Abstract / Excerpt:

This research aims to investigate the approach of applying the hidden Markov model in recognizing American Sign Language for the Xbox 360 Kinect device. Kinect's patent clearly states its capability to recognize American Sign Language and convert it to its spoken or test equivalent but has been excluded in the final release. OpenCV, a library of functions used for computer vision has been popular for manipulating digital images and videos thus essential for recognizing sign language due to its powerful libraries and functions focused on Computer Vision.

The research focuses on using OpenCV libraries to work with the Kinect in recognizing American Sign Language for two ways. One is for recording signing data to its text equivalent. Second is for interpreting signing to its text equivalent assuming the first has taken place. In this paper, the researchers would like to present a method for detecting and analyzing sign language through hand gesture recognition using Microsoft Kinect through contour detection and pattern matching.

Info

Source Institution	Ateneo de Davao University
Unit	Computer Studies
Authors	Lex Marion Bataller, Paolo Villanueva, Jerrico Sevilla
Page Count	11
Place of Publication	Davao City
Original Publication Date	March 1, 2012
Tags	American Sign Language, Microsoft Kinect

Preview