multiple resolvable group target tracking was considered in the frame of random finite sets. In particular, a group target model was introduced by combining graph theory with the labeled random finite sets (RFS). This accounted for dependence between group members. [...] Read more.

In this paper, multiple resolvable group target tracking was considered in the frame of random finite sets. In particular, a group target model was introduced by combining graph theory with the labeled random finite sets (RFS). This accounted for dependence between group members. Simulations were presented to verify the proposed algorithm. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

15 pages, 1364 KiB

Open AccessArticle

Double-Layer Cubature Kalman Filter for Nonlinear Estimation

by Feng Yang, Yujuan Luo and Litao Zheng

Sensors 2019, 19(5), 986; https://doi.org/10.3390/s19050986 - 26 Feb 2019

Cited by 9 | Viewed by 3366

Abstract

The cubature Kalman filter (CKF) has poor performance in strongly nonlinear systems while the cubature particle filter has high computational complexity induced by stochastic sampling. To address these problems, a novel CKF named double-Layer cubature Kalman filter (DLCKF) is proposed. In the proposed [...] Read more.

The cubature Kalman filter (CKF) has poor performance in strongly nonlinear systems while the cubature particle filter has high computational complexity induced by stochastic sampling. To address these problems, a novel CKF named double-Layer cubature Kalman filter (DLCKF) is proposed. In the proposed DLCKF, the prior distribution is represented by a set of weighted deterministic sampling points, and each deterministic sampling point is updated by the inner CKF. Finally, the update mechanism of the outer CKF is used to obtain the state estimations. Simulation results show that the proposed algorithm has not only high estimation accuracy but also low computational complexity, compared with the state-of-the-art filtering algorithms. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

18 pages, 2603 KiB

Open AccessArticle

Multi-Objective Optimization Based Multi-Bernoulli Sensor Selection for Multi-Target Tracking

by Yun Zhu, Jun Wang and Shuang Liang

Sensors 2019, 19(4), 980; https://doi.org/10.3390/s19040980 - 25 Feb 2019

Cited by 17 | Viewed by 4690

Abstract

This paper presents a novel multi-objective optimization based sensor selection method for multi-target tracking in sensor networks. The multi-target states are modelled as multi-Bernoulli random finite sets and the multi-Bernoulli filter is used to propagate the multi-target posterior density. The proposed method is [...] Read more.

This paper presents a novel multi-objective optimization based sensor selection method for multi-target tracking in sensor networks. The multi-target states are modelled as multi-Bernoulli random finite sets and the multi-Bernoulli filter is used to propagate the multi-target posterior density. The proposed method is designed to select the sensor that provides the most reliable cardinality estimate, since more accurate cardinality estimate indicates more accurate target states. In the multi-Bernoulli filter, the updated multi-target density is a multi-Bernoulli random finite set formed by a union of legacy tracks and measurement-updated tracks. The legacy track and the measurement-updated track have different theoretical and physical meanings, and hence these two kinds of tracks are considered separately in the sensor management problem. Specifically, two objectives are considered: (1) maximizing the mean cardinality of the measurement-updated tracks, (2) minimizing the cardinality variance of the legacy tracks. Considering the conflicting objectives simultaneously is a multi-objective optimization problem. Tradeoff solutions between two conflicting objectives will be derived. Theoretical analysis and examples show that the proposed approach is effective and direct. The performance of the proposed method is demonstrated using two scenarios with different levels of observability of targets in the passive sensor network. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

28 pages, 1386 KiB

Open AccessArticle

Optimal Shadowing Filter for a Positioning and Tracking Methodology with Limited Information

by Ayham Zaitouny, Thomas Stemler and Shannon Dee Algar

Sensors 2019, 19(4), 931; https://doi.org/10.3390/s19040931 - 22 Feb 2019

Cited by 10 | Viewed by 2965

Abstract

Positioning and tracking a moving target from limited positional information is a frequently-encountered problem. For given noisy observations of the target’s position, one wants to estimate the true trajectory and reconstruct the full phase space including velocity and acceleration. The shadowing filter offers [...] Read more.

