S-Cnn-based ship detection from high-resolution remote sensing images

Ruiqian Zhang, Jian Yao, Kao Zhang, Jiadong Zhang

Research output: Contribution to journalConference article

Abstract

Reliable ship detection plays an important role in both military and civil fields. However, it makes the task difficult with high-resolution remote sensing images with complex background and various types of ships with different poses, shapes and scales. Related works mostly used gray and shape features to detect ships, which obtain results with poor robustness and efficiency. To detect ships more automatically and robustly, we propose a novel ship detection method based on the convolutional neural networks (CNNs), called S-CNN, fed with specifically designed proposals extracted from the ship model combined with an improved saliency detection method. Firstly we creatively propose two ship models, the "V" ship head model and the "||" ship body one, to localize the ship proposals from the line segments extracted from a test image. Next, for offshore ships with relatively small sizes, which cannot be efficiently picked out by the ship models due to the lack of reliable line segments, we propose an improved saliency detection method to find these proposals. Therefore, these two kinds of ship proposals are fed to the trained CNN for robust and efficient detection. Experimental results on a large amount of representative remote sensing images with different kinds of ships with varied poses, shapes and scales demonstrate the efficiency and robustness of our proposed S-CNN-Based ship detector.

Original languageEnglish (US)
Pages (from-to)423-430
Number of pages8
JournalInternational Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences - ISPRS Archives
Volume41
DOIs
StatePublished - Jan 1 2016
Externally publishedYes
Event23rd International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences Congress, ISPRS 2016 - Prague, Czech Republic
Duration: Jul 12 2016Jul 19 2016

Fingerprint

neural network
Remote sensing
Ships
remote sensing
Ship models
efficiency
Neural networks
detection method
Military
detection
ship
lack
Detectors

Keywords

  • Convolutional neutral networks (CNN)
  • S-CNN
  • Ship detection
  • Ship model construction

ASJC Scopus subject areas

  • Information Systems
  • Geography, Planning and Development

Cite this

S-Cnn-based ship detection from high-resolution remote sensing images. / Zhang, Ruiqian; Yao, Jian; Zhang, Kao; Zhang, Jiadong.

In: International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences - ISPRS Archives, Vol. 41, 01.01.2016, p. 423-430.

Research output: Contribution to journalConference article

@article{8ed6193361ea44d4955e3e05e1bb1662,
title = "S-Cnn-based ship detection from high-resolution remote sensing images",
abstract = "Reliable ship detection plays an important role in both military and civil fields. However, it makes the task difficult with high-resolution remote sensing images with complex background and various types of ships with different poses, shapes and scales. Related works mostly used gray and shape features to detect ships, which obtain results with poor robustness and efficiency. To detect ships more automatically and robustly, we propose a novel ship detection method based on the convolutional neural networks (CNNs), called S-CNN, fed with specifically designed proposals extracted from the ship model combined with an improved saliency detection method. Firstly we creatively propose two ship models, the {"}V{"} ship head model and the {"}||{"} ship body one, to localize the ship proposals from the line segments extracted from a test image. Next, for offshore ships with relatively small sizes, which cannot be efficiently picked out by the ship models due to the lack of reliable line segments, we propose an improved saliency detection method to find these proposals. Therefore, these two kinds of ship proposals are fed to the trained CNN for robust and efficient detection. Experimental results on a large amount of representative remote sensing images with different kinds of ships with varied poses, shapes and scales demonstrate the efficiency and robustness of our proposed S-CNN-Based ship detector.",
keywords = "Convolutional neutral networks (CNN), S-CNN, Ship detection, Ship model construction",
author = "Ruiqian Zhang and Jian Yao and Kao Zhang and Jiadong Zhang",
year = "2016",
month = "1",
day = "1",
doi = "10.5194/isprsarchives-XLI-B7-423-2016",
language = "English (US)",
volume = "41",
pages = "423--430",
journal = "International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences - ISPRS Archives",
issn = "1682-1750",

}

TY - JOUR

T1 - S-Cnn-based ship detection from high-resolution remote sensing images

AU - Zhang, Ruiqian

AU - Yao, Jian

AU - Zhang, Kao

AU - Zhang, Jiadong

PY - 2016/1/1

Y1 - 2016/1/1

N2 - Reliable ship detection plays an important role in both military and civil fields. However, it makes the task difficult with high-resolution remote sensing images with complex background and various types of ships with different poses, shapes and scales. Related works mostly used gray and shape features to detect ships, which obtain results with poor robustness and efficiency. To detect ships more automatically and robustly, we propose a novel ship detection method based on the convolutional neural networks (CNNs), called S-CNN, fed with specifically designed proposals extracted from the ship model combined with an improved saliency detection method. Firstly we creatively propose two ship models, the "V" ship head model and the "||" ship body one, to localize the ship proposals from the line segments extracted from a test image. Next, for offshore ships with relatively small sizes, which cannot be efficiently picked out by the ship models due to the lack of reliable line segments, we propose an improved saliency detection method to find these proposals. Therefore, these two kinds of ship proposals are fed to the trained CNN for robust and efficient detection. Experimental results on a large amount of representative remote sensing images with different kinds of ships with varied poses, shapes and scales demonstrate the efficiency and robustness of our proposed S-CNN-Based ship detector.

AB - Reliable ship detection plays an important role in both military and civil fields. However, it makes the task difficult with high-resolution remote sensing images with complex background and various types of ships with different poses, shapes and scales. Related works mostly used gray and shape features to detect ships, which obtain results with poor robustness and efficiency. To detect ships more automatically and robustly, we propose a novel ship detection method based on the convolutional neural networks (CNNs), called S-CNN, fed with specifically designed proposals extracted from the ship model combined with an improved saliency detection method. Firstly we creatively propose two ship models, the "V" ship head model and the "||" ship body one, to localize the ship proposals from the line segments extracted from a test image. Next, for offshore ships with relatively small sizes, which cannot be efficiently picked out by the ship models due to the lack of reliable line segments, we propose an improved saliency detection method to find these proposals. Therefore, these two kinds of ship proposals are fed to the trained CNN for robust and efficient detection. Experimental results on a large amount of representative remote sensing images with different kinds of ships with varied poses, shapes and scales demonstrate the efficiency and robustness of our proposed S-CNN-Based ship detector.

KW - Convolutional neutral networks (CNN)

KW - S-CNN

KW - Ship detection

KW - Ship model construction

UR - http://www.scopus.com/inward/record.url?scp=84979536750&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84979536750&partnerID=8YFLogxK

U2 - 10.5194/isprsarchives-XLI-B7-423-2016

DO - 10.5194/isprsarchives-XLI-B7-423-2016

M3 - Conference article

AN - SCOPUS:84979536750

VL - 41

SP - 423

EP - 430

JO - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences - ISPRS Archives

JF - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences - ISPRS Archives

SN - 1682-1750

ER -