Image Drone

Applying machine learning techniques for automatic classification of drone images.

DRONING project

1. Exploratory analysis. A initial attempt at creating a training dataset and applying a simple Random Forest classifier.

2. Refining the classification scheme. Using mosiaced images and a more comprehensive training dataset to see how much detail can be reasonably extracted using a simple random forest algorithm.

Frisk Oslofjord project

1. Initial raster processing. Combining the RGB and multispectral datasets into a single 8-band image.

2. Image classification. Training and evaluating a Random Forest model, then applying it to predict substrate classes for the full Akerøya dataset.

3. Tuning and comparing different models. Initial code for calibrating and comparing a range of supervsied classification algorithms.