CN103559495A - Hyperspectral oil spilling information extraction method - Google Patents
Hyperspectral oil spilling information extraction method Download PDFInfo
- Publication number
- CN103559495A CN103559495A CN201310533895.0A CN201310533895A CN103559495A CN 103559495 A CN103559495 A CN 103559495A CN 201310533895 A CN201310533895 A CN 201310533895A CN 103559495 A CN103559495 A CN 103559495A
- Authority
- CN
- China
- Prior art keywords
- mnf
- oil film
- image
- eigenwert
- signal vector
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention discloses a hyperspectral oil spilling information extraction method which comprises the following steps: original remote sensing images are preprocessed; the preprocessed images are subjected to minimum noise fraction (MNF) to obtain MNF characteristic value images; various ground object target regions of interest are established to obtain a distribution diagraph of the regions of interest; differences between MNF characteristic value curves of all ground object targets are analyzed, optimal wavebands or waveband combination modes of all the ground object targets are determined and distinguished, and threshold values of all the ground object targets are determined and distinguished; a classification decision-making tree is established for the MNF characteristic value images to classify the MNF characteristic value images, and then oil spilling information is obtained. According to the hyperspectral oil spilling information extraction method, the preprocessed remote sensing images are subjected to MNF, an original data volume is reduced, the number of wavebands of the images is reduced, and therefore data dimensions are reduced. Moreover, influence of noise signals is eliminated through MNF, and the noise effect is weakened. By means of the hyperspectral oil spilling information extraction method, the data processing volume is reduced, the data processing speed is increased, tiny differences between categories can be identified, and identification accuracy is improved.
Description
Technical field
The present invention relates to ocean monitoring technologytechnologies field, relate in particular to a kind of high spectrum oil spilling information extracting method.
Background technology
The appearance of high-spectrum remote-sensing is a field technology revolution of remote sensing circle, and it originates from the multispectral romote sensing technology at the initial stage seventies, and it makes not detectable material in broadband remote sensing originally, in high-spectrum remote-sensing, can be detected.
In remote sensing monitoring oil film field, multispectral sensor, its spectral resolution is low, wave band number is few, usually there will be the phenomenon of the different spectrum of jljl, same object different images, has affected oil film target monitoring and identification.On the other hand, marine environment is complicated, can impact the spectral signature of oil film, reduces oil film accuracy of identification.At present, the spectroscopic data that the understanding of oil film spectrum response characteristic is mainly obtained based on sea trial (for a data), utilize field spectroradiometer to obtain oil film in the spectral signature of visible ray near-infrared band, the spectral signature of analyzing oil film changes and profit contrast rule.Spectroscopic data based on point contributes to be familiar with the spectral signature that oil is planted, but film distribution information cannot be provided, and also cannot obtain in time the information that oil is planted, and the substance that affects the emergent level of marine oil film improves.
High light spectrum image-forming data not only have the imaging function of traditional sensors, also can utilize spectral technique to obtain the spectral information of atural object simultaneously, thereby the spatial information and the spectral characteristic that are observed various atural objects are recorded.By the combination of spectral technique and imaging technique, at characteristics of image and two dimensions of spectral signature, observed object is processed and analyzed, effectively get rid of decoy, improve the accuracy of marine oil film identification.In prior art, utilizing remotely-sensed data to carry out oil spilling information extraction is mainly by wave band computing, improves the contrast of oil spilling and background image, thus identification oil spilling.The multiband feature having due to high spectrum image, and the redundance between wave band increases, and data volume is very large, utilizes conventional multispectral remote sensing sorting technique processing speed very slow.
Summary of the invention
The problems referred to above that exist for solving prior art, the present invention will design a kind of high spectrum oil spilling information extracting method that can reduce high-spectral data dimension and improve oil identification efficiency.
