Low-Altitude Aerial Video Surveillance via One-Class SVM Anomaly Detection from Textural Features in UAV Images

A Novel Transformer-Based IMU Self-Calibration Approach through On-Board RGB Camera for UAV Flight Stabilization

During flight, unmanned aerial vehicles (UAVs) need several sensors to follow a predefined path and reach a specific destination. To this aim, they generally exploit an inertial measurement unit (IMU) for pose estimation. Usually, in the UAV context, an IMU entails a three-axis accelerometer and a three-axis gyroscope. However, as happens for many physical devices, they can present some misalignment between the real value and the registered one. These systematic or occasional errors can derive from different sources and could be related to the sensor itself or to external noise due to the place where it is located. Hardware calibration requires special equipment, which is not always available. In any case, even if possible, it can be used to solve the physical problem and sometimes requires removing the sensor from its location, which is not always feasible. At the same time, solving the problem of external noise usually requires software procedures. Moreover, as reported in the literature, even two IMUs from the same brand and the same production chain could produce different measurements under identical conditions. This paper proposes a soft calibration procedure to reduce the misalignment created by systematic errors and noise based on the grayscale or RGB camera built-in on the drone. Based on the transformer neural network architecture trained in a supervised learning fashion on pairs of short videos shot by the UAV's camera and the correspondent UAV measurements, the strategy does not require any special equipment. It is easily reproducible and could be used to increase the trajectory accuracy of the UAV during the flight.

CUDM: A Combined UAV Detection Model Based on Video Abnormal Behavior

The widespread use of unmanned aerial vehicles (UAVs) has brought many benefits, particularly for military and civil applications. For example, UAVs can be used in communication, ecological surveys, agriculture, and logistics to improve efficiency and reduce the required workforce. However, the malicious use of UAVs can significantly endanger public safety and pose many challenges to society. Therefore, detecting malicious UAVs is an important and urgent issue that needs to be addressed. In this study, a combined UAV detection model (CUDM) based on analyzing video abnormal behavior is proposed. CUDM uses abnormal behavior detection models to improve the traditional object detection process. The work of CUDM can be divided into two stages. In the first stage, our model cuts the video into images and uses the abnormal behavior detection model to remove a large number of useless images, improving the efficiency and real-time detection of suspicious targets. In the second stage, CUDM works to identify whether the suspicious target is a UAV or not. Besides, CUDM relies only on ordinary equipment such as surveillance cameras, avoiding the use of expensive equipment such as radars. A self-made UAV dataset was constructed to verify the reliability of CUDM. The results show that CUDM not only maintains the same accuracy as state-of-the-art object detection models but also reduces the workload by 32%. Moreover, it can detect malicious UAVs in real-time.

DNS Request Log Analysis of Universities in Shanghai: A CDN Service Provider’s Perspective

Understanding the network usage patterns of university users is very important today. This paper focuses on the research of DNS request behaviors of university users in Shanghai, China. Based on the DNS logs of a large number of university users recorded by CERNET, we conduct a general analysis of the behavior of network browsing from two perspectives: the characteristics of university users’ behavior and the market share of CDN service providers. We also undertake experiments on DNS requests patterns for CDN service providers using different prediction models. Firstly, in order to understand the university users’ Internet access patterns, we select the top seven universities with the most DNS requests and reveal the characteristics of different university users. Subsequently, to obtain the market share of different CDN service providers, we analyze the overall situation of the traffic distribution among different CDN service providers and its dynamic evolution trend. We find that Tencent Cloud and Alibaba Cloud are leading in both IPv4 and IPv6 traffic. Baidu Cloud has close to 15% in IPv4 traffic, but almost no fraction in IPv6 traffic. Finally, for the characteristics of different CDN service providers, we adopt statistical models, traditional machine learning models, and deep learning models to construct tools that can accurately predict the change in request volume of DNS requests. The conclusions obtained in this paper are beneficial for Internet service providers, CDN service providers, and users.

The Use of Terrestrial and Maritime Autonomous Vehicles in Nonintrusive Object Inspection

Traditional nonintrusive object inspection methods are complex or extremely expensive to apply in certain cases, such as inspection of enormous objects, underwater or maritime inspection, an unobtrusive inspection of a crowded place, etc. With the latest advances in robotics, autonomous self-driving vehicles could be applied for this task. The present study is devoted to a review of the existing and novel technologies and methods of using autonomous self-driving vehicles for nonintrusive object inspection. Both terrestrial and maritime self-driving vehicles, their typical construction, sets of sensors, and software algorithms used for implementing self-driving motion were analyzed. The standard types of sensors used for nonintrusive object inspection in security checks at the control points, which could be successfully implemented at self-driving vehicles, along with typical areas of implementation of such vehicles, were reviewed, analyzed, and classified.

Remote Sensing Change Detection Based on Unsupervised Multi-Attention Slow Feature Analysis

With the development of big data, analyzing the environmental benefits of transportation systems by artificial intelligence has become a hot issue in recent years. The ground traffic changes can be overlooked from a high-altitude perspective, using the technology of multi-temporal remote sensing change detection. We proposed a novel unsupervised algorithm by combining the image transformation and deep learning method. The new algorithm for remote sensing images is named multi-attention slow feature analysis (ASFA). In this model, three parts perform different functions respectively. The first part records to the K-BoVW to classify the categories of the ground objects as a channel parameter. The second part is a residual convolution with multiple attention mechanisms including temporal, spatial, and channel attention. Feature extraction and updating are completed at this link. Finally, we put the updated features in the slow feature analysis to highlight the variant components which we want and then generate the change map visually. Experiments on three very high-resolution datasets verified that the ASFA has a better performance than four basic change detection algorithms and an improved SFA algorithm. More importantly, this model works well for traffic road detection and helps us analyze the environmental benefits of traffic changes.

Smart Cybersecurity Framework for IoT-Empowered Drones: Machine Learning Perspective

Drone advancements have ushered in new trends and possibilities in a variety of sectors, particularly for small-sized drones. Drones provide navigational interlocation services, which are made possible by the Internet of Things (IoT). Drone networks, on the other hand, are subject to privacy and security risks due to design flaws. To achieve the desired performance, it is necessary to create a protected network. The goal of the current study is to look at recent privacy and security concerns influencing the network of drones (NoD). The current research emphasizes the importance of a security-empowered drone network to prevent interception and intrusion. A hybrid ML technique of logistic regression and random forest is used for the purpose of classification of data instances for maximal efficacy. By incorporating sophisticated artificial-intelligence-inspired techniques into the framework of a NoD, the proposed technique mitigates cybersecurity vulnerabilities while making the NoD protected and secure. For validation purposes, the suggested technique is tested against a challenging dataset, registering enhanced performance results in terms of temporal efficacy (34.56 s), statistical measures (precision (97.68%), accuracy (98.58%), recall (98.59%), F-measure (99.01%), reliability (94.69%), and stability (0.73).