Citas bibligráficas
Clemente, C., (2024). Determinación del número de árboles en plantaciones forestales (Oxapampa) utilizando imágenes captadas con una aeronave pilotada remotamente [Universidad Nacional Agraria La Molina]. https://hdl.handle.net/20.500.12996/6293
Clemente, C., Determinación del número de árboles en plantaciones forestales (Oxapampa) utilizando imágenes captadas con una aeronave pilotada remotamente []. PE: Universidad Nacional Agraria La Molina; 2024. https://hdl.handle.net/20.500.12996/6293
@misc{renati/953628,
title = "Determinación del número de árboles en plantaciones forestales (Oxapampa) utilizando imágenes captadas con una aeronave pilotada remotamente",
author = "Clemente Escobar, Christhian",
publisher = "Universidad Nacional Agraria La Molina",
year = "2024"
}
This research focuses on the development and application of a methodology to determine the number of trees in forest plantations using images captured with the Zenmuse X5 visual sensor mounted on the Matrice 600 pro aircraft. The methodology initially consists of photogrammetric processing to obtain the products: Digital Surface Model (DSM), Digital Terrain Model (DTM) and orthomosaic. Then using the QGIS software and its SAGA GIS complement, the CHM (Canopy Height Model) was calculated with the objective of differentiating the trees by their height, this consisted of the difference between the DSM and DTM, to this result the Gaussian filter algorithm, and then the watershed segmentation algorithm until the generated points are located in the highest part of the CHM raster, which corresponds to the tree crown. To reduce the points generated in excess, a semi-automatic cleaning was carried out, obtaining as a result after this cleaning the number of trees detected by the proposed methodology. To validate the methodology, 16 subplots were used, in each subplot the number of trees was calculated by 3 different methods, counted manually in the field, counted visually in the orthomosaic and counted by the proposed methodology. The results were compared pair by pair and with the chi square goodness of fit statistic it was calculated that there is no significant difference between the distribution of the results of the 3 mentioned methods, in this way the proposal was validated for this type of evaluation. . Finally, the precision in each subplot was calculated, visually comparing the coincidences between the trees counted visually in the orthomosaic and the trees detected by the methodology and the result was an average precision equal to 75%, which is considered acceptable.
Este ítem está sujeto a una licencia Creative Commons Licencia Creative Commons