Identificação de Solo Exposto e Cupinzeiros em Pastagens Utilizando Deep Learning

ABSTRACTPasture degradation is a significant challenge in livestock farmingin Brazil, affecting the environmental and economic sustainabilityof the sector. Solutions that help manage pasture areas are crucialfor Brazilian agribusiness. In this context, this work presents theapplication of YOLO model of Deep Learning to identify exposedsoil, as well as indicators of pasture degradation, in this case, thenumber of termite mounds in each area. The image base used refersto pastures in Goiás and Mato Grosso.