CN115359363A

CN115359363A - Satellite remote sensing detection and identification method, system and equipment for plastic waste in coastal zone

Info

Publication number: CN115359363A
Application number: CN202210884133.4A
Authority: CN
Inventors: 李鹏; 周虹利; 林是聪; 王厚杰; 李振洪
Original assignee: Ocean University of China
Current assignee: Ocean University of China
Priority date: 2022-07-26
Filing date: 2022-07-26
Publication date: 2022-11-18

Abstract

The invention belongs to the field of plastic target remote sensing monitoring, and discloses a satellite remote sensing detection and identification method, a system and equipment for coastal plastic waste, which rely on a Google Earth Longine (GEE) platform to obtain Sentinel-2MSI multispectral remote sensing data, and form an editable research area image after preprocessing; making a land and seawater mask by utilizing the reflectivity difference of the near infrared band, and extracting floaters in the image; selecting the spectral reflectance of the extracted floating object pixel as a peak value limit for judging the plastic waste; according to the peak value conditions of different wave bands, eliminating interference pixels such as seaweeds and the like in the floating objects, picking up the pixels with plastic waste, and determining the actual area where the pixels are located. According to the technical scheme, the overall precision can reach 98%, and the F-score can reach 0.85. According to the method, appropriate reflectivity threshold values are designed for different wave bands according to the actual conditions of a research area, and the characteristics of peak values are applied to identify and distinguish marine floating plastics so as to achieve the purpose of extracting plastic floating objects in coastal zones.

Description

Satellite remote sensing detection and identification method, system and equipment for plastic waste in coastal zone

Technical Field

The invention belongs to the field of plastic target remote sensing monitoring, and particularly relates to a satellite remote sensing detection and identification method, system and equipment for plastic garbage in a coastal zone.

Background

At present, marine pollution has risen to become a global significant problem, wherein plastic pollution accounts for a large proportion and the harm is witnessed. Tens of thousands of plastic fragments are discarded every year all over the world, a large part of the plastic fragments flow to the ocean, and many marine animals die due to eating the plastic by mistake, and thousands of discarded fishing nets and plastic rings wrap some seals, birds and the like, so that the normal development and survival of the marine animals are influenced. Besides the components of plastics themselves, additives in plastics are also a concern for environmental pollution. In the face of the fact that plastics pose an irreparable hazard to the marine environment, marine plastic waste remediation is reluctant for global environments to be further destroyed, for human survival and sustainable development.

Compared with the traditional monitoring method, the remote sensing technology has the advantages of reducing the on-site investigation cost and expanding the monitoring range, and has absolute advantages in the aspects of marine plastic target identification, detection and monitoring. However, the marine floating plastic has the problems of small quantity, small size and the like, and is influenced by the power of seawater, the spectral characteristic noise of the plastic in water is higher than that of dry plastic, and the difficulty in identifying a plastic target is increased. Methods currently used for ground object classification are generally classified into supervised methods and unsupervised methods. In the absence of currently determinable marine plastic targets, there is no training data and many areas can only use unsupervised classification methods. The unsupervised classification method (such as K mean value, fuzzy C mean value and the like) does not need to train a sample, a plurality of wave bands and characteristic indexes are set as attributes, and only a small part of known real plastics can be distinguished; compared with unsupervised classification, the supervised classification method (such as support vector regression (SVM), semi-supervised fuzzy C-means and the like) can identify most plastics, but a plurality of wrong classification phenomena exist, and a certain number of samples are required by both the supervised classification methods. The naive Bayes (Bayesian) algorithm obtains a better effect in the identification of plastic targets at present, can use fewer samples for training, and can miss pixels possibly containing less plastics in different areas where the verification precision is relatively higher but various training samples are obtained.

In conclusion, the invention provides a method for effectively identifying a plastic target according to the spectral reflectivity difference of each waveband based on the spectral characteristics of the coastal zone based on the characteristics of few plastic samples, small size and the like in the current coastal zone.

Disclosure of Invention

Aiming at the problems in the prior art, the invention provides a satellite remote sensing detection and identification method, a system and equipment for plastic waste in a coastal zone.

