Artifacts reduction in VIR/Dawn data

Journal
Date Issued
2016
Author(s)
•
•
•
Ammannito, E.
•
Giardino, M.
•
•
•
DOI
Abstract
Remote sensing images are generally affected by different types of noise that degrade the quality of the spectral data (i.e., stripes and spikes). Hyperspectral images returned by a Visible and InfraRed (VIR) spectrometer onboard the NASA Dawn mission exhibit residual systematic artifacts. VIR is an imaging spectrometer coupling high spectral and spatial resolutions in the visible and infrared spectral domain (0.25-5.0 μm). VIR data present one type of noise that may mask or distort real features (i.e., spikes and stripes), which may lead to misinterpretation of the surface composition. This paper presents a technique for the minimization of artifacts in VIR data that include a new instrument response function combining ground and in-flight radiometric measurements, correction of spectral spikes, odd-even band effects, systematic vertical stripes, high-frequency noise, and comparison with ground telescopic spectra of Vesta and Ceres. We developed a correction of artifacts in a two steps process: creation of the artifacts matrix and application of the same matrix to the VIR dataset. In the approach presented here, a polynomial function is used to fit the high frequency variations. After applying these corrections, the resulting spectra show improvements of the quality of the data. The new calibrated data enhance the significance of results from the spectral analysis of Vesta and Ceres.
Volume
87
Issue
12
Start page
124501
Uri
Url
Issn Identifier
0034-6748
Ads BibCode
2016RScI...87l4501C
Rights
open.access
File(s)
Loading...
Thumbnail Image
Name

Carrozzo,2016,RSI,87,124501_[Artifacts reduction in VIR-Dawn data].pdf

Description
Pdf editoriale
Size

9.51 MB

Format

Adobe PDF

Checksum (MD5)

7135a31adc6227e9f555f69fac06cb54