This meant that the student network was predicting highly accurate for some burn locations, but not as accurate for others. Some images in burns such as Willamette V alley are more consistent and have a higher temporal resolution than the Sycan Marsh burn. Additionally, some imagery in FLAME 3 contains views of smoke and trees only, and no visible fire in the image. With a three-channel RGB color image only as input, and no distinct fire colors in the image, it may have proven difficult for the student network to segment the fire region. Some of these difficulties are visualized in Figure 3, rows b - e, reflecting not necessarily poor, but not ideal results. In summary, the overall sporadic nature and no visible flames of some of the burn imagery most likely caused lower quantitative IoU for the fire region (Class 1). Sample visual results for a test image from Willamette V alley for the teachers with DeepLabV3+ student network are shown in Figure 4. Table IV shows testing results with different teacher-student variants of the temperature predictions for the ground truth fire region pixels only.

The increasing accessibility of radiometric thermal imaging sensors for unmanned aerial vehicles (UAVs) offers significant potential for advancing AI-driven aerial wildfire management. Radiometric imaging provides per-pixel temperature estimates, a valuable improvement over non-radiometric data that requires irradiance measurements to be converted into visible images using RGB color palettes. Despite its benefits, this technology has been underutilized largely due to a lack of available data for researchers. This study addresses this gap by introducing methods for collecting and processing synchronized visual spectrum and radiometric thermal imagery using UAVs at prescribed fires. The included imagery processing pipeline drastically simplifies and partially automates each step from data collection to neural network input. Further, we present the FLAME 3 dataset, the first comprehensive collection of side-by-side visual spectrum and radiometric thermal imagery of wildland fires. Building on our previous FLAME 1 and FLAME 2 datasets, FLAME 3 includes radiometric thermal Tag Image File Format (TIFFs) and nadir thermal plots, providing a new data type and collection method. This dataset aims to spur a new generation of machine learning models utilizing radiometric thermal imagery, potentially trivializing tasks such as aerial wildfire detection, segmentation, and assessment. A single-burn subset of FLAME 3 for computer vision applications is available on Kaggle with the full 6 burn set available to readers upon request.