^{1}

^{2}

^{1}

^{2}

^{*}

^{1}

This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/3.0/).

This paper presents a new segmentation-based algorithm for oil spill feature extraction from Synthetic Aperture Radar (SAR) intensity images. The proposed algorithm combines a Voronoi tessellation, Bayesian inference and Markov Chain Monte Carlo (MCMC) scheme. The shape and distribution features of dark spots can be obtained by segmenting a scene covering an oil spill and/or look-alikes into two homogenous regions: dark spots and their marine surroundings. The proposed algorithm is applied simultaneously to several real SAR intensity images and simulated SAR intensity images which are used for accurate evaluation. The results show that the proposed algorithm can extract the shape and distribution parameters of dark spot areas, which are useful for recognizing oil spills in a further classification stage.

Oil spills from operational discharges and ship accidents always have calamitous impacts on the marine environment and ecosystems, even with small oil coverage volumes. Remote sensing solutions using space-borne or airborne sensors are playing an increasingly important role in monitoring, tracking and measuring oil spills and are receiving much more attention from governments and organizations around the world. Compared to airborne sensors, satellite sensors, with their large extent observation, timely data available and all weather operation, are proven to be more suitable for monitoring oil spills in marine environments, whilst the latter can be easily used to identify polluters and oil spill types but are of limited use due to costs and weather conditions [

Generally speaking, oil spill recognition includes three stages: dark spot detection, dark spot feature extraction and oil spill classification [

The paper is organized as follows: Section 2 describes in detail the algorithm developed in this paper. In Section 3, the proposed algorithm is tested by several simulated 4-looks SAR intensity images and applied to RADARSAT-1 SAR intensity images for oil spill extraction. Finally, Section 4 contains concluding remarks and perspectives for further research.

In a spatial statistic model, SAR backscatter energy can be characterized by a random field defined on a domain ^{2}, {_{i}_{i}, y_{i}_{i}, y_{i}_{i}, y_{i}_{i}_{1} and _{2} corresponding to the dark spot areas and its surroundings, respectively. To this end, _{j}_{j}_{j}_{j}_{j}_{j}_{j}_{i}_{i}, y_{i}_{j}_{i}, y_{i}_{j}_{1}, _{2}, _{1}, _{2}} be the set of parameters for Gamma distributions corresponding to dark spill regions and its surroundings, respectively. For a flexible and convenient tessellation, the Voronoi tessellation [_{j}, v_{j}_{j}, v_{j}_{j}, v_{j}_{j}, v_{j}_{j}_{j}_{j}, v_{j}_{j}, v_{j}

Given

Using a Bayesian paradigm, the inference about parameters {_{1})_{2})_{1})_{2}). Furthermore, assume that the scale (resp. shape) parameters are drawn from Gaussian distribution with mean _{β}_{α}_{β}_{α}_{β}σ_{β}_{α}_{α}_{j}_{j'}_{j}_{j}_{'}. For a polygon _{j}_{j}_{j}_{j}_{'}; _{j}_{'} ∼ _{j}_{j}_{j}_{'}; _{j}_{'} ∈ _{j}

Assume that the locations of generating points (_{j}, v_{j}

The posterior distribution defined in

In order to segment an SAR intensity image, it is necessary to simulate it from the posterior distribution in ^{*} for ^{*} = ^{*}(^{*} is higher than that of ^{*} with dimension matching, that is, |^{*}|. The appropriate acceptance probability for the proposed transition is given by:
^{*}) and ^{*} and ^{*})/∂(^{*}. The move types designed in this paper include: (1) updating Gamma distribution parameters; (2) updating the labels; (3) updating the positions of generating points; (4) birth or death of generating points.

