Visión artificial para el reconocimiento del lenguaje de señas.

The present research work aims to perform the implementation of a tool that facilitates the learning of sign language in real time, through the use of computer vision techniques and digital image processing. The methodology of programming employed is the methodology RAD (Rapid Application Development), due to the system was planned for a limited development time, and it was required the user interaction with the developers. For hand recognition algorithm was performed to pursue segmented objects through color, in the same way for prediction where the algorithm of classification are used the KNN (K-Nearest Neighbors) and SVM (Support Vector Machine) because they are frequently used in the development of this kind of systems, these two algorithms were combined to reduce the error rate and for getting a higher percentage of correct answers during the prediction. In the results the processes used to develop the system according to the stablished requirements initially by members of the Association of Deaf people of Loja named “Virgen del Cisne" are detailed. The importance of this research is in the employing of free software tools for the development of the project, so the system don´t require hardware or software components with powerful features, making it suitable for an efficient distribution.

• Auteur principal: Sivisapa Aguilera, Luis Germán (author)
• Format: bachelorThesis
• Langue: spa
• Publié: 2014
• Sujets: <VISIÓN ARTIFICIAL> <HERRAMIENTAS OPENCV> <INTELIGENCIA ARTIFICIAL> <ELEMENTOS PARA EMPLEO DE LA VISIÓN ARTIFICIAL> <LENGUAJES DE SEÑAS>
• Accès en ligne: http://dspace.unl.edu.ec/jspui/handle/123456789/14233