Abstract
Hyperspectral images enable the detection of targets due to the high spectral sampling. The latest generation of sensors also provides an unprecedented spatial resolution which is further exploited in this article to uncover hard to detect anomalies. In particular, we model and estimate the background building upon robust supervised linear unmixing. We benefit from the high resolution of the data to spatially constrain the background. This provides a novel framework for exploiting both the spectral and the energy variations created by the presence of unknown targets to detect them.
Original language | English |
---|---|
Title of host publication | 2018 IEEE Global Conference on Signal and Information Processing (GlobalSIP) |
Publisher | IEEE |
Pages | 963-967 |
Number of pages | 5 |
ISBN (Electronic) | 9781728112954 |
DOIs | |
Publication status | Published - 21 Feb 2019 |
Event | 6th IEEE Global Conference on Signal and Information Processing 2018 - Anaheim, United States Duration: 26 Nov 2018 → 29 Nov 2018 |
Conference
Conference | 6th IEEE Global Conference on Signal and Information Processing 2018 |
---|---|
Abbreviated title | Global SIP 2018 |
Country/Territory | United States |
City | Anaheim |
Period | 26/11/18 → 29/11/18 |
Keywords
- Anomaly detection
- Hyperspectral imaging
- Linear mixture model
ASJC Scopus subject areas
- Information Systems
- Signal Processing