_{k}, k_{k}_{k}, β_{k}_{k}^{*} and _{k}^{*} are Gaussian distributions with means _{k}_{k}_{αk}_{βk}_{k}^{*} ∼ _{k}., ε_{αk}_{k}^{*} ∼ _{k}, ε_{βk}_{k}^{*} and _{k}^{*} can be obtained by:
_{k}_{j}_{'} =

_{j}_{j}^{*}, which is uniformly drawn from {1, 2}. The acceptance probability for _{j}^{*} can be written as:
_{j}, v_{j}_{j}^{*}, _{j}^{*}) by drawing uniformly in _{j}_{j}_{j'}^{*}, NP_{j}^{*}}. The acceptance probability for the move turns out to be:

_{m}_{m}_{m}_{+1}, _{m}_{+1}) is drawn uniformly from _{m}_{+1}, _{m}_{+1}) be _{m}_{+1} and the set of labels of _{m}_{+1}'s neighbor polygons is _{m}_{+1}. The Voronoi tessellation is modified by the addition of this generating point from _{1}, …, _{j'}_{m}_{1}, …, _{j'}^{*}, …, _{m}, P_{m}_{+1}}. _{7}. It can be observed from _{7} include _{2}, _{4}, _{5} and _{6}, that is, _{7} = {2, 4, 5, 6}.

The new label _{m}_{+1} for polygon _{m}_{+1} is uniformly drawn from {1, 2}. It is evident that a birth or a death of generating point does not affect the Gamma distribution parameters in ^{*} = (^{*}, ^{*}, ^{*} = (_{1}, …, _{m}, L_{m}_{+1}), ^{*} = (_{m}_{+1}, _{m}_{+1})). The acceptance probability of the birth can be written as:
_{bm}_{m}, r_{dm}_{+1} = _{m}_{+1}/(_{m}_{+1}. The acceptance probability of a death of generating point is:

For any given proposal with acceptance probability

To estimate the parameter vector

To assess the accuracy of the proposed algorithm for extracting features of dark spots, two types of data, 4-looks SAR intensity images and simulated SAR intensity images which simulate 4-looks SAR intensity images, are used. It has been shown that multilook SAR intensity images can be modeled by Gamma distributions with fixed scale parameters, which are equal to the number of looks.

_{α}_{β}_{α}_{β}_{α}_{α}× μ_{α}_{α}_{β}^{*}^{t}, ε_{α}^{*}^{t}, ε_{β}^{*}^{*}^{t}^{t}

In this experiment, two evaluation schemes are used to assess the accuracy of extracted oil spill area quantitatively. First of all, some of the common measures are used for accuracy assessment, including error matrix, producer's accuracy, consumer's accuracy, overall and Kappa coefficient.

The entries in the matrix contain a count of pixels, which is based on the labels assigned to pixels in both the segmented image and the synthesizing images. For example, if a pixel is segmented to the oil spill region with the label 1 and actually belongs to the water region with label 2, it will be counted in the error matrix entry of column 1 and row 2, denoted by _{12}. The values of diagonal entries represent the count of correctly segmented pixels. The Table also lists row total (Σ_{r}_{s}

Except for the error matrix, the associated accuracy estimates are used, including the producer's accuracy, consumer's accuracy, overall accuracy and Kappa coefficient.

Another scheme for the accurate assessment of the developed segmentation algorithm is based on the degree to which the outlines of segmented homogeneous regions match their alternatives delineating the real regions, which is measured by the count of pixels of extracted outlines lying on the buffer zone around the real outlines of homogenous regions [

_{0} means the percent of the outlines exactly matching the outlines for the real oil spill regions in the synthesizing images. _{i}_{i}_{0} + _{1}+ … + _{i}

The results from above two accuracy analyses schemes manifest that the proposed algorithm extracts the shape and distribution parameters features of oil spill regions efficiently and effectively.

_{0} is drawn from the Poisson distribution with the mean 96 and the locations of _{0} generating points are drawn from _{j}_{0}}. It is found that there is no obvious impact of the initial segmentation on the finial segmentation result.

