Deep recurrent neural networks for land-cover classification using sentinel-1 insar time series

Shaojia Ge, Oleg Antropov, Weimin Su, Hong Gu, Jaan Praks

Research output: Chapter in Book/Report/Conference proceedingConference article in proceedingsScientificpeer-review

5 Citations (Scopus)

Abstract

To date, the potential of multitemporal interferometric SAR (InSAR) data in land-cover mapping has not been fully explored despite suitable time series increasingly acquired from SAR sensors. Here, we suggest to use an LSTM (Long Short Term Memory) based land-cover classifier to address this problem. Spatial context is preserved by using grey-level spatial dependencies and morphological profiles. Further, a 4-LSTM-based model was trained to capture the temporal dynamics of InSAR coherence. Altogether 39 Sentinel-1 interferometric coherence pairs acquired over Donana in Spain were used to evaluate the method performance. Achieved more than 90% overall accuracy indicates the strong potential of developed InSAR recurrent approach in improving differentiation between various land cover classes.

Original languageEnglish
Title of host publicationIGARSS 2019 - 2019 IEEE International Geoscience and Remote Sensing Symposium
PublisherIEEE Institute of Electrical and Electronic Engineers
Pages473-476
Number of pages4
ISBN (Electronic)978-1-5386-9154-0, 978-1-5386-9153-3
ISBN (Print)978-1-5386-9155-7
DOIs
Publication statusPublished - 2019
MoE publication typeA4 Article in a conference publication
Event39th IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2019 - Yokohama, Japan
Duration: 28 Jul 20192 Aug 2019

Conference

Conference39th IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2019
Country/TerritoryJapan
CityYokohama
Period28/07/192/08/19

Keywords

  • Coherence
  • GLCM
  • Land-cover mapping
  • LSTM
  • Morphological profile
  • Recurrent neural networks
  • Sentinel- 1
  • Synthetic aperture radar (SAR)
  • Time series

Fingerprint

Dive into the research topics of 'Deep recurrent neural networks for land-cover classification using sentinel-1 insar time series'. Together they form a unique fingerprint.

Cite this