RESUMO
Bee pollen is a hive product, resulting from floral pollen agglutination by worker bees and it is characterized by its excellent bioactive and nutritional composition. Currently, research is focused on bee pollen applications on food industry, because this product has been considered an excellent source of compounds for human nutrition. It is also important in some industries, where color and particle size are important characteristics for production. Due to the granular nature of bee pollen, conventional colorimetry does not allow describing color correctly; thus, digital image analysis is a better alternative. This technique could also allow classifying bee pollen according to its appearance beyond the color. Consequently, the aim of this work was to develop a novel methodology for image data processing to classify bee pollen as ingredient in food industry. Seven color groups in samples were established regarding harvest month and particle size. It was possible to calculate the percentage of each color group in all samples. This methodology also allowed selecting each fraction for different applications in food industry using colorimetry, granulometry and the relationship between both of them.
