|
Acadêmico(a): Renato Martins Lorenzi |
Título: SOFTWARE DE DETECÇÃO DE ESTRABISMO PARA DISPOSITIVOS MÓVEIS |
|
Abstract: |
This paper presents techniques as well as software specification and implementation to detect strabismus for mobile devices, in particular for the Android platform. Strabismus is a condition that affects the parallelism between the two eyes that point in different directions; may arise in babies adults for different reasons (VARELLA, 2014). To make the detection of this disease, one of the methods used is to Hirshberg test, which in this work the implementation for the detection of strabismus. To achieve this goal, image processing techniques are used. As main techniques were used the Haar feature-based cascade classifier and Hough transform. The Haar feature-based cascade classifier was used to detect the face region and eyes and Hough transform aimed to detect the region of the limbus and brightness of the eyes. The results obtained by the proposed methodology obtained 85.71% of accuracy in the analyzed patients. |
|
|