A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Automated Traffic Surveillance System with Aerial Camera Arrays Imagery: Macroscopic Data Collection with Vehicle Tracking
AbstractThe paper presents a novel computer vision-based traffic surveillance system capable of processing aerial imagery to track vehicles and their movements. The system uses a preprocessed 1-Hz image sequence with a coverage of 64.80 km2 (25 sq mi) from an aerial camera array mounted on an airplane. The unique characteristics of the input data make this work challenging. Heuristic and machine-learning approaches are combined and evaluated to detect and track vehicles for the purpose of collecting speed, density, and volume data for uninterrupted flow corridors, which are useful for big-data monitoring of traffic parameters over an entire 64.80 km2 (25 sq mi) area with a single sensor. The deep learning combined with speeded up robust features (SURF)-based approach is able to achieve over 94, 93, and 92% accuracies in speed, density, and volume estimates, respectively, on 50 s of data when compared with manually collected ground truth. It has 100% accuracy in measuring level of service (LOS) for the uninterrupted flow facilities tested. These evaluations were conducted for facilities of different levels of congestion as indicated by the different levels of service. With further research, improved preprocessing, and a higher frame rate, the accuracy of tracking vehicles can be improved, which will allow for other potential applications such as identification of erratic drivers and origin–destination studies.
Automated Traffic Surveillance System with Aerial Camera Arrays Imagery: Macroscopic Data Collection with Vehicle Tracking
AbstractThe paper presents a novel computer vision-based traffic surveillance system capable of processing aerial imagery to track vehicles and their movements. The system uses a preprocessed 1-Hz image sequence with a coverage of 64.80 km2 (25 sq mi) from an aerial camera array mounted on an airplane. The unique characteristics of the input data make this work challenging. Heuristic and machine-learning approaches are combined and evaluated to detect and track vehicles for the purpose of collecting speed, density, and volume data for uninterrupted flow corridors, which are useful for big-data monitoring of traffic parameters over an entire 64.80 km2 (25 sq mi) area with a single sensor. The deep learning combined with speeded up robust features (SURF)-based approach is able to achieve over 94, 93, and 92% accuracies in speed, density, and volume estimates, respectively, on 50 s of data when compared with manually collected ground truth. It has 100% accuracy in measuring level of service (LOS) for the uninterrupted flow facilities tested. These evaluations were conducted for facilities of different levels of congestion as indicated by the different levels of service. With further research, improved preprocessing, and a higher frame rate, the accuracy of tracking vehicles can be improved, which will allow for other potential applications such as identification of erratic drivers and origin–destination studies.
Automated Traffic Surveillance System with Aerial Camera Arrays Imagery: Macroscopic Data Collection with Vehicle Tracking
Zhao, Xi (author) / Dawson, Douglas / Sarasua, Wayne A / Birchfield, Stanley T
2017
Article (Journal)
English
BKL:
56.03
/
56.03
Methoden im Bauingenieurwesen
Local classification TIB:
770/3130/6500
| British Library Online Contents | 2017
Traffic data collection from aerial imagery
| IEEE | 2003
Traffic Data Collection from Aerial Imagery
| British Library Conference Proceedings | 2003