Positioning and tracking a moving target from limited positional information is a frequently-encountered problem. For given noisy observations of the target’s position, one wants to estimate the true trajectory and reconstruct the full phase space including velocity and acceleration. The shadowing filter offers a robust methodology to achieve such an estimation and reconstruction. Here, we highlight and validate important merits of this methodology for real-life applications. In particular, we explore the filter’s performance when dealing with correlated or uncorrelated noise, irregular sampling in time and how it can be optimised even when the true dynamics of the system are not known. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

19 pages, 8765 KiB

Open AccessArticle

A Three-Dimensional Hough Transform-Based Track-Before-Detect Technique for Detecting Extended Targets in Strong Clutter Backgrounds

by Bo Yan, Na Xu, Wen-Bo Zhao and Lu-Ping Xu

Sensors 2019, 19(4), 881; https://doi.org/10.3390/s19040881 - 20 Feb 2019

Cited by 29 | Viewed by 4385

Abstract

Hough Transform (HT), which has a low sensitivity to local faults and good ability in suppressing noise and clutters, usually applies to trajectory detection in a cluttered environment. This paper describes its application for detecting the trajectories of extended targets in three-dimensional measurements, [...] Read more.

Hough Transform (HT), which has a low sensitivity to local faults and good ability in suppressing noise and clutters, usually applies to trajectory detection in a cluttered environment. This paper describes its application for detecting the trajectories of extended targets in three-dimensional measurements, i.e., a two-dimensional positional information and its measuring time. For taking the full merits of a multi-scan, the measuring time is regarded as a variable for the time axis. This correspondence extends the HT to 3-dimensional data. Meanwhile, a three-dimensional accumulator matrix is built for the purpose of voting. The voting process is done in an iterative way by selecting the 3D-line with the most votes and removing the corresponding measurements in each step. The three dimensional Hough Transform-based extended target track-before-detect technique (3DHT-ET-TBD), proposed here, is suitable to track the extended target and non-extended target simultaneously and few false alarm trajectories arise. Both the real data and simulated data are exploited to evaluate its performance. Compared with the Gaussian Mixture Probability Hypothesis Density (GM-PHD) filter based methods and a 4DHT-TBD algorithm, the 3DHT-ET-TBD is a more promising approach for multi-extended target tracking problems due to its high efficiency and low computation, especially in situations where the noise and false alarms are considerably high but few measurements are generated by the extended targets. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

15 pages, 19556 KiB

Open AccessArticle

Data Association for Multi-Object Tracking via Deep Neural Networks

by Kwangjin Yoon, Du Yong Kim, Young-Chul Yoon and Moongu Jeon

Sensors 2019, 19(3), 559; https://doi.org/10.3390/s19030559 - 29 Jan 2019

Cited by 60 | Viewed by 9812

Abstract

With recent advances in object detection, the tracking-by-detection method has become mainstream for multi-object tracking in computer vision. The tracking-by-detection scheme necessarily has to resolve a problem of data association between existing tracks and newly received detections at each frame. In this paper, [...] Read more.

With recent advances in object detection, the tracking-by-detection method has become mainstream for multi-object tracking in computer vision. The tracking-by-detection scheme necessarily has to resolve a problem of data association between existing tracks and newly received detections at each frame. In this paper, we propose a new deep neural network (DNN) architecture that can solve the data association problem with a variable number of both tracks and detections including false positives. The proposed network consists of two parts: encoder and decoder. The encoder is the fully connected network with several layers that take bounding boxes of both detection and track-history as inputs. The outputs of the encoder are sequentially fed into the decoder which is composed of the bi-directional Long Short-Term Memory (LSTM) networks with a projection layer. The final output of the proposed network is an association matrix that reflects matching scores between tracks and detections. To train the network, we generate training samples using the annotation of Stanford Drone Dataset (SDD). The experiment results show that the proposed network achieves considerably high recall and precision rate as the binary classifier for the assignment tasks. We apply our network to track multiple objects on real-world datasets and evaluate the tracking performance. The performance of our tracker outperforms previous works based on DNN and comparable to other state-of-the-art methods. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

34 pages, 20190 KiB

Open AccessArticle

Behavioral Pedestrian Tracking Using a Camera and LiDAR Sensors on a Moving Vehicle

by Martin Dimitrievski, Peter Veelaert and Wilfried Philips

Sensors 2019, 19(2), 391; https://doi.org/10.3390/s19020391 - 18 Jan 2019

Cited by 57 | Viewed by 8661

Abstract

In this paper, we present a novel 2D–3D pedestrian tracker designed for applications in autonomous vehicles. The system operates on a tracking by detection principle and can track multiple pedestrians in complex urban traffic situations. By using a behavioral motion model and a [...] Read more.

