Reference spectral signature selection using density-based cluster for automatic oil spill detection in hyperspectral images

Delian Liu; Jianqi Zhang; Xiaorui Wang

doi:10.1364/OE.24.007411

1. Introduction

Oil spill occurs frequently because of drilling accident, transportation leakage, and natural leaking from ocean floor, etc. According to the European Space Agency (1998), 45% of the oil pollution comes from operative discharges from ships [1]. Especially, after the BP Deepwater Horizon oil spill [2, 3], great attentions are attracted from all around the world about the destructive effect of oil spill [4]. Hence, the detecting of oil spills becomes an urgent and desired work.

Currently, more attentions are focused on the use of satellite-based synthetic aperture radar (SAR) images to detect oil spills [5, 6]. In SAR images, the capillary waves backscattering of oil spill is different from that of sea background. When oil is spilled on the sea surface, it forms a thin layer on the sea surface. This layer decreases microwaves and thereby generates dark areas in the SAR images [7, 8]. The merit of SAR images is that they have very high ground resolution, which can as high as that of panchromatic remote sensing images. But SAR images may also have dark areas that are generated by other floating materials, which will lead to high false alarm of detection [9, 10]. Moreover, the pattern of oil spill may have very different features under different wind speed, and it is difficult to model. In addition, SAR remote sensing cannot find oil spills as wind speed is under or up a specific value [1].

Optical remote sensing is believed to have good performance in detecting oil spills under various conditions. Otremba et al. did extensive fundamental works about the optical characteristics of oil spill in seawater [11–15]. Lu et al. propose remarkable approaches to estimate oil slick thickness by optical remote sensing [16–19]. Hu et al. present excellent methods to detection oil spills by using visible infrared imaging radiometer (VIIRS) [20–22]. Among the airborne or satellite-borne remote sensing sensors, hyperspectral sensor acquires both the spatial and spectral information of a pixel, having high ability in discriminating different objects [23,24]. Hence, it has outstanding performance in detecting oil spills from seawater background [25–27]. To detect oil spill pixels from hyperspectral images, a reference spectral signature should be obtained at first. Unfortunately, the selection of reference spectral signature from hyperspectral images is still a challenge issue. In [28], Clark et al. measure the reflectance of oil in laboratory that is collected from the sea after oil spill, and utilize the measured spectral reflectance as a reference spectral signature. The reference spectral signature is then matched with the spectra of hyperspectral images that are corrected by atmosphere model. However, the measured spectral signature may be different from the spectra that are corrected by atmosphere correction model. Because atmosphere effect is so complicated that it cannot be modeled accurately. In some strong absorption bands, the compensated spectra have large deviation from its original spectral reflectance. Therefore, it is reasonable to extract reference spectral signatures directly from hyperspectral images. In [29, 30], the researchers use an un-mixing method to obtain the spectral signature of oil spills from hyperspectral images. The drawback of the approach is that the number of end-members should be given first [31, 32]. Moreover, after the un-mixing procedure the extracted end-members must be tested manually to examine which spectral signature comes from oil spills.

Recently, Rodriguez and Laio proposed a density-based cluster algorithm, published on Science [33], which can find clusters automatically and shows great adaptation under extreme conditions. This algorithm characterizes elements by their density and relative distance. The elements with high density and large relative distance are spotted as cluster centers. And their neighbors are clustered to the nearest cluster center automatically. The uniquely and profoundly important advantages of the algorithm are its simpleness and high efficiency. The algorithm is immediately applied in many practical applications, resulting in remarkable fruitions and much higher performance than conventional cluster algorithms [34–36]. For oil spill pixels, they close to each other in spectral domain, and are far away from the seawater pixels. This characteristic fits the requirement of the density-based cluster. Therefore, we introduce the density-based cluster algorithm to select a reference spectral signature from a hyperspectral image. After obtained the reference spectral signature of oil spill, the detection of oil spills can be achieved easily by introducing the framework of target detection in hyperspectral images.

This paper is organized as follows. In section 2, the framework of oil spill detection is given. In Section 3, the density-based cluster algorithm is modified to meet the requirement of oil spill detection. The background estimation procedure is shown in section 4. The execution of the proposed approach is illustrated in Section 5. In Section 6, the proposed approach is evaluated and discussed. Finally, the conclusion is provided in Section 7.

2. Framework of oil spill detection from hyperspectral images

The main task of an oil detection algorithm is to determine whether a pixel under test is an oil pixel or a seawater pixel. The mathematical framework of oil spill detection algorithm can be applied by that of target detection algorithms, which is primarily based on binary hypothesis testing. The optimum decision strategy of the binary hypothesis testing is to maximize the probability of detection (PD) while keeping the probability of false alarm (PFA) under a fixed value, which is known as the Neyman-Pearson criterion and is embodied in the likelihood ratio test [37,38].

