VisionAssist-Pi

This project integrates face recognition, ultrasonic distance measurement, and voice feedback to assist partially impaired individuals. Built using Python and Raspberry Pi, it uses a camera for real-time face recognition and an ultrasonic sensor to detect nearby obstacles, providing audio feedback for guidance.

Features

• Face Recognition: Identifies pre-trained individuals in real time.
• Obstacle Detection: Measures distance using an ultrasonic sensor and provides audio alerts for nearby obstacles.
• Voice Feedback: Uses espeak to provide spoken feedback in English.

---

Prerequisites

Ensure your Raspberry Pi is set up with the following dependencies:

• Python libraries:
  • face_recognition
  • picamera2
  • espeak
  • opencv-python
  • imutils
  • numpy
• Hardware components:
  • Raspberry Pi
  • Camera module
  • Ultrasonic sensor (HC-SR04)
  • Speaker for audio output

---

Usage

1. Capturing Photos for Face Recognition

To capture photos of individuals for face recognition:

• Run the following command: python image_capture.py
• Follow the on-screen instructions:
  • press SPACE to capture a photo.
  • Press Q to quit the photo capture session.
• Captured photos will be stored in the dataset/ folder under a subfolder named after the person.

2. Training the Face Recognition Model

Once photos have been captured:

• Run the training module to process the images and generate encodings: python model_training.py
• This will create a file named encodings.pickle, storing the facial encodings for recognition.
• Running the Smart Assistance System

3. To start the main system:

• Run the following command: python facial_recognition.py
• The system will:
  • Identify individuals using the trained face recognition model.
  • Detect nearby obstacles using the ultrasonic sensor.
  • Provide real-time audio feedback for identified faces and obstacle distance.
• Press Q to quit the system.

---

Screenshots

Real-time Image Capture

Model Training

Face Detection