Tag Detection Handler

The TagDetectionHandler (src/handlers/TagDetectionHandler.h & TagDetectionHandler.cpp) orchestrates all ArUco marker detection pipelines across active camera feeds.

---

1. Primary Responsibilities

1. Detector Lifecycle Management: Instantiates and initializes TagDetector worker threads for assigned cameras (eHeadMainCam, eRearCam).
2. Dual-Model Fusion: Configures detectors to run classical OpenCV ArUco decoding in parallel with LibTorch YOLO marker candidate detection.
3. Debug Overlay Streaming: Generates annotated image frames (RequestDetectionOverlayFrame()) containing marker bounding boxes, coordinate axes, IDs, and estimated distances for transmission to the Basestation GUI or WebRTC stream.
4. Synchronized Video Recording: Houses an internal RecordingHandler configured in RecordingType::eTagDetectionHandler mode to save annotated detection feeds to disk.

---

2. Managed Detectors

The handler provides access to detectors via the TagDetectors enumeration:

• TagDetectors::eHeadMainCam: Primary detector analyzing frames from the forward mast camera.
• TagDetectors::eRearCam: Secondary detector monitoring the rear camera feed when enabled.

---

3. Concurrency and Integration

• Each TagDetector runs as an independent AutonomyThread<void>.
• Frame pulling from CameraHandler is asynchronous.
• Detections are cached thread-safely in tagdetectutils::ArucoTag structs with creation timestamps, enabling TagDetectionChecker to evaluate lifetime persistence before triggering state machine transitions.

---

4. Usage Example

// Initialization in main.cpp
globals::g_pTagDetectionHandler = new TagDetectionHandler();
globals::g_pTagDetectionHandler->StartAllDetectors();
globals::g_pTagDetectionHandler->StartRecording();

// Querying detection overlay for UI streaming:
cv::Mat cvAnnotatedFrame = globals::g_pTagDetectionHandler->RequestDetectionOverlayFrame(
    TagDetectionHandler::TagDetectors::eHeadMainCam
);