Aplicação do LIDAR no inventário de florestas plantadas / Lidar application in planted forest inventory

AUTOR(ES)
FONTE

IBICT - Instituto Brasileiro de Informação em Ciência e Tecnologia

DATA DE PUBLICAÇÃO

05/08/2011

RESUMO

LIDAR (Light Detection and Ranging) has become an efficient tool for data extraction in forest inventory and other forest related studies. This works objective was to establish an efficient methodology for the use of this technology with the goal of counting the number of stems in a Eucalyptus plantation in South Bahia, Brazil, from the processing of raw data up to the results in counting the number stems. The efficiency of each intermediate step is of vital importance in the quality of the results, and thus were obtained in the following processing order: organization of the raw data (3D vectors); creation of canopy height surface through Triangular Irregular Network and Inverse distance weighted interpolation (IDW); local maximum algorithm for counting the number of steams with 3X3 and 5X5 windows. The obtained results were compared to a manual counting procedure executed in a Quickbird image. The automated counting procedure had an accuracy of 97.36% in the data set generated by IDW interpolation with 5X5 window counting procedure. This result exhibits the efficiency of the methodology and its potential for future applications.

ASSUNTO(S)

inventário florestal modelagem detecção de árvores local máxima recursos florestais e engenharia florestal forest inventory modeling single tree detection local maxima

ACESSO AO ARTIGO

Documentos Relacionados