Forest control of Yugra
Problem: Large areas of fires in the Khanty-Mansiysk Autonomous Okrug - Yugra and illegal logging of forests. Intelligent software and hardware complex for remote monitoring and early detection of fires, control of changes in forest lands in the Khanty-Mansiysk Autonomous Okrug - Yugra is designed to ensure the safety of citizens, natural resources, minimize damage in populated areas exposed to the threat of fire spreading to them from wildfires, as well as reduce the number of violations associated with illegal logging.
Artificial Intelligence And Digital Services
Artificial intelligence
From 6 months.
Khanty-Mansiysk Autonomous Okrug - Yugra
The video monitoring functionality, thanks to the use of artificial intelligence technologies, allows you to automatically determine the coordinates of smoke with an error of up to 150 m with a camera viewing radius of up to 40 kilometers, and notify emergency services about a fire. In this case, the video cameras make a full rotation around their axis within 10 minutes. The system implements the functionality of an intelligent system for monitoring changes in forest lands. Two neural network models of our own development provide the processes of processing and analyzing space images, recording and monitoring the measures taken and additional training of the system in accordance with the following main stages: detection of the presence of a new space image in the database; specialized pre-processing and unpacking of the original archives of space images; formation of multilayer multispectral images of space images based on primary data from the archive; formation of expected polygons of changes in forest lands by a trained neural network model; automatic verification of the identified changes in forest lands for compliance with the permitting documents posted in the FGIS LesEGAIS; recording and monitoring of measures taken for each identified violation using the AIS KND; additional training of the neural network based on the results of the control and supervisory activities carried out.
This solution is a winner of the "AI Leader" competition in 2023 (https://ai-awards.ru/award/3). Roscosmos summed up the results of the "Digital Space" competition for the practical use of remote sensing data (https://www.roscosmos.ru/39069/). Scientists from Yugra told what they are working on in the field of IT technologies (https://ugra-news.ru/article/chitayte_nad_chem_rabotayut_uchenye_yugry_v_napravlenii_it_tekhnologiy/).
Mapping forest felling areas in the Khanty-Mansiysk Autonomous Okrug using space images and a neural network approach / G. A. Kochergin, O. A. Baisalyamova, M. A. Kupriyanov [et al.] – EDN XIXQLA. Neural network model for segmentation of space images in monitoring factors of deforestation of the territory / A. V. Melnikov, G. A. Kochergin, V. R. Abbazov [et al.] – EDN AUZFZS.
