Regresar

Interpretation of ecuadorian sign language using computer vision

Abstract:

This work proposes an efficient and easily accessible computer vision system for the interpretation of the Ecuadorian alphabet and sign language (LSEC) numbers that can be used regardless of the environment and type of user. It begins with the acquisition of the image in the form of a video through a webcam, this will be the entrance to the processing stage. It is intended to use the hand of the user as a region of interest (ROI) that will then be analyzed and processed to combat the effects of light and other anomalies that may arise, then use Artificial Neural Networks (ANN) for learning the alphabet and LSEC numbers by using patterns obtained from different hands forming the letters of the alphabet and necessary numbers, has been used C + + with libraries OpenCV and the IDE Qtcreator.