A ground system for early forest fire detection based on infrared signal processing

Abstract

This article presents a ground remote automatic system for forest surveillance based on infrared signal processing applied to early fire detection. Advanced techniques, which are based on infrared signal processing, are used in order to process the captured images. With the aim of determining the presence or absence of fire, the system performs the fusion of different detectors that exploit different expected characteristics of a real fire, such as persistence and increase. Theoretical simulations and practical results are presented to corroborate the control of the probability of false alarm. Results in a real environment are also presented to authenticate the accuracy of the operation of the proposed system. In particular, some experiments have been done to evaluate the delay of the system (tens of seconds on average) in detecting a controlled ground fire in a range of 1–10 km. Moreover, temporary evolution of false alarms and true detections are presented to evaluate the long-term performance of the system in a real environment. We have reached a detection probability of 100% at a false alarm rate of around 1 × 10⁻⁹.

Acknowledgements

This work has been supported by Generalitat Valenciana, under grant GVEMP06/ 001, and by MEC under the FPU programme.