Drone Detection Research Articles

Drone Flight Detection at an Entrance to a Beehive Based on Audio Signals

Spotting a significant number of drones flying near the entrance of a beehive during late Spring could indicate the occurrence of swarming mood, as the the surge in drone presence is related to an overcrowded hive. Swarming refers to a natural reproductive process witnessed in honey bees, wherein half of the bee colony departs from their hive alongside the aging queen. In the paper, we propose an early swarming detection mechanism that relies on the behavior of the drones. The proposed method is based on audio signals registered in a close proximity to the beehive entrance. A comparative study was performed to find the most effective preprocessing method for the audio signals for the detection problem. We have compared the results for three different power spectrum density coefficients estimation methods, which are used as an input of an autoencoder neural network to discriminate drones from worker bees. Through simulations employing real-life signals, it has been demonstrated that drone detection based solely on audio signals is indeed feasible. The attained level of detection accuracy enables the creation of an efficient alarm system for beekeepers.

A Novel Approach to Detect Drones Using Deep Convolutional Neural Network Architecture.

Over the past decades, drones have become more attainable by the public due to their widespread availability at affordable prices. Nevertheless, this situation sparks serious concerns in both the cyber and physical security domains, as drones can be employed for malicious activities with public safety threats. However, detecting drones instantly and efficiently is a very difficult task due to their tiny size and swift flights. This paper presents a novel drone detection method using deep convolutional learning and deep transfer learning. The proposed algorithm employs a new feature extraction network, which is added to the modified YOU ONLY LOOK ONCE version2 (YOLOv2) network. The feature extraction model uses bypass connections to learn features from the training sets and solves the "vanishing gradient" problem caused by the increasing depth of the network. The structure of YOLOv2 is modified by replacing the rectified linear unit (relu) with a leaky-relu activation function and adding an extra convolutional layer with a stride of 2 to improve the small object detection accuracy. Using leaky-relu solves the "dying relu" problem. The additional convolution layer with a stride of 2 reduces the spatial dimensions of the feature maps and helps the network to focus on larger contextual information while still preserving the ability to detect small objects. The model is trained with a custom dataset that contains various types of drones, airplanes, birds, and helicopters under various weather conditions. The proposed model demonstrates a notable performance, achieving an accuracy of 77% on the test images with only 5 million learnable parameters in contrast to the Darknet53 + YOLOv3 model, which exhibits a 54% accuracy on the same test set despite employing 62 million learnable parameters.

ORION: Verification of drone trajectories via remote identification messages

With the widespread adoption of drones in daily life, next-generation smart cities need to establish highways, i.e., trajectories where drones can fly and operate safely. However, due to the untrusted nature of their ecosystem, drones might misbehave and take disallowed trajectories, e.g., to reduce the time to fly to a destination, reduce energy consumption, visit unauthorized areas, or disrupt operations of sensitive sites. In this paper, we address the cited problem by proposing ORION, a new framework for online drone trajectory verification. ORION requires one or more receivers distributed in a given area capable of receiving and analyzing standard Remote Identification (RID) messages emitted by operational drones. ORION compares the locations reported in such messages with the closest set of coordinates in the allowed trajectory. It raises an alarm if the distance between such locations exceeds a threshold calibrated offline. We validate the performance of ORION through data collected from both a real drone flight in Amsterdam (Netherlands) and taxi trajectories in Porto (Portugal), achieving a True Positive Ratio (correct detection of disallowed trajectories) up to 0.95 and a False Positive Ratio (incorrect detection of disallowed trajectories) up to 0.04. Our solution significantly outperforms existing approaches used for drone detection or time-series analysis. Finally, we also release the gathered data as open-source to foster future research.

LERFNet: an enlarged effective receptive field backbone network for enhancing visual drone detection

AbstractRecently, the world has witnessed a great increase in drone applications and missions. Drones must be detected quickly, effectively, and precisely when they are being handled illegally. Vision-based anti-drone systems provide an efficient performance compared to radar- and acoustic-based systems. The effectiveness of drone detection is affected by a number of issues, including the drone’s small size, conflicts with other objects, and noisy backgrounds. This paper employs enlarging the effective receptive field (ERF) of feature maps generated from the YOLOv6 backbone. First, RepLKNet is used as the backbone of YOLOv6, which deploys large kernels with depth-wise convolution. Then, to get beyond RepLKNet’s large inference time, a novel LERFNet is implemented. LERFNet uses dilated convolution in addition to large kernels to enlarge the ERF and overcome each other’s problems. The linear spatial-channel attention module (LAM) is used to give more attention to the most informative pixels and high feature channels. LERFNet produces output feature maps with a large ERF and high shape bias to enhance the detection of various drone sizes in complex scenes. The RepLKNet and LERFNet backbones for Tiny-YOLOv6, Tiny-YOLOv6, YOLOv5s, and Tiny-YOLOv7 are compared. In comparison to the aforementioned techniques, the suggested model’s results show a greater balance between accuracy and speed. LERFNet increases the MAP by $$2.8\%$$ 2.8 % , while significantly reducing the GFLOPs and parameter numbers when compared to the original backbone of YOLOv6.

