The work on forest ecosystem monitoring software developed by Tomsk Scientific Center together with their partners from OOO Indorsoft entered its implementation phase. Just recently a groundbreaking achievement has been made--the software has learnt to differentiate between the species compositions of tree stands.
"What we have is a self-learning artificial neural-network based on the program that can recognize tree species in UAV images and assess their condition without human assistance. It can conclusively pick out a dead tree among healthy ones, determine the height of a tree," the project manager and the head of IMCES s technology transfer office Aleksandr Myagkov said, "This software has no analogues in Russia."
Aleksandr Myagkov says that the algorithms for the analysis of IR images used in this program can easily spot and assess the insect damage caused by Polygraphus proximus or Siberian silk moth.
A unique experiment with controlled invasion of pine trees is being carried out within the framework of cooperation between IMCES, Tomsk Scientific Center and Indorsoft.