An Android application developed as a graduation project at Al-Azhar University, aiming to support early awareness and detection of strabismus (crossed eyes) using image analysis and machine learning.
Watch the app in action: 📺 Demo Video
-
📖 Informational Articles
Access articles about strabismus, common eye diseases, and general eye care practices. -
👁️ Strabismus Detection (Informational Only)
Use a camera-based system integrated with a trained deep learning model (VGG16) to analyze eye images and estimate potential strabismus misalignment.⚠️ Note: This feature is not a medical diagnosis and should not replace professional healthcare advice. -
🏋️ Eye Care Exercises & Guidelines
Includes recommended exercises and practical tips for maintaining healthy eye health.
- Android (Kotlin, XML)
- TensorFlow & Deep Learning (Python)
- Room Database for offline content storage
- VGG16 Convolutional Neural Network for image classification
- Created a custom dataset of 1008 labeled eye images.
- Tested multiple models:
- RESNET50 → 76% accuracy
- Custom CNN → 88% accuracy
- VGG16 (Selected) → 92.57% accuracy
- Onboarding Screens
- Home Screen (Detection, Articles, Exercises)
- Article View
- Exercise View
- Detection Screen (Image selection, result display)
- Expand educational resources and add multimedia content.
- Integrate teleconsultation features with ophthalmologists.
- Develop detection models for other eye diseases.
- Support wearable tech integration.
- Improve UI/UX and accessibility.
- Add multi-language support and localization.
This application is intended for educational and informational purposes only and should not be used as a substitute for professional medical evaluation and treatment.
- Asmaa Tharwt Ahmed
- Asmaa Elboghdady Elwehady
- Aisha Mahmoud Fathy
- Fatma Alzhraa Nagah Said
- Fatma Khaled Tayel
- Heba Farid El Sherbiny
- Walaa Elsayed Mohamed
Supervisor: Dr. Marwa Selim
For a detailed explanation, system design, model training, and implementation process, see the 📄 Graduation Project Book (PDF)