How hemanthreddykunduru/obj_detection_live_pc_screen works
A small Windows desktop app captures the live screen, runs YOLO oriented-box detection, draws labeled annotations, and presents them in a transparent always-on-top Tkinter overlay. All repository behavior is in one file; the graph separates its runtime stages, but source does not establish further subsystem boundaries.
Main entry points: detect_live_screen.py Detection loop
Components 6
Other components
- Live desktop
- Detection loop
detect_live_screen.py - Screen capture
detect_live_screen.py - YOLO inference
detect_live_screen.py - Annotation renderer
detect_live_screen.py - Tk overlay
detect_live_screen.py
Connections 6
- Detection loop → Screen capture: captures frame
- Live desktop → Screen capture: provides pixels
- Detection loop → YOLO inference: runs inference
- Detection loop → Annotation renderer: draws detections
- Detection loop → Tk overlay: schedules updates
- Annotation renderer → Tk overlay: supplies image