Computer-vision algorithms to steer powered wheelchairs

Malik Haddad, David Sanders, Giles Tewkesbury, Martin Clifford Langner, Sanar Fakhraddin Muhyaddin, Mohammed Ibrahim

Research output: Contribution to journalArticlepeer-review

Abstract

A novel approach to steer powered wheelchairs using Computer-Vision algorithms, a camera module and a Raspberry Pi is presented. When a powered wheelchair user made a movement, the new approach detects that movement and uses it to operate a powered wheelchair. A circuit is created connecting a camera module, relay and a Raspberry Pi. Python programming language is used to create a program that detects movement of a powered wheelchair user. Three different Computer-Vision algorithms are considered: Background-subtraction, Python-Imaging-Library and Open-source-Computer-Vision (OpenCV). The camera was directed towards user a body-part used for controlling a wheelchair. The new approach will detect and analyse movement of the body-part and activate a wheelchair motor accordingly. Two User Interfaces were created: a simple User Interface to operate the approach and a technical interface to control the sensitivity, frame rate, size of image captured and detection area, and duration of wheelchair motor activation. Practical testing showed that OpenCV provided the highest accuracy and sensitivity among the three algorithms. The new approach successfully detected users’ voluntary movements and translated those movements into driving-commands used to drive a wheelchair. Future clinical-trials will be conducted at Chailey Heritage Foundation.
Original languageEnglish
JournalJournal of Physics: Conference Series
Publication statusAccepted for publication - 1 Nov 2021
Event2nd International Symposium on Automation, Information and Computing 2021 - China, Beijing , China
Duration: 3 Dec 20216 Dec 2021
https://www.isaic-conf.com/#/

Keywords

  • UKRI
  • EPSRC
  • EP/S005927/1

Fingerprint

Dive into the research topics of 'Computer-vision algorithms to steer powered wheelchairs'. Together they form a unique fingerprint.

Cite this