A Hybrid CNN-Transformer Architecture for Semantic Segmentation of Radar Sounder data

Ghosh, Raktim; Bovolo, Francesca

doi:10.1109/igarss46834.2022.9883124

Cited by 2 publications

(1 citation statement)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Additional improvements to this clutter discrimination technique can be done by applying super-resolution techniques [20] to the radargrams before the co-registration step for improving the resolution of radar products. As future works, we plan to explore and test deep learning algorithms for automatic clutter features recognition [21] in orbital radar sounder products.…”

Section: Discussionmentioning

confidence: 99%

Clutter Detection and Surface/Subsurface Slope Determination by Combination of Repeat-Pass Sounder Orbits Applied to SHARAD Data

Raguso

Mastrogiuseppe²,

Seu³

2022

IEEE Geosci. Remote Sensing Lett.

View full text Add to dashboard Cite

Nadir-looking low-frequency radar sounders cannot easily resolve off-nadir surface returns from the subsurface nadir echoes. Cross-track surface echoes (also named "clutter") with time delays synchronized with subsurface returns are renowned for being a major challenge for scientists, as they can affect the analysis of orbital radar sounders data. We present a method for clutter discrimination and surface/subsurface slope estimation using data acquired from radar sounders in closely spaced repeated orbits configuration. The method takes advantage of cross-track signal migration to discriminate off-nadir clutter from subsurface signal returns received at the nadir. The migration of the off-nadir signals is also used to determine the clutter direction of arrival (DOA) as well as the surface/subsurface crosstrack slopes. The effectiveness of the method has been proven on the Mars Reconnaissance Orbiter (MRO)'s Shallow Radar (SHARAD) dataset and provides a proof-of-concept demonstration for the surface clutter discrimination when radar sounders repeated-passes data are available.

show abstract

Section: Discussionmentioning

confidence: 99%