Capability of Geometric Features to Classify Ships in SAR
Imagery
Haitao Lang
a,b
, Siwen Wu
a
, Quan Lai
c
, Li Ma
d
a
Department of Physics and Electronics, Beijing University of Chemical Technology, Beijing
100029, China;
b
Jiangsu Key Laboratory of Big Data Analysis Technology/B-DAT, Nanjing University of
Information Science & Technology, Nanjing, 210044, China;
c
Inner Mongolia Key Laboratory of Remote Sensing and Geography Information System,
Inner Mongolia Normal University, Hohhot, 010022, China;
d
China Three Gorges Corporation, Beijing, 100038, China.
ABSTRACT
Ship classification in synthetic aperture radar (SAR) imagery has become a new hotspot in remote sensing
community for its valuable potential in many maritime applications. Several kinds of ship features, such as
geometric features, polarimetric features, and scattering features have been widely applied on ship classification
tasks. Compared with polarimetric features and scattering features, which are subject to SAR parameters (e.g.,
sensor type, incidence angle, polarization, etc.) and environment factors (e.g., sea state, wind, wave, current,
etc.), geometric features are relatively independent of SAR and environment factors, and easy to be extracted
stably from SAR imagery. In this paper, the capability of geometric features to classify ships in SAR imagery
with various resolution has been investigated. Firstly, the relationship between the geometric feature extraction
accuracy and the SAR imagery resolution is analyzed. It shows that the minimum bounding rectangle (MBR)
of ship can be extracted exactly in terms of absolute precision by the proposed automatic ship-sea segmentation
method. Next, six simple but effective geometric features are extracted to build a ship representation for the
subsequent classification task. These six geometric features are composed of length (f
1
), width (f
2
), area (f
3
),
perimeter (f
4
), elongatedness (f
5
) and compactness (f
6
). Among them, two basic features, length (f
1
) and width
(f
2
), are directly extracted based on the MBR of ship, the other four are derived from those two basic features.
The capability of the utilized geometric features to classify ships are validated on two data set with different
image resolutions. The results show that the performance of ship classification solely by geometric features is
close to that obtained by the state-of-the-art methods, which obtained by a combination of multiple kinds of
features, including scattering features and geometric features after a complex feature selection process.
Keywords: geometric feature, feature selection, ship classification, synthetic aperture radar(SAR)
1. INTRODUCTION
Nowadays, synthetic aperture radar (SAR) offers a great potential in observing and monitoring the maritime
environment, thanks to its capability of wide swath coverage and the ability to operate in all weather conditions.
1
Knowledge about ship’s position and category may serve for a wide range of applications, from maritime traffic
safety, fisheries control, to border surveillance.
2, 3
Ship detection on SAR images has been intensively studied
and has been developed as a crucial component for many operational ocean monitoring systems, such as SUMO,
SIMONS, etc.
4
Recent years, new emitting of SAR sensors, such as TerraSAR-X, COSMO-SkyMed, Radarsat-2,
etc.,
5–7
with higher spatial resolutions provide additional assets for ship type recognition (ship classification).
Discriminate feature extraction is essential for carrying out ship classification. In the past few years, the
studies on features extraction in the SAR imagery to characterize ships have explosively increased. Several kinds
of ship features, such as geometric features which can reflect the basic shape difference between different types
of ships, polarimetric features which can reflect the different scattering mechanisms resulting from the different
Corresponding author: Haitao Lang, E-mail: langht@mail.buct.edu.cn
Image and Signal Processing for Remote Sensing XXII, edited by Lorenzo Bruzzone,
Francesca Bovolo, Proc. of SPIE Vol. 10004, 1000415 · © 2016 SPIE
CCC code: 0277-786X/16/$18 · doi: 10.1117/12.2241375
Proc. of SPIE Vol. 10004 1000415-1
Downloaded From: http://spiedigitallibrary.org/ on 11/03/2016 Terms of Use: http://spiedigitallibrary.org/ss/termsofuse.aspx