E. Kukawska et al., “Multitemporal Sentinel-2 data – remarks and observations,” 2017 9th International Workshop on the Analysis of Multitemporal Remote Sensing Images (MultiTemp), Brugge, 2017, pp. 1-4.

All the pre-processing algorithms are being improved constantly. The biggest challenge for the multitemporal analysis is to deal with errors caused directly by the chain of pre-processing of raw Sentinel-2 data to the level L1C – misregistration of pixels. Another problem to overcome while aggregating a series of classifications is the incorrect mask of clouds over artificial structures resulting from atmospheric correction performed with Sen2Cor software. Presented errors have direct influence on the overall accuracy of classifications performed in S2GLC project which are going to be implemented as a fully automatic process.

DOI: 10.1109/Multi-Temp.2017.8035212