The recent development of forest ecosystems – especially within the context of extensive areas of incidental felling in Czech republic during last years – indicates the necessity of future regeneration of forest stands on large areas together with subsequent large-area monitoring of quality of forest regeneration. The project aims on evaluation of possibilities of systematic monitoring of vitality and growth of both naturally and artificially regenerated forest stands, with the use of remote sensing methods.
The new methods represent a pronounced increase of work efficiency and substantial time reduction in comparison with current method of evaluation of forest regeneration success rate with the use of terrestrial walk through inspection.
In both project partner countries sets of research plots will be established with at least two different economically important forest species and with two different forest types. Images of the research plots will be captured with the use of unmanned aerial systems equipped with optical sensors. A visible range camera as well as a multispectral camera will be used for image acquisition. Acquired imagery will be processed with the use of photogrammetric method Structure from Motion. The result of data processing will be referenced three-dimensional point clouds and referenced orthomosaics of research plots. These products represent input data for automatic tree detection algorithms. Current algorithms will be modified so that they are able to process data from different forest stands and different species. For detected trees, position and dimension will be derived and the possibility of qualitative description (especially the health status) will be evaluated.
The new methods will be verified by comparison of derived parameters with ground truth data acquired from detailed terrestrial measurements. The results of the project will be published in a scientific journal as well as at a workshop prepared for the scientific and practical forestry community.