Automation in accession classification of Brazilian Capsicum germplasm through artificial neural networks

Germplasm classification by species requires specific knowledge on/of the culture of interest. Therefore, efforts aimed at automation of this process are necessary for the efficient management of collections. Automation of germplasm classification through artificial neural networks may be a viable and less laborious strategy. The aims of this study were to verify the classification potential of Capsicum accessions regarding/ the species based on morphological descriptors and artificial neural networks, and to establish the most important descriptors and the best network architecture for this purpose. Five hundred and sixty-four plants from 47 Brazilian Capsicum accessions were evaluated. Neural networks of multilayer perceptron type were used in order to automate the species identification through 17 morphological descriptors. Six network architectures were evaluated, and the number of neurons in the hidden layer ranged from 1 to 6. The relative importance of morphological descriptors in the classification process was established by Garson's method. Corolla color, corolla spot color, calyx annular constriction, fruit shape at pedicel attachment, and fruit color at mature stage were the most important descriptors. The network architecture with 6 neurons in the hidden layer is the most appropriate in this study. The possibility of classifying Capsicum plants regarding/ the species through artificial neural networks with 100 % accuracy was verified.

Automation in accession classification of Brazilian Capsicum germplasm through artificial neural networks

Germplasm classification by species requires specific knowledge on/of the culture of interest. Therefore, efforts aimed at automation of this process are necessary for the efficient management of collections. Automation of germplasm classification through artificial neural networks may be a viable and less laborious strategy. The aims of this study were to verify the classification potential of Capsicum accessions regarding/ the species based on morphological descriptors and artificial neural networks, and to establish the most important descriptors and the best network architecture for this purpose. Five hundred and sixty-four plants from 47 Brazilian Capsicum accessions were evaluated. Neural networks of multilayer perceptron type were used in order to automate the species identification through 17 morphological descriptors. Six network architectures were evaluated, and the number of neurons in the hidden layer ranged from 1 to 6. The relative importance of morphological descriptors in the classification process was established by Garson's method. Corolla color, corolla spot color, calyx annular constriction, fruit shape at pedicel attachment, and fruit color at mature stage were the most important descriptors. The network architecture with 6 neurons in the hidden layer is the most appropriate in this study. The possibility of classifying Capsicum plants regarding/ the species through artificial neural networks with 100 % accuracy was verified.(AU)

Sistema computacional para auxílio ao diagnóstico em exames de tuberculose animal / A computer software system to assist in the diagnosis of animal tuberculosis

The results obtained in evaluating the efficiency of a Neuro-Fuzzy System NEFCLASS (Neuro-Fuzzy Classification) in image classification of cattle tuberculosis, based on its texture features extracted using the wavelet transform are presented. For testing, images of animal tissues diagnosed with tuberculosis were used, as provided by the Tuberculosis Laboratory at the Instituto Biológico de São Paulo. The results of this study can serve as a basis for developing systems for diagnosis aimed at reducing human effort, by automating all or parts of the classification of images, helping lab technicians to sort amongst different pathologies.(AU)

Utilização da visão artificial para diagnóstico nutricional de nitrogênio, fósforo, potássio e manganês em milho / Use of the artificial vision for nutritional diagnosis of nitrogen, phosphorus, potassium and manganese in corn

Um sistema de visão artificial (SVA) para diagnose nutricional de milho, baseado em análise de imagens de folhas foi recentemente proposto pelo GCC-IFSC e Agrárias-FZEA/USP. O objetivo do estudo foi avaliar o estado nutricional do milho cultivado em casa de vegetação em solução nutritiva, com deficiência e suficiência nutricionais induzidas de nitrogênio (N), fósforo (P), potássio (K) e manganês (Mn) utilizando visão artificial, e posteriormente em campo visando validar o diagnóstico pelo sistema de visão artificial desenvolvido. As doses dos nutrientes foram constituídas pela omissão, 1/5, 2/5 e a dose completa, combinadas em três estádios de desenvolvimento do milho (V4, V7 e R1), com quatro repetições. O experimento foi individual para cada elemento. Em cada época foram coletadas imagens de folhas indicativas do estádio (FI), folhas velhas (FV) para o N, P e K e folhas novas para o Mn, que foram primeiramente digitalizadas em 1200 dpi, e em seguida encaminhadas para serem analisadas quimicamente. Também foram avaliadas nas plantas, as variáveis biométricas (altura, diâmetro do colmo e número de folhas) e determinar as produções de massa seca da parte aérea e do sistema radicular, além da determinação dos teores de nutrientes. A omissão de N, P e K proporcionaram deficiências nutricionais características nas folhas do milho, quando cultivado em casa de vegetação. As modificações nas folhas do milho, com as doses dos nutrientes estudadas, possibilitaram a obtenção de imagens necessárias para o desenvolvimento do SVA, em casa de vegetação. A utilização das imagens da casa de vegetação para treinar o SVA visando à validação de imagens do campo, gerou confusão na interpretação, levando a erros de classificação, entretanto, o uso desta tecnologia para diagnose nutricional do milho, tanto em casa de vegetação, como no campo, é promissora.

