Design of fuzzy supervised classification system for single-channel SAR images

Author

Fang, Su ; Wen, Hong ; Shiyi, Mao

Author_Institution

Dept. of Electr. Eng., Beijing Univ. of Aeronaut. & Astronaut., China

fYear

2001

fDate

2001

Firstpage

493

Lastpage

497

Abstract

A fuzzy supervised classification system for single-channel SAR images is designed. The system mainly consists of two parts: a fuzzy supervised partitioner and a fuzzy vector quantizer both of which employ fuzzy theory. The fuzzy partitioner is the key part that determines the classification accuracy level and the fuzzy vector quantizer is appended to improve the accuracy. The use of the system includes system training, data fuzzy analysis, result display and classification accuracy estimation. The classification result is compared with GML classification result to evaluate the system´s performance

Keywords

fuzzy systems; image classification; learning (artificial intelligence); radar imaging; synthetic aperture radar; vector quantisation; GML classification; classification accuracy; classification accuracy estimation; data fuzzy analysis; fuzzy supervised classification system design; fuzzy supervised partitioner; fuzzy theory; fuzzy vector quantizer; result display; single-channel SAR images; system performance evaluation; system training; Data analysis; Displays; Fuzzy systems; Image classification; Labeling; Maximum likelihood estimation; Pixel; Planets; Speckle; Synthetic aperture radar;

fLanguage

English

Publisher

ieee

Conference_Titel

Radar, 2001 CIE International Conference on, Proceedings

Conference_Location

Beijing

Print_ISBN

0-7803-7000-7

Type

conf

DOI

10.1109/ICR.2001.984754

Filename

984754