How mgagvani/apriltag-localization-frc works
Hardware Layer
• OAK Camera
• Jetson (Processing Unit)
• FRC Robot
• Network Switch
Software Components
Main Processing Pipeline:
• AprilTag Detection Module (detect_apriltag.py)
• SLAM Processing Module (slam_apriltag.py)
• Tag Localization Module (tag_localizer.py)
• Utilities (utils.py)
Data Flow Components
• Configuration Management (config_converter.py)
• Log Visualization (log_visualizer.py)
• Configuration Files (apriltag_configs/)
External Dependencies
• OpenCV
• DepthAI
• Spectacular AI SDK
• pupil_apriltags
• NetworkTables (FRC Communication)
• Input Layer |
• Processing Layer|
• Output Layer |
Input Layer should show
• OAK Camera
• Configuration Files
Processing Layer should show
• AprilTag Detection Pipeline
• SLAM Processing
• Calculation
Output Layer should show
• NetworkTables Communication
• Logging System
• Visualization Tools
Key Relationships to Highlight:
Data Flow
• Camera → Detection → SLAM → Position Calculation → NetworkTables
Configuration Management
• How apriltag_configs/ feeds into the main processing pipeline
Monitoring and Debugging
• How logging and visualization tools connect to the main pipeline
Special Considerations:
Real-time Processing
• Real-time nature of the system with appropriate symbols
• FPS constraints mentioned in the README (20-30 FPS)
Coordinate Systems
• A small coordinate system reference showing:
• Camera-relative coordinates
• Field-relative coordinates
• AprilTag positioning
Error Handling
• Fallback paths when AprilTags aren't detected
• Initialization requirements (AprilTag presence at startup)
• Blue: Core processing components
• Green: Input/sensor components
• Yellow: Configuration components
• Red: Critical paths/dependencies
• Gray: External services/dependencies
• Critical connection points
• Coordinate system references
Level of Detail:
• AprilTag detection pipeline
• SLAM processing details
• Network communication flow