Discrimination of Oil Slicks and Lookalikes in Polarimetric SAR Images Using CNN.

Jubai An,Hao Guo,Danni Wu

doi:10.3390/s17081837

Abstract

Oil slicks and lookalikes (e.g., plant oil and oil emulsion) all appear as dark areas in polarimetric Synthetic Aperture Radar (SAR) images and are highly heterogeneous, so it is very difficult to use a single feature that can allow classification of dark objects in polarimetric SAR images as oil slicks or lookalikes. We established multi-feature fusion to support the discrimination of oil slicks and lookalikes. In the paper, simple discrimination analysis is used to rationalize a preferred features subset. The features analyzed include entropy, alpha, and Single-bounce Eigenvalue Relative Difference (SERD) in the C-band polarimetric mode. We also propose a novel SAR image discrimination method for oil slicks and lookalikes based on Convolutional Neural Network (CNN). The regions of interest are selected as the training and testing samples for CNN on the three kinds of polarimetric feature images. The proposed method is applied to a training data set of 5400 samples, including 1800 crude oil, 1800 plant oil, and 1800 oil emulsion samples. In the end, the effectiveness of the method is demonstrated through the analysis of some experimental results. The classification accuracy obtained using 900 samples of test data is 91.33%. It is here observed that the proposed method not only can accurately identify the dark spots on SAR images but also verify the ability of the proposed algorithm to classify unstructured features.

Highlights

Marine oil slicks occur during the extraction and transportation of crude oil
Area Under Curve (AUC) of Convolutional Neural Network (CNN) and Artificial Neural Networks (ANN) were over 90%, and the AUC of CNN is greater than
The current study is to discriminate oil slicks from lookalikes in polarimetric Synthetic Aperture Radar (SAR) images based on the use of multi-feature fusion and a proposed CNN approach

Summary

Introduction

Marine oil slicks occur during the extraction and transportation of crude oil. The development of marine transportation and ocean development technologies has increased the possibility of oil accidents. Several phenomena (e.g., plant oil, oil emulsion, etc.) produce dark regions in SAR images These are called lookalikes and appear quite similar to oil slicks. The texture features of single polarimetric SAR images cannot fully describe the physical characteristics of the sea surface targets, which may cause misjudgment during the oil slick detection [1]. An algorithm is presented for discriminating oil slicks from lookalikes in SAR images This algorithm is based on CNN and multi-feature fusion. The proposed algorithm is run in the following steps: detection of dark spots in SAR images, extraction of features, analysis and selection of features, selection of Regions Of Interest (ROI) in feature images, and classification of dark spots into oil slicks or lookalikes. The proposed algorithm can effectively apply CNN to the classification of oil slicks and lookalikes in SAR images. The sea water temperature was in the range of 10 °C–17 °C

JuneOil

Dark Spots

Feature

Texture Features

Polarimetric Features

ThreeK-fold

Parameter Analysis of Boxcar Filter

The images of alpha of the data of theofGulf

Introduction to CNN

Classification of Oil Slicks and Lookalikes Based on CNN

The result of classification based fusion with the input patch

Comparison