Accéder directement au contenu Accéder directement à la navigation
Article dans une revue

An Electrooculography based Human Machine Interface for wheelchair control

Ajit Choudhari 1, * Prasanna Porwal 2 Venkatesh Jonnalagedda 1 Fabrice Meriaudeau 3 
* Auteur correspondant
3 Equipe IFTIM [ImViA - EA7535]
CHU Dijon - Centre Hospitalier Universitaire de Dijon - Hôpital François Mitterrand, UNICANCER/CRLCC-CGFL - Centre Régional de Lutte contre le cancer Georges-François Leclerc [Dijon], ImViA - Imagerie et Vision Artificielle [Dijon]
Abstract : This paper presents a novel single channel Electrooculography (EOG) based efficient Human–Machine Interface (HMI) for helping the individuals suffering from severe paralysis or motor degenerative diseases to regain mobility. In this study, we propose a robust system that generates control command using only one type of asynchronous eye activity (voluntary eye blink) to navigate the wheelchair without a need of graphical user interface. This work demonstrates a simple but robust and effective multi-level threshold strategy to generate control commands from multiple features associated with the single, double and triple voluntary eye blinks to control predefined actions (forward, right turn, left turn and stop). Experimental trials were carried out on the able-bodied and disabled subjects to validate the universal applicability of the algorithms. It achieved an average command detection and execution accuracy of 93.89% with information transfer rate (ITR) of 62.64 (bits/min) that shows the robust, sensitive and responsive features of the presented interface. In comparison with the established state of art similar HMI systems, our system achieved a better trade-off between higher accuracy and better ITR and while maintaining better performance in all qualitative and quantitative criteria. The results confirm that the proposed system offers a user-friendly, cost-effective and reliable alternative to the existing EOG-based HMI.
Type de document :
Article dans une revue
Liste complète des métadonnées
Contributeur : IMVIA - université de Bourgogne Connectez-vous pour contacter le contributeur
Soumis le : jeudi 12 décembre 2019 - 12:04:32
Dernière modification le : jeudi 4 août 2022 - 17:07:34




Ajit Choudhari, Prasanna Porwal, Venkatesh Jonnalagedda, Fabrice Meriaudeau. An Electrooculography based Human Machine Interface for wheelchair control. Biocybernetics and Biomedical Engineering, 2019, 39 (3), pp.673-685. ⟨10.1016/j.bbe.2019.04.002⟩. ⟨hal-02406980⟩



Consultations de la notice