Paper Info
Title
Spectral Image Utility Sensitivity to Image Preprocessing and Information Exploitation Parameters
Abstract
For a wide range of application areas, quantifying the ability of a spectral image to satisfy the informational requirements of an application task would be desirable. We propose that this metric may be termed the "image utility" and present a method for assessing the utility of spectral images for the subpixel target detection task. In this paper, we investigate the sensitivity of this utility metric to various image chain parameters. In particular, we examine the effect of small variations of preprocessing and target detection scenario parameters on the assessed utility of a spectral image. We offer a method of quantifying the sensitivity to facilitate a rank ordering of parameter sensitivities. This exploration constitutes an initial step towards gaining a fuller understanding of the key parameters that drive spectral image utility.
Year
DOI
Venue
2008
10.1109/IGARSS.2008.4779068
Geoscience and Remote Sensing Symposium, 2008. IGARSS 2008. IEEE International
Keywords
Field
DocType
image processing,information theory,object detection,sensitivity analysis,spectroscopy,image chain parameters,image preprocessing,information exploitation parameters,spectral image utility metric,subpixel target detection task,target detection scenario parameters,sensitivity analysis,spectral image utility,subpixel target detection
Information theory,Object detection,Computer vision,Image sensor,Pattern recognition,Computer science,Image processing,Hyperspectral imaging,Preprocessor,Pixel,Artificial intelligence,Subpixel rendering
Conference
Volume
ISBN
Citations 
2
978-1-4244-2808-3
2
PageRank 
References 
Authors
0.45
2
2
Name
Order
Citations
PageRank
Marcus S. Stefanou1393.21
John P. Kerekes219435.38