How naveen2002-ncg/sourcehubit-object-detection-and-tracking works
This is a local real-time object detection and tracking application. Its principal workflow reads frames from a webcam, video, or image; runs YOLO detection; optionally associates detections across frames; and displays or saves annotated results. The main processing and tracking paths are excerpted. The README documents additional controls and demo capabilities; configuration is present as a standalone module, but the available import index does not show it wired into the runtime.
Main entry points: object_detection_tracking.py Frame processing · tracker.py Object tracker · demo.py Interactive demos · utils.py IoU calculation
Components 13
Input and runtime
- CLI application
object_detection_tracking.py - Frame processing
object_detection_tracking.py
Detection and tracking
- YOLO detection
object_detection_tracking.py - Object tracker
tracker.py - Track state
tracker.py - IoU calculation
utils.py
Visualization and output
- Drawing and metrics
utils.py - Display or saved media
object_detection_tracking.py
Configuration and demos
- Runtime settings
config.py - Interactive demos
demo.py
Other components
- Operator
- Webcam or media
- YOLO model
Connections 12
- Operator → CLI application: runs
README.md - Webcam or media → Frame processing: provides frames
object_detection_tracking.py - CLI application → Frame processing: processes source
object_detection_tracking.py - Frame processing → YOLO detection: requests detections
object_detection_tracking.py - YOLO detection → YOLO model: runs inference
object_detection_tracking.py - Frame processing → Object tracker: updates tracks
object_detection_tracking.py - Object tracker → Track state: updates
tracker.py - Object tracker → IoU calculation: calculates overlap
tracker.py - Frame processing → Drawing and metrics: draws results
object_detection_tracking.py - Frame processing → Display or saved media: displays or saves
object_detection_tracking.py - Operator → Interactive demos: starts demos
README.md - Interactive demos → CLI application: uses detector
demo.py