In this paper, we present a novel 2D–3D pedestrian tracker designed for applications in autonomous vehicles. The system operates on a tracking by detection principle and can track multiple pedestrians in complex urban traffic situations. By using a behavioral motion model and a non-parametric distribution as state model, we are able to accurately track unpredictable pedestrian motion in the presence of heavy occlusion. Tracking is performed independently, on the image and ground plane, in global, motion compensated coordinates. We employ Camera and LiDAR data fusion to solve the association problem where the optimal solution is found by matching 2D and 3D detections to tracks using a joint log-likelihood observation model. Each 2D–3D particle filter then updates their state from associated observations and a behavioral motion model. Each particle moves independently following the pedestrian motion parameters which we learned offline from an annotated training dataset. Temporal stability of the state variables is achieved by modeling each track as a Markov Decision Process with probabilistic state transition properties. A novel track management system then handles high level actions such as track creation, deletion and interaction. Using a probabilistic track score the track manager can cull false and ambiguous detections while updating tracks with detections from actual pedestrians. Our system is implemented on a GPU and exploits the massively parallelizable nature of particle filters. Due to the Markovian nature of our track representation, the system achieves real-time performance operating with a minimal memory footprint. Exhaustive and independent evaluation of our tracker was performed by the KITTI benchmark server, where it was tested against a wide variety of unknown pedestrian tracking situations. On this realistic benchmark, we outperform all published pedestrian trackers in a multitude of tracking metrics. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

19 pages, 390 KiB

Open AccessArticle

Markov Chain Realization of Multiple Detection Joint Integrated Probabilistic Data Association

by Yuan Huang, Taek Lyul Song and Dae Hoon Cheagal

Sensors 2019, 19(1), 112; https://doi.org/10.3390/s19010112 - 30 Dec 2018

Cited by 4 | Viewed by 3604

Abstract

In multiple detection target tracking environments, PDA-based algorithms such as multiple detection joint integrated probabilistic data association (MD-JIPDA) utilize the measurement partition method to generate measurement cells. Thus, one-to-many track-to-measurements associations can be realized. However, in this structure, the number of joint data [...] Read more.

In multiple detection target tracking environments, PDA-based algorithms such as multiple detection joint integrated probabilistic data association (MD-JIPDA) utilize the measurement partition method to generate measurement cells. Thus, one-to-many track-to-measurements associations can be realized. However, in this structure, the number of joint data association events grows exponentially with the number of measurement cells and the number of tracks. MD-JIPDA is plagued by large increases in computational complexity when targets are closely spaced or move cross each other, especially in multiple detection scenarios. Here, the multiple detection Markov chain joint integrated probabilistic data association (MD-MC-JIPDA) is proposed, in which a Markov chain is used to generate random data association sequences. These sequences are substitutes for the association events. The Markov chain process significantly reduces the computational cost since only a few association sequences are generated while keeping preferable tracking performance. Finally, MD-MC-JIPDA is experimentally validated to demonstrate its effectiveness compared with some of the existing multiple detection data association algorithms. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

18 pages, 1838 KiB

Open AccessArticle

Strategy for Fault Diagnosis on Train Plug Doors Using Audio Sensors

by Yongkui Sun, Guo Xie, Yuan Cao and Tao Wen

Sensors 2019, 19(1), 3; https://doi.org/10.3390/s19010003 - 20 Dec 2018

Cited by 21 | Viewed by 4334

Abstract

As the only entry/exit for passengers getting on and off a train, the train plug door is of great importance to keep train operation safe and reliable. As signal processing technologies develop rapidly, taking the easy acquisition advantages of sound signals, a novel [...] Read more.

As the only entry/exit for passengers getting on and off a train, the train plug door is of great importance to keep train operation safe and reliable. As signal processing technologies develop rapidly, taking the easy acquisition advantages of sound signals, a novel fault diagnosis method for train plug doors using multi-scale normalized permutation entropy (MNPE) and an improved particle swarm optimization based multi-class support vector machine (IPSO-MSVM) is proposed. Firstly, sound samples are collected using high-precision audio sensor. In the features extraction process, a hybrid method blending empirical mode decomposition (EMD), multi-scale permutation entropy (MNPE) with Fisher discrimination criterion is utilized. First, EMD is used to decompose each sound signal into several intrinsic mode functions (IMFs) and a residue for stationary processing. Then, MNPE features are extracted from the IMFs. To obtain the most significant features, the Fisher discrimination criterion is further applied. To address the time-consuming defects of traditional grid based method for selecting the optimal parameters of multi-class SVM, an improved PSO (IPSO) is proposed. The superiority of the IPSO-MSVM model and the hybrid feature extraction method was tested on the collected sound samples by comparing to commonly applied methods. Results indicate the identification accuracy of the proposed method is highest, which reaches 90.54%, demonstrating its feasibility. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

19 pages, 1707 KiB

Open AccessArticle

Time-Matching Random Finite Set-Based Filter for Radar Multi-Target Tracking

by Defu Jiang, Ming Liu, Yiyue Gao, Yang Gao, Wei Fu and Yan Han

Sensors 2018, 18(12), 4416; https://doi.org/10.3390/s18124416 - 13 Dec 2018

Cited by 10 | Viewed by 3852

Abstract

The random finite set (RFS) approach provides an elegant Bayesian formulation of the multi-target tracking (MTT) problem without the requirement of explicit data association. In order to improve the performance of the RFS-based filter in radar MTT applications, this paper proposes a time-matching [...] Read more.

