Reduction of false alarm rate in automatic forest fire infrared surveillance systems

摘要: Abstract One of the main problems related to infrared remote sensing forest fires by automatic systems concerns rejection false alarms. The study spectral radiance emitted a biomass fire has been used define algorithms that permit separate from other sources considered as different behaviour medium (3–5 μm) and thermal (8–12 regions basis for definition new “figure” merit, so-called Fire Index (FI). An experimental burn test ability FI identify clearly fires. A second algorithm based on an intraband analysis, Mid-IR (MFI), proposed give complementary information FI. Evolution MFI values with distance IR calculated. Finally, validation both performed using two cameras working in windows.