The invention is realized in such a way that the satellite remote sensing detection and identification method for plastic garbage in coastal zones comprises the following steps:

the method comprises the following steps: acquiring and preprocessing Sentinel-2MSI multispectral remote sensing data of a research area by relying on a Google Earth Engine (GEE) platform, and forming an editable research area image;

step two: making land and seawater masks by utilizing visual judgment and near-infrared band reflectivity difference for extracting coastlines and floaters in the image preprocessed in the first step;

step three: selecting the spectrum information of the floating object pixel extracted in the step two as a peak value limit for judging the plastic waste;

step four: and (4) removing the seaweeds from the spectral information of the floating object pixels in the step three to obtain the actual area where the garbage is located.

In one embodiment, the second step: making a land and seawater mask by utilizing visual judgment and the difference of near-infrared band reflectivity for extracting a coastline and floaters in the image preprocessed in the step one, wherein the method specifically comprises the following steps:

manufacturing a land mask for cutting out a coastline in a research area in the remote sensing image;

a seawater mask is made for extracting relatively bright floating objects in the ocean region from the image after passing through the land mask.

In one embodiment, the step of making a seawater mask for extracting relatively bright floating objects in the region from the image is specifically as follows: the calculation characteristics refer to PI or FDI, PI can identify floating objects in a large range on the sea, and FDI can pick up floating fragments on a sub-pixel scale. Thus, displaying the computed image with a suitable grid rendering color library can determine the approximate location of the float:

PI＝R _rs,NIR /(R _rs,NIR +R _rs,red ) (1)

wherein Rrs, NIR is the spectral reflectance of the near infrared band, and Rrs, red is the spectral reflectance of the red band.

FDI＝R _rs,NIR -R' _rs,NIR (2)

Wherein Rrs, SWIR1 is short wave infrared 1 wave band spectral reflectivity, rrs, RE2 red side 2 wave band spectral reflectivity, and lambda is central wave band wavelength.

In one embodiment, the third step: by looking at the reflectivity of each band of a single point in the statistical image, 15% of the spectral reflectivity of the peak band is used as the peak limit that can be determined as plastic waste, in one embodiment, the fourth step excludes the seaweeds from the spectral information of the floating object pixels in the third step to obtain the actual area where the waste is located, which is specifically,

the pixels in the image are distinguished by judging the spectral reflectivity conditions of red edge wave bands (B5-B7) and near infrared wave bands (B8) in the spectrum of the floating object pixel,

when the light reflectivity of the spectrum of the floating object pixel is smaller than that of the pixel of the visible light and the near infrared band (B8) in the red side bands (B5-B7), the floating object pixel is regarded as plastic 1;

when the spectrum of the floating object pixel has peaks in B5 and B6 wave bands, the floating object pixel is regarded as plastic 2;

when the spectrum of the float pixel has peaks at B7 and B8, and the pixels not belonging to plastic 1 and plastic 2 are seaweeds.

Another object of the present invention is to provide an identification system suitable for the above method for detecting and identifying plastic garbage on coastal zones by satellite remote sensing, comprising:

the data preprocessing unit is used for acquiring and preprocessing multi-spectral remote sensing data of the Sentinel-2MSI in the research area and is used for forming an editable research area image;

the mask making unit is used for making land and seawater masks by utilizing visual judgment and near infrared band reflectivity difference and is used for extracting coastal zones and floaters in the coastal zones in the preprocessed images of the data preprocessing unit;

the data processing unit is used for selecting the spectral information of the floating object pixel extracted from the mask making unit as a peak value boundary for judging the plastic garbage and distinguishing plastic pixels and interference pixels such as seaweed and the like according to the existence condition of peak values of all wave bands;

another object of the invention is to provide a computer device comprising a memory and a processor, the memory storing a computer program which, when executed by the processor, causes the processor to perform the steps as described above:

the method comprises the following steps: acquiring and preprocessing multi-spectral remote sensing data of a research area Sentinel-2MSI, and forming an editable research area image;

step two: making a land and seawater mask by utilizing visual judgment and the difference of the reflectivity of the near infrared band, and extracting a coastline and a floater in the image preprocessed in the first step;

step four: and (4) removing the seaweeds from the spectral information of the floating object pixels in the step three to obtain the actual region where the garbage is located.

It is another object of the present invention to provide a computer-readable storage medium storing a computer program which, when executed by a processor, causes the processor to perform the steps as described above:

Another object of the present invention is to provide an information data processing terminal, which is adapted to implement the satellite remote sensing detection and identification method system for plastic waste in coastal zones as described above.