The random finite set (RFS) approach provides an elegant Bayesian formulation of the multi-target tracking (MTT) problem without the requirement of explicit data association. In order to improve the performance of the RFS-based filter in radar MTT applications, this paper proposes a time-matching Bayesian filtering framework to deal with the problem caused by the diversity of target sampling times. Based on this framework, we develop a time-matching joint generalized labeled multi-Bernoulli filter and a time-matching probability hypothesis density filter. Simulations are performed by their Gaussian mixture implementations. The results show that the proposed approach can improve the accuracy of target state estimation, as well as the robustness. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

29 pages, 3171 KiB

Open AccessArticle

Sensor Selection for Decentralized Large-Scale Multi-Target Tracking Network

by Feng Lian, Liming Hou, Bo Wei and Chongzhao Han

Sensors 2018, 18(12), 4115; https://doi.org/10.3390/s18124115 - 23 Nov 2018

Cited by 9 | Viewed by 3141

Abstract

A new optimization algorithm of sensor selection is proposed in this paper for decentralized large-scale multi-target tracking (MTT) network within a labeled random finite set (RFS) framework. The method is performed based on a marginalized δ-generalized labeled multi-Bernoulli RFS. The rule of weighted [...] Read more.

A new optimization algorithm of sensor selection is proposed in this paper for decentralized large-scale multi-target tracking (MTT) network within a labeled random finite set (RFS) framework. The method is performed based on a marginalized δ-generalized labeled multi-Bernoulli RFS. The rule of weighted Kullback-Leibler average (KLA) is used to fuse local multi-target densities. A new metric, named as the label assignment (LA) metric, is proposed to measure the distance for two labeled sets. The lower bound of LA metric based mean square error between the labeled multi-target state set and its estimate is taken as the optimized objective function of sensor selection. The proposed bound is obtained by the information inequality to RFS measurement. Then, we present the sequential Monte Carlo and Gaussian mixture implementations for the bound. Another advantage of the bound is that it provides a basis for setting the weights of KLA. The coordinate descent method is proposed to compromise the computational cost of sensor selection and the accuracy of MTT. Simulations verify the effectiveness of our method under different signal-to- noise ratio scenarios. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

18 pages, 2282 KiB

Open AccessArticle

Target Localization and Tracking by Fusing Doppler Differentials from Cellular Emanations with a Multi-Spectral Video Tracker

by Casey D. Demars, Michael C. Roggemann, Adam J. Webb and Timothy C. Havens

Sensors 2018, 18(11), 3687; https://doi.org/10.3390/s18113687 - 30 Oct 2018

Cited by 7 | Viewed by 3139

Abstract

We present an algorithm for fusing data from a constellation of RF sensors detecting cellular emanations with the output of a multi-spectral video tracker to localize and track a target with a specific cell phone. The RF sensors measure the Doppler shift caused [...] Read more.

We present an algorithm for fusing data from a constellation of RF sensors detecting cellular emanations with the output of a multi-spectral video tracker to localize and track a target with a specific cell phone. The RF sensors measure the Doppler shift caused by the moving cellular emanation and then Doppler differentials between all sensor pairs are calculated. The multi-spectral video tracker uses a Gaussian mixture model to detect foreground targets and SIFT features to track targets through the video sequence. The data is fused by associating the Doppler differential from the RF sensors with the theoretical Doppler differential computed from the multi-spectral tracker output. The absolute difference and the root-mean-square difference are computed to associate the Doppler differentials from the two sensor systems. Performance of the algorithm was evaluated using synthetically generated datasets of an urban scene with multiple moving vehicles. The presented fusion algorithm correctly associates the cellular emanation with the corresponding video target for low measurement uncertainty and in the presence of favorable motion patterns. For nearly all objects the fusion algorithm has high confidence in associating the emanation with the correct multi-spectral target from the most probable background target. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

19 pages, 6677 KiB

Open AccessArticle

Joint Detection and DOA Tracking with a Bernoulli Filter Based on Information Theoretic Criteria

by Guangpu Zhang, Ce Zheng, Sibo Sun, Guolong Liang and Yifeng Zhang

Sensors 2018, 18(10), 3473; https://doi.org/10.3390/s18103473 - 15 Oct 2018

Cited by 5 | Viewed by 2984

Abstract

In this paper, we study the problem of the joint detection and direction-of-arrival (DOA) tracking of a single moving source which can randomly appear or disappear from the surveillance volume. Firstly, the Bernoulli random finite set (RFS) is employed to characterize the randomness [...] Read more.

