| Popis: |
Tehnološki napredek omogoča optimizacijo opravil v različnih panogah, med katere sodi tudi kmetijstvo. Pršenje fitofarmacevtskih sredstev na vinsko trto varuje rastline pred škodljivci, povzročitelji bolezni in plevelom, vendar pa njihove učinkovine vsebujejo snovi, ki lahko negativno vplivajo na ljudi in okolje. Eden izmed načinov zmanjšanja in optimizacije njihove uporabe je preko prilagoditve programa pršenja glede na geometrijsko obliko, število in velikost listov vinske trte. Preko zajema slikovnega gradiva s pomočjo brezpilotnih letalnikov, je v kratkem času mogoče ustvariti tri dimenzionalen model vinograda. Ob zajemu podatkov so brezpilotni letalniki izpostavljeni različnim vplivom okolja, zato ima končni 3D model pogrešek. S podrobnim opisom okolja in ročno meritvijo dimenzij, smo primerjali med seboj več 3D modelov, katerih slikovno gradivo je bilo zajeto ob različnih razmerah z različnimi parametri. Iz primerjav smo ugotovili, da je metoda fotogrametrije ob primernih začetnih nastavitvah zajema primerna za izdelavo kompleksnega 3D modela vinske trte, ki ima dovolj visoko natančnost in ponovljivost. Technological advancements enable optimization of tasks in different sectors, one of those being agriculture. Spraying vine with phytopharmaceuticals protects a plant from pests, diseases and weeds, however ingredients of such chemicals could pose risk to people and environment. A way to reduce and optimize usage of phytopharmaceuticals is by regulating the amount of spraying by taking into account the geometrical shape, number and size of vine’s leaves. By capturing photographic data with the help of uncrewed aircraft systems, it is possible to create three dimensional model of the vineyard. Because of different environmental influences, which act on UAS, final 3D model has an intrinsic error. With detailed description of the environment and manual measurements, we created multiple 3D models, made from data with different capture variables. Results were compared and we found the following: the method of photogrammetry can be accurate and repeatable enough for creation of complex 3D model of a vine assuming the appropriate settings for given environment were set before capturing the data. |
