Poster De Conférence Année : 2026

Quantifying the impact of co-registration and satellite jitter filtering on the accuracy of multi-temporal elevation differences

Résumé

Multi-temporal satellite-based change detection analysis requires addressing biases related to coregistration and satellite jitter. This study compares two workflows combining two co-registration strategies with jitter filtering using Pléiades imagery from 2015 and 2023. The Time-SIFT approach shows improved horizontal accuracy and comparable vertical accuracy compared to the classical approach without requiring GCPs or pre-georeferenced data, reducing elevation difference dispersion from 1.71 to 1.32 m in terms of NMAD. After jitter correction, dispersions reach 1.13 m and 0.96 m NMAD for the classical and the Time-SIFT approaches, respectively. These results demonstrate that the Time-SIFT co-registration quality yields improved elevation difference after filtering.

Fichier principal
Vignette du fichier
NTRICS2026_poster_IB.pdf (2.31 Mo) Télécharger le fichier
NTRICS_2026_resubmitted_version.pdf (4.32 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Licence

Dates et versions

hal-05480252 , version 1 (27-01-2026)

Licence

Identifiants

  • HAL Id : hal-05480252 , version 1

Citer

Imen Brini, Denis Feurer, Riadh Tebourbi, Fabrice Vinatier, Riadh Abdelfattah. Quantifying the impact of co-registration and satellite jitter filtering on the accuracy of multi-temporal elevation differences. New Trends in Image, Communication, and Signal Processing, Jan 2026, Tunis, Tunisia. ⟨hal-05480252⟩
76 Consultations
50 Téléchargements

Partager

  • More