Abstract
Basing ourselves on a novel segmentation algorithm for multispectral images, we consider how to detect multipixel anomalous objects in image cubes where spectral information is available. In particular, we have developed several morphological filters to compensate for noise that may be present in the initial cube. We show that for different types of noise (Gaussian and speckle), a modification of our morphology technique allows us to better detect targets without an enhanced false-alarm result.
Original language | English |
---|---|
Article number | 023604 |
Journal | Optical Engineering |
Volume | 45 |
Issue number | 2 |
DOIs | |
State | Published - 1 Feb 2006 |
Keywords
- AVIRIS
- Anomaly detection
- Gaussian noise
- Hyperspectral imaging
- Morphological operation
- Speckle noise
- Spectral data
ASJC Scopus subject areas
- Atomic and Molecular Physics, and Optics
- General Engineering