A Systematic Review of UAV Structure and Monitoring Models for Forest Fire Detection

Highlights in Science Engineering and Technology, Год журнала: 2025, Номер 134, С. 81 - 87

Опубликована: Март 30, 2025

Forest fires pose a significant threat to ecosystems and economic development. In recent years, unmanned aerial vehicles (UAVs) have emerged as critical technology for forest fire monitoring due their high mobility, low cost, real-time surveillance capabilities. This paper provides systematic review of research progress on UAV-based monitoring, focusing three key aspects: hardware design, detection algorithm improvement, multi-sensor data fusion technologies. First, it summarizes optimization strategies UAV hardware, including sensor configurations, wing power supply improvements, aimed at enhancing flight stability environmental adaptability. Second, analyzes advancements in algorithms, particularly the performance enhancement lightweight modifications deep learning models, explores applicability high-noise environments. Finally, evaluates potential techniques improve accuracy by integrating temperature, smoke, image data. Despite advantages UAVs challenges remain, such limitations, trade-off between processing, complexity coordination. Future should focus optimization, development novel refinement integration further advance applications monitoring.

Язык: Английский

Intelligent Firefighting Technology for Drone Swarms with Multi-Sensor Integrated Path Planning: YOLOv8 Algorithm-Driven Fire Source Identification and Precision Deployment Strategy

Drones, Год журнала: 2025, Номер 9(5), С. 348 - 348

Опубликована: Май 3, 2025

This study aims to improve the accuracy of fire source detection, efficiency path planning, and precision firefighting operations in drone swarms during emergencies. It proposes an intelligent technology for based on multi-sensor integrated planning. The integrates You Only Look Once version 8 (YOLOv8) algorithm its optimization strategies enhance real-time detection capabilities. Additionally, this employs data fusion swarm cooperative path-planning techniques optimize deployment materials flight paths, thereby improving precision. First, a deformable convolution module is introduced into backbone network YOLOv8 enable flexibly adjust receptive field when processing targets, enhancing accuracy. Second, attention mechanism incorporated neck portion YOLOv8, which focuses feature regions, significantly reducing interference from background noise further recognition complex environments. Finally, new High Intersection over Union (HIoU) loss function proposed address challenge computing localization classification targets. dynamically adjusts weight various components training, achieving more precise classification. In terms visual sensors, infrared LiDAR sensors adopts Information Acquisition Optimizer (IAO) Catch Fish Optimization Algorithm (CFOA) plan paths coordinated swarms. By adjusting planning locations, can reach sources shortest possible time carry out operations. Experimental results demonstrate that improves by optimizing algorithm, algorithms, strategies. optimized achieved 94.6% small fires, with false rate reduced 5.4%. wind speed compensation strategy effectively mitigated impact material deployment. not only enhances but also enables rapid response scenarios, offering broad application prospects, particularly urban forest disaster rescue.

Язык: Английский

Wildfire Identification Based on an Improved MobileNetV3-Small Model

Forests, Год журнала: 2024, Номер 15(11), С. 1975 - 1975

Опубликована: Ноя. 8, 2024

In this paper, an improved MobileNetV3-Small algorithm model is proposed for the problem of poor real-time wildfire identification based on convolutional neural networks (CNNs). Firstly, a dataset constructed and subsequently expanded through image enhancement techniques. Secondly, efficient channel attention mechanism (ECA) utilised instead Squeeze-and-Excitation (SE) module within to enhance model’s speed. Lastly, support vector machine (SVM) employed replace classification layer model, with principal component analysis (PCA) applied before SVM reduce dimensionality features, thereby enhancing SVM’s efficiency. The experimental results demonstrate that achieves accuracy 98.75% average frame rate 93. Compared initial mean has been elevated by 7.23. designed in paper improves speed while maintaining accuracy, advancing development application CNNs field monitoring.

Язык: Английский