Desenvolvimento de uma base de dados de imagens digitais de madeira serrada de coníferas
Carregando...
Arquivos
Data
2013-08-16
Autores
Gomes, Roger Cristhian [UNESP]
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
A utilização de Pinus proveniente de plantações na indústria madeireira brasileira é expressiva e tem sido crescente nos últimos anos. Particularmente para a madeira serrada desse gênero, utilizada em larga escala na confecção de móveis e estruturas, há a necessidade imperiosa de sua classificação quanto aos defeitos, agravada pelo uso cada vez maior de madeiras advindas de florestas jovens. Essa realidade exige a classificação da madeira serrada quanto à sua qualidade, o que pode ser feito por meio de visão humana (análise visual desempenhada por graduadores humanos) ou de máquina. Essa classificação segue, no Brasil, as normas estabelecidas pela Associação Brasileira de Normas Técnicas (ABNT), porém as mesmas não contemplam, por meio de arquivos digitais, todas as variações que possam surgir nesse campo particular. Promovendo uma pesquisa nos meios digitais constatou-se a falta de material que demonstre, por meio de imagens digitalizadas, os defeitos e variações de defeitos que possam ocorrer durante a classificação visual de peças de madeira serrada de Pinus no dia a dia das serrarias e indústrias que utilizam essa madeira serrada como matéria prima, bem como para os pesquisadores que buscam referências a tais ocorrências em suas pesquisas. Este trabalho teve como objetivo principal a elaboração de uma base de dados de imagens digitais de amostras de madeira serrada de Pinus, apresentando uma gama significativa de defeitos e possíveis variações de defeitos encontrados em madeiras serradas de árvores de reflorestamento. Para o desenvolvimento dessa base de dados de imagens de tábuas de Pinus foram utilizadas árvores do Horto Florestal da cidade de Manduri - SP, sendo desdobradas tábuas, que foram fotografadas em um ambiente preparado (estúdio), em...
The use of pine from plantations in the Brazilian timber industry is significant and has been growing in recent years. Particularly for Pinus lumber, used extensively on making furniture and wooden structures, there is the imperative necessity of its classification for defects, aggravated by the increasing use of wood coming from young forests. This fact requires the classification of lumber for quality, which can be done by human eye (visual analysis performed by human graders) or machine. This classification follows, in Brazil, standards established by Brazilian Association of Technical Standards (ABNT), however they do not include, by means of digital files, all changes that may arise in this particular field. Ulterior research in digital media points the lack of material that present, by means of scanned images, the defects and variations of defects that may occur during the visual grading of sawn wood of Pinus in in-field practice on sawmills and industries, as well as for researchers seeking references to such occurrences in their research. This work aimed the development of a database of digital images of boards of Pinus, showing a significant range of defects and possible variations of defects found in this lumber from reforestation. For the development of this database trees from Horto Florestal city Manduri-SP were sawn into boards, which were photographed in a prepared environment (studio) in various stages of its processing (after primary sawing, drying and planing), generating a set of 382 board pictures. The boards finally were classified for defects (NBR 11700) being additionally cataloged images isolated of its major defects. Besides the modeling and implementation of the database it was also developed an information system, equipped with research tools for direct access to images of boards or specific defects and their ...
The use of pine from plantations in the Brazilian timber industry is significant and has been growing in recent years. Particularly for Pinus lumber, used extensively on making furniture and wooden structures, there is the imperative necessity of its classification for defects, aggravated by the increasing use of wood coming from young forests. This fact requires the classification of lumber for quality, which can be done by human eye (visual analysis performed by human graders) or machine. This classification follows, in Brazil, standards established by Brazilian Association of Technical Standards (ABNT), however they do not include, by means of digital files, all changes that may arise in this particular field. Ulterior research in digital media points the lack of material that present, by means of scanned images, the defects and variations of defects that may occur during the visual grading of sawn wood of Pinus in in-field practice on sawmills and industries, as well as for researchers seeking references to such occurrences in their research. This work aimed the development of a database of digital images of boards of Pinus, showing a significant range of defects and possible variations of defects found in this lumber from reforestation. For the development of this database trees from Horto Florestal city Manduri-SP were sawn into boards, which were photographed in a prepared environment (studio) in various stages of its processing (after primary sawing, drying and planing), generating a set of 382 board pictures. The boards finally were classified for defects (NBR 11700) being additionally cataloged images isolated of its major defects. Besides the modeling and implementation of the database it was also developed an information system, equipped with research tools for direct access to images of boards or specific defects and their ...
Descrição
Palavras-chave
Madeira serrada, Madeira - Banco de dados, Imagens digitais - Madeira, Conifera, Madeira - Classificação, Madeira - Identificação, Lumber
Como citar
GOMES, Roger Cristhian. Desenvolvimento de uma base de dados de imagens digitais de madeira serrada de coníferas. 2013. x, 95 f. Dissertação (mestrado) - Universidade Estadual Paulista“Julio de Mesquita Filho” , Faculdade de Ciências Agronômicas de Botucatu, 2013.