Application of remote sensing in detection of ocean oil pollution

J. T. Juang, K. S. Chen, C. F. Chen, L. Y. Chang

Research output: Contribution to journalArticlepeer-review

Abstract

It is well known that the increase of surface tension due to the presence of oil slicks causes the surface wave motion is depressed or even disappeared, thus making the surface electromagnetically smoother. Therefore, the radar backscattered energy is correspondingly decreased. These damping effects is now well understood and such effects enables the oil slicks to be discernible from the radar image. In this paper, we are concerned with the digital technique that effectively delineates the oil slicks pattern from the SAR image. The detected pattern allows us to estimate the coverage of oil spillage. The technique is based on the fact that the oil slicks make a gray value surface in the image which is a concave area with a certain size. In order to identify correctly these oil slicks and suppress the speckle noise and other natural phenomena, an image pyramid with multi-resolution layers is generated sequentially from the original image. Then a top-down approach, which applies both first and second order derivative operators, the Difference of Gaussian (DoG) and the Laplace of Gaussian (LoG), to the image pyramid, is used to detect oil patches. Two ERS-1 SAR images acquired on August 5, 1994 and September 5, 1994, around Taiwan, were used for testing. Results indicate that the proposed method depresses the speckle noise and other sea noise signals, and enhances the oil slicks pattern clearly. The robustness to the empirical parameters introduced in this scheme was also demonstrated.

Original languageEnglish
Pages (from-to)279-289
Number of pages11
JournalChemistry and Ecology
Volume14-15
Issue number1-4
DOIs
StatePublished - 1998

Keywords

  • Oil pollution
  • Remote sensing
  • SAR image

Fingerprint

Dive into the research topics of 'Application of remote sensing in detection of ocean oil pollution'. Together they form a unique fingerprint.

Cite this