Urban morpho-types classification from SPOT-6/7 imagery and Sentinel-2 time series

This paper aims at detecting several urban morpho-type classes out of complementary SPOT-6/7 imagery and Sentinel-2 time series and using existing Urban Atlas database. A bottom-up strategy is proposed. It first detects basic urban objects (buildings, roads, vegetation), that are used to calculate multi-scale morphological features. These features are then fed to a Random Forest classifier trained from samples out of Urban Atlas urban classes. Obtained results are optionally merged with a Random Forest classification based on Sentinel-2 time series. Obtained results are promising.