Project SCOUT: Interceptor Drone for Perimeter Defense

Azmain Yousuf; Siwei Cai; Knut Peterson; Lifeng Zhou; David Han

cs.ROSep 17, 2026

Project SCOUT: Interceptor Drone for Perimeter Defense

Authors: Azmain Yousuf, Siwei Cai, Knut Peterson, Lifeng Zhou, David Han

Organizations: Drexel University

Abstract

The rapid proliferation of unauthorized unmanned aerial vehicles (UAVs) has created a growing need for robust, jamming-resistant counter-UAV systems for perimeter defense. This paper presents SCOUT (Spatial Computation for Optimized UAV Tracking), a ROS-integrated onboard perception and control framework for real-time aerial defense against incoming UAVs. SCOUT performs visual detection, target association, track filtering, and control command generation directly onboard the defender UAV, without relying on external sensing infrastructure or ground-station computation. To provide stable control inputs, the perception pipeline combines TensorRT-accelerated drone detection with ByteTrack-based association and a lightweight track-retention state machine. The state machine rejects abrupt target jumps and maintains short-term target continuity during temporary detection degradation, reducing unstable control responses caused by false detections or target switching. We evaluate the proposed architecture through an integrated hardware deployment executing a planar ``goalkeeping'' interception strategy. In this setting, the defender UAV tracks the incoming target and adjusts its motion to maintain a blocking configuration near the protected boundary. Real-world flight results show that SCOUT maintains valid target detections for 92.2% of frames while operating at real-time onboard detection rates, demonstrating the feasibility of visual tracking and closed-loop control for UAV perimeter defense. A video demonstration of the end-to-end perimeter defense operation is available online. https://figshare.com/s/497befc033091fc0b84f

Figures & tables

Fig. 1: SCOUT Visual Perception Pipeline: Images from the defender UAV are processed by a TensorRT-accelerated detection model on the Jetson Orin NX. The resulting drone detections are temporally associated and geometrically filtered before being converted into velocity commands for closed-loop interception control.

Fig. 2: First-person view from the defender UAV. The detected target UAV is outlined in green, and the image-plane target-center offsets used for control are shown by the red and yellow markers.

Fig. 3: Invader Drone: DJI Neo (left) and Defender Drone: Custom Built (right)

Fig. 4: Overhead view of the indoor perimeter-defense test setup. The defender UAV operates near the protected boundary, while the manually piloted target UAV approaches from the attack side.

Model Variant	mAP@50	mAP@50-95
YOLO26n (pretrained)	0.878	0.600
YOLO26n (fine-tuned)	0.955	0.824

TABLE I: YOLO26n model performance for DJI Neo drone detection.

Configuration	Total Frames	Detections	False Positives
YOLO (Baseline)	1,827	1,195	30
YOLO + ByteTrack	1,827	1,195	30
YOLO + ByteTrack + Filter	1,827	1,164	0

TABLE II: Ablation Study of Tracking and Filtering Components

Fig. 5: Closed-loop image-center tracking performance during the active defense phase. The top plot shows the horizontal and vertical tracking errors in the control frame, while the bottom plot shows the corresponding issued lateral and vertical control commands. The defender maintains visual lock with RMSE values of $0.134$ horizontally and $0.107$ vertically in normalized image coordinates.

Metric	Value
Operation duration	104.53 s
Total messages	3326
Average publish Frame rate	31.82 FPS
Median message gap	30.29 ms
Maximum message gap	100.72 ms
Missing frames	29

TABLE III: Detection-stream performance during the representative real-world flight.

	PyTorch (ms)		TensorRT (ms)
Stage	Mean	Max	Mean	Max
Worker wake / image handoff	7.2	18.1	5.4	13.7
Preprocessing	4.1	5.8	4.8	6.3
Inference	26.1	33.6	7.9	10.6
NMS + ByteTrack association	7.1	12.3	2.5	10.8
Outlier filtering	$<0.1$	0.1	$<0.1$	$<0.1$

TABLE IV: Detector-side latency comparison: PyTorch vs. TensorRT on Jetson Orin NX.

Aerial Robotics

Aerial robotics studies the design, perception, control, and coordination of flying robots for navigation, inspection, transport, and physical interaction.

See all 329 papers

UAV-DETR: DETR for Anti-Drone Target Detection
Mar 24, 2026Jun Yang, Dong Wang, Hongxu Yin +2Aerial Object Detection Small Object Detection
School of Automation, NPU, Xian, ShaanXi
Decentralized UAV Swarms for Ground Target Protection in GPS- and Communication-Denied Environments
Jul 22, 2026Dimitria Silveria, Paulo Ricardo Marques de Araujo, Tiago Nascimento +1Swarm Robotics Multi-Robot Coordination
Department of Electrical and Computer Engineering, and the Ingenuity Labs Research Institute, Queen’s University, Kingston, ON, Canada · School of Computing and the Ingenuity Labs Research Institute, Queen’s University, Kingston, ON, Canada · Lab of Systems Engineering and Robotics (LASER), Department of Computer Systems, Universidade Federal da Para´ıba, and with the Department of Cybernetics, Czech Technical University in Prague, Prague 16629, Czechia
EventRadar: Long-Range Visual UAV Discovery through Spatiotemporal Event Sensing
Jun 9, 2026Zhiting Zhou, Xingchen Liu, Xinglin Yu +5Event-Based Vision UAV Detection
Shenzhen International Graduate School, Tsinghua University · Sun Yat-sen University · Harbin Institute of Technology, Shenzhen +3

Mar 24, 2026J/K move · Enter open · S save

UAV-DETR: DETR for Anti-Drone Target Detection

Jun Yang, Dong Wang, Hongxu Yin, Hongpeng Li+1

School of Automation, NPU, Xian, ShaanXi

Aerial Object Detection Small Object Detection Real-Time Object Detection Detection Transformer

Open paper AskPDF arXiv

Drone detection is pivotal in numerous security and counter-UAV applications. However, existing deep learning-based methods typically struggle to balance robust feature representation with computational efficiency. This challenge is particularly acute when detecting miniature drones against complex backgrounds under severe environmental interference. To address these issues, we introduce UAV-DETR, a novel framework that integrates a small-target-friendly architecture with real-time detection capabilities. Specifically, UAV-DETR features a WTConv-enhanced backbone and a Sliding Window Self-Attention (SWSA-IFI) encoder, capturing the high-frequency structural details of tiny targets while drastically reducing parameter overhead. Furthermore, we propose an Efficient Cross-Scale Feature Recalibration and Fusion Network (ECFRFN) to suppress background noise and aggregate multi-scale semantics. To further enhance accuracy, UAV-DETR incorporates a hybrid Inner-CIoU and NWD loss strategy, mitigating the extreme sensitivity of standard IoU metrics to minor positional deviations in small objects. Extensive experiments demonstrate that UAV-DETR significantly outperforms the baseline RT-DETR on our custom UAV dataset (+6.61% in mAP50:95, with a 39.8% reduction in parameters) and the public DUT-ANTI-UAV benchmark (+1.4% in Precision, +1.0% in F1-Score). These results establish UAV-DETR as a superior trade-off between efficiency and precision in counter-UAV object detection. The code is available at https://github.com/wd-sir/UAVDETR.

Abstract

Aerial Robotics

Explore similar work

UAV-DETR: DETR for Anti-Drone Target Detection

Decentralized UAV Swarms for Ground Target Protection in GPS- and Communication-Denied Environments

EventRadar: Long-Range Visual UAV Discovery through Spatiotemporal Event Sensing