Moving vehicle tracking based on improved tracking–learning–detection algorithm
This study addresses the tracking–learning–detection (TLD) algorithm for long‐term single‐target tracking of moving vehicle from video streams. The problems leading to tracking failures in existing TLD methods are discovered, and an improved TLD (ITLD) tracking algorithm is proposed which is more ro...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Wiley
2019-12-01
|
Series: | IET Computer Vision |
Subjects: | |
Online Access: | https://doi.org/10.1049/iet-cvi.2018.5787 |
_version_ | 1797684057288474624 |
---|---|
author | Enzeng Dong Mengtao Deng Jigang Tong Chao Jia Shengzhi Du |
author_facet | Enzeng Dong Mengtao Deng Jigang Tong Chao Jia Shengzhi Du |
author_sort | Enzeng Dong |
collection | DOAJ |
description | This study addresses the tracking–learning–detection (TLD) algorithm for long‐term single‐target tracking of moving vehicle from video streams. The problems leading to tracking failures in existing TLD methods are discovered, and an improved TLD (ITLD) tracking algorithm is proposed which is more robust to object occlusion and illumination variation. A square root cubature Kalman filter (SRCKF) is employed in the tracker of TLD to predict the position of the object when occlusion occurs. Besides, this study introduces fast retina keypoint (FREAK) feature into the tracker to alleviate the instability caused by illumination variation or scale variation. The overlap comparison and the normalised cross‐correlation coefficient (NCC) are introduced to the integrator of the TLD to obtain reliable bounding boxes with improved tracking precision. Experiments are conducted to compare the performance of the state‐of‐the‐art trackers and the proposed method, using the object tracking benchmark that includes 50 video sequences (OTB‐50) and TLD datasets. The experimental results show that the proposed ITLD outperforms on both tracking accuracy and robustness. The proposed method can track a moving vehicle even when it is temporally totally occluded. |
first_indexed | 2024-03-12T00:24:55Z |
format | Article |
id | doaj.art-334438bdee344deba15aeadd7cee199a |
institution | Directory Open Access Journal |
issn | 1751-9632 1751-9640 |
language | English |
last_indexed | 2024-03-12T00:24:55Z |
publishDate | 2019-12-01 |
publisher | Wiley |
record_format | Article |
series | IET Computer Vision |
spelling | doaj.art-334438bdee344deba15aeadd7cee199a2023-09-15T10:41:13ZengWileyIET Computer Vision1751-96321751-96402019-12-0113873074110.1049/iet-cvi.2018.5787Moving vehicle tracking based on improved tracking–learning–detection algorithmEnzeng Dong0Mengtao Deng1Jigang Tong2Chao Jia3Shengzhi Du4Tianjin Key Laboratory For Control Theory and Applications in Complicated Systems, Tianjin University of TechnologyTianjin300384People's Republic of ChinaTianjin Key Laboratory For Control Theory and Applications in Complicated Systems, Tianjin University of TechnologyTianjin300384People's Republic of ChinaTianjin Key Laboratory For Control Theory and Applications in Complicated Systems, Tianjin University of TechnologyTianjin300384People's Republic of ChinaTianjin Key Laboratory For Control Theory and Applications in Complicated Systems, Tianjin University of TechnologyTianjin300384People's Republic of ChinaDepartment of Electrical EngineeringTshwane University of TechnologyPretoria0001South AfricaThis study addresses the tracking–learning–detection (TLD) algorithm for long‐term single‐target tracking of moving vehicle from video streams. The problems leading to tracking failures in existing TLD methods are discovered, and an improved TLD (ITLD) tracking algorithm is proposed which is more robust to object occlusion and illumination variation. A square root cubature Kalman filter (SRCKF) is employed in the tracker of TLD to predict the position of the object when occlusion occurs. Besides, this study introduces fast retina keypoint (FREAK) feature into the tracker to alleviate the instability caused by illumination variation or scale variation. The overlap comparison and the normalised cross‐correlation coefficient (NCC) are introduced to the integrator of the TLD to obtain reliable bounding boxes with improved tracking precision. Experiments are conducted to compare the performance of the state‐of‐the‐art trackers and the proposed method, using the object tracking benchmark that includes 50 video sequences (OTB‐50) and TLD datasets. The experimental results show that the proposed ITLD outperforms on both tracking accuracy and robustness. The proposed method can track a moving vehicle even when it is temporally totally occluded.https://doi.org/10.1049/iet-cvi.2018.5787illumination variationsquare root cubature Kalman filterimproved tracking precisionTLD datasetsITLDtracking accuracy |
spellingShingle | Enzeng Dong Mengtao Deng Jigang Tong Chao Jia Shengzhi Du Moving vehicle tracking based on improved tracking–learning–detection algorithm IET Computer Vision illumination variation square root cubature Kalman filter improved tracking precision TLD datasets ITLD tracking accuracy |
title | Moving vehicle tracking based on improved tracking–learning–detection algorithm |
title_full | Moving vehicle tracking based on improved tracking–learning–detection algorithm |
title_fullStr | Moving vehicle tracking based on improved tracking–learning–detection algorithm |
title_full_unstemmed | Moving vehicle tracking based on improved tracking–learning–detection algorithm |
title_short | Moving vehicle tracking based on improved tracking–learning–detection algorithm |
title_sort | moving vehicle tracking based on improved tracking learning detection algorithm |
topic | illumination variation square root cubature Kalman filter improved tracking precision TLD datasets ITLD tracking accuracy |
url | https://doi.org/10.1049/iet-cvi.2018.5787 |
work_keys_str_mv | AT enzengdong movingvehicletrackingbasedonimprovedtrackinglearningdetectionalgorithm AT mengtaodeng movingvehicletrackingbasedonimprovedtrackinglearningdetectionalgorithm AT jigangtong movingvehicletrackingbasedonimprovedtrackinglearningdetectionalgorithm AT chaojia movingvehicletrackingbasedonimprovedtrackinglearningdetectionalgorithm AT shengzhidu movingvehicletrackingbasedonimprovedtrackinglearningdetectionalgorithm |