MorseVision

Eye Blink Based Morse Code Communication System

---

Project Overview

MorseVison is a real-time assistive communication system that allows users to generate text using eye blinks, which are interpreted as Morse code signals.

The system:

• Detects eye blinks using Computer Vision
• Converts blink duration into Morse code (dot/dash)
• Translates Morse into readable text
• Speaks the output using Text-to-Speech

It is designed primarily for:

• Physically challenged individuals
• Paralyzed patients
• ALS patients
• Hands-free communication use cases

---

Problem Statement

Many individuals with motor disabilities cannot communicate verbally or through typing. Traditional assistive technologies are expensive and hardware-dependent.

Problem:

How can we create a low-cost software-based communication system using just a webcam?

Solution:

Use eye blink detection + Morse code interpretation to create a communication bridge.

---

Technologies Used

Technology	Purpose
Python	Core programming language
OpenCV	Real-time video capture & processing
Haar Cascade / Eye Detection	Eye localization
Custom Blink Detector	Blink duration classification
Morse Interpreter	Dot/dash decoding
pyttsx3 / TTS Engine	Voice output

---

Project Architecture

Camera Input
     ↓
Eye Detection (OpenCV)
     ↓
Blink Detection Logic
     ↓
Blink Duration Analysis
     ↓
Morse Code Generator
     ↓
Text Interpreter
     ↓
Text-to-Speech Output

---

Project Folder Structure

MorseVison/
│
├── main.py
├── requirements.txt
│
├── src/
│   ├── eye_detector.py
│   ├── blink_detector.py
│   ├── morse_interpreter.py
│   ├── tts.py
│
└── assets/

---

1. Eye Detector (eye_detector.py)

Purpose: Detect left and right eye from video frame.

Process:

• Capture frame
• Convert to grayscale
• Use Haar Cascade classifier
• Return detected eye coordinates

Output:

frame, left_eye, right_eye

---

2. Blink Detector (blink_detector.py)

Purpose: Determine whether eye is open or closed.

Logic:

• Measure eye aspect ratio (EAR) or pixel threshold

• If eye closed for:

  • Short duration → DOT (.)
  • Long duration → DASH (-)

Key Variables:

• blink_start_time
• blink_duration
• DOT_THRESHOLD
• DASH_THRESHOLD

---

3. Morse Interpreter (morse_interpreter.py)

Purpose: Convert dot/dash sequence into alphabet.

Example:

.-   → A
-... → B

Internal Dictionary:

MORSE_CODE_DICT = {
    ".-": "A",
    "-...": "B",
}

---

4. Text-to-Speech (tts.py)

Purpose: Convert decoded text to voice.

Uses:

pyttsx3.init()
engine.say(text)
engine.runAndWait()

---

5. Main Controller (main.py)

Controls:

• Camera loop
• Module coordination
• Frame processing
• Exit handling

---

How It Works (Step-by-Step)

1. Webcam starts capturing

2. Eyes are detected in frame

3. Blink duration is calculated

4. Blink classified as:

   • Dot
   • Dash

5. Morse sequence stored

6. When pause detected → decode character

7. Text displayed

8. TTS speaks output

---

System Requirements

• Python 3.8+
• Webcam
• Windows/Linux/macOS
• OpenCV installed

---

Installation Guide

1. Clone the Project

git clone https://github.com/yourusername/MorseVison.git

cd MorseVison

2. Create Virtual Environment

   python -m venv venv
   

   Activate it:

   • Linux / macOS

     source venv/bin/activate
     

   • Windows

     venv\Scripts\activate
     

3. Run it using python

   python -m src.camera
   

---

Performance & Limitations

Works well when:

• Proper lighting
• Face clearly visible
• Camera stable

Limitations:

• Sensitive to lighting
• No ML-based blink classification
• No GUI yet
• No calibration mode

---

Future Possible Improvements

Upcoming features

---

1. AI-Based Blink Classification

Instead of threshold-based detection:

• Train CNN model
• Classify open/closed eyes
• Improve accuracy

---

2. Real-Time Blink Timing Visualization

• Graph blink duration
• Show dot/dash visually
• Add debugging panel

---

3. Calibration Mode

Allow user to:

• Set custom blink thresholds
• Auto-detect natural blink timing

---

4. Word Prediction System

Add:

• Basic NLP word suggestions
• Autocomplete
• Faster sentence building

---

5. Android Version

• Convert to mobile app
• Use CameraX
• On-device ML

---

6. Emergency Mode

• Triple long blink → Send SOS
• Connect to API for SMS/Email alert

---

7. Cloud Integration

• Store communication logs
• Remote monitoring dashboard

---

8. GUI Interface

• PyQt or Tkinter interface
• Sentence builder UI
• History panel

---

9. Noise Filtering

• Ignore involuntary blinks
• Use time-based buffer logic

---

Links

• Source Code

---