BS final-year project

Building segmentation from fused imagery

Automated extraction of urban features from fused satellite and aerial imagery with deep learning.

Final-year project at the Institute of Geo-Information and Earth Observation, PMAS Arid Agriculture University Rawalpindi, supervised by Dr Hasan Ali Baig, carried out with Muhammad Bilal. An end-to-end pipeline in Keras and fastai compared U-Net, SSD and Mask R-CNN. U-Net reached an F1 score of 0.89.

The repository holds the code and a ready-to-use dataset, and stays as the record of that work. A new version is planned.

Related