, Maila Claro, L. Santos, Flávio Henrique, Alcilene Sousa, Nayara Moura
Univeridade Federal do Piauí. Picos-PI Brazil
Glaucoma is an asymptomatic disease and is considered the second greater cause of blindness in the world, it attacks the optic nerve causing irreversible damage to the visual field. Although there is no cure there exist treatments, which can be facilitated from a early diagnosis. Thus, this work consists of methods for feature extraction in retinal images from the use of descriptors for the automatic detection of glaucoma. These methods are constructed from the use of bag of words, which are robust forms of representation, and the texture descriptors Binary Pattern Place and Grey Level Co-occurrence Matrix (GLCM).