Smart Glove

Human beings have natural ability to see, listen and interact with its surrounding. Unfortunately, there are some people in the society which do not have ability to use their senses to the best extent possible. As communication is a fundamental aspect of human life, such people depend on the other m

Project Title

Smart Glove

Project Area of Specialization

Wearables and Implantable

Project Summary

Human beings have natural ability to see, listen and interact with its surrounding. Unfortunately, there are some people in the society which do not have ability to use their senses to the best extent possible. As communication is a fundamental aspect of human life, such people depend on the other means of communication like sign language (a non-verbal form of intercourse). To bridge this communication gap, it is important to have an advance sign-language detection and gesture recognition system for people in the community when they try to engage in interaction with normal public that do not understand sign-language. It is a basic necessity for every human being to share their feelings and emotions without facing any restrictions. Therefore, an effort has been made to develop a smart glove using different hardware modules and software tools for real-time gesture recognition. It is known that every person hand has unique shape and size, we aimed to design a device that could provide reliable translations regardless of those differences. The system will be capable of recognizing 25 real-time  hand gesture and its translation into speech (audible sound) and visual text. To make this communication possible several components are integrated on the glove like sensors, microcontroller and other hardware components for real time transmission of  hand gestures to Machine Learning models in python script for its translation. For system efficiency, a huge amount of data is collected and trained over different machine learning models which further experimentally examined and compared in order to select the best model among all based on the accuracy rates.

Project Objectives

The main objective of this project  is to create an easy wearable device which help speech-impaired people to convey their ideas and messages to normal public without remaining restricted in a small social circle that understands sign language. To bridge this communication gap so that this community can share their feelings and emotions withoutfacing any restrictions when try to interact with external environment. We aimed to design a device “Smart Glove” that could provide reliable translations regardless of  differences in different hand sizes and shapes using different hardware modules and software tools for real-time gesture translation into audible sound and visual text.

Project Implementation Method

The system design of 'Smart Glove' is categorized into to two major portions i.e. hardware and software implementation:

1. Hardware Implementation:

It includes all necessary components required in building data glove such as : flex sensors, mpu 6050, arduino board and Bluetooth module. MPU 6050 and flex sensors are used to detect the hand position in space and the value by which the hand fingers are bend in order to recognize hand gesture of the user. Arduino board is attached to integrate these numeric values and send it to the PC using Bluetooth module.

       2. Software Implementation

On receiving analog data from Bluetooth, it is applied to the trained Machine learning models for its translation into speech and text in python script. Software Tools used in this process for dataset collection and translation of real-time gesture are: Tera Term, Anaconda, spyder and Arduino IDE.

 ‘Smart Glove’ is designed to recognize 25 real time ASL (American Sign Language) hand gestures. To train our machine learning models, dataset for these 25 gestures is collected manually up to 13000 entries with different hand positions in space.

Benefits of the Project

Some of the benefits for our system are as follows:

1. Bridge the communication gap between normal and speech-impaired community.
2. Provide free of speech for disabled community.
3. Limit the restrictions for speech-impaired people for expressing themselves.
4. Capable of translating 22/23 real-time hand gestures correctly with the accuracy of 85/90 %.
5.  Can be applied to many applications such as: controlling of home appliances, movement of wheel-chair, gaming in virtual reality just by the movement of hands.

Technical Details of Final Deliverable

Following are the major technical details for final deliverable :

1. At first the system is implemented via wired connections that simply requires serial communication between the glove and the PC. But the problem arises, when the person is not in contact with PC through wire or at some distant places. Another module is added to the system i.e. Bluetooth is introduced to make the system  wireless to reduce the limitation of our system.
2. When the analog data from glove for its translation is transmitted to the PC, all the values from each finger and hand position should be synchronized for the correct recognition. To resolve this problem, we have added a delay of 1 or 2 seconds for reading data input that makes the gesture translation after 3 or 4 seconds after its complete processing.

The final deliverance of our system includes: Smart Glove is capable of translating 22/23 real-time hand gestures correctly wirelessly with a processing time of 2-3 seconds minimum.

Final Deliverable of the Project

HW/SW integrated system

Type of Industry

Education , Others

Technologies

Wearables and Implantables

Sustainable Development Goals

Reduced Inequality

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com