Λ (x) = \frac{f (x; θ_{1} | H_{1} = oil)}{f (x; θ_{0} | H_{0} = seawater)} ≷_{H_{0}}^{H_{1}} η

As shown in Eq. (1), the probability of observing x under the null hypothesis is f (x|H₀), and the probability of observing x under the alternative hypothesis is f (x|H₁). The desired PFA is achieved by setting the threshold η to an appropriate level.

For target detection, the parameters used for conventional algorithms are commonly the reference spectral signature of targets, mean and covariance of background, and target detection algorithms. Thus, for oil spill detection, it is necessary to estimate the reference spectral signature of oil spills, the mean vector and covariance of seawater background, and an appropriate detection algorithm.

3. Extraction of reference spectral signature of oil spill

3.1. Density-based cluster algorithm

Density-based cluster algorithm is a very simple and high efficient cluster algorithm, having excellent adaption in many applications. It is proposed based on the idea that cluster centers have a higher density than their neighbors, and the cluster centers are isolated from each other. It can automatically identify cluster centers, and assign every element to its closest center. Density-based cluster algorithm only depends on the distance between each two elements. Since oil spill and seawater have different spectral signatures, a proper distance between each two spectral signatures can be designed to make that the reference spectral signature of oil spills has the highest density in all oil spill spectral signatures.

In hyperspectral images, the reference spectral signature of oil spill appears on the vertex of the convex polygon in spectral space. It cannot fit the requirement of density-based cluster algorithm, a new distance measurement should be designed to make the reference spectral signature close to all spectral signatures of the oil spill pixels.

Let x(m,n) = [x(m,n,1),x(m,n,2),⋯,x(m,n,k),⋯,x(W,H,N_d)] denote a spectral signature of a pixel in position (m,n). N_d denotes the number of spectral channels. The width and height of a hyperspectral image are W and H respectively. Let i = nW + m, the distance of two spectral signatures is designed as follows

d (i, j) = \frac{x (i) \cdot x (j)}{‖ x (i) ‖ ‖ x (j) ‖}

This definition is based on the assumption that the reference spectral signature of oil spill has small distance from the oil contaminated pixels and has large distance from the seawater pixels.

In [33], Rodriguez and Laio define a local density ρ(i)

ρ (i) = \sum_{j} χ [(d (i, j)]

Where χ(x) is

χ (x) = {\begin{matrix} 1, & x < d_{c} \\ 0, & otherwise \end{matrix}

d_c is called cutoff distance. It is found by sorting all the distances. The distance that poses on the 2% of the distances is the cutoff distance. For practical applications, the authors suggest that

χ (x) = \exp [- {(\frac{x}{d_{c}})}^{2}]

In [33], Rodriguez and Laio also define another parameter called δ (i). It is measured by computing the minimum distance between element i and any other elements with higher density:

δ (i) = \min_{j : ρ (j) > ρ (i)} [d (i, j)]

For the element that has highest density, δ (i) = maxj[d(i, j)]. δ (i) is much larger than the typical nearest neighbor distance only for elements that are local or global maxima in the density. Therefore, cluster centers are recognized as elements for which the value of δ (i) is anomalously large.

3.2. Down-sampling hyperspectral scene

The density-based cluster algorithm is designed based on the distances of each two elements. For a hyperspectral image with the size of W × H, the distances, required to be calculated, is W ×H ×(W ×H −1)/2. Usually, the size of a hyperspectral image is very large, so the number of distances required to be calculated is much larger. Suppose the size of a hyperspectral image is 800 × 400, the number of distances required to be calculated is 1.152 × 10¹¹. This is a very large number, which will cost much time to process the cluster procedure.

To save time consuming, we first down-sample hyperspectral images to reduce the number of distances required to calculate, and extract a rough reference spectral signature in low resolution. Then refine the rough result to the original resolution.

For a hyperspectral image, it is down-sampled to a small size. Let x(m_d,n_d) denotes a spectral signature in low resolution.

x (m_{d}, n_{d}) = \frac{1}{W_{d}^{2}} \sum_{l_{m} = 1, l_{n} = 1}^{W_{d}, W_{d}} x (m + l_{m}, n + l_{n})

Where m_d = ⌊m/W_d⌋, n_d = ⌊n/W_d⌋, (m_d,n_d) denotes pixel position in low resolution. W_d is the width of down-sample window. It is suggested that down-sampling a hyperspectral scene to the size of 2000 pixels is suitable.