ISAR Imaging for Drone Detection Based on Backprojection Algorithm Using Millimeter-Wave Fast Chirp Modulation MIMO Radar

ISAR Imaging for Drone Detection Based on Backprojection Algorithm Using Millimeter-Wave Fast Chirp Modulation MIMO Radar

Laser-based drone vision disruption with a real-time tracking system for privacy preservation

The capabilities of drones are increasing every day, as is the ease with which civilians can buy and fly them. Most drones are equipped with a camera that is used by a point-of-view operator and, at the same time, can be used for image capture. The use of drones creates a threat to privacy whereby anyone who can fly a drone can take pictures without permission. This study aims to create a 2-axis tracker system that can recognize a drone and locate the position of the drone camera so that a laser beam can track and dazzle the drone camera. The depth-sensing camera is used to localize the part of the target corresponding to the drone’s camera and is created using the YOLOv5 algorithm as a deep-learning detector model. The drone’s camera range and position are challenging to detect due to its small size. Our adaptive detection method combines drone detection and drone camera detection. The depth-sensing camera provides input in the form of a three-coordinate axis from the target. If only the drone is detected, a predictive algorithm can determine the camera’s position for illumination with the laser. Alternatively, if the drone camera is detected, the laser can follow the target’s movement more quickly. In this study, a green (520 nm) laser module with adjustable power is used to investigate factors that affect the dazzling range. The computer vision detection algorithm can detect and localize the position of the drone camera up to 500 cm with a confidence level of more than 65%. If the target is in the center of the field of view, the accuracy of the target position can reach 98%. The tracker can follow the drone’s movement from 2 m/s to 4 m/s with a maximum error of 1.9 cm from the center point of the drone camera for close range. For long range, the maximum error is 6.2 cm. A laser power of 23.5 mW at 500 cm distance is found to be sufficient to dazzle and track drone cameras.

Tiny drone object detection in videos guided by the bio-inspired magnocellular computation model

Detecting drones in infrared videos is highly desired in many realistic scenarios, e.g., unauthorized drone monitoring around airports. Nevertheless, automated drone detection is rather challenging when the targets appear as tiny objects (≤10×10 pixels) against complex backgrounds. Conventional object detection algorithms, which mainly use static visual features, can hardly distinguish tiny objects from undesired artefacts in complex backgrounds. To alleviate this problem, we learn from the early biological visual pathway (including the parvocellular and magnocellular pathways), which process static and motion information simultaneously. Therefore, we propose a magnocellular inspired method for video tiny-object detection (Magno-VTOD) that integrates both static and motion visual information. The Magno-VTOD firstly employs a retinal magnocellular computation model to extract the motion strength of moving objects. The motion responses are then used to enhance the areas of the flying tiny drones effectively and efficiently, thereby facilitating the subsequent target detection procedure. We implement the video tiny-object detection method based on the widely adopted deep neural networks guided by the magnocellular computation model. Experimental results obtained on the large-scale Anti-UAV dataset (304451 video frames) validate that the proposed Magno-VTOD method significantly outperforms the competing state-of-the-art object detection methods on the tiny drone detection task. Particularly, the AP value is increased by 15.4% for tiny object detection, and by 17.1%/13.7% against wood/mountain backgrounds.

Multistatic Passive Radar for drone detection based Random Finite State

Considering the implication of radar sensors in our daily life and environment. Localizing and identifying drones are becoming a research with a greater focus in recent years. Consequently, when an unmanned aerial vehicle is used with bad intention, this can lead to a serious public safety and privacy probem. This study investigate pratical use of spectrum range for multistatic passive radar (MSPR) signal processing . Firstly, signal processing is performed after MSPR sensing detection range ,this include multipath energy detection, reference signal extraction, and receiving antenna configuration. Secondly, based on the MSPR nature, the reference signal is extracted and analyzed. In addition,taking into consideration the vulnerability of passive radar when comes to a moving target localization and real time tracking detection ,a novel method for spectrum sensing and detection which relies on Gaussian filter is proposed. The main goal is to optimize the use of the reference signal extracted with minimum interference as the shared reference signal in spectrum sensing. This will improve the system detection capability and spectrum access. Finally, a recursive method based on Bernoulli random filter is proposed, this takes consideration of drone’s present and unknown states based on time. Moreover, a system is developed meticulously to track and enhance detection of the target. A careful result of the experiment demonstrated that spectral detection can be achieved accurately even when the drone is moving while chasing its position. It shows that Cramer Rao lower error bounds remains significantly within 3% range.

