Boosting Multi-Vehicle Tracking with a Joint Object Detection and Viewpoint Estimation Sensor
In this work, we address the problem of multi-vehicle detection and tracking for traffic monitoring applications. We preset a novel intelligent visual sensor for tracking-by-detection with simultaneous pose estimation. Essentially, we adapt an Extended Kalman Filter (EKF) to work not only with the d...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2019-09-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/19/19/4062 |
| _version_ | 1817992813891026944 |
|---|---|
| author | Roberto J. López-Sastre Carlos Herranz-Perdiguero Ricardo Guerrero-Gómez-Olmedo Daniel Oñoro-Rubio Saturnino Maldonado-Bascón |
| author_facet | Roberto J. López-Sastre Carlos Herranz-Perdiguero Ricardo Guerrero-Gómez-Olmedo Daniel Oñoro-Rubio Saturnino Maldonado-Bascón |
| author_sort | Roberto J. López-Sastre |
| collection | DOAJ |
| description | In this work, we address the problem of multi-vehicle detection and tracking for traffic monitoring applications. We preset a novel intelligent visual sensor for tracking-by-detection with simultaneous pose estimation. Essentially, we adapt an Extended Kalman Filter (EKF) to work not only with the detections of the vehicles but also with their estimated coarse viewpoints, directly obtained with the vision sensor. We show that enhancing the tracking with observations of the vehicle pose, results in a better estimation of the vehicles trajectories. For the simultaneous object detection and viewpoint estimation task, we present and evaluate two independent solutions. One is based on a fast GPU implementation of a Histogram of Oriented Gradients (HOG) detector with Support Vector Machines (SVMs). For the second, we adequately modify and train the Faster R-CNN deep learning model, in order to recover from it not only the object localization but also an estimation of its pose. Finally, we publicly release a challenging dataset, the GRAM Road Traffic Monitoring (GRAM-RTM), which has been especially designed for evaluating multi-vehicle tracking approaches within the context of traffic monitoring applications. It comprises more than 700 unique vehicles annotated across more than 40.300 frames of three videos. We expect the GRAM-RTM becomes a benchmark in vehicle detection and tracking, providing the computer vision and intelligent transportation systems communities with a standard set of images, annotations and evaluation procedures for multi-vehicle tracking. We present a thorough experimental evaluation of our approaches with the GRAM-RTM, which will be useful for establishing further comparisons. The results obtained confirm that the simultaneous integration of vehicle localizations and pose estimations as observations in an EKF, improves the tracking results. |
| first_indexed | 2024-04-14T01:31:18Z |
| format | Article |
| id | doaj.art-af266947dd06450e9e18a83cc6f0a544 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-04-14T01:31:18Z |
| publishDate | 2019-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-af266947dd06450e9e18a83cc6f0a5442022-12-22T02:20:12ZengMDPI AGSensors1424-82202019-09-011919406210.3390/s19194062s19194062Boosting Multi-Vehicle Tracking with a Joint Object Detection and Viewpoint Estimation SensorRoberto J. López-Sastre0Carlos Herranz-Perdiguero1Ricardo Guerrero-Gómez-Olmedo2Daniel Oñoro-Rubio3Saturnino Maldonado-Bascón4GRAM, Department of Signal Theory and Communications, University of Alcalá, 28805 Alcalá de Henares, SpainGRAM, Department of Signal Theory and Communications, University of Alcalá, 28805 Alcalá de Henares, SpainBBVA Next Technologies, 28050 Madrid, SpainNEC Labs Europe, Kurfürsten-Anlage 36, 69115 Heidelberg, GermanyGRAM, Department of Signal Theory and Communications, University of Alcalá, 28805 Alcalá de Henares, SpainIn this work, we address the problem of multi-vehicle detection and tracking for traffic monitoring applications. We preset a novel intelligent visual sensor for tracking-by-detection with simultaneous pose estimation. Essentially, we adapt an Extended Kalman Filter (EKF) to work not only with the detections of the vehicles but also with their estimated coarse viewpoints, directly obtained with the vision sensor. We show that enhancing the tracking with observations of the vehicle pose, results in a better estimation of the vehicles trajectories. For the simultaneous object detection and viewpoint estimation task, we present and evaluate two independent solutions. One is based on a fast GPU implementation of a Histogram of Oriented Gradients (HOG) detector with Support Vector Machines (SVMs). For the second, we adequately modify and train the Faster R-CNN deep learning model, in order to recover from it not only the object localization but also an estimation of its pose. Finally, we publicly release a challenging dataset, the GRAM Road Traffic Monitoring (GRAM-RTM), which has been especially designed for evaluating multi-vehicle tracking approaches within the context of traffic monitoring applications. It comprises more than 700 unique vehicles annotated across more than 40.300 frames of three videos. We expect the GRAM-RTM becomes a benchmark in vehicle detection and tracking, providing the computer vision and intelligent transportation systems communities with a standard set of images, annotations and evaluation procedures for multi-vehicle tracking. We present a thorough experimental evaluation of our approaches with the GRAM-RTM, which will be useful for establishing further comparisons. The results obtained confirm that the simultaneous integration of vehicle localizations and pose estimations as observations in an EKF, improves the tracking results.https://www.mdpi.com/1424-8220/19/19/4062traffic monitoring sensorvehicle trackingvehicle detectiontracking by detectionviewpoint estimationsmart city |
| spellingShingle | Roberto J. López-Sastre Carlos Herranz-Perdiguero Ricardo Guerrero-Gómez-Olmedo Daniel Oñoro-Rubio Saturnino Maldonado-Bascón Boosting Multi-Vehicle Tracking with a Joint Object Detection and Viewpoint Estimation Sensor Sensors traffic monitoring sensor vehicle tracking vehicle detection tracking by detection viewpoint estimation smart city |
| title | Boosting Multi-Vehicle Tracking with a Joint Object Detection and Viewpoint Estimation Sensor |
| title_full | Boosting Multi-Vehicle Tracking with a Joint Object Detection and Viewpoint Estimation Sensor |
| title_fullStr | Boosting Multi-Vehicle Tracking with a Joint Object Detection and Viewpoint Estimation Sensor |
| title_full_unstemmed | Boosting Multi-Vehicle Tracking with a Joint Object Detection and Viewpoint Estimation Sensor |
| title_short | Boosting Multi-Vehicle Tracking with a Joint Object Detection and Viewpoint Estimation Sensor |
| title_sort | boosting multi vehicle tracking with a joint object detection and viewpoint estimation sensor |
| topic | traffic monitoring sensor vehicle tracking vehicle detection tracking by detection viewpoint estimation smart city |
| url | https://www.mdpi.com/1424-8220/19/19/4062 |
| work_keys_str_mv | AT robertojlopezsastre boostingmultivehicletrackingwithajointobjectdetectionandviewpointestimationsensor AT carlosherranzperdiguero boostingmultivehicletrackingwithajointobjectdetectionandviewpointestimationsensor AT ricardoguerrerogomezolmedo boostingmultivehicletrackingwithajointobjectdetectionandviewpointestimationsensor AT danielonororubio boostingmultivehicletrackingwithajointobjectdetectionandviewpointestimationsensor AT saturninomaldonadobascon boostingmultivehicletrackingwithajointobjectdetectionandviewpointestimationsensor |