Download PDFOpen PDF in browser

Controlling External Devices by Using Speech and Simple Eye Movements

11 pagesPublished: May 14, 2020

Abstract

In Vietnam, the number of people with limb defects or diseases that affect the ability of limbs is increasing [1]. They were reduced or incapable of living, unable to walk, to run, and utilities around them became useless. Therefore, it is imperative to build a patient support system in controlling external devices. Recognizing the problem, the author has studied, researched, and built a wheelchair control system based on speech recognition and eye movement. In this study, the eye movement system is at the simple level: glance left and right. The system uses an image processing method to locate the pupil. The external device for testing is a car model that control by Arduino connected via Bluetooth. Experimental evaluation results show that the eye movement system achieves high accuracy of over 70%. However, the eye movement system by image processing takes a lot of time to process for a command. All of the above shows the feasibility of developing an external control device for people with disabilities.

Keyphrases: arduino, control, eye movement, limb defect, speech recognition

In: Tich Thien Truong, Trung Nghia Tran, Quoc Khai Le and Thanh Nha Nguyen (editors). Proceedings of International Symposium on Applied Science 2019, vol 3, pages 231-241.

BibTeX entry
@inproceedings{ISAS2019:Controlling_External_Devices_Using,
  author    = {Duy Long Dao and Thanh An Le and Anh Tai Tran and Minh Huong Nguyen Thi},
  title     = {Controlling External Devices by Using Speech and Simple Eye Movements},
  booktitle = {Proceedings of International Symposium on Applied Science 2019},
  editor    = {Tich Thien Truong and Trung Nghia Tran and Quoc Khai Le and Thanh Nha Nguyen},
  series    = {Kalpa Publications in Engineering},
  volume    = {3},
  publisher = {EasyChair},
  bibsource = {EasyChair, https://easychair.org},
  issn      = {2515-1770},
  url       = {/publications/paper/pkBP},
  doi       = {10.29007/jk28},
  pages     = {231-241},
  year      = {2020}}
Download PDFOpen PDF in browser