Towards Reliable Identification and Tracking of Drones Within a Swarm

Drone swarms consist of multiple drones that can achieve tasks that individual drones can not, such as search and recovery or surveillance over a large area. A swarm’s internal structure typically consists of multiple drones operating autonomously. Reliable detection and tracking of swarms and individual drones allow a greater understanding of the behaviour and movement of a swarm. Increased understanding of drone behaviour allows better coordination, collision avoidance, and performance monitoring of individual drones in the swarm. The research presented in this paper proposes a deep learning-based approach for reliable detection and tracking of individual drones within a swarm using stereo-vision cameras in real time. The motivation behind this research is in the need to gain a deeper understanding of swarm dynamics, enabling improved coordination, collision avoidance, and performance monitoring of individual drones within a swarm. The proposed solution provides a precise tracking system and considers the highly dense and dynamic behaviour of drones. The approach is evaluated in both sparse and dense networks in a variety of configurations. The accuracy and efficiency of the proposed solution have been analysed by implementing a series of comparative experiments that demonstrate reasonable accuracy in detecting and tracking drones within a swarm.

Monitoring GPS‐collared moose by ground versus drone approaches: efficiency and disturbance effects

Efficient wildlife management requires precise monitoring methods, for example to estimate population density, reproductive success, and survival. Here, we compared the efficiency of drone (equipped with a RGB camera) and ground approaches to detect and observe GPS‐collared female moose Alces alces and their calves. We also quantified how drone (n = 42) and ground (n = 41) approaches affected moose behavior and space use (n = 24 individuals). The average time used for drone approaches was 17 min compared to 97 min for ground approaches, with drone detection probability being higher (95% of adult female moose and 88% of moose calves) compared to ground approaches (78% of adult females and 82% of calves). Drone detection success increased at lower drone altitudes (50–70 m). Adult female moose left the site in 35% of drone approaches (with > 40% of those moose becoming disturbed once the drone hovered < 50 m above ground) compared to 56% of ground approaches. We failed to find short‐term effects (3 h after approaches) of drone approaches on moose space use, but moose moved > fourfold greater distances and used larger areas after ground approaches (compared to before the approaches had started). Similarly, longer‐term (24 h before and after approaches) space use did not differ between drone approaches compared to days without known disturbance, but moose moved comparatively greater distances during days of ground approaches. In conclusion, we could show that drone approaches were highly efficient to detect adult moose and their calves in the boreal forest, being faster and less disturbing than ground approaches, making them a useful tool to monitor and study wildlife.

Unauthorized Drone Detection Using Deep Learning

Abstract: The growing concerns over drone misuse in national airspace, including activities like drug smuggling and privacy violations, demand advanced drone detection systems. Traditional methods struggle to accurately identify drones among various airborne objects. YOLO v8, renowned for its efficient single-pass object detection, presents a sophisticated solution. Through training on annotated datasets featuring drones in diverse environments and fine-tuning with aerial imagery, YOLO v8 excels at recognizing drones amidst complex backgrounds. Its multi-scale detection capabilities enable it to detect drones of varying sizes, crucial for different distances from the camera. Real-time prediction ensures timely detection, vital for immediate responses to unauthorized intrusions. Integration with drone tracking systems enhances situational awareness, facilitating proactive measures. Furthermore, employing adaptive thresholding and post-processing techniques like Kalman filtering refines drone detection accuracy, reducing false positives. YOLO v8 thus emerges as a potent tool in mitigating risks associated with drone misuse, bolstering defenses against unauthorized activities in national airspace.

