Detect humans in real-time with lightning-fast YOLOv8 no touch, no tags, just AI vision. Perfect for smart surveillance, people counting, and intelligent automation at your fingertips.
The Human Scanner Application is a real-time human detection tool using AI and computer vision. It scans video input (from a webcam or camera) and identifies human presence using the YOLOv8 model, which is fast and accurate. This application is useful for security surveillance, crowd monitoring, people counting, and smart automation systems.
🎥 Live Human Detection using YOLOv8 object detection.
⚡ Real-Time Processing on webcam feed.
📍 Draws Bounding Boxes around detected humans.
🧠 Optimized for Speed and Accuracy with modern YOLO architecture.
📁 Optional Logging of time and number of humans detected.
💻 Works offline on any system with a webcam and Python installed.
- Install OpenCV (for video capture and image processing)
pip install opencv-python
- Install NumPy (for numerical operations with arrays and matrices)
pip install numpy
- Install Ultralytics (to use YOLOv8 for human detection)
pip install ultralytics
Once installed, you can test it with:
yolo help
requirements.txt(Copy and save this in your project folder):
opencv-python
numpy
ultralytics
- How to Use It:
Once the file is saved, you (or anyone using your project) can install all the required libraries at once by running:
pip install -r requirements.txt
- The camera is turned on via
cv2.VideoCapture(0). - The YOLOv8 model loads and starts scanning each video frame.
- It detects humans in real-time and draws bounding boxes.
- The user sees the result live in a window.
- (Optional) Human count or detection logs can be saved for later use.
On your webcam feed:
A bounding box appears on each detected person.
It may show:
Detected Humans: 3
Optional: Save detections to a CSV like:
| Timestamp | Count |
|---|---|
| 2025-06-26 10:00AM | 4 |
| 2025-06-26 10:05AM | 2 |
| Feature | Our Application | Others |
|---|---|---|
| Uses latest YOLOv8 model | ✅ Yes | ❌ Older versions |
| Real-time on webcam | ✅ Yes | ✅/❌ Sometimes |
| Lightweight & fast | ✅ Yes | ❌ Slower |
| Works offline | ✅ Yes | ❌ Often cloud-only |
| Easy to customize | ✅ Yes | ❌ Less flexible |
- Add person tracking & movement heatmaps.
- Integrate with CCTV systems for real-time alerts.
- Show real-time person count dashboards.
- Gender, age, and posture detection using more models.
- Use for intruder detection in restricted areas.
Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.
Project Owner: Priyanshi Chaudhary
Email: [email protected]