NASA Logo

NTRS

NTRS - NASA Technical Reports Server

Back to Results
Filtering Interpolators for Image Comparison AlgorithmsComparing two or more images, either by differencing or ratioing, is important to many remote sensing problems. Because the pixel sample points for the images are (almost) always separated by some nonzero shift, a resampling, or interpolation, process must be performed if one image is to be accurately compared to another. Considered in Fourier space, an interpolator acts as a filter that attenuates some frequencies (usually high) of the image. Thus, when the shifted and unshifted images are compared, the former has been filtered, while the latter has not; the effect of this difference is called interpolation error. The key idea of this paper is to apply a filter to the unshifted image that matches the filtering effect of applying the interpolator to the shifted image, thereby drastically reducing interpolation error. The resulting interpolators, called filtering interpolators, are derived and discussed in detail elsewhere. Basic results will be given in this presentation.
Document ID
19930019608
Acquisition Source
Legacy CDMS
Document Type
Conference Paper
Authors
Lucke, R. L.
(Naval Research Lab. Washington, DC, United States)
Stocker, A. D.
(Space Computer Corp. Santa Monica, CA., United States)
Date Acquired
September 6, 2013
Publication Date
January 1, 1993
Publication Information
Publication: Lunar and Planetary Inst., Workshop on Advanced Technologies for Planetary Instruments, Part 1
Subject Category
Earth Resources And Remote Sensing
Accession Number
93N28797
Distribution Limits
Public
Copyright
Work of the US Gov. Public Use Permitted.
No Preview Available