Programa de Pós-Graduação em Computação Aplicada - PPCA/NDAE/Tucuruí
URI Permanente desta comunidadehttps://repositorio.ufpa.br/handle/2011/9398
Navegar
Navegando Programa de Pós-Graduação em Computação Aplicada - PPCA/NDAE/Tucuruí por Assunto "Agricultura familiar"
Agora exibindo 1 - 1 de 1
- Resultados por página
- Opções de Ordenação
Item Acesso aberto (Open Access) Estimativa da produção de uma lavoura através de imagens digitais capturadas por veículo aéreo não tripulado (VANT)(Universidade Federal do Pará, 2018-10-08) SEREJO, Gerson Lima; GOMES, Ana Claudia da Silva; http://lattes.cnpq.br/9898138854277399; SANTOS, Viviane Almeida dos; http://lattes.cnpq.br/1489376127395764The use of Unmanned Aerial Vehicles (UAV) is becoming an important accessible tool for small to medium sized agribusiness. Its application supports the execution of complex and laborious activities, as well as promotes new studies and challenges for the field to assist the farmer's decision making. The County of Tucuruí, in the State of Pará Brazil, is part of a region that concentrates a great amount of rural properties characterized by being of family agriculture. The objective of this work is to present an exploratory study for applying steps of Digital Image Processing (DIP) and Computational Vision (CV) in images captured by a UAV to obtain the quantification of cassava seedlings and, consequently, harvest of this crop in a farm of the county. The scientific contribution of this study corresponds to the results obtained from the application of 4 vegetation indices: ExG, ExR, (ExG-ExR) and MaxG. The MaxG index presented the best result, counting 91% of the seedlings, in the best case, with an accuracy of 70%. The ExR index was more appropriate for counting the seedlings in initial stages of germination. The index (ExG-ExR) allowed the estimation with unsupervised thresholding, which improves the development of CV systems for this purpose. The ExG index surprised us with the lowest performance for the studied context, counting 58% of the seedlings, in the worst case, with accuracy of 73%. As practical contributions to the farmer, this study made it possible to raise awareness of the importance of forecasting the harvest to better plan the negotiation of production, later plantings and the search for resources to increase the mechanized area of the crop. Further indepth research needs to be conducted to confirm these findings.