To achieve these goals, technical scheme of the present invention is as follows: a kind of high spectrum oil spilling information extracting method, comprises the following steps:
A, pre-service: original remote sensing images are carried out to pre-service, and described pre-service comprises radiant correction, atmospheric correction and mask process, to improve the sharpness of image, obtains pretreatment image;
B, MNF conversion: pretreatment image is carried out to the separated conversion of minimal noise, obtain MNF eigenwert image; Described MNF is the abbreviation of the English MinimumNoiseFraction of the separated conversion of minimal noise;
If the original signal vector z that i band image of high spectrum image forms
iby the noise-free signal vector s under ideal state
iwith noise signal vector n
iform noise-free signal vector s
iwith noise signal vector n
iuncorrelated, z
ican be expressed as:
z
i=s
i+n
i
Wherein, i=1,2,, L, L is wave band number;
First by low-pass filtering, from original signal vector z, isolate noise signal vector n, then obtain respectively the covariance matrix Σ of z and n
zand Σ
n, Z=(z wherein
1, z
2,, z
l), N=(n
1, n
2,, n
l);
Calculate
Σ
zeigenvalue λ
iwith corresponding vectorial u
i, suppose that eigenwert meets λ
1>=λ
2>=λ
l, make U=(u
1,, u
l), after MNF conversion, consequential signal vector is Y, MNF conversion finally may be defined as Y=U
tz; By the image that after MNF conversion, consequential signal vector Y forms, it is MNF eigenwert image;
C, extraction region of interest: in steps A gained pretreatment image, select seawater, heavy oil film, intermediate gauge oil film, thin oil film and very thin oil film as ground object target, set up seawater, heavy oil film, intermediate gauge oil film, thin oil film and very thin oil film region of interest, obtain region of interest distribution plan;
D, MNF eigenvalue graph is analyzed: on MNF eigenwert image and region of interest distribution plan, find seawater, heavy oil film, intermediate gauge oil film, thin oil film and very thin oil film region of interest correspondence position on MNF eigenwert image, extract the MNF eigenvalue graph of correspondence position pixel on MNF eigenwert image, and analyze the difference of each ground object target MNF eigenvalue graph, determine and distinguish seawater, heavy oil film, intermediate gauge oil film, the best band of thin oil film and very thin oil film or band combination mode, determine and distinguish seawater, heavy oil film, intermediate gauge oil film, the threshold value of thin oil film and very thin oil film,
E, set up decision tree: the analysis according to step D to MNF eigenvalue graph, MNF eigenwert image is set up to categorised decision tree and classify, obtain oil spilling information;
Described set up the method that categorised decision tree classifies and comprise the following steps: according to the definite differentiation seawater of step D, heavy oil film, intermediate gauge oil film, thin oil film and very thin oil film best band or band combination mode and definite threshold value, the MNF eigenwert image of take is input picture, the described threshold value of take is set up binary tree as Rule of judgment, thereby form corresponding categorised decision, sets.
Compared with prior art, the present invention has following beneficial effect:
Due to the present invention to pre-service after remote sensing images first carried out MNF conversion, original data volume is compressed, reduced the wave band number of image, thereby reduced data dimension.And by MNF, convert the impact of having rejected noise signal, weakened noise effect.Therefore the present invention has not only reduced data processing amount, improves data processing speed, and can identify the fine difference between classification, improves accuracy of identification.
Accompanying drawing explanation
3, the total accompanying drawing of the present invention, wherein:
Fig. 1 is process flow diagram of the present invention.
Fig. 2 is the categorised decision tree that the present invention builds according to MNF eigenwert wave spectrum.
Fig. 3 is minimal noise separation characteristic value curve of the present invention.
Embodiment
Below in conjunction with accompanying drawing, the invention will be further described.As shown in Figure 1, first the present invention carries out pre-service to airborne visible ray/Infrared Imaging Spectrometer (AirborneVisibleInfraredImagingSpectrometer, AVIRIS) data, comprises radiant correction, atmospheric correction, mask process etc.; Then utilize MNF transfer pair image data to carry out dimensionality reduction and denoising, thereby obtain MNF image, this image top n wave band has been concentrated information most in former data, and residue wave band mostly is noise information, does not participate in subsequent treatment; For identifying more exactly each ground object target, extract MNF end member eigenvalue graph; By the analysis to MNF eigenvalue graph, set up the categorised decision tree (as shown in Figure 2) based on MNF, thereby realize the extraction of oil spilling information.