By combining all the technical schemes, the invention has the advantages and positive effects that:

the method is based on a Google Earth Engine (GEE) platform, the GEE can rapidly process huge images in batches, and in part of existing algorithms in the platform, image information can be acquired and calculated through a webpage editor. The GEE platform provides a Sentinel-2A grade product, derived from the atmospheric Bottom (BOA) reflectance image of a 1C grade product. Since 3 months 2018, class 2A products were produced systematically in the european terrestrial sector and extended globally in 12 months 2018.

From literature, scientific reports, news articles on the serious pollution of marine waste, it can be speculated that plastic waste in many areas is often aggregated with algae or marine foams. Therefore, starting from the absorption and reflection characteristics of 10 bands in the Sentinel-2MSI image, the method first identifies the spectral characteristics of objects that would interfere with the identification of plastics:

(1) sea water: water has a high absorbance in the near infrared (NIR, 833 nm) and short wavelength near infrared (SWIR, 1610 to 2202 nm), strongly impairing the spectral reflectance at these wavelengths, with the emissivity of clear water being very high, close to 1.

(2) Seaweed: reflecting green (B3, 560nm) and red-side band (B5-B7, 700-780 nm) light;

(3) foaming: the reflectivity is highest in a green and red (B4, 665nm) visible light wave band, and is smaller in a near infrared wave band. Plastics of different colors have different reflectivities in the visible band, mostly showing a minimum in the red band (B4) and a maximum in the near infrared band (B8). The average spectral curves of the plastic targets artificially laid on the coastal zone in the remote sensing images by the PLPs 2018 and the PLPs 2019 (figure 1).

Drawings

FIG. 1 is a spectral plot of a plastic provided by an embodiment of the present invention;

FIG. 2 is a flow chart for identifying plastic targets based on differences in spectral reflectance provided by an embodiment of the present invention;

FIG. 3 (a) is a pixel diagram corresponding to process step (a) of the flowchart for identifying plastic objects based on differences in spectral reflectance provided by an embodiment of the present invention;

FIG. 3 (b) is a pixel diagram corresponding to processing step (b) of the flowchart for identifying plastic objects based on differences in spectral reflectance provided by an embodiment of the present invention;

FIG. 3 (c) is a pixel diagram corresponding to process step (c) of the flowchart for identifying plastic objects based on differences in spectral reflectance provided by an embodiment of the present invention;

FIG. 3 (d) is a schematic pixel diagram corresponding to the processing step (d) of the flowchart for identifying plastic objects according to the difference in spectral reflectance according to the embodiment of the present invention;

FIG. 3 (e) is a schematic pixel diagram corresponding to the processing step (e) of the flowchart for identifying plastic objects according to the difference in spectral reflectance according to the embodiment of the present invention;

FIG. 3 (f) is a pixel diagram corresponding to process step (f) of the flowchart for identifying plastic objects based on differences in spectral reflectance according to an embodiment of the present invention;

FIG. 4 is a schematic diagram of an area for artificial placement of real plastic (Greek Tsakia beach) provided by an embodiment of the present invention;

fig. 5 (a) is a processing result applied to a research area where 3 real plastic targets are arranged in 2018, 6, 7 days in the plastic project PLP2018 of the university of einzel.

Fig. 5 (b) is a processing result of a research area applied to laying real plastics in 2019, 4 month and 18 day of the plastic project PLP2019 of the university of einzel.

Fig. 5 (c) is a processing result of a research area for laying real plastics in 2019, 5 month and 3 days in 2019, which is applied to the plastic project PLP2019 of the university of love.

Fig. 5 (d) is a processing result of a research area applied to laying real plastics in 2019, 5, 18 th of the plastic project PLP2019 of the university of einzel.

Fig. 5 (e) is a processing result of a research area applied to laying real plastics in 2019, 6, 7 and 6 of the plastic project PLP2019 of the university of einzel.

Fig. 6 is a flowchart illustrating steps of a method for detecting and identifying plastic garbage on the coast by satellite remote sensing according to an embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

Aiming at the problems in the prior art, the invention provides a satellite remote sensing detection and identification method for plastic waste in a coastal zone, and the invention is described in detail below with reference to the accompanying drawings.

