Real-Time Human Height Measurement using YOLOv8 Pose + ArUco Marker (Tilt-Corrected)

This project measures a person's height in feet & inches (or cm) in real time using:

• YOLOv8 Pose for human keypoint detection
• ArUco marker for accurate distance and camera pitch estimation
• Camera calibration + perspective correction

Tested and working with Python 3.11

---

Requirements

Python 3.11 

1. Create and Activate Virtual Environment

Create virtual environment

py -3.11 -m venv height-env

Activate it

Windows:

env\Scripts\Activate

macOS / Linux:

source env/bin/activate

2. Install Dependencies

Run this command once inside the activated environment:

pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118

pip install opencv-python numpy ultralytics

(If you don't have a CUDA GPU, use CPU-only torch: pip install torch torchvision torchaudio)

3. Download YOLOv8 Pose Model (Automatic)

The script will automatically download yolov8n-pose.pt on first run.

4. Print & Place ArUco Marker

Marker ID: 0
Dictionary: 6x6_250
You can generate it here: https://chev.me/arucogen/

5. First Run - Camera Calibration

On first launch, the script will ask you to calibrate using a chessboard (9×6). Print a standard chessboard pattern and follow on-screen instructions.