Protótipo para inspeção visual automática da qualidade de feijão: concepção, desenvolvimento e análise de custos

• Ano de defesa: 2012
• Autor(a) principal: Pires, Alexandre Cezar
• Orientador(a): Araújo, Sidnei Alves de
• Banca de defesa: Bertevello, Luiz Carlos , Farias, Thiago Michel de Brito , Pereira, Fabio Henrique
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Nove de Julho
• Programa de Pós-Graduação: Programa de Pós-Graduação de Mestrado e Doutorado em Engenharia de Produção
• Departamento: Engenharia
• País: BR
• Palavras-chave em Português: feijão de qualidade; padrões brasileiros; inspeção visual; visão computacional
• Palavras-chave em Inglês: beans; Brazilian standards; visual inspection; computer vision; quality
• Área do conhecimento CNPq: ENGENHARIAS::ENGENHARIA DE PRODUCAO
• Link de acesso: http://bibliotecatede.uninove.br/tede/handle/tede/201

The bean is an important source of energy and protein, and is present in the daily diet of the Brazilian people. As most of food products, visual properties of beans are important mainly to help the choice of consumers. In Brazil, the quality inspection of food products is usually done manually by following rules and procedures of the Ministry of Agriculture Livestock and Supplies. In manual processes of quality inspection usually occur some problems such as high cost and lack of standardization of results. Thus, automated processes in order to reduce costs, standardize results and also preserve the physical integrity of workers who work in these activities are desirable. In this paper a prototype to classify beans according to their visual qualities has been developed. This prototype was composed by a conveyor belt which carries out the bean seeds to the box of image acquisition. At the top of acquisition box there is a camera to capture the images. The acquired images were processed on a PC Pentium dual core 2 duo, which displays on its screen a response to the quality of the analyzed sample of beans. To this end, a software based on computer vision techniques was implemented. According to performed experiments, the prototype can be used to enhance the quality control of Brazilian beans.