Using Projection Pursuit in Multispectral Image Analysis,
Abstract
Principal components analysis is already used in multispectral image analysis to reduce the number of spectral dimensions. We, propose to use projection pursuit to find interesting combinations of spectral variates that produce images that enhance, contrast differences between differing land-use types. We develop a 3-dimensional moment index based on Jones and Sibson's index for projection into 2-dimensions.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1992
- Accession Number
- ADP007216
Entities
People
- G. P. Nason
- Robin Sibson
Organizations
- University of Bath