_{1,2} and _{1,2} corresponding to the segmented dark spot and sea regions.

It can be observed that the histograms of intensities in each segmented homogenous region are coincident with the Gamma distribution curves with

This paper presents a new segmentation-based approach to the feature extraction of oil spills from SAR intensity images, including their area and distribution parameters. The proposed segmentation algorithm is statistical region-based, which combines the Voronoi tessellation, Bayesian inference and reversible Jump Markov chain Monte Carlo (RJMCMC) methods. The Voronoi tessellation has been widely used to characterize models of many natural phenomena or processes in crystallography, metallography, physics, astrophysics, biology, ecology, geology, geography,

In this paper the technique is introduced to design a region-based segmentation algorithm for oil spill feature extraction. By region, it means that Voronoi tessellation is used to partition the image domain into sub-regions (polygons) corresponding to components of oil spill regions or their surroundings (waters). Therefore, the segmentation of SAR intensity image for the purposes of oil spill feature extraction is completed by labeling those polygons as oil spills or water and thus modeled as a label field. By statistical, it means that each region (oil spill or water) is statistically homogenous, which is characterized by a Gamma distribution. Under the Bayesian inference paradigm, the label field for segmentation and distributing parameter can be expressed as a posterior conditional on a given SAR intensity image. The RJMCMC

method is employed to simulate the conditional posterior distribution. The results of testing the proposed approach on both real and synthesizing SAR intensity images demonstrate that it can extract the area and distribution parameters for oil spills with high accuracy and is promising.

This work was supported by the Key Laboratory of Marine Oil Spill Identification and Damage Assessment Technology (No. 201211) and National Natural Science Foundation of China (No. 41271435; No. 41301479).

The authors declare no conflict of interest.

(_{1}–_{6} corresponding to generating points (_{1}, _{1})–(_{6}, _{6}); (_{1}–_{7} formed by adding the generating point (_{7}, _{7}).

Results of (

Changes of (

Histogram and curves of gamma distributions with real and estimated model parameters.

Results of (

Extracted outlines overlaid on the buffer zones around the outlines of real regions.

Real 4-looks SAR intensity images in which the dark areas are oil spills.

Final partition (

The histogram for two segmented homogeneous regions and gamma distribution curves with estimated parameters.

Constants used in the experiment.

_{α} |
_{α} |
_{β} |
_{β} |
_{α} |
_{β} |
|||
---|---|---|---|---|---|---|---|---|

96 | 1 | 4 | 1 | 32 | 8 | 0.5 | 1 | 4,000 |

Estimated model parameters and their errors.

_{α} |
_{β} | |||
---|---|---|---|---|

C1 | 4.04 | 1.0 | 27.37 | 2.25 |

C2 | 3.95 | 1.25 | 18.56 | 3.11 |

Error matrix and statistical measurements.

_{1} |
_{2} |
_{s} |
||
---|---|---|---|---|

_{1} |
39,908 | 1,448 | 41,356 | 94.13 |

_{2}. |
971 | 23,209 | 24,180 | 97.61 |

Σ_{r} |
40,879 | 24,657 | 65,536 | Overall accuracy (%) |

User's accuracy (%) | 95.98 | 96.95 | 0.92 | 96.3 |

Percentage of extracted outlines in buffer zones.

_{0}(%) |
_{1}/_{1}(%) |
_{2}/_{2}(%) |
_{3}/_{3}(%) |
_{4}/_{4}(%) |
---|---|---|---|---|

35.7 | 39.6/75.3 | 15.6/90.9 | 5.7/96.6 | 1.7/98.3 |

Estimated shape and scale parameters.

_{1} |
_{2} |
_{1} |
_{2} | |
---|---|---|---|---|

a | 5.00 | 3.22 | 28.27 | 23.11 |

b | 4.06 | 2.44 | 28.64 | 34.02 |

c | 4.72 | 2.67 | 31.02 | 32.44 |