This paper reports a visible and thermal drone monitoring system that integrates deep-learning-based detection and tracking modules. The biggest challenge in adopting deep learning methods for drone detection is the paucity of training drone images especially thermal drone images. To address this issue, we develop two data augmentation techniques. One is a model-based drone augmentation technique that automatically generates visible drone images with a bounding box label on the drone's location. The other is exploiting an adversarial data augmentation methodology to create thermal drone images. To track a small flying drone, we utilize the residual information between consecutive image frames. Finally, we present an integrated detection and tracking system that outperforms the performance of each individual module containing detection or tracking only. The experiments show that, even being trained on synthetic data, the proposed system performs well on real-world drone images with complex background. The USC drone detection and tracking dataset with user labeled bounding boxes is available to the public.

A novel design of an integrated system using Synthetic Aperture Radar (SAR) image and Automatic Identification System (AIS) data is proposed in this paper for the purpose of identifying ships at sea. TerraSAR-X® (SpotLight mode) images and AIS data collected over Incheon Port (Korea) and Tokyo Bay (Japan) were used on different dates. Four main steps for integration of SAR and AIS based ships can be identified, namely: ‘Time Matching’ to retrieve the respective Dead Reckoning (DR) position of the ships at SAR image acquisition times; ‘Position Matching’ based on a nearest neighbourhood re-sampling method with compensation of position shift; ‘Size Matching’ and ‘Speed Matching’. Under each of the matching criteria, the measurement error in each of the matching criteria was found to be less than 20% and the SAR extracted ship's hull boundaries were presented on a screen to display the system results. The results of this study will contribute to the design a Near-Real-Time (NRT) operational system for ship detection, identification, and classification by SARs in different data acquisition modes over various geographical locations at different acquisition times. This novel integrated system design will provide a most important preliminary step towards integration based on ships' hull monitoring in order to recognize ‘friend’ and ‘foe’ ship targets over a huge oceanic region and would be useful for coast guards as an early warning system.

The use of low-cost GPS receivers and antennas for attitude determination can significantly reduce the overall hardware system cost. Compared to the use of high performance GPS receivers, the carrier phase measurements from low-cost equipment are subject to additional carrier phase measurement errors, such as multipath, antenna phase centre variation and noise. These error sources, together with more frequent cycle slip occurrences, severely deteriorate attitude determination availability, reliability and accuracy performance. This paper presents the investigation of a low-cost GPS/gyro integration system for attitude determination. By employing the dead reckoning sensor type, the ambiguity search region can be specifically defined as a small cube to enhance the ambiguity resolution process. A Kalman filter is implemented to fuse the rate gyro data with GPS carrier phase measurements. The quality control system based on innovation sequences is used to identify cycle slip occurrences and incorrect inter-antenna vector solutions. The availability of the integrated system also improves with respect to the GPS standalone system since the attitude parameters can be estimated using the angular rate measurements from rate gyros during GPS outages. The low-cost hardware used to design and test the integrated system consists of CMC Allstar receivers with the OEM AT575-70 antennas and Murata ENV-05D-52 piezoelectric vibrating rate gyroscopes. Tests in the urban area demonstrated that the introduction of rate gyros in a GPS-based attitude determination system not only effectively decreased the noise level in the estimated attitude parameters but coasted the attitude output during GPS outages and also significantly improved the system reliability.