Rastreamento de embarcações em imagens satelitais utilizando metodologia Multicritério para a priorização em tarefas de busca e salvamento Tracking vessels in satellite images using Multicriteria methodology for prioritizing search and rescue tasks
Between 2017 and 2018, 160 million bags of 60kg of coffee were produced in the world. Currently, Brazil occupies the position of largest producer of coffee in the world, with 31.9% of total production. In order to classify a sample of post-harvest coffee beans, this work was based on the comparison between two computational techniques. The first one is the use of Convolutional Neural Networks (CNN) and the second one is the combination of Digital Image Processing methods to extract features to feed an Artificial Neural Network of the Perceptron Multi-layered type (MLP) with the goal of classifying coffee beans. As a result, an overall average accuracy of 96.60% was obtained for CNN, greatly improving on the accuracy obtained by MLP, which was 78.13%. Resumo: Entre os anos de 2017 e 2018 foram produzidas 160 milhões de sacas de 60kg de café no mundo. Atualmente, o Brasil ocupa a posição de maior produtor de café do mundo, com 31,9% da produção total. A fim de classificar uma amostra de grãos de café pós-colheita, este trabalho baseou-se na comparação entre duas técnicas computacionais. A primeiraé a utilização de Redes Neurais Convolucionais (CNN) e a segundaé a junção de métodos de Processamento Digital de Imagens (PDI) para extrair características visando alimentar uma Rede Neural Artificial (RNA) do tipo Perceptron Multicamadas (MLP) com o objetivo de se obter a classificação dos grãos de café. Como resultado, obteve-se uma acurácia média global de 96.60% para a CNN, muito superiorà acurácia obtida pelo método de PDI com MLP, que foi de 78.13%.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.