Object Detection System

A real-time object detection system using YOLOv8 to identify and track vehicles, people, and traffic lights in videos. The system includes:

• YOLOv8 model for object detection
• Traffic light color detection
• Statistical analysis of detected objects

Note: Full IEEE format documentation is available in object_detection_ieee_paper.md

Features

• Real-time Object Detection: Identifies vehicles, people, and traffic lights
• Traffic Light Color Recognition: Analyzes traffic light status (Red, Yellow, Green)
• Statistics Tracking: Counts and records objects per frame
• Performance Metrics: Monitors processing speed and detection rates
• Video Processing: Processes and saves annotated videos with detection overlays

Requirements

• Python 3.7+
• PyTorch
• OpenCV 4.5+
• NumPy 1.20+
• Ultralytics YOLO

Installation

1. Clone this repository:

   git clone https://github.com/yourusername/object-detection.git
   cd object-detection
   

2. Install the required packages:

   pip install -r requirements.txt
   

3. Ensure the YOLOv8 model file (yolov8n.pt) is in the project root directory

Usage

Process a video with object detection

python main.py --mode process --input your_video.mp4 --output result.mp4

Additional options

--no-display     # Disable real-time display during processing
--no-stats       # Disable statistics saving

How It Works

The system uses a deep learning-based approach for object detection:

Object Detection:

• Uses YOLOv8 pretrained model
• Identifies vehicles (cars, trucks, buses, bikes)
• Detects people and traffic lights
• Draws bounding boxes with class labels and confidence scores

Traffic Light Analysis:

• Detects traffic light status (Red, Yellow, Green)
• Uses HSV color space analysis for light state detection

Statistics:

• Tracks number of vehicles and people per frame
• Calculates average and maximum object counts
• Generates statistics report at completion
• Overlays real-time stats on processed video

Project Structure

• main.py: Main script for video processing
• utils/object_detector.py: Object detection implementation
• utils/config.py: Configuration settings
• requirements.txt: Required Python packages
• object_detection_ieee_paper.md: IEEE format documentation

Examples

Run the detector on included test videos:

python main.py --mode process --input test.mp4 --output output.mp4

License

MIT License