Object detection system for identifying ships and boats in aerial imagery using deep learning. The project uses a dataset of 621 annotated images in PASCAL VOC format with bounding box annotations for ship localization in satellite and drone imagery.
The detection pipeline was implemented using YOLO, processing aerial images to localize and classify vessels. Model performance was evaluated on an annotated validation set to assess detection accuracy and bounding box precision.
Python, YOLO, PASCAL VOC annotations, Kaggle dataset