作者: Stefanie Adam , Annelies De Backer , Steven Degraer , Jaak Monbaliu , Erik A. Toorman
DOI: 10.1016/S1568-2692(08)80011-3
关键词: Remote sensing (archaeology) 、 Similarity (geometry) 、 Reliability (statistics) 、 Cube 、 Remote sensing 、 Hyperspectral imaging 、 Sediment 、 Geography 、 Data set 、 Linear map
摘要: Abstract In this paper an automated method for hyperspectral image classification is proposed. The based on a linear transformation of each spectrum in the cube. Different sediment types and land covers were classified using two dimensions transformed data space. methodology applied to images IJzermonding mudflat, acquired by Compact Airborne Spectrographic Imager (CASI) 2001 2003. Comparable results obtained standard employed processing. similarity between was 82 65% 2003, respectively. superiority proposed user-friendly lies its autonomy, reliability objectivity. uses underlying statistical information set, while mainly expert knowledge.