Weight estimation models for commercial Pinus radiata wood in small felling stands based on UAV-LiDAR data

Development of a convolutional neural network to accurately detect land use and land cover

“…High land fragmentation and small landholdings are two prominent features in the study area, and it is important to highlight how these conditions can complicate the development of methodologies for land use classification [14] .…”

“…These areas are characterized by houses on large properties, which is why various constructions (houses, garages, annexes, etc.) are combined with gardens, orchard areas and more green areas [14] . This confusion is common in other studies like Li et al and Zhang et al [ 19 , 20 ] since the high density of the residential class when training can cause errors in less dense areas.…”