Project Title: Third Eye – Aid For Blind.
Introduction:
This project aims to develop a complete portable aid(Raspberry Pi) for blind pedestrians and deal with problems in existing systems efficiently. The system designed will detect an object or obstacle
using ultrasonic sensors and gives audio instructions for guidance. It also gives information
about people looking or waving to them using face recognition.
Assumption and Constraints
- Maximum range of an ultrasonic sensor is 4m.
- Model won’t be able to work accurately in the crowded area.
- GPS is a satellite-based radio navigation system, so it can track predefined locations only.
- Raspberry pi camera module has 5-megapixel resolution and it is capable of only taking video and pictures not sound.
- Object detection and face detection, recognition has limitation over accuracy.
Conceptual framework:
Features Provided:
- It can detect objects with the help of Ultrasonic sensor.
- It can detect the name of the object with the help of YOLO algorithm.
- It can detect the face with the help Raspberry pi camera use with face detection and recognition process.
- It can tell the location of the person with the help of gps used.
Diagrams: Relevant diagrams like ER-diagram, UML diagrams, etc
Output of Face Detection:
Output of YOLO algorithm:
Use Case diagram:
System Architecture:
Activity Diagram:
Algorithms
- YOLO (You Only Look Once) algorithm use to detect object in background in real time.
- Face recognition using OpenCV and haar-cascades classifiers.
- Tensorflow used.
Tools Used:
- Used Raspberry pi to make it portable.
- Use train model to detect objects.
- Python Ide.
- Linux terminal.
Application:
The World Health Organization (WHO) reported that there are 285 million visually-impaired people worldwide.
Among these individuals, there are 39 million who are totally blind. The product developed is light in weight, compact hence does not cause fatigue to the user. So this helps the blind person to be self-dependent till particular extent by giving info of surrounding objects, person, and help in keeping the track of current location.
Team :
Chandan Vanwari
Charvi Jain
Akhil Bansal
Note: This project idea is contributed for ProGeek Cup 2.0- A project competition by GeeksforGeeks.