Video driven fire spread forecasting (f) using multi-modal LWIR and visual flame and smoke data

Being able to model and forecast the fire can help emergency services to work more efficiently and save lives. However, the calculations with current fire modeling techniques, such as CFD and zone models, still take too long and valuable time is often lost. Using the video driven fire spread forecasting methodology proposed in this paper, which is able to give real-time information about the state of the environment, model-based predictions of the future state of a fire can be improved and accelerated. By combining the information about the fire from models and real-time multi-modal LWIR and visual flame and smoke data an estimate of the fire can be produced that is better than could be obtained from using the model or the data alone.

► Novel video driven fire spread forecasting methodology.
► Automatic extraction of real-time information about the (fire) environment.
► Multi-modal video based improvement and acceleration of model-based predictions.
► Real-time multi-modal LWIR and visual fire analysis.