3.3. Band feature extraction

The density-based cluster algorithm defines a density and a delta parameter to find cluster centers. The delta parameter represents the gap of different clusters. However, oil diffuses in seawater, thus the transition region of oil and seawater is not clear, which makes the delta parameter becomes invalid. To address the issue, we replace the delta parameter with a new designed parameter based on the spectral features of oil spill. Fig. 1 is the spectral signature of oil spill, cloud, and seawater.

Fig. 1 Reflectance of oil, cloud, and seawater.

Download Full Size | PDF

As shown in Fig. 1, oil spill has some obvious features in 1.2μm and 1.73μm that are generated by C-H bond. Whereas, cloud has high reflectance in both visible and infrared bands. Seawater only has high reflectance before 0.8μm. In [39], Clark et al. proposed a band feature extraction method that has been shown excellent performance in identifying planet materials. In this paper, the band feature extract method is introduced to extract the band features of oil spectral signatures.

Suppose the spectral signature of oil in the selected band is denoted by x_bo(k). The continuum of x_bo(k) is denoted by x_cbo(k), as shown in Fig. 1. The continuum-removed oil spectral signature in the band is

x_{n b o} (k) = \frac{x_{b o} (k)}{x_{c b o} (k)}

Similarly, a continuum-removed observed spectral signature in the band is

x_{n b} (k) = \frac{x_{b} (k)}{x_{c b} (k)}

where x_cb(k) is the continuum of the observed spectral signature in the band.

A strength factor of the oil spectral signature and the observed spectral signature is

b = \frac{\sum x_{n b} x_{n b o} - (\sum x_{n b} \sum x_{n b o}) / N_{f}}{\sum x_{n b o}^{2} - {(\sum x_{n b o})}^{2} / N_{f}}

where N_f denotes the number of spectral channels in the feature band. Another strength factor is

b_{g} = \frac{\sum x_{n b} x_{n b o} - (\sum x_{n b} \sum x_{n b}_{o}) / N_{f}}{\sum x_{n b}^{2} - {(\sum x_{n b})}^{2} / N_{f}}

Thus, the correlation coefficient of the oil spectral signature and the observed spectral signature is

α = b b_{g}

Let the slope of the continuum of the observed spectral signature is denoted by k_s, we defined a slope factor

β = \frac{1}{1 + {(\frac{k_{s} - k_{o s}}{k_{t}})}^{4}}

where k_os is the slope of the continuum of the oil spectral signature in the feature band. k_t is the cutoff slope of the continuum of oil spectral signature.

Therefore, the band feature of an observed spectral signature is

f_{m} = α β

For a spectral signature, the integral band feature f_b can be calculated by multiplying all the band feature in each selected spectral band.

f_{b} = \prod_{m} f_{m}

where f_b ≤ 1. The larger the f_b is, the larger probability the spectral signature is an oil spectral signature.

For the Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) hyperspectral images, the selected bands are 1.2μm and 1.73μm.

Therefore, a decision graph can be generated by using the density and band feature of a spectral signature. As indicated in Eq. (14), if a spectral signature is an oil spectral signature, the f_m of the spectral signature is large. Combing with the characteristic of ρ(i), the oil spectral signatures appear on the top right of the decision graph. In contrast, the points of seawater appear on the bottom left of the decision graph. Fig. 2 is a decision graph.

Fig. 2 A decision graph and its corresponding spectral signature.

Download Full Size | PDF

As shown in Fig. 2, the points that have large density and large f_b correspond to the spectral signatures that have obvious absorption bands which are formed by oil spill. So the point on the top right of the decision graph can be extract as a rough reference spectral signature.

3.4. Identification of spectral signature of oil spill

Normalize ρ(i)

ρ_{n} (i) = \frac{ρ (i) - ρ_{\min}}{ρ_{\max} - ρ_{\min}}

where ρ_max and ρ_min are the maximum and minimum of ρ(i), respectively.

The point on the top right corner of the decision graph can be identified as reference spectral signature.

In [33], Rodriguez and Laio suggest a integrated parameter to identify cluster centers by multiplying ρ and δ. Thus

f_{c} (i) = ρ_{n} (i) f_{b}

f_{m} = max f_{c} (i)

f_{o s} = {\begin{matrix} 1, & f_{m} \geq τ_{s p} \\ 0, & f_{m} < τ_{s p} \end{matrix}

The pixel that f_os corresponds is the selected pixel in low resolution.

