This is a submission repository for European Association Of Geoscientist & Engineers (EAGE) Earth-Hack Hackathon 2019
We try to analyze and extract rock properties by classifying the rock properties and location corresponding to the core images using transfer learning.
- Akmal
- Hasif
- Sidiq
- Syuhada
Machine Learning Library
- Install all the library required
pip3 install flask torch torchvision fastai scikit-learn numpy scipy pandas
- Download the trained models and data from google drive
https://1drv.ms/f/s!ArJJyxURU-h_h5QECCEsR1SG6cGZdQ
files to src/satellite/modelshttps://1drv.ms/f/s!ArJJyxURU-h_h5QDNLfWbj2uhY54aw
files to src/satellite/train
- Run the script to initialize the server
python3 server.py
- To access the application
http://localhost:5000
- To access the API, send an image named 'file' as a multipart/form-data POST request to
http://localhost:5000/image
The hack is governed by Creative Commons, CC BY. You can:
- Copy
- Adapt or Modify
- Redistribute (Publish, Display, Publicly Perform or Communicate the Work
- License to others