This is the result of the Agile 2017 Paris Hackathon, a weekend of coding among friends with geophysical and computational skills.
We aim to help seismic interpreters in detecting features of interest within seismic data using Machine Learning techniques.
Haar Cascade classifiers are commonly used to identify faces or object in photos and video streams. We have adopted the very same approach for seismic data by taking advantage of OpenCV library routines. So far we have trained Haar Cascade to detect seismic traps and faults (as well as beer bottles!), but we plan to train our model on other seismic features in the near future.
If you want to detect seismic features in your awesome stacks, upload your image on our website and we can do it for you!