To select the spectral signature in original resolution, the pixel that is selected in the low resolution is enlarged to the original resolution. It will correspond to some pixels in original resolution. The final reference spectral signature is extracted by calculating the density of the pixels in original resolution. The selected pixel in original resolution is

ρ_{m} = \max_{l_{m} = 1, l_{n} = 1}^{W_{d}, W_{d}} [ρ (m_{f} + l_{m}, n_{f} + l_{n})]

Where m_f = m_l ×W_d, n_f = n_l ×W_d, (m_l,n_l) is the selected pixel position in the low resolution. The pixel that has the largest density is the extracted pixel. The spectral signature that the pixel corresponds is the final selected reference spectral signature.

4. Background parameter estimation

As discussed in Section 2, the detection of oil spill requires the parameters of background. In many target detection algorithms, the parameters of background are estimated using a whole hyperspectral image, the disturbance of targets is neglected. This does not lead to large deviations, because targets only occupy a few pixels in a hyperspectral image. However, for oil spill detection, oil spills may cover the large part of a hyperspectral scene. Using the whole scene to estimate the parameters of background will lead to large deviations. Hence, the estimation of background parameters should be restricted on the seawater regions.

As shown in Fig. 1, the reflectance of seawater is much smaller than that of clouds and oil in 1.0-2.5μm bands. Thus, the seawater can be segmented by thresholding reflectance in the bands

r_{b} = \frac{1}{K_{r} - K_{l} + 1} \sum_{k = K_{l}}^{K_{r}} x (k)

l_{b} = {\begin{matrix} 1, & r_{b} \geq τ_{r} \\ 0, & r_{b} < τ_{r} \end{matrix}

where τ_r is set to be about 0.1. Here, the threshold should be select as low as possible. This can make the segmented region are only water regions, even the transition regions of oil and seawater are also excluded.

After thresholding, the parameter of seawater regions can be calculated.

m_{b} = \frac{1}{N_{b}} \sum_{l_{b} (i) = 1} x (i)

C_{b} = \frac{1}{N_{b} - 1} \sum_{l_{b} (i) = 1} [x (i) - m_{b}] {[x (i) - m_{b}]}^{T}

where N_b denotes the number of pixels of seawater background.

In marine environment, seawater regions occupy the main part of a hyperspectral image, it is easy to segment seawater regions from a hyperspectral image.

5. Oil spill detection

The oil detection algorithm used here is borrowed from the target detection algorithms that are designed and evaluated in the last two decades. In [38, 40, 41], Manolakis et al evaluated many conventional target detection algorithms and suggested that the adaptive cosine estimator (ACE) algorithm [42,43] had better detection performance than the other evaluated algorithms. In this paper, we introduce the ACE algorithm for the oil spill detection.

Every pixel in a hyperspectral image relates to an area of sea surface. The reflectance of that pixel is the function of the reflectance of the materials in the area. This process is denoted by the very famous model called linear mixture model. The model assumes that the reflectance of a pixel is linearly mixed with the reflectance of the materials covered by the pixel. A reasonable assumption is that the abundance of each material is approximately the fraction of the pixel that the material occupies.

x = \sum_{p = 1}^{N_{p}} α_{p} s_{p}

where α_p denotes the abundant of the pth materials. This model means that the α_p must satisfy two constraints for all pixels:

\sum_{p = 1}^{N_{p}} α_{p} = 1

α_{p} \geq 0

This two equation are known as the sum-to-one and non-negativity constraints.

Consider a pixel that is completely filled by a target material. Under the mixing constraints, the reflectance of the pixel are generated by a target. This type of target is called a full pixel target. Whereas a pixel that is not completely filled by a target material. This type of target is called a subpixel target. The pixel contains some combination of the target signature and other material signatures.

Based on the discussion in Section 2, the detection of oil spill can be processed as a hypothesis testing problem where there are two competing hypotheses [42,43].

{\begin{cases} H_{0} : x = n_{b} \\ H_{1} : x = s α + σ n_{b} \end{cases}

Here σ denotes the amount of background covered area in H₁ hypotheses. n_b corresponds to the background modeled as a multivariate normal distribution with zero mean and covariance matrix

n_{b} \sim N (0, C_{b})

Consider s an N_d dimension vector (lexicographically ordered) of reference spectral signature. The test statistic of ACE algorithm is [42,43]

T_{A C E}^{2} = \frac{{(s^{T} C_{b}^{- 1} x)}^{2}}{(s^{T} C_{b}^{- 1} s) (x^{T} C_{b}^{- 1} x)}

If $T_{A C E}^{2} > η$ , the pixel under test is oil spill, otherwise background. Where η is determined by probability of false alarm. C_b is the covariance of seawater background. To fit the requirement of Eq. (29), a demean process should be executed before applying of Eq. (30) [37,38].

After extracted the reference spectral signature, each spectral signatures of a hyperspectral image is tested by using Eq. (30) to determine whether it comes from oil or not.

The procedure of our approach is executed as shown in Fig. 3.