Main scheme and effect description section:

a satellite remote sensing detection and identification method for plastic waste in coastal zones comprises the following steps:

step S01: acquiring and preprocessing Sentinel-2MSI multispectral remote sensing data of a research area by relying on a Google Earth Engine platform, and forming an editable research area image;

step S02: making a land mask according to an existing or visually drawn coastline vector file, extracting a coastline, setting a near infrared band reflectivity threshold value, and making a seawater mask for extracting floaters in an image;

step S03: selecting the spectral information of the floating object pixel extracted in the step S02 as a peak value limit for judging the plastic waste;

step S04: algae are excluded from the spectral information of the float pixels of step S03 to derive the actual area where the trash is located.

The method for identifying the plastic garbage in the coastal zone by using the multispectral remote sensing data based on the known spectral characteristics of the marine plastics and other floaters has the overall precision of 98% and the F-score of 0.85. According to the method, appropriate reflectivity threshold values are designed for different wave bands according to the actual conditions of a research area, and the characteristics of peak values are applied to identify and distinguish the marine floating plastics so as to achieve the purpose of extracting plastic floating objects in coastal zones.

The first embodiment is as follows:

acquiring images, acquiring multispectral remote sensing data, and preprocessing. Taking the S2 satellite remote sensing data as an example, the L1C data before 2018 needs to be subjected to atmospheric correction through a Sen2Cor plug-in of SNAP, the Google Earth Engine (GEE) can accept and process tif format images uploaded by users, but the images subjected to atmospheric correction by SNAP are single images of each waveband, so that the processed remote sensing data can be subjected to waveband fusion by using image processing software to form an image containing information of all wavebands, and data processing can be performed uniformly.

Step (b) fabricating a land mask. Land masks can be used for cutting out the coastal zones in the research area by drawing cutting surfaces by visual observation and utilizing clip functions. The method comprises the following steps that a clip function is required to be adopted for an image uploaded to GEE from the outside to obtain a research area for subsequent editing; the cutting surface can be directly drawn by utilizing the existing remote sensing data of the GEE platform, so that a coastal zone research area is obtained, and the operation speed is accelerated.

And (c) manufacturing a seawater mask. Using Sentinel-2 as raw data, both floating plastics and plants reflected light at a central wavelength of approximately 833nm, which was located in the Sentinel-2 MSI. In addition to the above spectral features, seaweed may also absorb SWIR (900-1700 nm) light relative to the average spectrum at 1610nm for seawater and plastics, but it is not excluded that the cause of this relatively abnormal phenomenon is atmospheric correction. And (3) looking up the B8 wave band image processed in the step (2), finding a brighter aggregate relative to the background (seawater), and adjusting the GEE brightness display parameter Gamma to make the floater clearer. Clicking the pixels of the seawater can check the image information of the single point, compiling codes to set a brightness threshold value, and extracting floating object pixels.

It should be further noted that the computation features refer to PI, which can identify a large range of floating objects at sea, or FDI, which can pick up floating debris on a sub-pixel scale. Thus, displaying the computed image with a suitable grid rendering color library can determine the approximate location of the float:

PI＝R _rs,NIR /(R _rs,NIR +R _rs,red ) (4)

wherein R is _rs,NIR Is the spectral reflectance, R, of the near infrared band _rs,red The spectral reflectance in the red band.

FDI＝R _rs,NIR -R' _rs,NIR (5)

Wherein R is _rs,SWIR1 Is short wave infrared 1 band spectral reflectance, R _rs,RE2 Red edge 2 band spectral reflectance, λ is central band wavelength.

It is further noted that looking at the spectral information of the float pixels, the limits that can be considered as peaks are determined. The highest spatial resolution of the resampled S2 data is 10m, and the proportion of plastic material in a pixel is usually small, so the overall spectral characteristics of the pixel can be affected by other materials. The spectral characteristics of a known real plastic target in seawater can be analyzed: when a pixel is covered with a complete plastic, the spectral information of the pixel has a light reflectivity in the red-side band (B5-B7) that is substantially smaller than in the near infrared and visible bands. When other materials (typically seaweed) are mixed in the plastic pixel, a distinct peak occurs in the red-edge band.

