We propose an effective vehicle tracker and speed estimation method from Unmanned Aerial Vehicles (UAVs) videos that can be deployed on UAV-embedded edge devices. Our tracker uses segmentation-derived vehicle regions to initialise a MOSSE tracker. This enables road operators to make multipurpose use of segmentation outputs while still being able to track the vehicles across frames. The vehicle speed is estimated using flight parameters derived from the UAV's flight computer and the vehicle displacement across frames. We trained CABiNet on the UAVid urban segmentation benchmark dataset and finetuned it on a dataset collected at our study site. A mean Intersection over Union (mIoU) of 0.73 was obtained for the vehicle class. Our segmentation-initialised MOSSE tracker was evaluated on the VisDrone Multi-Object Tracking (MOT) benchmark dataset and compared against traditional methods that utilise object regions for tracker initialisation. Our approach yielded a Multi-Object Tracking Precision (MOTP) of 0.872 compared to 0.830 when using YOLOv4. Our vehicle speed estimations approach was evaluated using a privately collected ground truth vehicle speed dataset. Our approach yielded a Root Mean Square Error (RMSE) between 3.42 and 16.12 km/hr across different flight configurations. Finally, our approach was deployed on an NVIDIA Jetson Xavier NX edge device and could be executed at 8 Frames Per Second (FPS). The results indicate that our approach is a simple yet fast alternative to traditional tracking methods while producing multipurpose segmentation information.

Consumer-grade Unmanned Aerial Vehicles (UAVs) are poorly suited to monitor complex scenes where multiple analysis tasks need to be carried out in real-time and in parallel to fulfil time-critical requirements. Therefore, we developed an innovative UAV agnostic system that is able to carry out multiple road infrastructure monitoring tasks simultaneously and in real-time. The aim of the paper is to discuss the system design considerations and the performance of the processing pipeline in terms of computational strain and latency. The system was deployed on a unique typology of UAV and instantiated with realistic placeholder modules that are of importance for infrastructure inspection tasks, such as vehicle detection for traffic monitoring, scene segmentation for qualitative semantic reasoning, and 3D scene reconstruction for large-scale damage detection. The system was validated by carrying out a trial on a highway in Guadalajara, Spain. By utilizing edge computation and remote processing, the end-to-end pipeline, from image capture to information dissemination to drone operators on the ground, takes on average 2.9 s, which is sufficiently quick for road monitoring purposes. The system is dynamic and, therefore, can be extended with additional modules, while continuously accommodating developments in technologies, such as IoT or 5G.

This monitoring system aims at increasing the resilience of the road infrastructures and ensuring reliable network availability under unfavourable conditions, such as extreme weather, landslides, and earthquakes. The main target is to combine downscaled climate change scenarios (applied to road infrastructures) with simulation tools (structural/geotechnical) and actual data (from existing and novel sensors), so as to provide the operators with an integrated tool able to support more effective management of their infrastructures at planning, maintenance and operation level. Towards this, the proposed framework aims to use high resolution modelling data for the determination and the assessment of the climatic risk of the selected, transport infrastructures and associated expected damages, use existing SHM data (from accelerometers, strain gauges etc.) with new types of sensor-generated data (computer vision) to feed the structural/geotechnical simulator, utilize tailored weather forecasts (combining seamlessly all available data sources) for specific hot-spots, providing early warnings with corresponding impact assessment in real time; develop improved multi-temporal, multi-sensor UAV, computer vision and machine learning-based damage diagnostic for diverse transport infrastructures; design and implement a Holistic Resilience Assessment Platform environment as an innovative planning tool that will permit a quantitative resilience assessment through an end-to-end simulation environment, running “what-if” impact/risk/resilience assessment scenarios. The effects of adaptation measures can be investigated by changing the hazard, exposure and vulnerability input parameters; design and implement a Common Operational Picture, including an enhanced visualisation interface and an Incident Management System. The integrated platform (and its sub-modules) will be validated in two real case studies in Spain and in Greece.

We present MultEYE, a traffic monitoring system that can detect, track, and estimate the velocity of vehicles in a sequence of aerial images. The presented solution has been optimized to execute these tasks in real-time on an embedded computer installed on an Unmanned Aerial Vehicle (UAV). In order to overcome the limitation of existing object detection architectures related to accuracy and computational overhead, a multi-task learning methodology was employed by adding a segmentation head to an object detector backbone resulting in the MultEYE object detection architecture. On a custom dataset, it achieved 4.8% higher mean Average Precision (mAP) score, while being 91.4% faster than the state-of-the-art model and while being able to generalize to different real-world traffic scenes. Dedicated object tracking and speed estimation algorithms have been then optimized to track reliably objects from an UAV with limited computational effort. Different strategies to combine object detection, tracking, and speed estimation are discussed, too. From our experiments, the optimized detector runs at an average frame-rate of up to 29 frames per second (FPS) on frame resolution 512 × 320 on a Nvidia Xavier NX board, while the optimally combined detector, tracker and speed estimator pipeline achieves speeds of up to 33 FPS on an image of resolution 3072 × 1728. To our knowledge, the MultEYE system is one of the first traffic monitoring systems that was specifically designed and optimized for an UAV platform under real-world constraints.

We present an unsupervised deep learning approach for post-disaster building damage detection that can transfer to different typologies of damage or geographical locations. Previous advances in this direction were limited by insufficient qualitative training data. We propose to use a state-of-the-art Anomaly Detecting Generative Adversarial Network (ADGAN) because it only requires pre-event imagery of buildings in their undamaged state. This approach aids the post-disaster response phase because the model can be developed in the pre-event phase and rapidly deployed in the post-event phase. We used the xBD dataset, containing pre- and post- event satellite imagery of several disaster-types, and a custom made Unmanned Aerial Vehicle (UAV) dataset, containing post-earthquake imagery. Results showed that models trained on UAV-imagery were capable of detecting earthquake-induced damage. The best performing model for European locations obtained a recall, precision and F1-score of 0.59, 0.97 and 0.74, respectively. Models trained on satellite imagery were capable of detecting damage on the condition that the training dataset was void of vegetation and shadows. In this manner, the best performing model for (wild)fire events yielded a recall, precision and F1-score of 0.78, 0.99 and 0.87, respectively. Compared to other supervised and/or multi-epoch approaches, our results are encouraging. Moreover, in addition to image classifications, we show how contextual information can be used to create detailed damage maps without the need of a dedicated multi-task deep learning framework. Finally, we formulate practical guidelines to apply this single-epoch and unsupervised method to real-world applications.

Abstract. Degradation and damage detection provides essential information to maintenance workers in routine monitoring and to first responders in post-disaster scenarios. Despite advance in Earth Observation (EO), image analysis and deep learning techniques, the quality and quantity of training data for deep learning is still limited. As a result, no robust method has been found yet that can transfer and generalize well over a variety of geographic locations and typologies of damages. Since damages can be seen as anomalies, occurring sparingly over time and space, we propose to use an anomaly detecting Generative Adversarial Network (GAN) to detect damages. The main advantages of using GANs are that only healthy unannotated images are needed, and that a variety of damages, including the never before seen damage, can be detected. In this study we aimed to investigate 1) the ability of anomaly detecting GANs to detect degradation (potholes and cracks) in asphalt road infrastructures using Mobile Mapper imagery and building damage (collapsed buildings, rubble piles) using post-disaster aerial imagery, and 2) the sensitivity of this method against various types of pre-processing. Our results show that we can detect damages in urban scenes at satisfying levels but not on asphalt roads. Future work will investigate how to further classify the found damages and how to improve damage detection for asphalt roads.