Wildland Fire Detection and Monitoring Using a Drone-Collected RGB/IR Image Dataset

Abstract

Current forest monitoring technologies including satellite remote sensing, manned/piloted aircraft, and observation towers leave uncertainties about a wildfire's extent, behavior, and conditions in the fire's near environment, particularly during its early growth. Rapid mapping and real-time fire monitoring can inform in-time intervention or management solutions to maximize beneficial fire outcomes. Drone systems' unique features of 3D mobility, low flight altitude, and fast and easy deployment make them a valuable tool for early detection and assessment of wildland fires, especially in remote forests that are not easily accessible by ground vehicles. In addition, the lack of abundant, well-annotated aerial datasets -in part due to unmanned aerial vehicles' (UAVs') flight restrictions during prescribed burns and wildfires -has limited research advances in reliable data-driven fire detection and modeling techniques. While existing wildland fire datasets often include either color or thermal fire images, here we present (1) a multimodal UAV-collected dataset of dual-feed side-by-side videos including both RGB and thermal images of a prescribed fire in an open canopy pine forest in Northern Arizona and (2) a deep learning-based methodology for detecting fire and smoke pixels at accuracy much higher than the usual single-channel video feeds. The collected images are labeled to "fire" or "no-fire" frames by two human experts using side-by-side RGB and thermal images to determine the label. To provide context to the main dataset's aerial imagery, the included supplementary dataset provides a georeferenced pre-burn point cloud, an RGB orthomosaic, weather information, a burn plan, and other burn information. By using and expanding on this guide dataset, research can develop new data-driven fire detection, fire segmentation, and fire modeling techniques.INDEX TERMS Data-driven fire detection, prescribed fire, fire modeling, fire data, unmanned aerial vehicle (UAV), deep learning.

Precision-Boosted Forest Fire Target Detection via Enhanced YOLOv8 Model

“…Simultaneously, we plan to optimize the dataset to enhance the accuracy of fire detection. Chen et al [36] proposed a method utilizing a multi-modal dataset collected by drones, achieving high accuracy in detecting fire and smoke pixels by collecting dual-channel videos containing RGB and thermal images. Furthermore, they provided rich auxiliary data, such as georeferenced point clouds, orthomosaics, and weather information to offer more comprehensive contextual information.…”

SWIFT: Simulated Wildfire Images for Fast Training Dataset

“…It combines four other datasets, BowFire [17], Flickr-FireSmoke [42], Flickr-Fire [42], SmokeBlock [43]. FLAME2 [12] is also a public dataset from prescribed wildfires in 2021 in the canopy pine forest of northern Arizona. It includes a large amount of aerial images extracted from video recorded by a Mavic 2 Enterprise Advanced dual RGB/IR camera.…”

“…In addition, there is the human error risk. Machine learning methods for autolabeling have been developed, but they carry a lower-than-ideal accuracy for their labeling, triggering false alarms, especially with tricky scenarios such as sunsets [12]. This factor helps in understanding why many images are missing their fire-zone description.…”

“…Additionally, the quality of these datasets depends on the methodology employed to generate data, with the best-case scenario being a prescribed fire and setup of cameras and drones to record as much data as possible, similar to what was employed during the prescribed fire in Arizona to generate FLAME2 [12]. Prescribed fires are not easily arranged, requiring many aspects to be solved before the permission to start one is received.…”

“…Consequently, as a response to these recent events, there has been a strong effort to develop ways to combat and monitor this natural disaster. Recently, DL (deep learning) methods have been used to process images and videos collected from satellites [8,9], surveillance cameras [10], and unmanned air vehicles such as drones [11,12]. These models have shown high performance, proving their effectiveness for the detection, segmentation, and classification of wildfires at different stages [13][14][15].…”