The artificial intelligence fights against wildfires in Republic of Belarus

On March 13, 2019 the ITV-AxxonSoft company reported that in the Public forestry and landscape institution (GLHU) "Miloshevichsky Forestry" located in the Lelchitsky region of the Gomel region the system of early detection of wildfires on the basis of intellectual video surveillance passed test.

The test system founded on the Intelligence platform using neural network detectors of smoke includes three rotary cameras combined with thermal imagers and installed on fire towers at the height of 35 meters. RUP "Belgosles" acted as the customer of a system.

Cameras automatically rotate around a vertical axis by 360 degrees, fixing the smoke rising over the wood. The neural network which is specially trained at similar stages defines availability of smoke, distinguishing it from clouds, fog or dust. In good weather the range of detection can reach 50 km.

On the monitor video from cameras and thermal imagers and the interactive map on which sectors of observation of cameras are designated at the moment is displayed. At smoke detection, together with the sound notification and an output of the disturbing message to the monitor, to email of staff of forest area the screenshot of the card and a frame which was alarming goes. Accuracy of localization of the center of ignition is several hundred meters that is a good indicator for thousands of hectares of forest lands.

The software of ITV - AxxonSoft replaced a system which did not use neural network methods of the analysis of images, had the smaller range and accuracy of detection of smoke, gave more false alarms.

According to information for March, 2019 a system is implemented only in Miloshevichsky forestry. It is going to connect to a system all forestry of the country. In ITV - AxxonSoft consider that it will increase reliability of control, having reduced influence of a human factor. Besides, the intelligent system will demand the smaller number of operators, it will not be necessary to monitor constantly the monitor, and it will be necessary only to react to alarms quickly. From 580 fire towers and masts in forestries of Republic of Belarus 80% are equipped with surveillance cameras. It means that the base for implementation of an intelligent system of early detection of wildfires is ready.