Ship Detection on Sea Surface Based on Visual Saliency

DING Peng; ZHANG Ye; JIA Ping; CHANG Xu-ling; LIU Rang

doi:10.3969/j.issn.0372-2112.2018.01.018

您当前的位置：

首页 >

文章列表页 >

Ship Detection on Sea Surface Based on Visual Saliency

更新时间：2025-07-16

- Ship Detection on Sea Surface Based on Visual Saliency
- Acta Electronica Sinica Vol. 46, Issue 1, Pages: 127-134(2018)
- 作者机构：
  
  1. 中国科学院长春光学精密机械与物理研究所中科院航空光学成像与测量重点实验室,吉林,长春,130033
  2. 中国科学院大学,北京,100049
  3. 中国科学院长春光学精密机械与物理研究所中科院航空光学成像与测量重点实验室,吉林,长春,130033
  4. 中国科学院大学,北京,100049
- 作者简介：
- 基金信息：
- DOI：10.3969/j.issn.0372-2112.2018.01.018
  CLC： TP391.4
- Published Online：25 January 2018，
  
  Published：2018
- 稿件说明：
移动端阅览
DING Peng, ZHANG Ye, JIA Ping, et al. Ship Detection on Sea Surface Based on Visual Saliency[J]. Acta Electronica Sinica, 2018, 46(1): 127-134.
DOI：

DING Peng, ZHANG Ye, JIA Ping, et al. Ship Detection on Sea Surface Based on Visual Saliency[J]. Acta Electronica Sinica, 2018, 46(1): 127-134. DOI： 10.3969/j.issn.0372-2112.2018.01.018.

摘要

海面舰船检测技术有着特殊的军用以及民用意义，为了在宽广且环境复杂多变的海面上快速、高效、精确地检测到舰船目标，本文提出了一种基于多特征、多尺度视觉显著性的海面舰船检测方法.该方法充分利用了四元数图像可在多个通道上同时进行操作，节省操作时间，并保证不同尺度特征之间关联性的特点；除此之外，该方法还利用人眼对不用大小的图像关注目标不同的特点对图像进行尺度大小变换以避免漏检.该方法先用顶帽算法对原图进行简单的图像预处理以抑制云层、油污的干扰；其次提取多种特征构成四元数图像进行舰船目标显著性检测；在得到显著图后利用OTSU分割算法确定舰船所在的区域，并在原图上标定、提取舰船目标.通过在多种海面情况下分别进行实验分析，实验结果表明该算法可以排除云、雾、油污等干扰，精确、快速地检测到舰船目标，真正率达96.52%，虚警率低至2.11%，相较于他显著性检测算法在舰船检测方面有明显的优势.

Abstract

The ship detection technology is of special significance in both military and civilian level. In order to detect ship targets quickly

efficiently and accurately in a wide and complex sea surface environment

a new method for ship detection based on multi-features and multi-scale visual saliency is proposed. This method makes full use of features of the hyper-complex images which can be operated simultaneously in a number of channels

save operation time

and guarantee the characteristics of different scale characteristics. First

the method uses top-hat algorithm for image preprocessing of the original image to suppress the interference of clouds and oil. Secondly

a variety of features are extracted to form hyper-complex images to detect the significance of ship targets. When we get the last saliency map

we segment ships to ensure the target location by using the OTSU algorithm

and then we mark the ship target in the original image. We make the experimental analysis in several sea conditions

and experimental results show that the algorithm can eliminate fog

cloud and grease interfering with accurate detection of ship targets. In this algorithm

true rate meets 96.52% and the false alarm rate is as low as 2.11%. Compared to other saliency detection algorithm in ship detection

this algorithm has obvious advantages.

关键词

Keywords

references

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Error Concealment for Frame Losses in MVD Video Using Saliency

Invariant Features Based Ship Detection Model for Multi-source Remote Sensing Images

Review of the Methods for Salient Object Detection Based on Deep Learning

Extracting Attention Information Algorithm Based onContrast Sensitivity and Markov Chain

Visual Saliency and Text Confidence Map Based Background Suppression for Scene Text

Related Author

LU Yu

HUANG Xiao-feng

YIN Hai-bing

ZHOU Yang

CUI Jin-peng

GAO Xin-bo

WANG Nan-nan

YANG Xi

Related Institution

School of Telecommunication Engineering, Hangzhou Dianzi University

State Key Laboratory of Integrated Services Networks, Xidian University

Chongqing Key Laboratory of Image Cognition, Chongqing University of Posts and Telecommunications

College of Information Engineering， Jiangxi University of Science and Technology

College of Electronic and Information Engineering， Nanjing University of Aeronautics and Astronautics

⁰