RESULTS
Re-photographing: solutions were developed for overlapping historical photos with current situations in town. This shows how city corners changed and give impressions of urban modifications or happened hazards.
Colorization: using neural networks, colors were inpainted in historical photos for better visualization and communication purposes.
3D effects: historical images were enhanced with 3D effects based on deep learning solutions for visualization needs and better understanding.
4D city modelling: using historical maps and machine learning techniques, unavailable building heights were predicted and maps were converted in 3D cities for studies, visualization and urban analyses purposes.
Recent developments: 3D city modelling from cadastral data to investigate and visualize the city growth in the last 50 years (1973-2016).
