Sign language translator

Thumbnail Image
Heng, Chin Yang
Journal Title
Journal ISSN
Volume Title
Research Projects
Organizational Units
Journal Issue
Conflict occurs when there is miscommunication between two parties where both the parties are not able to interact efficiently. This could happen between disabilities and normal people as most people are not familiar with sign language. As time goes by, the gap between them will become wider resulting in more problems including disrespect, misunderstanding and poor teamwork. In this project, a sign language translation system is proposed to ease the communication problem among people. A sign language can be translated through image processing, data glove or a sign language interpreter. Among all the approaches, image processing is chosen by making use of depth images to be translated into English text. For this case, Kinect, a depth sensor, is required to provide the depth data for UP board, a single board computer, to be processed. Dynamic gestures are trained to create database for American Sign Language based on one person. An executable application is written in C# programming language by using Microsoft Visual Studio Community 2017 to process the sign language translation. The proposed system improves the portability of sign language translator with the usage of single board computer. Furthermore, it can translate sign language in real-time and provide high accuracy at the same time. All the results are discussed based on a real prototype.
HV2474.H46 2018
Sign language , Translator