Can we fire all the geologists?

No. DeepCore reduces manual operations and improves core description speed. User receives information which is ready for import to any database. The geologist improves description speed and efficiency.

What determines the accuracy of a prediction?

DeepCore is an application created using artificial convolutional neural networks trained on a huge amount of data. Each outcome contains a probabilistic prediction. The prediction accuracy is determined by the number and variety of examples that were available for model training.

Are there any requirements to the size of the core columns?

We recommend loading the data with a minimum core column width in the image of 400 px for 10 cm core and 250 px for 6 cm. It is approximately a 100 dpi photo. 

What data formats are supported?

The acceptable formats for data uploading are JPEG, PNG, TIFF. 

What does the accuracy of description depend on?

Accuracy of description depends on how well the primary data base was prepared (variety and accuracy of description).

Is it possible that DeepCore does not recognize some core material?

DeepCore is not just a database – it is a geologist's assistant. The product allows you to extract automatically core from photos, depth reference and classify them.

Is DeepCore a database for images?

No. All the algorithms used in DeepCore work separately from the databases.

Do algorithms of DeepCore compare current images to images from the database?

Yes. In this case, the system assigns the most similar available class mark it as low-probability detection and allow an expert to correct it.

Is it possible that DeepCore does not recognize some core material?

Write to us at sales@petroleum.digital

The trial period is over. Can I extend?

If you have any doubts – just contact us hi@petroleum.digital