In pretreated AVIRIS data, choose known typical feature target type as region of interest, comprised seawater, heavy oil film, intermediate gauge oil film, thin oil film and very thin oil film (sheen).Image after pre-service is associated with MNF image, determine the position of seawater, heavy oil film, intermediate gauge oil film, thin oil film and very thin oil film.
As shown in Figure 3, choose front 25 wave bands of MNF image, extract above MNF eigenwert wave spectrum curve corresponding to several typical feature targets.This curve shows the increase along with band number, and MNF eigenwert goes to zero gradually, and the quantity of information of contained use is fewer more backward to show wave band.By the analysis to MNF eigenvalue graph, visible seawater, heavy oil film, intermediate gauge oil film, thin oil film and very thin oil film differ greatly in the MNF of each wave band eigenwert, not only can obviously distinguish seawater and heavy oil film and intermediate gauge oil film, and can distinguish thin oil film and water body by the computing of the first wave band and the second wave band, by triband eigenwert, distinguish very thin oil film and water body.
The above; it is only preferred forms of the present invention; but protection scope of the present invention is not limited to this; anyly be familiar with those skilled in the art in the technical scope that the present invention discloses; according to technical scheme of the present invention and inventive concept thereof, be equal to replacement or changed, within all should being encompassed in protection scope of the present invention.
Claims (1)
1. a high spectrum oil spilling information extracting method, is characterized in that: comprise the following steps:
A, pre-service: original remote sensing images are carried out to pre-service, and described pre-service comprises radiant correction, atmospheric correction and mask process, to improve the sharpness of image, obtains pretreatment image;
B, MNF conversion: pretreatment image is carried out to the separated conversion of minimal noise, obtain MNF eigenwert image; Described MNF is the abbreviation of the English MinimumNoiseFraction of the separated conversion of minimal noise;
If the original signal vector z that i band image of high spectrum image forms
iby the noise-free signal vector s under ideal state
iwith noise signal vector n
iform noise-free signal vector s
iwith noise signal vector n
iuncorrelated, z
ican be expressed as:
z
i=s
i+n
i
Wherein, i=1,2,, L, L is wave band number;
First by low-pass filtering, from original signal vector z, isolate noise signal vector n, then obtain respectively the covariance matrix Σ of z and n
zand Σ
n, Z=(z wherein
1, z
2,, z
l), N=(n
1, n
2,, n
l);
Calculate
Σ
zeigenvalue λ
iwith corresponding vectorial u
i, suppose that eigenwert meets λ
1>=λ
2>=λ
l, make U=(u
1,, u
l), after MNF conversion, consequential signal vector is Y, MNF conversion finally may be defined as Y=U
tz; By the image that after MNF conversion, consequential signal vector Y forms, it is MNF eigenwert image;
C, extraction region of interest: in steps A gained pretreatment image, select seawater, heavy oil film, intermediate gauge oil film, thin oil film and very thin oil film as ground object target, set up seawater, heavy oil film, intermediate gauge oil film, thin oil film and very thin oil film region of interest, obtain region of interest distribution plan;
D, MNF feature curve analysis: on MNF eigenwert image and region of interest distribution plan, find seawater, heavy oil film, intermediate gauge oil film, thin oil film and very thin oil film region of interest correspondence position on MNF eigenwert image, extract the MNF eigenvalue graph of correspondence position pixel on MNF eigenwert image, and analyze the difference of each ground object target MNF eigenvalue graph, determine and distinguish seawater, heavy oil film, intermediate gauge oil film, the best band of thin oil film and very thin oil film or band combination mode, determine and distinguish seawater, heavy oil film, intermediate gauge oil film, the threshold value of thin oil film and very thin oil film,
