How kongquan98/silearnplatform works
Main Components
• Mobile Client (Android Application)
• Business Logic & Controllers: Java classes that handle navigation, state management, and the calling of recognition or learning modules.
• Sign Language Recognition Engine
• Camera Integration: Capture via CameraActivity (and its “Back” variant) feeding frames for processing.
• Image Processing & Feature Detection: Integration with OpenCV library (provided as a separate module, “openCVLibrary3413”).
• External Services & Infrastructure
• Firebase: Authentication and image storage services for user management and data persistence. Indicate the Firebase Auth and Storage modules with arrowed interactions with the mobile client.
• (Optional) Python Script for Model Training: Even if not part of the deployed runtime, mention that models were generated externally using Python.
• Data Assets & Resources
• Assets: Show that various supporting files like .txt files, multimedia files (images, videos), and model files are provided in the assets folder.
• Native Libraries: JNI libraries under “jnilibs” that could be tied to performance-critical methods in recognition processing.
Relationships & Data Flows
• From the UI components (e.g., CameraActivity) to the Sign Recognition Engine.
• Within the recognition module, show a flow: Camera input → OpenCV processing → TFLite Inference → Output recognition result.
• Firebase’s role as a backend service for user authentication (login, registration) and for storing/retrieving images or associated data.
Architectural Patterns & Technologies
• Library Integration: Since the OpenCV integration is done through a dedicated module, highlight it as an imported library. Similarly, show the TensorFlow Lite engine as a key part of the image recognition processing.