Tree Vitality Assessment Using Computer Vision and Determination of Vegetation Spectral Indices

Monitoring tree vitality is an important part of tree assessment in an urbanized landscape. Early identification and solution of vitality problems is important to maintain healthy trees and improve the quality of the urban environment. Determining vitality is a difficult task because trees in an urbanized environment face various stress factors, such as soil compaction, lack of water, air pollution, etc. Various methods are used to determine the vitality of trees, which are based primarily on visual evaluation directly in the field. A simple and useful tool for monitoring the vitality of plants, including trees in the urban environment is the use of the Normalized Difference Vegetation Index (NDVI), the determination of which is based on scanning the territory, and its use can play an important role in building and using a digital twin of the city. This index makes it possible to track changes over time and quickly identify areas where trees are under stress or dying. NDVI can be conveniently calculated from a georeferenced orthophotomosaic, onto which tree crown contours can be mapped. NDVI pixels inside the tree can indicate the tree’s vitality, and NDVI of greenery outside the tree contours can indirectly determine soil quality. Tree contours are obtained from Terrestrial Laser Scanning (TLS). The paper deals with the tree vitality assessment in the studied area of the city center of Žiar nad Hronom (Slovakia) using TLS.