MÉTODO COMPUTACIONAL PARA AVALIAÇÃO DO CRESCIMENTO RADICULAR DA CULTURA DA SOJA

• Ano de defesa: 2013
• Autor(a) principal: Name, Márcio Hosoya
• Orientador(a): Fonseca, Adriel Ferreira da
• Banca de defesa: Zagonel, Jeferson , Sanches, Ionildo José
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: UNIVERSIDADE ESTADUAL DE PONTA GROSSA
• Programa de Pós-Graduação: Programa de Pós Graduação Computação Aplicada
• Departamento: Computação para Tecnologias em Agricultura
• País: BR
• Palavras-chave em Português: PDI; Sistema radicular; Glycine max L. (Merr.); Raiz
• Palavras-chave em Inglês: PDI; Root system; Glycine max L. (Merr.); Root
• Área do conhecimento CNPq: CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: http://tede2.uepg.br/jspui/handle/prefix/166

The evaluation of the root system is important for better understanding of the effects of nutrient management on soil and plant mineral nutrition. However, this review has been a complex, comprehensive and conducive to sampling error and measurement activity. In this context, digital image processing and image analysis can help as they have been used in solutions of similar problems. The objective of this study was to develop a computational method to assist in the evaluation of root growth for washed samples of soybean, reducing the time spent. The results were compared with traditional methods (intersection line and fresh mass), and SAFIRA software Embrapa. The proposed method has been developed in Java platform with the OpenCV library supply through the plug-in JavaCV. The method validation was performed by comparing images (300 dpi) of samples of copper wires, 10, 20 and 50 mm in length, with values obtained manually using a caliper, obtaining coefficients of variation (CV) between 0.01 and 2.99%. Relations among the estimated lengths with the proposed method and those obtained by the traditional method yielded CVs ranging between 0.10 and 2.10 %, which were better than the average SAFIRA (5.11% < CV < 49.45 %) at 300 dpi images. The method also reduced the time to obtain the attributes of roots in more than 53%. Thus, the proposed method showed to be effective, especially for measures of length and area samples are washed and is recommended for studies of roots of soybean.