Fire Dynamic Vision for tracking fire and plume behavior across scales: Case Studies from Tall Timbers and the Jones Center from 2019-2023

Sample data and Fire Dynamic Vision (FDV) code for experiments in the paper: Sagel, D., & Quaife, B. (2025). Fire dynamic vision: Image segmentation and tracking for multi-scale fire and plume behavior. Environmental Modelling & Software, 185, 106286. https://doi.org/10.1016/j.envsoft.2024.106286This repository contains infrared and visual video data for three experiments presented in the paper listed above. In particular, the datasets areInfrared dataset of fire spread (infrared_fire.mat)Visual video of fire spread (visual_fire.mp4)Visual video of plume evolution (visual_plume.mp4 or image sequence within 'visual_plume_frames' directory)Each video dataset has a corresponding file with sample FDV code that performs the processing steps described in our paper (infrared_fire.m, visual_fire.m, visual_plume.m) and a shared file that visualizes all output and results from the FDV code (visualize.m). The code files contain suggested values from the paper, but you are invited to experiment with the datasets and try your own! The output of the code includes images of plume/fire boundaries, statistical characterizations of velocities and burn times, inpainted images, and labelled datasets. The 'Examples' directory contains sample output from each of the above files.These codes and data are identical to https://github.com/quaife/FDV_Data (commit 19da028be9de6f2f8b83c672d0caf07514968de0).