The ALFA project uses Zenodo as its open research data repository, in order to grant Open Access to scientific publications. Check out our Zenodo community ALFA!

Expand all | Collapse all

Abstract: As maritime smuggling is being combatted more effectively, the criminal “modus operandi” consists more frequently of using small aircraft and drones for drug transport. To address this issue, we report our efforts to develop a system capable of accurately tracking suspicious flying objects and identifying them on video streams. Our solution consists in coupling classical computer vision with deep learning to perform tracking and object detection. A discrete Kalman filter is used to predict the location of each object being tracked while the Hungarian algorithm is used to match objects between successive frames. Whenever a potential target is considered suspicious the input images are zoomed and fed into a deep learning pipeline that separates images into the classes aircraft, drones, birds or clouds. A literature survey indicates that this problem with important applications is yet to be fully explored.


expand Tracking and Classification of Aerial Objects
Marcia Baptista, Luis Fernandes, Paulo Chaves
Abstract: Unauthorized drone flying can prompt disruptions in critical facilities such as airports or railways. To prevent these situations, we propose a surveillance system that can sense malicious and/or illicit aerial targets. The idea is to track moving aerial objects using a static camera and when a tracked object is considered suspicious, the camera zooms in to take a snapshot of the target. This snapshot is then classified as an aircraft, drone, bird or cloud. In this work, we propose the classical technique of two-frame background subtraction to detect moving objects. We use the discrete Kalman filter to predict the location of each object and the Jonker-Volgenant algorithm to match objects between consecutive image frames. A deep residual network, trained with transfer learning, is used for image classification. The residual net ResNet-50 developed for the ILSVRC competition was retrained for this purpose. The performance of the system was evaluated with positive results in real-world conditions. The system was able to track multiple aerial objects with acceptable accuracy and the classification system also exhibited high performance.


expand Research on the Applications of Physically Unclonable Functions within the Internet of Things
Martin Deutschmann; Lejla Iriskic; Sandra-Lisa Lattacher; Mario Münzer; Felix Stornig; Oleksandr Tomashchuk
Abstract: This paper presents a overview over research activities in the field of Physically Unclonable Functions (PUFs) that reach into the domain of the Internet of Things (IoT) and summary of their outcomes. After brief introduction PUF technology and its basic properties, paper analyses one of most common PUF types, SRAM PUF, on selected set of commodity hardware. The paper also presents the authentication scheme, based on SRAM PUF, for resource-constrained embedded devices especially designed to be employed in the domain of the Internet of Things (IoT).


expand Radar Classifier for Small Manned Air Targets
Gilles Prémel-Cabic; Jacco J.M. de Wit; Miguel Caro Cuenca
Abstract: The ALFA project aims at timely detection, tracking, classification, and intent assessment of LSS targets. The system relies on a heterogeneous sensor suite, including radar. The objective of the radar component is sector surveillance including target classification. Since the revisit time needs to be short, classification must be done with very short time-on-target. Based on measurements, three suitable features for classification of two relevant target classes, i.e., small aircraft and helicopters, have been developed. These features exploit the targets’ micro-Doppler characteristics and their evolution over time. Best classification performance is obtained by using a combination of these features and by considering the variation of the features’ distributions depending on the signal-to-noise ratio.


expand A modular localization system combining passive RF detection and passive radar
Markus Krueckemeier, Fabian Schwartau, Joerg Schoebel
Abstract: This paper presents a completely passive system for detection and localization of small aircraft and UAVs. The system makes use of the fact that almost any such target emits some kind of RF emission, either by active transmissions or by passive reflection of other sources. Active transmissions can for example be caused by telemetry or video downlinks to a remote control or some kind of unintended emission like a mobile phone carried by a passenger. These transmissions can be identified by means of passive detection. At the same time, passive reflections of signals radiated by illuminators of opportunity like FM radio stations can be used to build a multistatic passive radar.


expand Synchronization of multiple USRP SDRs for coherent receiver applications
Markus Krueckemeier; Fabian Schwartau; Carsten Monka-Ewe; Joerg Schoebel
Abstract: This paper describes the necessary means to combine multiple Ettus Research USRP X310 software-defined radios to a multichannel coherent receiver for direction-of-arrival (DoA) and passive radar applications. The requirements to combine several software-defined radios to a multichannel coherent receiver are examined in general. In particular the requirement of phase coherence necessitates a closer look on the receiver synchronization, since the straightforward approach of synchronizing the systems with a common reference clock will in most cases lead to phase ambiguities between the channels. The mechanism inducing these phase ambiguities between several systems that are phase-locked to a common reference is discussed in detail. Results regarding the achieved phase stability and a preliminary measurement demonstrating the DoA capabilities of the system are shown.


expand Automatic threat evaluation for border security and surveillance
Bert van den Broek; Jos van der Velde; Michiel van den Baar; Loek Nijsten; Rob van Heijster
Abstract: We present a study of border surveillance systems for automatic threat estimation. The surveillance systems should allow border control operators to be triggered in time so that adequate responses are possible. Examples of threats are smuggling, possibly by using small vessels, cars or drones, and threats caused by unwanted persons (e.g. terrorists) crossing the border. These threats are revealed by indicators which are often not exact and evidence for these indicators incorporates significant amounts of uncertainty. This study is linked to the European Horizon 2020 project ALFA, which focuses on the detection and threat evaluation of low flying objects near the strait of Gibraltar. Several methods are discussed to fuse the indicators while taking the uncertainty into account, including Fuzzy Reasoning, Bayesian Reasoning, and Dempster-Shafer Theory. In particular the Dempster-Shafer Theory is elaborated since this approach incorporates evaluation of unknown information next to uncertainty. The method is based on belief functions representing the indicators. These functions show a gradual increase or decrease of the suspiciousness depending on input parameters such as object speed, size etc. The fusion methods give two output values for each track: a suspect probability and an uncertainty value. The complete dynamic risk assessment of detected flying objects is evaluated by the automatic system and targets with probabilities exceeding a certain threshold and appropriate uncertainty values are presented to the border control operators.