According to IBGE (Brazilian Institute of Geo- graphy and Statistics), severe vision impairment is one of the most frequent physical deficiency in the Brazilian population. Thus, the development of technology aiming to facilitate the insertion of people with different degrees of vision deficiency into the society is very important and of great social impact. This paper presents AndroidEye, a computer vision system proposed for Android mobile devices that can be operated by voice commands and is able to recognize face from video captured by the device's embedded camera. The identity of the recognized person is informed to the system user through headphones. The proposed and implemented system was assessed on MoBio, a video face database. Experimental results showed that for mobile devices with reduced memory and processing power skin-color based method is more suitable for face detection than the Viola-Jones method, while LBP based method is more appropriated for face recognition than PCA and ASM methods.
Unicluster: Uma Proposta de ICP para PAD
Implementação de Algoritmos Genéticos Paralelos Aplicados à Fármacos
Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web:
Buscar na Web