Pool-Detection

1. We have used and trained the yolo v3 based on DarkNet53 architecture to find bounding box around the swimming pools in satellite images.
2. We have used convert.py file to generate the bounding boxes labels from PASCAL VOC annotations file for the given satellite images.
3. We have modified the last layer of architecture to make output class as pool only.

4. Download the model weights from the here.