https://github.com/bee1997/people-counting-in-out
People Counting System that tracks and counts individuals entering and exiting a predefined region
https://github.com/bee1997/people-counting-in-out
computer-vision machine-learning opencv yolo11
Last synced: 6 months ago
JSON representation
People Counting System that tracks and counts individuals entering and exiting a predefined region
- Host: GitHub
- URL: https://github.com/bee1997/people-counting-in-out
- Owner: Bee1997
- License: mit
- Created: 2024-12-22T10:38:24.000Z (10 months ago)
- Default Branch: main
- Last Pushed: 2024-12-22T10:57:21.000Z (10 months ago)
- Last Synced: 2025-04-09T00:07:12.598Z (6 months ago)
- Topics: computer-vision, machine-learning, opencv, yolo11
- Language: Python
- Homepage:
- Size: 9.35 MB
- Stars: 1
- Watchers: 1
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# People-Counting-In-Out
This project implements a **People Counting System** that tracks and counts individuals entering and exiting a predefined region in a video. It uses **Ultralytics' Object Counter solution** with a YOLO model (`yolo11n.pt`). The system is ideal for applications in malls, offices, hospitals, and other facilities where monitoring foot traffic is essential.
## Key Features
- **Real-Time People Counting**: Tracks and counts people crossing a designated region in the video, providing separate counts for "In" and "Out" movements.
- **Region-Based Tracking**: Supports rectangular regions for monitoring specific areas.
- **Annotated Video Output**: Saves the output video with annotated bounding boxes, region markers, and counts.
- **Performance**: Processes each frame in approximately **~200ms**, including:
- Reading the video frame
- Object detection
- Tracking and counting objects crossing the region
- Annotating the frame and writing to the output video
## Applications
- **Malls**: Monitor customer foot traffic for better management and analysis.
- **Offices**: Track employee or visitor flow for security and resource allocation.
- **Hospitals**: Count individuals entering and exiting specific zones for efficient crowd management.
## System Performance
- **Average Processing Time**: ~200ms per frame (includes frame reading, detection, tracking, counting, annotation, and writing output).
- **System Specs**:
- Processor: Intel Core i5-6300U @ 2.40GHz
- RAM: 8GB
- **No GPU** was used for this project.
## Example Output
Here is a snapshot from the annotated output video:
The annotated video shows the bounding boxes, counts, and the designated region for tracking.
## Prerequisites
- Python 3.x
- Video file (`demo.mp4`) for processing.
- Custom YOLO model file (`yolo11n.pt`).
## Setup
1. Clone the repository:
```bash
git clone https://github.com/your-username/people-counting.git
cd people-counting
```
2. Install the required dependencies:
```bash
pip install -r requirements.txt
```
3. Add the input video file (demo.mp4) to the repository.
4. Ensure the custom YOLO model file (yolo11n.pt) is in the correct location.
## Running the Code
To process the video and generate the output:
```bash
python run.py
```
### Output
- Annotated Video: The processed video will be saved as demo_out.avi.
- Processing Time: Average processing time per frame is printed in the terminal after execution.
## Future Enhancements
- Real-Time Processing: Integrate with live camera feeds for real-time counting.
- Improved Performance: Optimize processing time for faster frame rates.
- Multi-Region Tracking: Extend support for monitoring multiple regions simultaneously.
- Data Analytics: Log and analyze traffic data for insights and predictions.
## Acknowledgments
This project uses:
- Ultralytics Solutions for object detection and tracking.
- OpenCV for video processing
## Contact
For questions or contributions, feel free to contact me or open an issue in the repository.