| |
| Acadêmico(a): Marco Aurélio Caramori |
| Título: Protótipo de Software para Leitura Labial |
| |
| Abstract: |
Lip Reading is a technique to extract information from lips, tongue or face in order to provide extra information during speech and is widely applied by hearing impaired humans. Making this task computably possible requires basically three steps: acquiring the images from speech, extracting the features and finally classifying new samples. This work presents a prototype of lip reading software using the ratio of change from three lip measures over time. The images are captures through a video camera attached to a computer. The features are extracted using processing image techniques, while the classification is attained through a simple Bayes filter algorithm. The prototype was tested with a data set collected from a single individual, resulting in a success rate of eighty per cent.
|
| |
|
