Computer Engineering and Applications ›› 2011, Vol. 47 ›› Issue (18): 195-197.

• 图形、图像、模式识别 • Previous Articles     Next Articles

Oil spills identification in SAR image using Mahalanobis distance

ZHOU Hui1,CHEN Peng2   

  1. 1.Department of Computer Science,Dalian Neusoft Information University,Dalian,Liaoning 116023,China
    2.College of Environmental Science and Engineering,Dalian Maritime University,Dalian,Liaoning 116023,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2011-06-21 Published:2011-06-21

利用Mahalanobis距离的SAR图像溢油识别

周 慧1,陈 澎2   

  1. 1.大连东软信息学院 计算机科学与技术系,辽宁 大连 116023
    2.大连海事大学 环境科学与工程学院,辽宁 大连 116023

Abstract: A method on feature vector of the oil spills identification in SAR images is proposed,The advantages of Synthetic Aperture Radar(SAR) are useful,which can work on day and night and all weather conditions for high resolution monitoring of oil spills.The Mahalanobis distance approach has been developed in order to evaluate the probability that a dark area is a slick on SAR images.The experimental verification,a small number of characteristic values has been selected in the paper for that oil spills is significant.The algorithm of Mahalanobis distance discrimination is clear,and the accuracy rate reaches 96% or more.Compare to other oil spills determine methods,this approach appends less conditions and is conducive to computer programming.

Key words: Synthetic Aperture Radar(SAR) image, oil spill, feature vector, Mahalanobis distance

摘要: 利用合成孔径雷达(SAR)可在日夜及全天候条件下进行高分辨率溢油监测的优点,提出了一种基于特征向量的SAR图像溢油识别方法。在暗区边界确定的SAR图像中进行量算,得到特征向量,并采用Mahalanobis距离对目标物进行识别。经实验验证,选取的特征值数量合理,且对于判定是否为溢油效果明显;利用Mahalanobis距离进行判别,算法清晰,且准确率达到96%以上;与其他溢油判定方法相比,附加条件较少,且利于在计算机上编程实现。

关键词: 合成孔径雷达(SAR)图像, 溢油, 特征向量, Mahalanobis距离