For a hyperspectral image, it is first down-sampled to a small size of between 2000 and 5000 pixels. This number is not a necessary requirement. It is determined by the performance of the used computer. The down-sampling process is to reduce the number of spectral signatures to make the following clustering process can be applied.
The improved density-based cluster is applied on the down-sampled image. Distances of each two spectral signatures of the down-sampled image are calculated using Eq. (2), with which the density parameter is obtained. The physical feature parameter of every spectral signature is calculated by using Eq. (14). Using the density and physical feature parameter, a decision graph is generated. The point on the top right of the graph is selected as the rough result.
After clustering, a rough spectral signature is identified. And the corresponding pixel is found. The down-sampled image is enlarged to the original resolution. Thus the identified pixel relates to some pixels in original resolution. All related pixels are tested using the density parameter, the pixel that has the largest density is the final selected pixel, hence the final selected spectral signature is identified.
A reflectance threshold for background segmentation is estimated according to the physical property of clear water. To ensure the segmented regions only come from seawater, the reflectance threshold should be selected as small as possible.
The reflectance of each pixel is calculated by averaging reflectance in bands between 1.5μm and 2.5μm. If the reflectance of the pixel is below the reflectance threshold, the pixel is labeled 1, or else it is labeled 0.
For all pixels that are labeled 1, we calculate their mean vector and covariance matrix.
Use the selected spectral signature as prior signal, substitute the reference spectral signature and the covariance matrix into Eq. (30), the oil spill detection result can be obtained.

Fig. 3 Execution of the proposed reference spectral signature selection approach.

Download Full Size | PDF

6. Experimental result and discussion

To validate our proposed reference spectral signature selection approach, AVIRIS data is used as test samples. AVIRIS is a proven instrument in the realm of earth remote sensing. It is a unique optical sensor that delivers calibrated images of the upwelling spectral radiance in 224 contiguous spectral channels with wavelengths from 400 to 2500 nanometers [44]. In this paper, six hyperspectral subimages are select to test the performance of the proposed approach. Four hyperspectral subimages have oil spills that were captured by AVIRIS after Gulf oil spill on May 18, 2010, overflying on the ER-2 aircraft at an altitude of 9,000 m. Two hyperspectral subimages has no oil spills that are also selected to test the reliability of our approach. The data is downloaded from the website of Jet Propulsion Laboratory. The six hyperspectral images are shown in false color generated using the 5th, 24th, 38th band of the images. The AVIRIS radiance data were converted to surface reflectance data using the FLAASH model in ENVI software. The data in strong absorption bands is discarded. The false color images of the data are shown in Fig. 4.

Fig. 4 False color images of test hyperspectral images. (a) Scene 1, (b) Scene 2, (c) Scene 3, (d) Scene 4, (e) Scene 5, (f) Scene 6.

Download Full Size | PDF

As shown in Fig. 4(a), oil spills spread in seawater on the top of the scene. The bright areas relates to the high content of oil. The dark areas are the transition of oil spills and seawater. In Fig. 4(b), there are some clouds in the scene. This is a disturbance for oil spill detection. In Fig. 4(c), the oil spills are covered by heavy smoke, which is a strong interference for oil spill detection. Figure 4(d) has strong sun glint belt on the top of the scene. Fig. 4(e) only has clear water and seashore. Figure 4(f) has heavy algae bloom along the island. Without prior information of oil, it is difficult to detection oil spills from the hyperspectral images.

The scenes were first down-sampled to the 1/64 of original resolution to apply the improved density-based cluster algorithm. After down-sampling, the number of pixels in the six images becomes to 5000. This can highly reduce the burden of the following clustering procedure. We then calculate the distances of each two spectral signatures. By using the distances, the density parameter is obtained. Next, the band feature of oil spill is calculated. By using the density and band feature, decision graphs are generated. Rough spectral signatures can be extracted from the top right of the decision graphs. The decision graphs and the selected points are shown in Fig. 5.

Fig. 5 Decision graphs and selected points. (a) Scene 1, (b) Scene 2, (c) Scene 3, (d) Scene 4, (e) Scene 5, (f) Scene 6.

Download Full Size | PDF

The red points in Figs. 5(a)–5(d) are the selected points in low resolution. They are on the top right of the decision graph and are distinct from the others points, which make them easily to extract. While, Fig. 5(e)–5(f) have points only on the bottom left of the graphs. By using the integral equation, no point are extracted as oil spectral signature. Although there are difference interferences in the scenes 1-6, our approach can extract the true oil spill spectral signature. The extracted oil pixels are indicated by a white arrow in Fig. 4. The selected oil spectral signatures are shown in Fig. 6.

