Situation Monitor AI — Real-Time Object Detection Dashboard

A webcam dashboard (like the screenshot) built with OpenCV + YOLOv8, showing live bounding boxes, labels, FPS, confidence, and a "Detected: ..." status bar.

Setup

python -m venv venv
venv\Scripts\activate        # Windows
# source venv/bin/activate   # Mac/Linux

pip install -r requirements.txt

Run

python main.py

• First run downloads yolov8n.pt automatically (~6MB).
• Press q to quit the window.
• Edit CAMERA_INDEX in main.py if you have multiple webcams.
• Set TARGET_CLASSES = {"person", "cell phone"} in main.py to only show those two classes (default detects all 80 COCO classes).
• Edit CLASS_COLORS to change box colors per class.

Files

• main.py — full app (detection + dashboard UI)
• requirements.txt — dependencies

Notes

• Uses YOLOv8n (nano) for speed; swap MODEL_NAME = "yolov8s.pt" for higher accuracy at the cost of FPS.
• All UI elements (header, sidebar, footer, heatmap-style panel) are drawn manually with OpenCV so you can freely restyle colors/fonts/layout.