E, set up decision tree: the analysis according to step D to MNF eigenvalue graph, MNF eigenwert image is set up to categorised decision tree and classify, obtain oil spilling information;
Described set up the method that categorised decision tree classifies and comprise the following steps: according to the definite differentiation seawater of step D, heavy oil film, intermediate gauge oil film, thin oil film and very thin oil film best band or band combination mode and definite threshold value, the MNF eigenwert image of take is input picture, the described threshold value of take is set up binary tree as Rule of judgment, thereby form corresponding categorised decision, sets.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310533895.0A CN103559495A (en) | 2013-10-31 | 2013-10-31 | Hyperspectral oil spilling information extraction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310533895.0A CN103559495A (en) | 2013-10-31 | 2013-10-31 | Hyperspectral oil spilling information extraction method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103559495A true CN103559495A (en) | 2014-02-05 |
Family
ID=50013736
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310533895.0A Pending CN103559495A (en) | 2013-10-31 | 2013-10-31 | Hyperspectral oil spilling information extraction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103559495A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104102926A (en) * | 2014-04-24 | 2014-10-15 | 上海海事大学 | Sea surface oil slick identification method |
| CN104122233A (en) * | 2014-07-29 | 2014-10-29 | 大连海事大学 | Selection method of hyperspectral detection channel for crude oil films with different thickness on sea surface |
| CN105424616A (en) * | 2015-11-26 | 2016-03-23 | 青岛市光电工程技术研究院 | Multispectral camera for ocean oil spill monitoring and imaging processing method |
| CN105844298A (en) * | 2016-03-23 | 2016-08-10 | 中国石油大学(华东) | High spectral oil overflow image classification method based on Fuzzy ARTMAP neural network |
| CN106767454A (en) * | 2016-12-02 | 2017-05-31 | 大连海事大学 | A kind of water-surface oil film thickness measurement system and method based on spectral reflectivity feature |
| CN111597751A (en) * | 2020-03-24 | 2020-08-28 | 自然资源部第一海洋研究所 | Crude oil film absolute thickness inversion method based on self-expansion depth confidence network |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2010052530A1 (en) * | 2008-11-05 | 2010-05-14 | Ecoserv Remote Observation Centre Co. Ltd. | Multi-polarization combined radar-radiometer system |
| CN102507474A (en) * | 2011-10-28 | 2012-06-20 | 大连海事大学 | Method and system for identifying oil spilling target of ship |
| CN102880856A (en) * | 2012-08-22 | 2013-01-16 | 国家海洋局第二海洋研究所 | Method for automatically identifying sea-surface oil spill in aviation hyperspectral remote sensing mode based on spectral characteristic difference of oil and water |
-
2013
- 2013-10-31 CN CN201310533895.0A patent/CN103559495A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2010052530A1 (en) * | 2008-11-05 | 2010-05-14 | Ecoserv Remote Observation Centre Co. Ltd. | Multi-polarization combined radar-radiometer system |
| CN102507474A (en) * | 2011-10-28 | 2012-06-20 | 大连海事大学 | Method and system for identifying oil spilling target of ship |
| CN102880856A (en) * | 2012-08-22 | 2013-01-16 | 国家海洋局第二海洋研究所 | Method for automatically identifying sea-surface oil spill in aviation hyperspectral remote sensing mode based on spectral characteristic difference of oil and water |
Non-Patent Citations (2)
| Title |
|---|
| ANDREW A. GREEN 等: "《A Transformation for Ordering Multispectral Data in Terms of Image Quality with Implications for Noise Removal》", 《IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING》, vol. 26, no. 1, 31 January 1988 (1988-01-31), XP007902504, DOI: doi:10.1109/36.3001 * |
| 刘丙新: "《基于高光谱特征的水上油膜提取与分析研究》", 《中国博士学位论文全文数据库 工程科技Ⅰ辑 》, 15 September 2013 (2013-09-15) * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104102926A (en) * | 2014-04-24 | 2014-10-15 | 上海海事大学 | Sea surface oil slick identification method |
