ARGO-E GROUP's Machine Learning Division has developed a Python application that assists engineers or scientists without any programming skills, to classify images easily and efficiently.

The app was developed as part of a rock slope characterization project in collaboration with Professor Dimitrios Zekkos research group at UC Berkeley, where thousands of photos are used as input and are quickly classified. For the purposes of the specific  project, images were classified in four categories:

1. Good rock;
2. Poor or weathered rock and colluvium;
3. Active landslides or instability;
4. Terraces or flatter ground conditions; and
5. Other.

The use of the app is shown in the video below. The user-friendly interface allows users to create clusters within each image and then label the relevant clusters in the appropriate category. The output is a new, cropped image that is labeled. These labeled photos can then be used to train and test Machine Learning or Deep Learning algorithms.

If you have an interest in using such an app please contact ARGO-E GROUP, or Prof. Zekkos.