Detecção e mapeamento de vegetação daninha em áreas de cultivo de cana-de-açúcar por meio de modelagem geoespacial e técnicas de sensoriamento remoto

• Autor(a) principal: Gromboni, João Felipe [UNESP]
• Data de Publicação: 2018
• Tipo de documento: Trabalho de conclusão de curso
• Idioma: por
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://hdl.handle.net/11449/203681; http://www.athena.biblioteca.unesp.br/exlibris/bd/capelo/2019-06-17/000916589.pdf
• topic: Ervas daninhas; Cana-de-açúcar - Cultivo; Sensoriamento remoto; Vegetação - Mapeamento; Agricultura de precisão; Geoprocessamento

The presence of weeds in sugar cane cultivation areas can trigger off the 85% drop in the productive potential of the fields, and the activities involved in the control of these infestations represent 8% of the total cost of implementing canebrake. It is observed that current methods of mapping of this information lack compatible techniques to data collection with the speed and scale with which these infestations are manifested in the crop. In this context, the present research proposal aims to establish a methodological procedure to detect and map the presence of weed vegetation in sugarcane fields by means of orbital images of medium spatial resolution. Thus, the strategy adopted to lead the work was to characterize the vegetation index variations (NDVI) of the field targets, with special emphasis on the phenological development phases of healthy sugarcane. The temporal series of the NDVI values were simulated from orbital images (Landsat, Sentinel, CBERS e IRS), seeking to detect the anomalies in the expected behavior of the temporal profile of the vegetation. The outputs showed that the best phases for weed detection are between the cane sprouting period and the beginning of the growth phase, when the soil presents a relation of only 18% of corresponding green leaf to the sugar cane, and with low NDVI values. The maturation period also appeared to be adequate, since the percentage of dry leaf in the soil covers 63% of the area. Therefore in these conditions, it was observed that areas of weed infestation can be detected due to the high vigor of phytomass, contrasting with the conditions of sugar cane