An artificial vision system (AVS) for nutrient diagnosis of corn, based on analysis of images of leaves was recently proposed by SCG-IFSC and Agrarian-FZEA/USP. The objective was evaluate the nutritional status of maize grown in a greenhouse in nutrient solution with induced nutritional deficiency and sufficiency of nitrogen (N), phosphorus (P), potassium (K) and manganese (Mn) using artificial vision, and correlate the results obtained with foliar analysis, and then the field in order to validate the diagnosis by artificial vision system developed. Doses of nutrients were established by omission, 1/5, 2/5 and full dose, combined into three developmental stages of corn (V4, V7 and R1), with four replications. The experiment was for each individual element. Images of leaves were collected in each epoch indicating the stage (FI), old leaves (FV) for N, P and K and Mn for new leaves, which were first scanned at 1200 dpi, then sent to be analyzed chemically. The biometric variables (height, stem diameter and number of leaves) were also evaluated, and the dry matter production of shoots and roots was determined, besides the determination of nutritional content. The omission of N, P and K caused typical nutrient deficiencies provided in the leaves of maize when grown in a greenhouse. The changes in the leaves of maize, with doses of nutrients studied, allowed the imaging necessary for training SVA in a greenhouse. The use of images of the greenhouse to train the SVA aiming to validate images of the field has led to confusion in the interpretation leading to errors of classification, however the use of this technology for nutrient diagnosis of corn, both in the greenhouse and in the field, is promising.

Utilização da visão artificial para diagnóstico nutricional de nitrogênio, fósforo, potássio e manganês em milho / Use of the artificial vision for nutritional diagnosis of nitrogen, phosphorus, potassium and manganese in corn

Um sistema de visão artificial (SVA) para diagnose nutricional de milho, baseado em análise de imagens de folhas foi recentemente proposto pelo GCC-IFSC e Agrárias-FZEA/USP. O objetivo do estudo foi avaliar o estado nutricional do milho cultivado em casa de vegetação em solução nutritiva, com deficiência e suficiência nutricionais induzidas de nitrogênio (N), fósforo (P), potássio (K) e manganês (Mn) utilizando visão artificial, e posteriormente em campo visando validar o diagnóstico pelo sistema de visão artificial desenvolvido. As doses dos nutrientes foram constituídas pela omissão, 1/5, 2/5 e a dose completa, combinadas em três estádios de desenvolvimento do milho (V4, V7 e R1), com quatro repetições. O experimento foi individual para cada elemento. Em cada época foram coletadas imagens de folhas indicativas do estádio (FI), folhas velhas (FV) para o N, P e K e folhas novas para o Mn, que foram primeiramente digitalizadas em 1200 dpi, e em seguida encaminhadas para serem analisadas quimicamente. Também foram avaliadas nas plantas, as variáveis biométricas (altura, diâmetro do colmo e número de folhas) e determinar as produções de massa seca da parte aérea e do sistema radicular, além da determinação dos teores de nutrientes. A omissão de N, P e K proporcionaram deficiências nutricionais características nas folhas do milho, quando cultivado em casa de vegetação. As modificações nas folhas do milho, com as doses dos nutrientes estudadas, possibilitaram a obtenção de imagens necessárias para o desenvolvimento do SVA, em casa de vegetação. A utilização das imagens da casa de vegetação para treinar o SVA visando à validação de imagens do campo, gerou confusão na interpretação, levando a erros de classificação, entretanto, o uso desta tecnologia para diagnose nutricional do milho, tanto em casa de vegetação, como no campo, é promissora. (AU)

An artificial vision system (AVS) for nutrient diagnosis of corn, based on analysis of images of leaves was recently proposed by SCG-IFSC and Agrarian-FZEA/USP. The objective was evaluate the nutritional status of maize grown in a greenhouse in nutrient solution with induced nutritional deficiency and sufficiency of nitrogen (N), phosphorus (P), potassium (K) and manganese (Mn) using artificial vision, and correlate the results obtained with foliar analysis, and then the field in order to validate the diagnosis by artificial vision system developed. Doses of nutrients were established by omission, 1/5, 2/5 and full dose, combined into three developmental stages of corn (V4, V7 and R1), with four replications. The experiment was for each individual element. Images of leaves were collected in each epoch indicating the stage (FI), old leaves (FV) for N, P and K and Mn for new leaves, which were first scanned at 1200 dpi, then sent to be analyzed chemically. The biometric variables (height, stem diameter and number of leaves) were also evaluated, and the dry matter production of shoots and roots was determined, besides the determination of nutritional content. The omission of N, P and K caused typical nutrient deficiencies provided in the leaves of maize when grown in a greenhouse. The changes in the leaves of maize, with doses of nutrients studied, allowed the imaging necessary for training SVA in a greenhouse. The use of images of the greenhouse to train the SVA aiming to validate images of the field has led to confusion in the interpretation leading to errors of classification, however the use of this technology for nutrient diagnosis of corn, both in the greenhouse and in the field, is promising. (AU)