A Fast and Accurate Approach to Multiple-Vehicle Localization and Tracking from Monocular Aerial Images
In this work we present a novel end-to-end solution for tracking objects (i.e., vessels), using video streams from aerial drones, in dynamic maritime environments. Our method relies on deep features, which are learned using realistic simulation data, for robust object detection, segmentation and tra...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-12-01
|
| Series: | Journal of Imaging |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2313-433X/7/12/270 |
| _version_ | 1797503349611823104 |
|---|---|
| author | Daniel Tøttrup Stinus Lykke Skovgaard Jonas le Fevre Sejersen Rui Pimentel de Figueiredo |
| author_facet | Daniel Tøttrup Stinus Lykke Skovgaard Jonas le Fevre Sejersen Rui Pimentel de Figueiredo |
| author_sort | Daniel Tøttrup |
| collection | DOAJ |
| description | In this work we present a novel end-to-end solution for tracking objects (i.e., vessels), using video streams from aerial drones, in dynamic maritime environments. Our method relies on deep features, which are learned using realistic simulation data, for robust object detection, segmentation and tracking. Furthermore, we propose the use of rotated bounding-box representations, which are computed by taking advantage of pixel-level object segmentation, for improved tracking accuracy, by reducing erroneous data associations during tracking, when combined with the appearance-based features. A thorough set of experiments and results obtained in a realistic shipyard simulation environment, demonstrate that our method can accurately, and fast detect and track dynamic objects seen from a top-view. |
| first_indexed | 2024-03-10T03:49:20Z |
| format | Article |
| id | doaj.art-fa9e54c4dd9d45bc82715c4949457023 |
| institution | Directory Open Access Journal |
| issn | 2313-433X |
| language | English |
| last_indexed | 2024-03-10T03:49:20Z |
| publishDate | 2021-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Journal of Imaging |
| spelling | doaj.art-fa9e54c4dd9d45bc82715c49494570232023-11-23T09:00:46ZengMDPI AGJournal of Imaging2313-433X2021-12-0171227010.3390/jimaging7120270A Fast and Accurate Approach to Multiple-Vehicle Localization and Tracking from Monocular Aerial ImagesDaniel Tøttrup0Stinus Lykke Skovgaard1Jonas le Fevre Sejersen2Rui Pimentel de Figueiredo3Department of Electrical and Computer Engineering, Aarhus University, Nordre Ringgade, 18000 Aarhus, DenmarkDepartment of Electrical and Computer Engineering, Aarhus University, Nordre Ringgade, 18000 Aarhus, DenmarkDepartment of Electrical and Computer Engineering, Aarhus University, Nordre Ringgade, 18000 Aarhus, DenmarkDepartment of Electrical and Computer Engineering, Aarhus University, Nordre Ringgade, 18000 Aarhus, DenmarkIn this work we present a novel end-to-end solution for tracking objects (i.e., vessels), using video streams from aerial drones, in dynamic maritime environments. Our method relies on deep features, which are learned using realistic simulation data, for robust object detection, segmentation and tracking. Furthermore, we propose the use of rotated bounding-box representations, which are computed by taking advantage of pixel-level object segmentation, for improved tracking accuracy, by reducing erroneous data associations during tracking, when combined with the appearance-based features. A thorough set of experiments and results obtained in a realistic shipyard simulation environment, demonstrate that our method can accurately, and fast detect and track dynamic objects seen from a top-view.https://www.mdpi.com/2313-433X/7/12/270object detectionmultiple object trackingconvolutional neural networks |
| spellingShingle | Daniel Tøttrup Stinus Lykke Skovgaard Jonas le Fevre Sejersen Rui Pimentel de Figueiredo A Fast and Accurate Approach to Multiple-Vehicle Localization and Tracking from Monocular Aerial Images Journal of Imaging object detection multiple object tracking convolutional neural networks |
| title | A Fast and Accurate Approach to Multiple-Vehicle Localization and Tracking from Monocular Aerial Images |
| title_full | A Fast and Accurate Approach to Multiple-Vehicle Localization and Tracking from Monocular Aerial Images |
| title_fullStr | A Fast and Accurate Approach to Multiple-Vehicle Localization and Tracking from Monocular Aerial Images |
| title_full_unstemmed | A Fast and Accurate Approach to Multiple-Vehicle Localization and Tracking from Monocular Aerial Images |
| title_short | A Fast and Accurate Approach to Multiple-Vehicle Localization and Tracking from Monocular Aerial Images |
| title_sort | fast and accurate approach to multiple vehicle localization and tracking from monocular aerial images |
| topic | object detection multiple object tracking convolutional neural networks |
| url | https://www.mdpi.com/2313-433X/7/12/270 |
| work_keys_str_mv | AT danieltøttrup afastandaccurateapproachtomultiplevehiclelocalizationandtrackingfrommonocularaerialimages AT stinuslykkeskovgaard afastandaccurateapproachtomultiplevehiclelocalizationandtrackingfrommonocularaerialimages AT jonaslefevresejersen afastandaccurateapproachtomultiplevehiclelocalizationandtrackingfrommonocularaerialimages AT ruipimenteldefigueiredo afastandaccurateapproachtomultiplevehiclelocalizationandtrackingfrommonocularaerialimages AT danieltøttrup fastandaccurateapproachtomultiplevehiclelocalizationandtrackingfrommonocularaerialimages AT stinuslykkeskovgaard fastandaccurateapproachtomultiplevehiclelocalizationandtrackingfrommonocularaerialimages AT jonaslefevresejersen fastandaccurateapproachtomultiplevehiclelocalizationandtrackingfrommonocularaerialimages AT ruipimenteldefigueiredo fastandaccurateapproachtomultiplevehiclelocalizationandtrackingfrommonocularaerialimages |