Multiple Object Tracking with Kernelized Correlation Filters in Urban Mixed Traffic

Yang, Yukun; Bilodeau, Guillaume-Alexandre

doi:10.1109/crv.2017.18

Cited by 22 publications

(22 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Earlier, object detection was done by separating the background (static part) from the objects (dynamic/moving part). Background subtraction was a major method used to accomplish this task [2,3,4]. Gaussian Mixture Model based background subtraction [5], Hybrid support vector machine (SVM with extended Kalman filter) [6] and particle filter-based trackers [7] were also used.…”

Section: Fig 1 Artificial Intelligence and Subdomainsmentioning

confidence: 99%

Intelligent Traffic Management System

Singh*¹,

Unadkat²,

Kanani³

2019

IJRTE

View full text Add to dashboard Cite

Urbanization has presented opportunities of progress which has attracted people from rural areas to the cities thus leading to mass migration. This migration has been going on for decades all around the globe and has reached a point of saturation. The area of the city remains the same but the population density has increased multiple times. Commuting for work is a scene of chaos on the roads. Though there are modes of public transport, roadway is the major mode of commute and the load on roadways is ever increasing due to the rise in population. There is hardly any scope to expand the area of the roadways. The rise in the number of vehicles each year has saturated the capacity the roads were built to carry. This leads to congestion and long hours of traffic on a daily basis which tests the patience of citizens. This provokes the daily commuters to violate the traffic rules which may sometimes amount to grave accidents. Even on Highways, the empty roads entice drivers to experience the thrill of speed overlooking the fact that they are putting themselves at risk. There have been regulations imposed to reduce the chance of an accidents by implementing rules and levying heavy fines on traffic violations. Traffic cameras have been installed all around the city to monitor for traffic violations and get hold of violators. With the technological advancements to store and process large chunks of data efficiently using techniques like Deep Learning and Computer Vision, this paper proposes an automated system to detect Traffic Violations using YOLOv3 to detect and track vehicles and save a snapshot in case a violation is committed.

show abstract

Section: Fig 1 Artificial Intelligence and Subdomainsmentioning

confidence: 99%

Intelligent Traffic Management System

Singh*¹,

Unadkat²,

Kanani³

2019

IJRTE

View full text Add to dashboard Cite

show abstract

“…In our work, we present a simple foreground matching algorithm to improve the accuracy of feature pairs. Then, we determine a frame-wide homography for every target based on a multi-target tracking method [ 13 ]. These contributions make our proposed framework more precise and have a more extensive range of applications.…”

Section: Related Workmentioning

confidence: 99%

“…For aligning non-planar scenes, we assign a reservoir for each target. Under the circumstances, multi-target tracking [ 13 ] is needed to distinguish different reservoirs. Following that, the homography of a target is estimated with all matches in its assigned reservoir.…”

Section: Proposed Frameworkmentioning

confidence: 99%

See 1 more Smart Citation

An Automatic Multi-Target Independent Analysis Framework for Non-Planar Infrared-Visible Registration

Sun

Zhang

et al. 2017

Sensors

View full text Add to dashboard Cite

In this paper, we propose a novel automatic multi-target registration framework for non-planar infrared-visible videos. Previous approaches usually analyzed multiple targets together and then estimated a global homography for the whole scene, however, these cannot achieve precise multi-target registration when the scenes are non-planar. Our framework is devoted to solving the problem using feature matching and multi-target tracking. The key idea is to analyze and register each target independently. We present a fast and robust feature matching strategy, where only the features on the corresponding foreground pairs are matched. Besides, new reservoirs based on the Gaussian criterion are created for all targets, and a multi-target tracking method is adopted to determine the relationships between the reservoirs and foreground blobs. With the matches in the corresponding reservoir, the homography of each target is computed according to its moving state. We tested our framework on both public near-planar and non-planar datasets. The results demonstrate that the proposed framework outperforms the state-of-the-art global registration method and the manual global registration matrix in all tested datasets.

show abstract

“…Recently, DCFs have been successfully explored in single object tracking applications [32] due to the high computational efficiency. To investigate how correlation filters can improve multiple human tracking, [33] and [34] have been proposed to apply multiple single object trackers based on the Kernelized Correlation Filters (KCFs) in parallel for fast tracking. In [35], authors proposed to integrate correlation filters (CFs) and a confidence-based relative motion network to perform a two-step data association to track multiple objects, where CFs are employed as a verifying step to confirm the target estimates.…”

Section: Introductionmentioning

confidence: 99%

Multi-Level Cooperative Fusion of GM-PHD Filters for Online Multiple Human Tracking

Angelini

Chambers

et al. 2019

IEEE Trans. Multimedia

View full text Add to dashboard Cite

In this paper, we propose a multi-level cooperative fusion approach to address the online multiple human tracking problem in a Gaussian Mixture Probability Hypothesis Density (GM-PHD) filter framework. The proposed fusion approach consists essentially of three steps. Firstly, we integrate two human detectors with different characteristics (full-body and bodyparts), and investigate their complementary benefits for tracking multiple targets. For each detector domain, we then propose a novel Discriminative Correlation Matching (DCM) model, and fuse it with spatio-temporal information to address ambiguous identity association in the GM-PHD filter. Finally, we develop a robust fusion center with virtual and real zones to make a global decision based on preliminary candidate targets generated by each detector. This center also mitigates the sensitivity of missed detections in the Generalized Covariance Intersection (GCI) fusion process, thereby improving the fusion performance and tracking consistency. Experiments on the MOTChallenge Benchmark demonstrate the proposed method achieves improved performance over other state-of-the-art RFS based tracking methods.

show abstract

Multiple Object Tracking with Kernelized Correlation Filters in Urban Mixed Traffic

Cited by 22 publications

References 14 publications

Intelligent Traffic Management System

Intelligent Traffic Management System

An Automatic Multi-Target Independent Analysis Framework for Non-Planar Infrared-Visible Registration

Multi-Level Cooperative Fusion of GM-PHD Filters for Online Multiple Human Tracking

Contact Info

Product

Resources

About