Real-time Object Detection with YOLOv5

View on GitHub

A real-time object detection application leveraging YOLOv5 that processes webcam input to identify up to 80 distinct object categories including people, vehicles, and animals at live frame rates.

Key Features

  • Live webcam detection - Approximately 20 FPS performance
  • CPU-compatible - Works without GPU (tested on MacBook Pro M4)
  • 80 object classes - Detects person, car, dog, cat, laptop, cell phone, and more
  • Easy setup - Automated environment initialization

Detectable Objects

The model can identify a wide range of objects including:

  • People and body parts
  • Vehicles (car, bicycle, motorcycle, bus, truck)
  • Animals (dog, cat, bird, horse)
  • Electronics (laptop, cell phone, TV)
  • Furniture and household items
  • And many more...

Quick Start

# Clone the repository
git clone https://github.com/feraandrei1/ultralytics-yolov5-ai-video-recognition.git

# Run the setup script
./run.sh

Technologies

  • Python 3.6+
  • YOLOv5 (Ultralytics framework)
  • OpenCV for video processing
  • Virtual environment for dependency management

Platform Support

  • macOS with Python environment
  • Camera access required
  • Troubleshooting guidance included for camera permissions

This project demonstrates YOLO inference on video and serves as a great starting point for computer vision applications.