Uma Abordagem para Reconhecimento de Placas de Perigo para Pessoas com Deficiência Visual

According to the Brazilian Institute of Geography and Statistics,there are about 6.5 million visually impaired in Brazil. A frequentissue is the insertion of them in hazardous environments signaledby signs only. Solutions must be produced to improve the life qualityof visually impaired whether at work or in public places. In thissense, the present paper presents an application of danger signrecognition aimed at people with visual impairment through asupervised learning classifier called Haar Cascade. Furthermore,the paper reports preliminary results regarding effectiveness ofthe proposed system applied in a set of images. Overall, for themajority of the samples, the proposed system achieved a promisingscore and this can be a good alternative for studies that aim visuallyimpaired inclusion in dangerous environments.