Fig. 6 Selected spectral signatures. (a) Scene 1, (b) Scene 2, (c) Scene 3, (d) Scene 4.

Download Full Size | PDF

The oil spectral signatures are the solid line in Fig. 6. Their waveform are similar to that of the oil spectral signature in Fig. 1. Both of them have obvious absorption bands in 1.2μm and 1.73μm. This results indicated the selected pixels are the oil pixels in the scenes.

To estimate the parameters of background, the seawater regions are segmented by using the radiance in bands of 1.0μm–2.5μm. The regions that have oil spills and clouds are excluded from the scene. Seawater are labeled as background. The parameters of background are calculated by using the segmented areas. We next use ACE detector to detect oil spills. The results are shown in Fig. 7

Fig. 7 Oil spill detection results. (a) Scene 1, (b) Scene 2, (c) Scene 3, (d) Scene 4.

Download Full Size | PDF

Fig. 7(a) shows that the oil spills on the top of the scene have high response. The thicker the oil spill is, the higher the response is. In contrast, seawater background has low response. The result in Fig. 7(b) is similar to that in Fig. 7(a). Figure 7(c) only has high response on the thick oil spills, because oil spills are covered by thick smoke, which severely reduce the detection performance. Figure 7(d) has high response on the oil spills area, while the sun glint areas have low response. The results show the proposed reference spectral signature selection approach has high performance in extracting oil spectral signatures. By using a proper threshold, the oil spill area can be obtained.

7. Conclusion

In this paper, we propose an reference spectral signature selection approach for automatic oil spill detection. Based on the framework of target detection algorithm in hyperspectral image. We first employ the density-based cluster to select specific spectral signature from hyperspectral images. We then estimate the parameter of background according to the reflectance of seawater in the infrared band. And the conventional ACE algorithm is adopted to achieve oil spill detection. Finally, the proposed approach is tested via two practical hyperspectral images that are collected during the Horizon Deepwater oil spill. The experimental results show that our new approach can automatically detect oil spill from hyperspectral images and has good detection performance.

Acknowledgments

This work is supported by the National Natural Science Foundation of China (61301290) and the Fundamental Research Funds for the Central Universities (NSIY151410, NSIZ011401).

References and links

1. C. Brekke and A. H. Solberg, “Oil spill detection by satellite remote sensing,” Remote Sensing of Environment 95(1), 1–13 (2005). [CrossRef]

2. I. Mezić, S. Loire, V. A. Fonoberov, and P. Hogan, “A new mixing diagnostic and Gulf oil spill movement,” Science 330(6003), 486–489 (2014). [CrossRef]

3. I. Leifer, W. J. Lehr, D. Simecek-Beatty, E. Bradley, R. Clark, P. Dennison, Y. Hu, S. Matheson, C. E. Jones, B. Holt, M. Reif, D. A. Roberts, J. Svejkovsky, G. Swayze, and J. Wozencraft, “State of the art satellite and airborne marine oil spill remote sensing: Application to the BP Deepwater Horizon oil spill,” Remote Sensing of Environment 124, 185–209 (2012). [CrossRef]

4. G. E. Machlis and M. K. McNutt, “Scenario-building for the deepwater horizon oil spill,” Science 329(5995), 1018–1019 (2010). [CrossRef] [PubMed]

5. A.-B. Salberg, O. Rudjord, and A. H. S. Solberg, “Oil spill detection in hybrid-polarimetric SAR images,” IEEE Transactions on Geoscience and Remote Sensing 52(10), 6521–6533 (2014). [CrossRef]

6. O. Garcia-Pineda, I. MacDonald, C. Hu, J. Svejkovsky, M. Hess, D. Dukhovskoy, and S. L. Morey, “Detection of Floating Oil Anomalies From the Deepwater Horizon Oil Spill With Synthetic Aperture Radar,” Oceanography 26(2), 124–137 (2013). [CrossRef]

7. Y. Li and J. Li, “Oil spill detection from SAR intensity imagery using a marked point process,” Remote Sensing of Environment 114, 1590–1601 (2010). [CrossRef]

8. O. Garcia-Pineda, I. MacDonald, X. Li, C. Jackson, and W. Pichel, “Oil spill mapping and measurement in the Gulf of Mexico with textural classifier neural network algorithm (TCNNA),” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 6(6), 2517–2525 (2013). [CrossRef]

9. Y. Shu, J. Li, H. Yousif, and G. Gomes, “Dark-spot detection from SAR intensity imagery with spatial density thresholding for oil-spill monitoring,” Remote Sensing of Environment 114, 2026–2035 (2010). [CrossRef]

