Tire scan
Read out all tire data with just one photo
Team
3 AI developers
Project duration
7 months
Product success
Integration into worldwide online shops
Tire detection by photo
Together with the European market leader in online tire trading - Delticom AG - we have developed an AI-based tire recognition system. This enables tire data such as brand, dimension, profile, DOT, load & speed index to be read out for any car tire using just one photo. Our tire detection works with any camera - both for smartphones and within permanently installed systems. You can also find more information in our blog.
And here, too, an innovative, joint product emerged from an internal project, which is now also being made available to the entire market
Our partner :
Our services: Consulting, AI development, deployment, continuous development and support
Technologies used : CNN, RNN, Beam Search, REST API, Kubernetes, Google Cloud Deployment, GPU multi-worker setup
Reliable
Detection with over 93% accuracy
Fast
Just 2 seconds instead of 2 minutes
Simple
Can be quickly integrated into software and websites via API
And how does it work?
Tyre data from a plant, as often used in the warehousing process, as well as mounted on the passenger car and under different angles are suitable as input
Create photo
Our segmentation network reliably identifies the tire on the image and returns the relevant area.
AI-based tire detection
The tire is calculated "smoothly" - with the highest possible resolution and minimal distortion. This is the input for our AI-based OCR - regardless of the original image.
Image is processed by AI service
The AI-based OCR recognizes the labels on the pre-processed image and outputs them sorted by manufacturer, speed index, etc. - even for dirty or worn tires. Can also be integrated directly into existing IT systems as a REST API.
OCR for the output
For more information
or information about our services
.png)