Boat-detection-CNN

Boat detection using a Convolutional Neural Netwrok

Using satellite images of San Francisco Bay Area from Planet's API a CNN was trained to identify boats.

• The dataset contains 40,000 images
• The confusion matrix was used to analyze metrics, priotizing the sensitiviy over precision.

Documentation

De file code.ipynb contains a the code as well as explanation for each part (load data, clean data, CNN creation, CNN training,...)