"Height Detection Model" is an application that leverages computer vision and pose estimation technologies to measure human height using a standard webcam. This project combines OpenCV for image processing and MediaPipe for accurate pose estimation, providing a non-intrusive way to measure height digitally.
To estimate distance and set a calibration factor, a depth estimation model must be used.
In this case, we are using Depth-anything, an open-source Hugging Face model.
- Real-time height measurement using a webcam.
- Automated voice instructions for user guidance.
- Calibration using a reference image with a known distance.
Check out the demonstration of the project on YouTube to see how it works in action: Output Video.
Ensure your system meets the following requirements:
- Python 3.x
- Webcam or any standard camera compatible with OpenCV.
git clone https://github.com/zamalali/Height-Detection.git
cd Height-DetectionBefore running the application, you need to capture a reference image (Ref_image.jpg) of a person standing at a known distance from the camera. This image is crucial for the algorithm to calibrate and accurately measure heights by comparison.
- Capture Reference Image: Take a photo of a person standing at a precisely measured distance from the camera. Ensure the full body is visible and positioned centrally in the frame.
- Save Reference Image: Name this image
Ref_image.jpgand save it in the project directory.
To start the application, run the ex.py script:
python ex.py