Dark Spot Detection in SAR Images of Oil Spill Using Segnet

Abstract

Damping Bragg scattering from the ocean surface is the basic underlying principle of synthetic aperture radar (SAR) oil slick detection, and they produce dark spots on SAR images. Dark spot detection is the first step in oil spill detection, which affects the accuracy of oil spill detection. However, some natural phenomena (such as waves, ocean currents, and low wind belts, as well as human factors) may change the backscatter intensity on the surface of the sea, resulting in uneven intensity, high noise, and blurred boundaries of oil slicks or lookalikes. In this paper, Segnet is used as a semantic segmentation model to detect dark spots in oil spill areas. The proposed method is applied to a data set of 4200 from five original SAR images of an oil spill. The effectiveness of the method is demonstrated through the comparison with fully convolutional networks (FCN), an initiator of semantic segmentation models, and some other segmentation methods. It is here observed that the proposed method can not only accurately identify the dark spots in SAR images, but also show a higher robustness under high noise and fuzzy boundary conditions.

A Deep-Learning Framework for the Detection of Oil Spills from SAR Data

“…Recent deep-learning-based methods [9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28] that used CNN structures for both automated feature extraction, as well as classification of SAR images have relied on the use of patches to reduce the background concentration in the tested images. Pre-trained models, such as ResNet 101, VGG-16, and GAN networks as in [10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28] or multi-level CNN networks as in [9], were introduced to classify patches with modest performance (i.e., precision and Dice score). In [23], authors obtained improved results via introducing the use of VGG-16 and dark batch generation algorithm on spaceborne SAR images.…”

“…Further, Qia et al proposed a 3D model for short-term and long-term oil spill paths caused by the Sanchi tanker [19]. Guo et al introduced the use of SegNet to segment oil spills represented as dark spots in SAR images [14] with an accuracy of 93% under high noise. Further, Li et al introduced the use of polarimetric SAR filters (e.g., Boxcar, Refined Lee, and Lopez filters) to extract respective polarimetric SAR features and feed the features to a stacked autoencoder [15].…”

Special Features on Intelligent Imaging and Analysis

“…Oil spill detection on the ocean surface has been a hot issue, and a refined technique is required to detect dark spots on SAR (synthetic aperture radar) images. A deep convolution neural network, called Segnet has been tested for this application by Guo et al [8]. It has the basic framework of encoder and decoder for image semantic segmentation.…”

Instance Segmentation for Large, Multi-Channel Remote Sensing Imagery Using Mask-RCNN and a Mosaicking Approach

“…This approach achieves state-of-the-art results in different applications in remote sensing digital image processing [5]: pan-sharpening [6][7][8][9]; image registration [10][11][12][13], change detection [14][15][16][17], object detection [18][19][20][21], semantic segmentation [22][23][24][25], and time series analysis [26][27][28][29]. The classification algorithms applied in remote sensing imagery uses spatial, spectral, and temporal information to extract characteristics from the targets, where a wide variety of targets show significant results: clouds [30][31][32][33], dust-related air pollutant [34][35][36][37] land-cover/land-use [38][39][40][41], urban features [42][43][44][45], and ocean [46][47][48][49], among others.…”