Project Description

Indoor Navigation System Using WiFi Localization For Visually Impaired People

Asmaa Obaid Altairsh
Computer Engineering Student
Fatima Mahmoud Said
Computer Engineering Student
Supervised By:
Dr. Mohammed Almeer
Dr. Wael Elmedany
 Assistant Professors
Abstract

The fundamental objective of this project is to design a system that can assist the visually impaired people while walking through an unknown indoor environment, as there is no proper assignment provided In addition, utilizing the existing infrastructure of the indoor environment with no need for additional tools The design comprises the following components raspberry pi 3 model B and headphone Localization is made using fingerprint technology along with K Nearest Neighbor (KNN) traditional machine learning algorithm that determines the user’s location Moreover, the user can determine his desired destination as the system will navigate him to it The headphone acts as the user interface

Objectives
  1. Design a system that can assist the visually impaired peoplewhile walking through an unknown indoor environment.
  2. Generalize the system to navigate the user from any locationin the covered area with no restriction for a starting point.
  3. Apply Artificial Intelligence in localization stage.
  4. Take advantage of the existing WiFi infrastructure of the indoor environment with no additional cost.
Methods/Diagrams/Figures
Results

A prototype of the proposed system has been implemented and tested, and found to have met all the project objectives The system was able to localize the user with an error of 0 to 2 meters The navigation process was operating successfully and smoothly via proper voice instruction corresponding to the position The system was able to recognize the user’s speech to determined the desired location.

Conclusion and Future Work
  1. Increase the robustness of the system in term of dealing with obstacles and alterations in the radio distribution of the place.
  2. Apply the dead reckoning technique, where the system willbe able to expect the user s current position from his previous location using accelerometer and gyroscope sensor.
  3. Apply the system to an autonomous vehicles to help a larger category of users
View Poster Online