Inspiration
There exists no crack annotation tool for analyzing crack images.
What it does
CRAT is a cross platform preprocessing and annotation tool for cracks in materials. It allows the user to isolate cracks from the materials using machine learning algorithms and annotate them with other useful parameters. It also allows manual overrides by the user in case the annotated data is estimated incorrectly. The image of the crack is saved along with the data annotated for the crack. Since we have the crack shape as well as the material image (since cracks are removed from it) , the preprocessed data could be used as input to more sophisticated learning models to study cracks. The cross platform nature of the application makes it possible to use this tool on any device (mobiles, tablets and PCs) thus making it more convenient to annotate images on the go.
How we built it
The application is developed using Python and Kivy.
Challenges we ran into
Defining the problem to solve. Being Neuroengineering students, crack analysis was something we were unfamiliar with
Accomplishments that we're proud of
We were able to pose a solution after much brainstorming. App under development. Once developed it would be cool.
What we learned
We will be learning app development with Kivy
What's next for CRAT
Can be deployed as a preprocessing tool for more advanced machine learning for crack analysis
Built With
- kivy
- python
Log in or sign up for Devpost to join the conversation.