10. Y. Li, Y. Zhang, J. Chen, and H. Zhang, “Improved compact polarimetric SAR quad-pol reconstruction algorithm for oil spill detection,” IEEE Geoscience and Remote Sensing Letters 11(6), 1139–1142 (2014). [CrossRef]

11. Z. A. Otremba, “Modeling of the light transfer in a water column polluted with oil suspension,” Journal Of The European Optical Society-Rapid Publications 8, 13067 (2013). [CrossRef]

12. Z. Otremba, O. Zielinski, and C. Hu, “Optical contrast of oil dispersed in seawater under windy conditions,” Journal Of The European Optical Society-Rapid Publications 8, 13051 (2013). [CrossRef]

13. Z. Otremba, “Oil droplets as light absorbents in seawater,” Optics Express 15(14), 8592–8597 (2007). [CrossRef] [PubMed]

14. Z. Otremba and J. Piskozub, “Modelling the bidirectional reflectance distribution function (BRDF) of seawater polluted by an oil film,” Optics Express 12(8), 1671–1676 (2004). [CrossRef] [PubMed]

15. Z. Otremba and J. Piskozub, “Modeling the remotely sensed optical contrast caused by oil suspended in the sea water column,” Optics Express 11(1), 2–6 (2003). [CrossRef] [PubMed]

16. Y. Lu, X. Li, Q. Tian, G. Zheng, S. Sun, Y. Liu, and Q. Yang, “Progress in Marine Oil Spill Optical Remote Sensing: Detected Targets, Spectral Response Characteristics, and Theories,” Marine Geodesy 36(3), 334–346 (2013). [CrossRef]

17. Y. Lu, Q. Tian, X. Wang, G. Zheng, and X. Li, “Determining oil slick thickness using hyperspectral remote sensing in the Bohai Sea of China,” International Journal Of Digital Earth 6(1), 76–93 (2013). [CrossRef]

18. Y. Lu, X. Li, Q. Tian, and W. Han, “An optical remote sensing model for estimating oil slick thickness based on two-beam interference theory,” Optics Express 20(22), 24496–24504 (2012). [CrossRef] [PubMed]

19. Y. Lu, Q. Tian, and X. Li, “The remote sensing inversion theory of offshore oil slick thickness based on a two-beam interference model,” Science China-Earth Sciences 54(5), 678–685 (2011). [CrossRef]

20. J. Zhao, M. Temimi, H. Ghedira, and C. Hu, “Exploring the potential of optical remote sensing for oil spill detection in shallow coastal waters-a case study in the Arabian Gulf,” Optics Express 22(11), 13755–13772 (2014). [CrossRef] [PubMed]

21. C. Hu, S. Chen, M. Wang, B. Murch, and J. Taylor, “Detecting surface oil slicks using VIIRS nighttime imagery under moon glint: a case study in the Gulf of Mexico,” Remote Sensing Letters 6(4), 295–301 (2015). [CrossRef]

22. M. Wang and C. Hu, “Extracting Oil Slick Features From VIIRS Nighttime Imagery Using a Gaussian Filter and Morphological Constraints,” IEEE Geoscience And Remote Sensing Letters 12(10), 2051–2055 (2015). [CrossRef]

23. N. U. Chowdhury, A. A. Sakla, and M. S. Alam, “Oil spill detection in ocean environment via ultrasonic imaging and spectral fringe-adjusted joint transform correlation,” Optical Engineering 52(8), 083109 (2013). [CrossRef]

24. R. F. Kokaly, B. R. Couvillion, J. M. Holloway, D. A. Roberts, S. L. Ustin, S. H. Peterson, S. Khanna, and S. C. Piazza, “Spectroscopic remote sensing of the distribution and persistence of oil from the Deepwater Horizon spill in Barataria Bay marshes,” Remote Sensing of Environment 129, 210–230 (2013). [CrossRef]

25. R. N. Clark, G. A. Swayze, I. Leifer, K. E. Livo, R. Kokaly, T. Hoefen, S. Lundeen, M. Eastwood, R. O. Green, N. Pearson, C. Sarture, I. McCubbin, D. Roberts, E. Bradley, D. Steele, T. Ryan, and R. Dominguez, “A method for quantitative mapping of thick oil spills using imaging spectroscopy,” Open-File Report1020–1167, U.S. Department of the Interior, U.S. Geological Survey (2010).

26. R. F. Kokaly, T. M. Hoefen, K. E. Livo, G. A. Swayze, I. Leifer, I. B. McCubbin, M. L. Eastwood, R. O. Green, S. R. Lundeen, C. M. Sarture, D. Steele, T. Ryan, E. S. Bradley, and D. A. Roberts, “A rapid method for creating qualitative images indicative of thick oil emulsion on the oceans surface from imaging spectrometer data,” Open-File Report1020–1107, U.S. Departmentof the Interior, U.S. Geological Survey (2010).

