Real-time Object Detection with YOLO and Webcam

This project implements real-time object detection using YOLOv3 (You Only Look Once) and a webcam feed, built with Python and OpenCV. It detects objects from 80 different classes defined in the COCO dataset and displays bounding boxes with labels and confidence scores.

Features

• Real-time object detection using webcam feed
• Detects 80 different object classes from the COCO dataset
• Displays bounding boxes with class labels and confidence scores
• Uses YOLOv3 pre-trained model

Prerequisites

Before running the project, ensure you have the following:

Software Requirements

• Python 3.6+
• OpenCV (pip install opencv-python)
• NumPy (pip install numpy)

Required Files

1. yolov3.weights - Pre-trained weights file (~237MB)
2. yolov3.cfg - YOLO configuration file
3. coco.names - File containing 80 class names

Download these files:

• yolov3.weights
• yolov3.cfg
• coco.names

Place all downloaded files in the same directory as the Python script.

Installation

1. Clone this repository:

git clone <repository-url>
cd <repository-name>