Mostrar registro simples

dc.contributor.authorSilva, Matheus Nunespt_BR
dc.contributor.otherMachado, Álvaro Muriel Limapt_BR
dc.contributor.otherLingnau, Christelpt_BR
dc.contributor.otherUniversidade Federal do Paraná. Setor de Ciências Agrárias. Programa de Pós-Graduação em Engenharia Florestalpt_BR
dc.date.accessioned2019-06-28T17:56:47Z
dc.date.available2019-06-28T17:56:47Z
dc.date.issued2011pt_BR
dc.identifier.urihttps://hdl.handle.net/1884/27398
dc.descriptionOrientadora : Profa. Dra. Christel Lingnaupt_BR
dc.descriptionCoorientador : Prof. Dr. Álvaro Muriel Lima Machadopt_BR
dc.descriptionDissertação (mestrado) - Universidade Federal do Paraná, Setor de Ciências Agrárias, Programa de Pós-Graduação em Engenharia Florestal. Defesa: Curitiba, 16/12/2011pt_BR
dc.descriptionBibliografia: fls. 124-139pt_BR
dc.descriptionÁrea de concentração : Manejo florestalpt_BR
dc.description.abstractResumo: O uso de dados laser scanner terrestre é uma ferramenta importante para coleta de dados tridimensionais. Esta dissertação teve como objetivo utilizar dados obtidos por meio de varredura simples e múltipla para a mensuração de árvores validando os resultados com valores paramétricos e comparar com os resultados de dados dendrométricos medidos de forma tradicional. A coletada de dados tradicionais, laser scanner terrestre e paramétricos foi realizada em 12 árvores de Pinus spp.. As medidas tradicionais foram diâmetros ao longo do fuste medidos com a árvore derrubada com suta e fita métrica e altura total com hipsômetro Vertex III. As medidas paramétricas foram a altura total das árvores medidas com trena após a derrubada, fotografias digitais classificadas de cortes transversais em alturas especificas e volume determinado por xilômetro das três primeiras toras de um metro. O processamento de dados laser consistiu na filtragem manual das árvores para obtenção dos fustes e a criação e programação de algoritmos para determinação das variáveis dendrométricas. Para obtenção da altura total das árvores aplicou-se um único método de processamento laser tanto para varreduras múltiplas como simples. Testaram-se quatro métodos de obtenção de diâmetros pelo processamento laser (Distância Máxima, Circunferência Ótima, Área Laser e Área Laser Modificado), sendo que os quatro foram aplicados nas varreduras múltiplas e três nas varreduras simples totalizando um total de 9 combinações de métodos e modelos de varredura. Os volumes das três primeiras toras foram determinados pelo método de Smalian, Huber e Newton e comparados com a cubagem rigorosa por medidas tradicionais e com o xilômetro. Os resultados indicaram que alguns pontos da árvore são problemáticos para a filtragem como é o caso da base e das copas. A ocorrência de vento no momento da aquisição de dados é um fator limitante principalmente em varreduras múltiplas. As alturas obtidas pelo processamento laser, utilizando dados de varredura simples ou múltipla, foram superiores aos obtidos pelo hipsômetro Vertex III. Porém, não existiram diferenças estatísticas (Tukey ? = 0,05) entre os resultados laser, medidas tradicionais e o valor paramétrico. Para os diâmetros e áreas transversais quando comparados com a área transversal real apenas um dos métodos laser propostos diferiu estatisticamente dos resultados das medidas tradicionais e do valor paramétrico. Conclui-se que a partir da altura onde existe um adensamento de copas tornou-se inviável a modelagem com o laser. Para os volumes, o método de Huber foi o mais eficiente para as primeiras toras utilizando dados laser. Os métodos de Newton e Smalian não são indicados para a volumetria da primeira tora, pois esta sofre muita influência da filtragem da base das árvores. Estatisticamente, eliminando o método de Smalian e Newton para a primeira tora, as medidas laser, tradicionais e paramétricas são iguais pelo teste de Tukey a 5% de significância. Embora em geral não existam diferenças estatísticas entre varreduras simples e múltiplas em comparação com medidas tradicionais e paramétricas, os resultados de varredura múltipla tem menor erro absoluto e menor variância de resultados sendo indicado o uso principalmente para obtenção de diâmetro e áreas transversais nas partes mais altas das árvores.pt_BR
dc.description.abstractAbstract: The use of terrestrial laser scanner data is an important tool for collecting three-dimensional data. This paper aims to use data obtained from single and multiple scan for measuring trees, validating the results with parametric values and comparing them with traditionally measured data. The data collected from traditional, parametric and terrestrial laser scanner measuring were taken from 12 Pinus spp. trees. The traditional measurement were the diameters taken along the stem of overturned trees by both tape measure and caliper; and the height, taken through a Vertex III hypsometer. The parametric measures were the total height of the trees, that were measured by tape after the overthrow; classified digital photographs from cross-sections taken at determined heights; and their volumes, determined by means of xylometer from their first three one meter logs. The laser data processing consisted of a manual filtering in order to obtain the tree stems and the creation and programming of algorithms for the determination of dendrometric variables. With the aim to obtain the total height of the trees, it was applied a single laser processing method for both simple and multiple scans. Four methods were tested to obtain the diameters by laser processing; the four of them were applied on multiple scans and three of them on simple scan, comprising a total of nine combinations of methods and sweeping models. The volumes of the first three logs were determined by the method of Smalian, Huber and Newton and compared with the scaling obtained by traditional measurements and xylometer. The results indicated that some points of the tree are problematic for the filtering, such as the base and the canopy. The occurrence of winds at the time of data acquisition is a limiting factor, particularly in multiple scans. The heights obtained by laser scan processing, using data from single or multiple scan, were better than those obtained by the Vertex III hypsometer. However, there were no statistical differences (Tukey á = 0,05) between the laser output, the traditional measurements and the parametric value. For the diameters and cross-sectional areas, when compared to the real cross-sectional area, only one of the proposed laser methods differed significantly from the results of the traditional measurements and the parametric value. We conclude that from the height where there is a densification of crowns, it became unviable the modeling by means of terrestrial laser scanner. For the volumes, the Huber method is the most efficient for the first log data using laser. The methods of Newton and Smalian are not prescribed for the volumetry of the first log, because it suffers a strong influence from the base filtering. Statistically, rejecting the Smalian and Newton method for the first log, laser measurements, traditional and parametric values are equal according to the Tukey test at 5% significance level. Although in general there are not statistical differences between single and multiple scans compared to traditional measurements and parameters, the results of multiple scans presents a lesser absolute error and lesser variance of results, being indicated mainly for the obtention of the diameters and cross-section areas of the higher parts of trees.pt_BR
dc.format.extent130f. : il. [algumas color.], grafs., tabs.pt_BR
dc.format.mimetypeapplication/pdfpt_BR
dc.languagePortuguêspt_BR
dc.relationDisponível em formato digitalpt_BR
dc.subjectArvores - Mediçãopt_BR
dc.subjectSistemas de varredurapt_BR
dc.subjectDendrometriapt_BR
dc.subjectPinheiropt_BR
dc.subjectLaserspt_BR
dc.titleAplicação de laser scanner terrestre para determinação de variáveis dendrométricas em Pinus SPPpt_BR
dc.typeDissertaçãopt_BR


Arquivos deste item

Thumbnail

Este item aparece na(s) seguinte(s) coleção(s)

Mostrar registro simples