Image-to-image co-registration is one of the preprocessing steps needed for the analysis of satellite time series. This chapter presents a new approach where all the available images are simultaneously co-registered, overcoming the limits of traditional techniques. This method was tested on the flood and landslide that occurred in Valtellina (northern Italy) during summer of 1987, resulting in the large rockslide of Val Pola. A data set made up of 13 medium-resolution satellite images collected with Landsat-4 and Landsat-5 Thematic Mapper over a period of 30 years was automatically processed. Results showed that the new approach can provide subpixel accuracy close to manual measurements, which today are considered the most accurate method for image registration. The multi-image co-registration method also demonstrated to be atmospheric resistant and robust against land-cover changes, snow, and cloud cover.
A new Approach to Satellite Time Series Co-registration for Landslide Monitoring
BARAZZETTI, LUIGI;GIANINETTO, MARCO;SCAIONI, MARCO
2015-01-01
Abstract
Image-to-image co-registration is one of the preprocessing steps needed for the analysis of satellite time series. This chapter presents a new approach where all the available images are simultaneously co-registered, overcoming the limits of traditional techniques. This method was tested on the flood and landslide that occurred in Valtellina (northern Italy) during summer of 1987, resulting in the large rockslide of Val Pola. A data set made up of 13 medium-resolution satellite images collected with Landsat-4 and Landsat-5 Thematic Mapper over a period of 30 years was automatically processed. Results showed that the new approach can provide subpixel accuracy close to manual measurements, which today are considered the most accurate method for image registration. The multi-image co-registration method also demonstrated to be atmospheric resistant and robust against land-cover changes, snow, and cloud cover.File | Dimensione | Formato | |
---|---|---|---|
CH12_Satellite_time_series.pdf
Accesso riservato
Descrizione: full paper
:
Publisher’s version
Dimensione
712.68 kB
Formato
Adobe PDF
|
712.68 kB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.