Low-Cost Finger Gesture Recognition System for Disabled and Elderly People

Abstract

Gesture recognition is increasingly being used to design systems that assist the disabled people and elderly people, who struggle with motor control and motor functionality. In this paper, we propose a device which helps people with severe motor disabilities, however, have full functionality in their fingers even if they cannot move or lift their arms. The proposed device tracks the movement of the fingers of the users and classifies them as gestures. The classified gestures can be used to control various electrical devices such as switches, electrical appliances, wheelchairs, etc. and can also be used to signal the need for human assistance. The device is robust, easy to use, highly accurate, and low power consuming. Therefore, it can be used for an extended period of time and is reliable for helping the disabled people and elderly people.

Existing System

Smart home system for disabled people via wireless Bluetooth.

Proposed System

The proposed device uses an array of phototransistors (PT), as light sensors. We will track the movement of the four fingers of the users, all except thumb, for gestures. The user raising one or more fingers from rest position, i.e., all of the fingers aligned as in flat hand to one or more fingers being raised, is classified as a gesture. The output of the PT varies with the movement of fingers. By monitoring the changes in the response of an array of PTs, we can thus track the movements of fingers placed in front of the PTs. Our contributions in this paper as follows: We propose a low-cost and low-resource finger gestures device for disabled and elderly people. The device can detect a small set of simple gestures, accurately and robustly. Our proposed devices utilize COTS low-cost and limited resource components. We present the underlying working principle of the gesture recognition device and also the algorithm used to classify the finger gestures. We present our findings in testbed experiment and point out the scope for future improvements.

Architecture

Gesture Recognition System

Flow chart of the finger gesture recognition algorithm