How frezzie-cats/capstone_mobile_app works
The tree suggests a Flutter mobile app with a camera screen, on-device TensorFlow Lite inference, result display, and a database helper. The README is only the default Flutter-project introduction, and no application source was excerpted, so the workflow and component relationships below are inferred from filenames; dashed edges mark those unverified links.
Main entry points: tflite_service.dart TFLite service · main.dart App entry · camera_screen.dart Camera screen · result_screen.dart Result screen
Components 8
App flow
- App entry
lib/main.dart - Camera screen
lib/screens/camera_screen.dart - Result screen
lib/screens/result_screen.dart
Image inference
- TFLite service
lib/services/tflite_service.dart - Binary model
assets/models/stage1_mobilenetv2_binary.tflite - Four-class model
assets/models/stage2_efficientnetb0_4class.tflite
Local data
- Database helper
lib/database/db_helper.dart
Other components
- User
Connections 7
- User → App entry: opens
- App entry → Camera screen: shows
- Camera screen → TFLite service: submits image
- TFLite service → Binary model: uses model
- TFLite service → Four-class model: uses model
- TFLite service → Result screen: returns prediction
- Result screen → Database helper: stores result