It is further noted that algae and other interfering pixels are excluded from the spectral information of the float pixels of step (c) to derive the actual area where the trash is located, which is specifically:

and (c 1) regarding the floating object as plastic 1 when the light reflectivity of the spectrum of the floating object pixel in the red side wave band (B5-B7) is smaller than that of the pixel in the visible light and near infrared wave band (B8). It is further noted that the proportion of plastic in the pixel is large. When marine plastics are identified in an actual environment, algae plants are contained in a pixel, and the reflectivity of the pixel to red light (B4) can be influenced, so that the light reflectivity of red-edge wave bands (B5-B7) is selected to be smaller than that of the red light (B4) in the actual environment;

step (c 2) when the spectrum of the floating object pixel has peak values in B5 and B6 wave bands, the floating object pixel is regarded as plastic 2, and it needs to be further pointed out that the wave band is regarded as having the peak value when the difference value of the reflectivity of the peak wave band and the adjacent wave band is greater than or equal to 15% of the reflectivity of the wave band;

step (c 3) when the spectrum of the floating object pixel has peaks at B7 and B8, and the pixels not belonging to plastic 1 and plastic 2 are seaweeds, it should be further noted that the pixels having peaks at B7 and B8 are selected. The seaweed has high reflectivity in B8 wave band as floating object, the general spectral feature of the plant is that the reflectivity increased along with the wavelength exists in the red edge wave band, therefore, the peak value can also appear in B7, and the pixels belonging to the step (c 3) and not belonging to the step (c 2) and the step (c 1) are selected and regarded as the seaweed (plant).

Verifying the area with real plastic, checking the coincidence condition of the identified plastic pixel and the real plastic pixel, and properly adjusting the brightness limit value and the peak limit value by checking the reflectivity of a spectrum curve of the real plastic pixel in a near infrared band (B8), the peak condition of red edge bands (B5-B7) and the reflectivity difference between the peak band and an adjacent band.

The experimental demonstration is as follows:

due to the factors that the marine plastic target has small reflectivity, is easily influenced by the atmosphere, has small sample quantity, has small reflectivity difference with other ground objects and the like, the results obtained by a common machine classification method are common. Some scholars specialize the traditional machine classification method according to the characteristics of floating plastics, but the influence of the region is large. The method is relatively flexible and is less limited by the sample, and can analyze each limit value more accurately under the condition that a given sample exists, but can also analyze the floating object to obtain the method.

The method performs precision analysis according to the following method:

TABLE 1 confusion matrix

The overall accuracy is correctly identified as the proportion of plastics in the total, and the F-score gives consideration to the accuracy and the recall ratio, is the weighted average of the two, and ranges from 0 to 1. In the case of better model performance, the overall accuracy tends to 100%, while F-score tends to 1. The overall accuracy and calculation formula for F-score can be expressed as:

where TP refers to the number of identified plastic pixels that are real plastic, FP refers to the number of identified plastic pixels that are not real plastic, FN refers to the number of real plastic pixels that are not identified as plastic pixels, and TN refers to the number of pixels that are identified as non-plastic that are non-plastic.

Data for verifying the classification method are derived from plastic items PLP2018 and PLP2019, 5 sceneinel-2 images are shared, the research area is located at a certain beach, and plastic pixels in the images are all plastic targets artificially placed on the sea surface (shown in figure 2). The plurality of pixels used for classification in the verification comprise all known pixels of the plastic target and non-plastic pixels, and the non-plastic pixels are randomly selected.

TABLE 2 confusion matrix based on validation set data for spectral feature classification

From the confusion matrix of table 2, it can be seen that the overall accuracy of the reflectance feature classification method used in the research area where the presence of plastics has been determined can reach 98.9%, and the F-score can reach 0.85, so that the method has reliability for the identification of marine floating plastics. In practical applications, the precision of practical marine floating plastics may be reduced because the spectral reflectance of the near infrared band may be lower than that of artificially fabricated plastic targets.

It should be noted that the embodiments of the present invention can be realized by hardware, software, or a combination of software and hardware. The hardware portion may be implemented using dedicated logic; the software portions may be stored in a memory and executed by a suitable instruction execution system, such as a microprocessor or specially designed hardware. It will be appreciated by those skilled in the art that the apparatus and methods described above may be implemented using computer executable instructions and/or embodied in processor control code, for example such code provided on a carrier medium such as a diskette, CD-or DVD-ROM, a programmable memory such as read-only memory (firmware) or a data carrier such as an optical or electronic signal carrier. The apparatus and its modules of the present invention may be implemented by hardware circuits such as very large scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., or by software executed by various types of processors, or by a combination of hardware circuits and software, e.g., firmware.

