Open Access2D comb feature for analysis of ship classification in high‐resolution SAR imagery
Author(s) -
Leng Xiangguang,
Ji Kefeng,
Zhou Shilin,
Xing Xiangwei,
Zou Huanxin
Publication year - 2017
Publication title -
electronics letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.375
H-Index - 146
ISSN - 1350-911X
DOI - 10.1049/el.2016.4598
Subject(s) - feature (linguistics) , remote sensing , high resolution , geology , pattern recognition (psychology) , artificial intelligence , computer science , contextual image classification , resolution (logic) , image (mathematics) , philosophy , linguistics
A new feature named ‘2D comb’ to improve ship classification is proposed. The proposed feature presents added value to distinguish between container ship, tank ship and cargo ship. It is based on radar cross‐section (RCS) statistic of the ship target related to the ship structure. Besides, related local RCS is proposed to classify three kinds of ships. Experimental results based on TerraSAR‐X images show that the proposed feature can abstract and describe the ship structure in high‐resolution synthetic aperture radar (SAR) imagery. It establishes relationship between RCS and ship structure, which is very useful to distinguish different kinds of ships.