Rwanda Eye

Rwanda Eye is a bilingual (English + Kinyarwanda) AI vision desktop app built with Python and OpenCV.

It provides 3 real-time modules:

• Plant Doctor: detects plant disease and gives treatment guidance.
• Road Inspector: detects potholes, estimates severity, and simulates district reporting.
• Accessibility Oracle: estimates emotion/age/gender and supports voice guidance mode.

The app also includes:

• Voice commands (English and Kinyarwanda)
• Spoken feedback
• Synthetic alert sounds
• Session report overlay

How It Works

main.py starts the app and shows the menu UI. From there, users can open modules by keyboard or voice.

Core flow:

1. App starts VoiceEngine, SoundEngine, HUD, and report generator.
2. A background voice thread listens for commands like plant, road, face, report, menu, quit.
3. Selected module opens camera/video input and runs live analysis.
4. Stats are tracked in session_stats.
5. Report view summarizes total scans, detections, reports, and impact.

Detection behavior

• If model files exist in models/, YOLO/DeepFace are used.
• If model files are missing, the app falls back to demo/simulated mode so you can still run and test the UI.

Expected optional model files:

• models/plant_disease.pt
• models/pothole.pt

Project Structure

• main.py: app entry point and module routing
• modules/plant_doctor.py: plant disease module
• modules/road_inspector.py: road damage module
• modules/accessibility.py: accessibility analysis module
• voice_engine.py: speech output + voice command listening
• sound_engine.py: generated sound effects
• hud.py: shared visual overlays
• report_generator.py: end-of-session report screen
• translations/kinyarwanda.py: translation map

Requirements

• Python 3.10+ (project currently has a Python 3.12 venv)
• Webcam (for Plant/Accessibility modules)
• Microphone (for voice commands)
• Speaker/audio output (for voice + alerts)

Python dependencies are listed in requirements.txt.

Setup and Run

1. Create and activate a virtual environment

python3 -m venv .venv
source .venv/bin/activate

2. Install dependencies

pip install -r requirements.txt

3. Run the app

python main.py

Controls

Keyboard

• 1: Plant Doctor
• 2: Road Inspector
• 3: Accessibility Oracle
• r: Show session report
• m: Return to menu (inside modules)
• q or Esc: Quit/exit current module
• b: Toggle blind mode (Accessibility module)

Voice commands

English: plant, road, face, report, help, menu, stop, quit

Kinyarwanda keywords supported: tangira, hagarara, raporo, nfasha, inzira, ubuso, soma

Notes

• First run may be slower if heavy ML packages are installed.
• gTTS may need internet for Kinyarwanda voice synthesis.
• If camera/mic/audio is unavailable, the app still starts but related features may degrade gracefully.

WhatsApp.Video.2026-04-15.at.20.06.20.mp4