Hand Gesture Recognition Based on Image Processing for Supporting Deaf People

EasyChair Preprint 3040

Abstract

One of the important problems our society is facing is that deaf and dumb people find difficulties in communicating with normal people who don’t understand sign language. Even though sign language is the best way for deaf and dumb people to communicate with each other. Sign language is also used a little by normal people because it is an expressive and natural way for communicating. The motivation for our study is the improvement of accessibility to public information announcements for deaf and less hearing people. The whole idea is to build a service robot that enables communications between speech-hearing impaired individual and a normal person. In this paper, we will present two tasks, the first is how to translate sign language to speech, the second is focusing in the image processing algorithm to achieve hand gestures using depth sensor and then convert to speech. Several sign language visualization methods were evaluated, in order to perform this study a machine gesture translation system. It was concluded that system is suitable service-delivery platforms for sign language machine translation systems.

Keyphrases: deaf people, image processing, service robot, sign language

@booklet{EasyChair:3040,
  author    = {Phan Gia Luan and Tuong Phuoc Tho and Nguyen Truong Thinh},
  title     = {Hand Gesture Recognition Based on Image Processing for Supporting Deaf People},
  howpublished = {EasyChair Preprint 3040},
  year      = {EasyChair, 2020}}