Skip to content

Repository files navigation

SenseNet - Unwrapping the Future (Phase II Report)

Project Overview

"SenseNet - Unwrapping the Future" is an engineering project focused on developing a multi-faceted system to assist older individuals and those with visual impairments or postural issues. This Phase II report details the project's progress and objectives, aiming to improve health management for the elderly, enhance healthcare outcomes, reduce healthcare burdens, and promote independence.

Project Phases

The SenseNet project is structured into three main phases:

  1. Phase I (Posture Correction for all ages): Addresses common postural deformities to maintain a healthy lifestyle.
  2. Phase II (SenseNet for Older Generation): Prevents Kyphosis and acts as a "secondary eye" to assist with vision and movement, specifically targeting presbyopia and cataracts.
  3. Phase III (SenseNet for Visually Impaired): Aims for high accuracy in vision and movement assistance, integrating smart glasses and a spine device to create a sensory network.

System Design Architecture

The system design for SenseNet is comprised of two primary modules:

Eye Module

This module integrates several advanced functionalities:

  • Facial Identification: To recognize individuals.
  • Object Detection: To identify objects in the environment.
  • Path Detection: To assist with navigation.
  • Google Map API Integration: For location-based services and guidance.

Posture Detection Module

This module focuses on corrective measures for posture:

  • Sensor-based Alerts and Suggestions: Utilizes IoT-based sensors to detect poor posture and provide real-time feedback.

Working Principle

SenseNet leverages deep learning for vision tasks and IoT-based sensors for posture correction:

  • Deep Learning: Employed for facial recognition, object detection, and path detection.
  • PoseNet: Utilized for accurate posture tracking.

Hardware and Software Requirements

Hardware

  • Flex sensors
  • Arduino UNO
  • Buzzers
  • LEDs

Software

  • Python
  • OpenCV
  • TensorFlow
  • Flask
  • Siamese network
  • YOLO algorithm

Results and Discussion

The report details the software implementation using technologies like Python, OpenCV, TensorFlow, Flask, Siamese network, and the YOLO algorithm. Hardware implementation is discussed across three phases: a flex sensor approach, a buzzer-based approach, and an LED-based approach, including code snippets and output examples.

Individual Contributions

The project involved a team with diverse roles:

  • AI Developer
  • Hardware Integration Specialist
  • IoT Engineer (e.g., Member 8: Amit Rathore - worked with microcontroller code, hardware assembly, flex sensors, LEDs, and research analysis)
  • NLP Specialist
  • Computer Vision Engineer

Authors

  • Avrodeep Saha (20BAI10041)
  • Chahak Garg (20BAI10280)
  • Varun Ram S (20BAC10038)
  • Jigar Sharma (20BAC10025)
  • Rachit Goyal (20BCG10066)
  • Om Mani Tripathi (20BCG10076)
  • Prerna Singhal (20BCG10087)
  • Amit Rathore (20BCE11115)

University

  • VIT Bhopal University

Report Submission Date

  • May 2023

About

This was a part of the Community project which I worked along with 8 other peers from different branches in VIT Bhopal. The aims to assist the handicapped and needy in posture correction, object detection and path assistance.

Topics

Resources

Stars

0 stars

Watchers

1 watching

Forks

Releases

Packages

Contributors

Languages