Project: VisionAid AI – Smart Navigation for the Visually Impaired

Problem

Millions of visually impaired people face difficulties navigating unfamiliar places safely due to obstacles, traffic, and poor accessibility.

Solution

VisionAid AI is a mobile app that uses AI and the phone's camera to detect obstacles, read text aloud, recognize objects, and guide users with voice instructions in real time.

Key Features

🎥 Real-time obstacle detection using the camera
🗣️ Voice navigation with spoken directions
📝 OCR to read signs, menus, and documents aloud
🚦 Traffic light and crosswalk detection
🆘 SOS button that shares the user's live location with emergency contacts
🌐 Offline mode for basic navigation and object detection

Tech Stack

Frontend: Flutter
Backend: Firebase
AI: Google Gemini API (for scene understanding), TensorFlow Lite or YOLO (for object detection)
Maps: Google Maps API
OCR: Google ML Kit

Target Users

Visually impaired individuals
Elderly people
Accessibility organizations

Impact

Improves independence and safety
Makes navigation easier in unfamiliar places
Supports inclusive technology and accessibility

Future Enhancements

Smart glasses integration
Indoor navigation using Bluetooth beacons
Public transport assistance
Multilingual voice support

Demo Flow

User opens the app.
The camera scans the surroundings.
AI detects obstacles and speaks: "Obstacle ahead. Move slightly left."
The app reads nearby signs aloud.
If the user feels unsafe, they tap SOS, and their live location is shared with emergency contacts.