27. C. Hu, X. Li, W. G. Pichel, and F. E. Muller-Karger, “Detection of natural oil slicks in the NW Gulf of Mexico using MODIS imagery,” Geophysical Research Letters 36(1), L01604 (2009). [CrossRef]

28. R. S. Rand, R. N. Clark, and K. E. Livo, “Feature-based and statistical methods for analyzing the Deepwater Horizon oil spill with AVIRIS imagery,” Proc. SPIE 8158, 81580N (2011). [CrossRef]

29. D. Sykas, V. Karathanassi, C. Andreou, and P. Kolokoussis, “Oil spill thickness estimation using unmixing methods,” in Proceedings of 3rd Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing, (2011), pp. 1–4.

30. P. Sidike, J. Khan, M. Alam, and S. Bhuiyan, “Spectral unmixing of hyperspectral data for oil spill detection,” Proc. SPIE 8498, 84981B (2012). [CrossRef]

31. X. Li, X. Jia, L. Wang, and K. Zhao, “On spectral unmixing resolution using extended support vector machines,” IEEE Transactions on Geoscience and Remote Sensing 53(9), 4985–4996 (2015). [CrossRef]

32. X. Jia and L. Wang, “Fuzzy assessment of spectral unmixing algorithms,” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 7(6), 1947–1955 (2014). [CrossRef]

33. A. Rodriguez and A. Laio, “Clustering by fast search and find of density peaks,” Science 344(6191), 1492–1496 (2014). [CrossRef] [PubMed]

34. D. Yu, X. Ma, Y. Tu, and L. Lai, “Both piston-like and rotational motions are present in bacterial chemoreceptor signaling,” Scientific Reports 5, 8640 (2015). [CrossRef] [PubMed]

35. H. Yu and Y.-S. Lin, “Toward structure prediction of cyclic peptides,” Physical Chemistry Chemical Physics 17(6), 4210–4219 (2015). [CrossRef] [PubMed]

36. K. Sun, X. Geng, and L. Ji, “Exemplar component analysis: A fast band selection method for hyperspectral imagery,” IEEE Geoscience and Remote Sensing Letters 12(5), 998–1002 (2015). [CrossRef]

37. D. Manolakis, D. Marden, and G. A. Shaw, “Hyperspectral image processing for automatic target detection applications,” Lincoln Laboratory Journal 14(1), 79–116 (2003).

38. D. Manolakis and G. Shaw, “Detection algorithms for hyperspectral imaging applications,” IEEE Signal Processing Magazine 19(1), 29–43 (2002). [CrossRef]

39. R. N. Clark, G. A. Swayze, K. E. Livo, R. F. Kokaly, S. J. Sutley, J. B. Dalton, R. R. McDougal, and C. A. Gent, “Imaging spectroscopy: Earth and planetary remote sensing with the usgs tetracorder and expert systems,” Journal of Geophysical Research: Planets (1991–2012) 108(E12) 5131 (2003). [CrossRef]

40. D. G. Manolakis, “Overview of algorithms for hyperspectral target detection: theory and practice,” Proc. SPIE 4725, 202–215 (2002). [CrossRef]

41. D. Manolakis, R. Lockwood, T. Cooley, and J. Jacobson, “Is there a best hyperspectral detection algorithm?” Proc. SPIE 7334, 733402 (2009). [CrossRef]

42. S. Kraut and L. Scharf, “The CFAR adaptive subspace detector is a scale-invariant GLRT,” IEEE Transactions on Signal Processing 47(9), 2538–2541 (1999). [CrossRef]

43. S. Kraut, L. Scharf, and L. McWhorter, “Adaptive subspace detectors,” IEEE Transactions on Signal Processing 49(1), 1–16 (2001). [CrossRef]

44. W. M. Porter and H. T. Enmark, “A system overview of the airborne visible/infrared imaging spectrometer (AVIRIS),” Proc. SPIE 0834, 22–31 (1987). [CrossRef]

Reference spectral signature selection using density-based cluster for automatic oil spill detection in hyperspectral images

Abstract

1. Introduction

2. Framework of oil spill detection from hyperspectral images

3. Extraction of reference spectral signature of oil spill

3.1. Density-based cluster algorithm

3.2. Down-sampling hyperspectral scene

3.3. Band feature extraction

3.4. Identification of spectral signature of oil spill

4. Background parameter estimation

5. Oil spill detection

6. Experimental result and discussion

7. Conclusion

Acknowledgments

References and links

Cited By

Figures (7)

Equations (30)

Optics Express