| CN104102926B (en) * | 2014-04-24 | 2017-03-22 | 上海海事大学 | Sea surface oil slick identification method |
| CN104122233A (en) * | 2014-07-29 | 2014-10-29 | 大连海事大学 | Selection method of hyperspectral detection channel for crude oil films with different thickness on sea surface |
| CN105424616A (en) * | 2015-11-26 | 2016-03-23 | 青岛市光电工程技术研究院 | Multispectral camera for ocean oil spill monitoring and imaging processing method |
| CN105844298A (en) * | 2016-03-23 | 2016-08-10 | 中国石油大学(华东) | High spectral oil overflow image classification method based on Fuzzy ARTMAP neural network |
| CN106767454A (en) * | 2016-12-02 | 2017-05-31 | 大连海事大学 | A kind of water-surface oil film thickness measurement system and method based on spectral reflectivity feature |
| CN111597751A (en) * | 2020-03-24 | 2020-08-28 | 自然资源部第一海洋研究所 | Crude oil film absolute thickness inversion method based on self-expansion depth confidence network |
| CN111597751B (en) * | 2020-03-24 | 2023-10-24 | 自然资源部第一海洋研究所 | Crude oil film absolute thickness inversion method based on self-expanding depth confidence network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103559495A (en) | Hyperspectral oil spilling information extraction method | |
| CN107992891B (en) | Multispectral remote sensing image change detection method based on spectral vector analysis | |
| CN108765359B (en) | Fusion method of hyperspectral remote sensing image and full-color image based on JSK model and NSCT technology | |
| Uezato et al. | A novel endmember bundle extraction and clustering approach for capturing spectral variability within endmember classes | |
| Dhumal et al. | Classification of Crops from remotely sensed Images: AnOverview | |
| CN113052216B (en) | Oil spill hyperspectral image detection method based on two-way graph U-NET convolutional network | |
| Jawak et al. | Advancement in landcover classification using very high resolution remotely sensed 8-band WorldView-2 satellite data | |
| Tripathi et al. | Comparative evaluation threshold parameters of Spectral Angle Mapper (SAM) for mapping of Chhabadiya Talc Minerals, Jahajpur, Bhilwara, India using Hyperion hyperspectral remote sensing data | |
| Huynh et al. | Hyperspectral imaging for skin recognition and biometrics | |
| Teo et al. | Object-based land cover classification using airborne lidar and different spectral images | |
| CN111047616A (en) | Remote sensing image landslide target constraint active contour feature extraction method | |
| CN110619370A (en) | Hyperspectral image super-pixel local linear embedding dimension reduction method | |
| da Silva et al. | Multispectral image classification using multilayer perceptron and principal components analysis | |
| Bowles et al. | An optical real-time adaptive spectral identification system (ORASIS) | |
| Ilehag et al. | Revisiting existing classification approaches for building materials based on hyperspectral data | |
| Ghosh et al. | Fractal-based dimensionality reduction of hyperspectral images | |
| CN112418156B (en) | Unsupervised classification-based optical remote sensing image sea surface ship target detection method | |
| Bue et al. | An adaptive similarity measure for classification of hyperspectral signatures | |
| Chen et al. | Evaluation and comparison of dimensionality reduction methods and band selection | |
| Misman et al. | UAV-based hyperspectral imaging system for tree species identification in tropical forest of Malaysia | |
| CN111325269B (en) | Hyperspectral image classification evaluation method based on three-dimensional receiver working characteristic curve | |
| Borah et al. | Dimension reduction of hyperion data for improving classification performance—An assessment | |
| Rosario | Algorithm development for hyperspectral anomaly detection | |
| Cimtay et al. | A novel derivative-based classification method for hyperspectral data processing | |
| Rong et al. | Hyperspectral image classification based on joint superpixel-constrained and weighted sparse representation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20140205 |