Real-time lightweight drone detection model: Fine-grained Identification of four types of drones based on an improved Yolov7 model

Recently, the rapid progress of artificial intelligence has enhanced the human-robot relationship through the development of several autonomous robots; such as drones. The overwhelming rise of drones has brought both relevant advantages and threatening risks into our daily lives. They are strongly deployed to achieve hard tasks and reach critical areas within a short period of time. In the past few years, drones' anarchic and malicious use has generated many incidents and accidents, causing far-reaching impacts. Due to their dual-edged nature, they are regarded as potential threats. Recognizing the type of flying drones represents a challenging task for anti-drone systems to reinforce airspace safety and security. To this end, we propose a novel lightweight model able to differentiate between the main types of drones in real-time using appropriate modules. To find the optimal compromise between size, performance and speed, a lightweight model is proposed based on Yolov7 with the incorporation of CNeB and C3C2 modules as well as the Re- Parameterization Decoupled (RePD) head structure. Further, a CNeB module is used to reduce the model’s size and improve the model’s performance, a C3C2 module to enhance the feature extraction and fusion and a Re- Parameterization Decoupled RePD to improve both the efficacy and accuracy of the prediction part while reducing the inference time. The experimental results show that the Yolov7-CNeB-C3C2-RePD meets the anti-drone requirements by improving the detection performance, minimizing the inference speed and reducing the model size. In comparison to other models, the developed has provided high results reaching 91.9% precision, 90.5% recall, 95% mAP@0.5, and 20.3 ms during the inference stage, which outperforms the other algorithms in the benchmark. Furthermore, the model has an average detection speed of 0.003 ms per image and a capacity to process 27.88 frames per second, thus satisfying the real-time requirements.

Human Detection in Flood Using Drone

Drowning people in India approximately around 38000 peoples per year leads to dead finally because of, we have insufficient water rescue or timely emergency response to search and rescue team during emergency, also the lack of information to the rescue team about the drowning people place. We should believe that a few seconds' difference could have saved a person’s life. The timely information conveyed to the rescue team is also an important criterion for drowning to dead rate being very higher At first, we make a dataset, which contains many human targets at sea. Then, we improve the algorithm In the feature extraction network, we use the residual module with channel attention mechanism. Finally, on the settings of the raspberry pi Pico with GPS and GSM, we use a linear transformation method to deal with the python generated by clustering algorithm. The detection accuracy of the improved algorithm for human targets at sea is improved, which has a good detection effect. The drone with detecting and alerting with voice message to the Rescue Team at remote end with required all details about the drowning people make sense for faster rescue and save as the highest accuracy. The camera detection of the rescue Drone had a proper in that the range of the active camera and the speed of the video with Wi-Fi to the control room also optimal for the detection to work properly.

Drones Detection Using a Fusion of RF and Acoustic Features and Deep Neural Networks.

The use of drones has recently gained popularity in a diverse range of applications, such as aerial photography, agriculture, search and rescue operations, the entertainment industry, and more. However, misuse of drone technology can potentially lead to military threats, terrorist acts, as well as privacy and safety breaches. This emphasizes the need for effective and fast remote detection of potentially threatening drones. In this study, we propose a novel approach for automatic drone detection utilizing the usage of both radio frequency communication signals and acoustic signals derived from UAV rotor sounds. In particular, we propose the use of classical and deep machine-learning techniques and the fusion of RF and acoustic features for efficient and accurate drone classification. Distinct types of ML-based classifiers have been examined, including CNN- and RNN-based networks and the classical SVM method. The proposed approach has been evaluated with both frequency and audio features using common drone datasets, demonstrating better accuracy than existing state-of-the-art methods, especially in low SNR scenarios. The results presented in this paper show a classification accuracy of approximately 91% at an SNR ratio of -10 dB using the LSTM network and fused features.

Detection of Drones using SDR as a RADAR

With the evolution of technology, the application of drones in Fourth Generation Warfare has increased manifolds. Drones offer many advantages as they are inexpensive to operate, can fly for longer periods, and are less vulnerable to enemy fire. This makes them an ideal, inconspicuous, and low-cost weapon for both conventional and non-state actors to carry out targeted attacks on valuable targets. Detection of drones is a major challenge as they offer minimum Radio signature and low Radar cross-section. A credible method to detect incoming drones is by employing Software Defined Radio as a RADAR

