Home / Current Issue / Paper 1717896
Eye Controlled Mouse Cursor for Physically Disabled Individual using CNN
Subject area: Science,Engineering and Technology · Area of research: CNN
DOI: https://doi.org/10.64388/IREV9I11-1717896
Abstract
This paper presents an eye-controlled mouse cursor system designed to assist physically disabled individuals in interacting with computers using natural facial gestures. The proposed system utilizes a standard webcam to capture real-time video input and employs Convolutional Neural Networks (CNNs) for accurate detection and classification of facial gestures. Eye movements are tracked to control cursor position, while specific gestures—including deliberate single eye blink, mouth opening, and both eyes closed—are mapped to left-click, right-click, and click-and-drag actions respectively. The system integrates multiple modules: face detection using pre-trained models, eye and mouth region extraction through facial landmarks, CNN-based state classification for eyes and mouth, and gaze tracking for cursor movement. Temporal filters and confirmation mechanisms are incorporated to distinguish intentional gestures from involuntary actions such as natural blinking or speaking, thereby minimizing false triggers. The system operates in real- time under varying lighting conditions and head orientations without requiring specialized hardware or physical attachments. Experimental results demonstrate high accuracy in gesture recognition and smooth cursor control, offering an affordable, non-invasive, and intuitive assistive technology solution. This work contributes to enhancing digital accessibility and independence for individuals with motor impairments, promoting greater inclusion in education, employment, and daily computer-based activities.
Keywords
Eye-Controlled Mouse, Assistive Technology, Convolutional Neural Networks, Facial Gesture Recognition, Human-Computer Interaction, Computer Vision, Physically Disabled Individuals
How to cite this paper
@article{1717896,
author = {M. Rajamani, M Balamurugan, R Surya},
title = {Eye Controlled Mouse Cursor for Physically Disabled Individual using CNN},
journal = {Iconic Research And Engineering Journals},
year = {2026},
volume = {9},
number = {11},
pages = {2322-2328},
issn = {2456-8880},
url = {https://www.irejournals.com/formatedpaper/1717896.pdf},
abstract = {This paper presents an eye-controlled mouse cursor system designed to assist physically disabled individuals in interacting with computers using natural facial gestures. The proposed system utilizes a standard webcam to capture real-time video input and employs Convolutional Neural Networks (CNNs) for accurate detection and classification of facial gestures. Eye movements are tracked to control cursor position, while specific gestures—including deliberate single eye blink, mouth opening, and both eyes closed—are mapped to left-click, right-click, and click-and-drag actions respectively. The system integrates multiple modules: face detection using pre-trained models, eye and mouth region extraction through facial landmarks, CNN-based state classification for eyes and mouth, and gaze tracking for cursor movement. Temporal filters and confirmation mechanisms are incorporated to distinguish intentional gestures from involuntary actions such as natural blinking or speaking, thereby minimizing false triggers. The system operates in real- time under varying lighting conditions and head orientations without requiring specialized hardware or physical attachments. Experimental results demonstrate high accuracy in gesture recognition and smooth cursor control, offering an affordable, non-invasive, and intuitive assistive technology solution. This work contributes to enhancing digital accessibility and independence for individuals with motor impairments, promoting greater inclusion in education, employment, and daily computer-based activities.},
keywords = {Eye-Controlled Mouse, Assistive Technology, Convolutional Neural Networks, Facial Gesture Recognition, Human-Computer Interaction, Computer Vision, Physically Disabled Individuals},
month = {May},
doi = {https://doi.org/10.64388/IREV9I11-1717896}
}