In this paper, we study the problem of the joint detection and direction-of-arrival (DOA) tracking of a single moving source which can randomly appear or disappear from the surveillance volume. Firstly, the Bernoulli random finite set (RFS) is employed to characterize the randomness of the state process, i.e., the dynamics of the source motion and the source appearance. To increase the performance of the detection and DOA tracking in low signal-to-noise ratio (SNR) scenarios, the measurements are obtained directly from an array of sensors and allow multiple snapshots. A track-before-detect (TBD) Bernoulli filter is proposed for tracking a randomly on/off switching single dynamic system. Secondly, since the variances of the stochastic signal and measurement noise are unknown in practical applications, these nuisance parameters are marginalized by defining an uninformative prior, and the likelihood function is compensated by using the information theoretic criteria. The simulation results demonstrate the performance of the filter. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

26 pages, 1795 KiB

Open AccessArticle

Distributed Space Debris Tracking with Consensus Labeled Random Finite Set Filtering

by Baishen Wei and Brett Nener

Sensors 2018, 18(9), 3005; https://doi.org/10.3390/s18093005 - 7 Sep 2018

Cited by 10 | Viewed by 3273

Abstract

Space debris tracking is a challenge for spacecraft operation because of the increasing number of both satellites and the amount of space debris. This paper investigates space debris tracking using marginalized

δ

-generalized labeled multi-Bernoulli filtering on a network of nodes consisting of [...] Read more.

Space debris tracking is a challenge for spacecraft operation because of the increasing number of both satellites and the amount of space debris. This paper investigates space debris tracking using marginalized

δ

-generalized labeled multi-Bernoulli filtering on a network of nodes consisting of a collection of sensors with different observation volumes. A consensus algorithm is used to achieve the global average by iterative regional averages. The sensor network can have unknown or time-varying topology. The proposed space debris tracking algorithm provides an efficient solution to the key challenges (e.g., detection uncertainty, data association uncertainty, clutter, etc.) for space situational awareness. The performance of the proposed algorithm is verified by simulation results. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

Other

Jump to: Research

13 pages, 3059 KiB

Open AccessLetter

Extended Target Tracking and Feature Estimation for Optical Sensors Based on the Gaussian Process

by Haoyang Yu, Wei An and Ran Zhu

Sensors 2019, 19(7), 1704; https://doi.org/10.3390/s19071704 - 10 Apr 2019

Cited by 7 | Viewed by 3045

Abstract

A problem of tracking surface shape-shifting extended target by using gray scale pixels on optical image is considered. The measurement with amplitude information (AI) is available to the proposed method. The target is regarded as a convex hemispheric object, and the amplitude distribution [...] Read more.

A problem of tracking surface shape-shifting extended target by using gray scale pixels on optical image is considered. The measurement with amplitude information (AI) is available to the proposed method. The target is regarded as a convex hemispheric object, and the amplitude distribution of the extended target is represented by a solid radial function. The Gaussian process (GP) is applied and the covariance function of GP is modified to fit the convex hemispheric shape. The points to be estimated on the target surface are selected reasonably in the hemispheric space at the azimuth and pitch directions. Analytical representation of the estimated target extent is provided and the recursive process is implemented by the extended Kalman filter (EKF). In order to demonstrate the algorithm’s ability of tracking complex shaped targets, a trailing target characterized by two feature parameters is simulated and the two feature parameters are extracted with the estimated points. The simulations verify the validity of the proposed method with compared to classical algorithms. Full article

(This article belongs to the Special Issue Multiple Object Tracking: Making Sense of the Sensors)

► Show Figures

Figure 1

Journal Menu

Journal Browser

Multiple Object Tracking: Making Sense of the Sensors

Share This Special Issue

Special Issue Editor

Special Issue Information

Keywords

Benefits of Publishing in a Special Issue

Published Papers (25 papers)

Research

Other

Further Information

Guidelines

MDPI Initiatives

Follow MDPI