INOV has been developing, building and testing innovative systems for fire detection in the forest and industrial environments based on the LIDAR (Light Detection And Ranging) technology. In LIDAR applications an intense pulse of laser radiation of short duration is emitted to the atmosphere. When the pulse encounter a solid obstacle or a smoke plume, a part of the emitted radiation is reflected back to the LIDAR. This retroreflected radiation is captured by the light-gathering optics and analysed by appropriate methods for identifying and locating the target.This type of detection method is known as active, as opposed to the passive methods based on the analysis of video or infrared images. The advantages of the active detection are: relative independence on the weather conditions, possibility of night operation, accurate fire location and excellent ability to reject false alarms.
The system uses an advanced artificial neural network method to distinguish the signatures of smoke plumes from signals associated with other phenomena. The LIDAR detectors can be used without additional equipment or in conjunction with video cameras (telesurveillance), which allows to combine the advantages of both detection techniques. In for the latter case we designed a system in which the video camera shares the optical system of the LIDAR detector; here the image analysis is invoked for the preliminary detection while the LIDAR is used for confirmation (dramatically reducing the false alarm probability) and precise location of the fire.
More recently we developed a method for studying the health of growing plants by their fluorescence induced by coherent or incoherent radiation. This system is planned to be used for evaluating the forest-fire impact in vegetation.