UNet For Crack Detection In Concrete

UNet For Crack Detection In Concrete

Segments cracks in a photo of concrete with a U-Net model: you upload an image of a cracked surface and the demo marks the cracks. For engineers, inspectors and students.

Made by Vizuara on Hugging Face. These notes are Mindket's, written from its Space page and README; the app's own files and license are on Hugging Face.

About

What it is: This Space runs a U-Net model that finds cracks in pictures of concrete. You upload an image of a cracked surface and the app segments it, which means it marks the pixels that belong to the cracks. The model weights ship inside the Space.

Who it suits: Site engineers, inspectors and students who want to see crack segmentation on a photo without setting up a model.

Industry | Who uses it | How they use it | Example
Construction | Site engineers | Mark cracks in a photo of concrete | Upload a photo of a cracked wall
Infrastructure | Inspectors | Look over inspection photos for cracks | Upload a picture of a damaged surface
Education | Students | See how U-Net segmentation works on a real image | Upload an image and view the marked result

People who need a structural safety verdict can skip it, since it segments an image and does not judge a building.

Where to use it:

• A browser: the Space runs on Hugging Face with nothing to install
• Windows, macOS or Linux: run a copy of the Space with Python

Use case | Who | What they get back
Mark cracks | Engineers | A segmentation of the cracks in the photo
Check inspection photos | Inspectors | A marked image for each upload
Learn segmentation | Students | A working U-Net example with its model code

How to set it up:

• Python with pip on Windows, macOS or Linux
• The Python packages listed in requirements.txt
• Nothing at all if you only use the Space in a browser

1. Install the Hugging Face hub package.
2. Download the Space's files into a folder.
3. Move into that folder.
4. Install Gradio 5.43.1 and the packages from requirements.txt.
5. Start the app.

pip install -U huggingface_hub
hf download Vizuara/UNet-for-crack-detection-in-concrete --repo-type space --local-dir UNet-for-crack-detection-in-concrete
cd UNet-for-crack-detection-in-concrete
pip install gradio==5.43.1
pip install -r requirements.txt
python app.py

How to use it: Open the Space page in a browser, or start your own copy with the command below. Then upload an image of cracked concrete to get its segmentation.

python app.py

Example: You upload a photo of a cracked concrete wall. The app runs the U-Net model on it and returns a segmentation that shows where the cracks are.

For newcomers and experts: If you are new, open the Space in a browser and upload one clear photo of a cracked surface. Nothing needs installing, and the result shows what the model picked out.

For experts, the repository holds the model definition in unet_model.py and a weights file named unet_weights_v2.pth, with app.py as the Gradio front end. The Space also names the Vizuara/unet-crack-segmentation model.

You upload a concrete photo
Gradio passes it to app.py
The U-Net model segments the cracks
You see the segmentation

Good to know:

• License: mit
• Cost to run: none named
• Data it reaches: the image you upload, and the Vizuara/unet-crack-segmentation model it names
• Not for: none stated in the README

On Hugging Face, 13 November 2026

• Likes: 1
• Built with: gradio
• License: mit

Links

• Space: https://huggingface.co/spaces/Vizuara/UNet-for-crack-detection-in-concrete
• Files: https://huggingface.co/spaces/Vizuara/UNet-for-crack-detection-in-concrete/tree/main

Try it

Open the Space in your browser, or copy it to your own Hugging Face account from its page. Nothing in this kit runs by itself.