Abstract

It has been found that morphological information contained in a wing can be used to identify the species of certain insects, including bees. This paper uses the information contained in the photograph of a wing to discriminate the specie of native stingless bees. A discriminator algorithm based on SIFT is proposed. Specimens used in the research were collected directly from the nests in different geographic areas of Costa Rica and correspond to 18 species of 10 genders. The proposed algorithm was validated with 436 images with 81% success in gender discrimination and 76% for the species.

Keywords: Stingless bees; classification; image processing; SIFT descriptors; clustering