GROUP MEMBERS:

• Mercy Yankey
• Osen Mac-iriase
• Andre Ebu
• Saidamir Osimov
• Aayush Rajak

Technologies used

• Python 3.12
• MediaPipe Tasks (vision.object_detector)
• OpenCV
• Flask (for REST API)
• NumPy & CSV for data handling
• EfficientDet Lite2 model (.tflite)

Example Objects Detected

=======

Example Objects Detected (in class)

• Person
• Cell phone
• Laptop
• TV
• Chair
• Cup

To run in terminal:

python3 object_detector_webcam.py --model efficientdet_lite2.tflite

REST API

Run in terminal:

python3 api_server.py

In another terminal:

curl -X POST http://127.0.0.1:5050/start
curl http://127.0.0.1:5050/status
curl http://127.0.0.1:5050/detections
curl -X POST http://127.0.0.1:5050/stop

Logging Feature

• Logs webcam start/stop events to system.log
• Logs each detection (timestamp, object name, confidence score) to detections_log.csv
• Files are created automatically when the program runs