Geoscan Bor software for taxing forests from aerial photographs from drones

Main article: Big Data

2025: Presentation of the Bor complex

Geoskan Group of Companies has developed the Bor software complex with artificial intelligence for taxing forests from aerial photographs from drones. The complex automatically processes images, creates vector maps of the location of trees and generates reports on the breed composition and wood reserves. The developer announced this on September 25, 2025.

𝓲

Фото: Geoscan

With the help of neural networks trained on Geoskan's own datacets, the software performs automatic segmentation of forest lands, determines the dry and tree crown boundaries, classifies them by breed composition and calculates the taxational characteristics of the allocation. Based on the data received, Bor generates reports on wood reserves and vector maps that can be loaded into GPS receivers, third-party programs, geoinformation systems and web portals.

If there are previously unknown neural networks of rocks within the boundaries of the isolation, the program can blindly classify trees using machine learning methods. Groups of similar trees are combined into layers, and the operator assigns the breed.

The Bor software package was tested in a pilot project to determine the characteristics of forests in the south of Sakhalin Island. Comparison of aerial survey data with forest management data showed that the results are within the margin of error.

The work of neural networks in complex with ALS gives accuracy comparable in quality to the result of a continuous list and complies with the requirements of the Forest Management Instruction. At the same time, you do not need to spend time walking through often hard-to-reach areas and manually processing data. "Bor" will become an effective tool for forest management and for the development of new territories, "said Arseniy Afanasenko, head of the development department of the Geoskan Group of Companies.