The above description is only for the purpose of illustrating the present invention and the appended claims are not to be construed as limiting the scope of the invention, which is intended to cover all modifications, equivalents and improvements that are within the spirit and scope of the invention as defined by the appended claims.

Claims

1. A satellite remote sensing detection and identification method for plastic waste in coastal zones is characterized by comprising the following steps:

the method comprises the following steps: acquiring and preprocessing Sentinel-2MSI multispectral remote sensing data of a research area by relying on a Google Earth Engine platform, and forming an editable research area image;

step two: making a land mask according to an existing or visually drawn coastline vector file, extracting a coastline, setting a near-infrared band reflectivity threshold value, and making a seawater mask for extracting floaters in an image;

step three: selecting and judging peak value limits of all wave bands of the plastic waste according to the spectrum information of the floating object pixels extracted in the second step;

2. The coastal zone plastic waste satellite remote sensing detection and identification method according to claim 1, characterized in that the second step: utilizing the difference of the reflectivity of the near infrared band to manufacture a land and seawater mask for extracting a coastline and floaters in the image preprocessed in the step one, which comprises the following specific steps:

making a land mask for cutting out a coastal zone and an ocean in a research area from the remote sensing image;

a seawater mask is made for extracting relatively bright surface floaters from the remote sensing image.

3. The method for the satellite remote sensing detection and identification of plastic wastes in coastal zones as claimed in claim 2, wherein said step of making a seawater mask for extracting relatively bright floating objects in the area from the multi-spectrum remote sensing data comprises: and calculating a characteristic index PI or FDI, wherein PI can identify the floating objects in a large range on the sea, and FDI can pick up floating fragments on a sub-pixel scale. Thus, displaying the computed image with a suitable grid rendering color library, the approximate position of the float can be determined:

PI＝R _rs,NIR /(R _rs,NIR +R _rs,red ) (1)

wherein R is _rs,NIR Is the spectral reflectance, R, of the near infrared band _rs,red Spectral reflectance in the red band.

FDI＝R _rs,NIR -R' _rs,NIR (2)

4. The method for the satellite remote sensing detection and identification of plastic waste in coastal zones as claimed in claim 1, wherein the steps of three: by checking the reflectivity of each wave band of a single point in the statistical image, 15% of the spectral reflectivity of the peak wave band is used as a peak limit which can be judged as plastic waste.

5. The method for the satellite remote sensing detection and identification of plastic garbage on the coastal zone according to claim 4, wherein in the fourth step, the seaweeds are excluded from the spectral information of the floating object pixels in the third step to obtain the actual area where the garbage is located, and the method specifically comprises the following steps:

judging the condition that the red side wave bands (B5-B7) and the near infrared wave band (B8) in the spectrum of the floating object pixel can be identified as peak values,

when the spectrum of the float pixels has an additional peak in the B5 or B6 band in addition to the peak in the near infrared band (B8), these pixels are considered to be plastic 2;

when the spectrum of the floating object pixel has peaks only at B7 and B8, the floating object pixel is regarded as a seaweed.

6. Identification system suitable for the method for the remote satellite detection and identification of plastic waste coastal zones according to any of claims 1 to 5,

the mask making unit is used for making a land mask and a seawater mask by utilizing the difference between the visual judgment and the near infrared band reflectivity, and is used for extracting a coastal zone and floaters in the coastal zone in the image after the pretreatment of the data pretreatment unit;

and the data processing unit is used for selecting the spectral information of the floating object pixel extracted from the mask making unit as a peak value limit for judging the plastic garbage and distinguishing the plastic pixel from interference pixels such as seaweed and the like according to the existence condition of peak values of all wave bands.

7. A computer arrangement, characterized in that the computer arrangement comprises a memory and a processor, the memory storing a computer program, which, when executed by the processor, causes the processor to carry out the steps of any of claims 1-5:

8. A computer-readable storage medium storing a computer program which, when executed by a processor, causes the processor to perform the steps of any of claims 1-5:

9. An information data processing terminal, characterized in that the information data processing terminal is used for realizing the satellite remote sensing detection and identification method system of plastic garbage in coastal zones as claimed in claim 6.