Classification of Genetics Based on Machine Learning Algorithms: A Review

The multifaceted applications of drones in addressing humanitarian challenges, enhancing governance services, medical assistance, and security considerations. Drones are showcased as adaptable and swift responders in conflict or disaster-affected areas, mitigating risks for humanitarian workers and delivering crucial supplies to remote locations. The integration of digital technology in governance services is discussed, emphasizing transparency, efficacy, and reduced corruption. The study also introduces a taxonomy for GPS- guided drones in medical supply delivery, highlighting challenges in accuracy and cost reduction. Drones' wide- ranging potential uses, from police operations to advertising and shipment transportation, are outlined. A comprehensive evaluation of drone security, from consumer drones to military systems, is provided, along with preventive suggestions. Machine learning algorithms for drone detection and classification, showcasing the proposed DDI system's capability to accurately identify intruding drones and their operational modes. machine learning in the context of used drones, focusing on detection and classification. The study assesses various machine learning algorithms, including image processing, sound analysis, and RF signal-based techniques, to identify and classify drones effectively. Data from diverse sensors are utilized for feature extraction, employing algorithms such as Deep Neural Networks, Support Vector Machines, and deep belief networks. The proposed DDI system adopts an RF-based approach and integrates a Deep Learning algorithm for precise detection and identification of used or intruding drones.

Securing Your Airspace: Detection of Drones Trespassing Protected Areas.

Unmanned Aerial Vehicle (UAV) deployment has risen rapidly in recent years. They are now used in a wide range of applications, from critical safety-of-life scenarios like nuclear power plant surveillance to entertainment and hobby applications. While the popularity of drones has grown lately, the associated intentional and unintentional security threats require adequate consideration. Thus, there is an urgent need for real-time accurate detection and classification of drones. This article provides an overview of drone detection approaches, highlighting their benefits and limitations. We analyze detection techniques that employ radars, acoustic and optical sensors, and emitted radio frequency (RF) signals. We compare their performance, accuracy, and cost under different operating conditions. We conclude that multi-sensor detection systems offer more compelling results, but further research is required.

The Sound of Surveillance: Enhancing Machine Learning-Driven Drone Detection with Advanced Acoustic Augmentation

In response to the growing challenges in drone security and airspace management, this study introduces an advanced drone classifier, capable of detecting and categorizing Unmanned Aerial Vehicles (UAVs) based on acoustic signatures. Utilizing a comprehensive database of drone sounds across EU-defined classes (C0 to C3), this research leverages machine learning (ML) techniques for effective UAV identification. The study primarily focuses on the impact of data augmentation methods—pitch shifting, time delays, harmonic distortion, and ambient noise integration—on classifier performance. These techniques aim to mimic real-world acoustic variations, thus enhancing the classifier’s robustness and practical applicability. Results indicate that moderate levels of augmentation significantly improve classification accuracy. However, excessive application of these methods can negatively affect performance. The study concludes that sophisticated acoustic data augmentation can substantially enhance ML-driven drone detection, providing a versatile and efficient tool for managing drone-related security risks. This research contributes to UAV detection technology, presenting a model that not only identifies but also categorizes drones, underscoring its potential for diverse operational environments.

Drone detection with radio frequency signals and deep learning models

The widespread use of drones raises security, environmental, privacy, and ethical issues; therefore, effective detection by drones is important. There are several methods for detecting drones, such as wireless signal detection, photoelectric detection, radar detection, and sound detection. However, these detection methods are not accurate enough to identify drones for use. To solve this question, more robust drone detection method are needed. In addition, for different types of drones and application scenarios, different technical means need to be used for detection and identification. Based on 2-class ,4-class and 10-class problems on an open ratio frequency (RF) signal dataset, we compared the drone detection and classification performances of different machine learning with deep learning models and multi-task models which is proposed by combining different RF methods with Convolutional neural networks (CNNs). Our experimental results show that the XGBoost model achieved the latest results on this groundbreaking dataset, with 99.96% accuracy for 2-class problem, 92.31% accuracy for 4-class problem, and 74.81% accuracy for 10-class problem, which exhibits the best performance for drone detection and classification.

Digital twin of multi-model drone detection system on Airsim for RF and vision modalities

Drones have become more prevalent in recent years and are used for both beneficial and malicious purposes. As a result, protecting restricted areas from unauthorized drone activities has become crucial. However, some researchers face challenges in developing drone detection systems due to the high costs of necessary equipment. This paper presents an innovative solution by creating an Airsim Graphical User Interface (GUI) tool compatible with the Unreal Engine. This tool enables the simulation of drone flights and creation of image and radio frequency (RF) datasets for drone detection in a simulation environment. Our approach involves modeling the measurement devices such as cameras to capture image data and software defined radio (SDR) receiver to capture RF signals as raw in-phase and quadrature (IQ) data. Moreover, users can manage automated route planning for drones, recording configurations, and different cameras and RF configurations. Researchers can now generate datasets with various images and RF configurations without the need for physical drones, cameras, or SDRs, enabling experimentation with different drone detection models. Furthermore, we proposed models for drone detection systems by using generated datasets from the proposed dataset generation system.