Relação diamétrica utilizando sistema de inferência neuro-fuzzy

Abstract

Since the beginning of the establishment of planted forests in Brazil research has been developed by companies and universities which has brought considerable progress in the Brazilian forestry sector, among the objects of research stands out activities that aim to decrease the sampling intensity in inventories. For a long time the height was the variable of interest in which methods were developed to estimate it, however with the advance of the use of remote sensing tools and the implementation of LiDAR in inventories the variable of interest became the diameter and for such there are still few alternatives. In this case the opportunity arises to use artificial intelligence to establish diameter estimates from tree heights. In this study the efficiency of the hybrid systems for diameter estimation is proven, but there is a need to have data that represents the validation data. When this does not occur, the results tend to be more imprecise.