Lia Nara Balta QuintaWillian Paraguassu AmorimMarcelo Henriques de CarvalhoMarney Pascoli CeredaHemerson Pistori

Detecção de Nós Maliciosos em Redes de Sensores sem Fio

In the production of honey some pollen grains are deposited on the bees. On returning to the hive, the grains of the species fall into the product and can be viewed through a microscope. Currently for identifying the product, it is necessary to count the pollen grains present in the honey.This work aims to use techniques of computer vision and artificial intelligence in microscopic images, to classify pollen grains present in the honey. To perform the classification of pollen grains was applied a technique called Optimal-Path Forest (OPF). We extracted attributes of color (RGB and HSB), shape (form factor, roundness, aspect ratio and density) and texture (co-occurrence matrix and interaction maps). The OPF was compared with traditional classifiers C4.5, SVM and KNN. According to Friedman test can be affirmed that there is no difference between the classifiersin relation to the result of classification and that the SVM has more training time than the other classifiers.

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato:
     Mantida por: