CN114694035A

CN114694035A - Multi-angle polarized satellite cloud detection method integrating spatial and spectral characteristics

Info

Publication number: CN114694035A
Application number: CN202210217237.XA
Authority: CN
Inventors: 李正强; 侯梦雨; 谢一凇; 乔瑞; 谢艳清; 伽丽丽; 史正
Original assignee: Aerospace Information Research Institute of CAS
Current assignee: Aerospace Information Research Institute of CAS
Priority date: 2022-03-07
Filing date: 2022-03-07
Publication date: 2022-07-01

Abstract

The invention provides a multi-angle polarized satellite cloud detection method integrating spatial and spectral characteristics, which comprises the following steps: acquiring observation data of a multi-angle polarization imager pixel at N-layer angles; according to observation data of multi-angle polarization imager pixels at each layer angle, according to corresponding land cloud identification results, ocean cloud identification results, glaring ocean cloud identification results, mixed pixel identification results or uncertain pixel identification results of the multi-angle polarization imager pixels at each layer angle, cloud detection results of the multi-angle polarization imager pixels at each layer angle are obtained, the times that the multi-angle polarization imager pixels are identified as clouds, the times that the multi-angle polarization imager pixels are identified as clear sky and the times that the multi-angle polarization imager pixels are identified as uncertain pixels at N layer angles are determined, and whether the multi-angle polarization imager pixels are clouds or not is determined according to the times that the multi-angle polarization imager pixels are identified as clouds, the times that the multi-angle polarization imager pixels are identified as clear sky and the times that the multi-angle polarization imager pixels are identified as uncertain pixels at N layer angles.

Description

Multi-angle polarized satellite cloud detection method integrating spatial and spectral characteristics

Technical Field

The invention relates to the technical field of atmospheric science, in particular to a multi-angle polarized satellite cloud detection method integrating spatial and spectral characteristics.

Background

The satellite remote sensing has the characteristics of instantaneous imaging, large information detection range, continuous space coverage and high spatial resolution, can meet the requirements of a plurality of different fields such as ecology, environment, resources, climate and the like on qualitative and quantitative detection of key parameters of the earth, makes up the defects of the traditional labor-consuming and labor-consuming observation method, is one of the most effective means for realizing rapid detection of a plurality of circle layers such as earth land, ocean, atmosphere and the like at present, and becomes a research hotspot and an important development direction in recent decades. With the development and application of remote sensing technology, the requirement of quantitative remote sensing on the quality of remote sensing images is higher and higher, and cloud is used as condensation of water vapor in the atmosphere, and the condensation state of fine water drops or ice crystals covers 50-70% of the area of the surface of a landed ball, so that a large number of blind areas occur in remote sensing images, the loss of ground object information is caused, and the inversion accuracy of various parameters of the earth surface and the atmosphere is seriously reduced. The accurate separation of cloud and clear sky pixels is a necessary basis for the quantitative inversion of various parameters of the atmosphere and the earth surface by satellite remote sensing.

Cloud detection algorithms for satellite images generally use spectral characteristics of clouds to perform detection, and multispectral sensors such as a mode-resolution Imaging spectrometer (MODIS) use spectral characteristics of the clouds in visible and infrared bands. Compared with the traditional multispectral satellite observation, the multi-angle polarization is a remote sensing means with more comprehensive detection capability and more dimensions, the multi-angle observation greatly improves the observation data quantity, the cloud characteristics are described under different visual angles, the polarization observation is sensitive to the atmosphere, and more atmospheric information can be focused.

The detection method based on the texture and the spatial characteristic of the cloud and the spectral characteristic method belong to a threshold value method essentially, only the detection basis is different, the basis of the spectral characteristic method for carrying out cloud detection on the remote sensing image is a radiation value, namely radiation information, and the basis of the texture and the spatial characteristic method for carrying out cloud detection is the spatial information of the image. With the improvement of the spatial resolution of the remote sensing image, the texture features play an increasingly important role in the processing process of the remote sensing image, and the existence of the cloud in the remote sensing image obviously increases the spatial variability of the radiance, so that the cloud detection by using the texture and the spatial characteristics of the cloud is an effective way. The texture feature method is a cloud detection algorithm mainly based on the difference of spatial information features (reflecting the gray level property and the spatial relationship of images) of cloud and other ground feature images, and at present, the algorithm is mainly applied to remote sensing image data with medium and high resolution, low-resolution images are fused with different types of earth surfaces and cloud pixels due to the fact that pixel resolution is low, and texture features are difficult to distinguish.

The POLDER-3 carried by a French PARASOL satellite launched in 12 months in 2004 is a representative sensor with multi-angle polarization detection capability, and the POLDER-3 stably runs in orbit for about 9 years and finishes the detection task in 10 months in 2013. In 2018, in 5 months, high-score five-number satellites in high-score series in China are successfully transmitted in a satellite transmission center, and a multi-angle polarization imager (DPC) carried on the satellite is similar to the POLDER-3 and is an advanced sensor with multi-band, multi-angle, strength and polarization comprehensive detection capability. Many scholars develop a series of satellite methods based on spectral characteristics for cloud detection based on POLDER-3 and DPC data, and do not fully utilize the spatial information characteristics of cloud and other ground feature images. Meanwhile, compared with POLDER-3, the spatial resolution of DPC is improved from 6km multiplied by 7km to 3.3km multiplied by 3.3km, and the possibility is provided for the cloud detection method based on the spatial texture with higher requirement on the spatial resolution. How to develop a set of cloud detection method for comprehensively utilizing spectral information and spatial information based on a multi-angle polarized satellite sensor independently developed in China, so that the accuracy of remote sensing images is improved, a necessary basis is provided for quantitative remote sensing, and the method is a problem which needs to be solved urgently.

Disclosure of Invention

The invention provides a multi-angle polarized satellite cloud detection method integrating spatial and spectral characteristics, which is used for solving the problem that no cloud detection method suitable for a domestic multi-angle polarized satellite sensor DPC/GF-5 exists in the prior art.

In a first aspect, the invention provides a multi-angle polarized satellite cloud detection method integrating spatial and spectral characteristics, comprising:

acquiring observation data of a multi-angle polarization imager pixel at N-layer angles;

according to observation data of multi-angle polarization imager pixels at each layer angle, acquiring corresponding land cloud identification results, ocean cloud identification results, dazzling ocean cloud identification results, mixed pixel identification results or uncertain pixel identification results of the multi-angle polarization imager pixels at each layer angle, and according to the corresponding land cloud identification results, ocean cloud identification results, dazzling ocean cloud identification results, mixed pixel identification results or uncertain pixel identification results of the multi-angle polarization imager pixels at each layer angle, acquiring corresponding cloud detection results of the multi-angle polarization imager pixels at each layer angle, wherein the cloud detection results comprise clouds, clear sky and uncertain pixels;

determining the times of the multi-angle polarization imager pixels being identified as clouds, the times of the multi-angle polarization imager pixels being identified as clear sky and the times of the multi-angle polarization imager pixels being identified as uncertain pixels under N layers of angles according to the corresponding cloud detection results of the multi-angle polarization imager pixels under each layer of angles, and determining whether the multi-angle polarization imager pixels are clouds or not according to the times of the multi-angle polarization imager pixels being identified as clouds, the times of the multi-angle polarization imager pixels being identified as clear sky and the times of the multi-angle polarization imager pixels being identified as uncertain pixels under the N layers of angles;

the land cloud identification result corresponding to each layer angle of the multi-angle polarization imager pixel is determined according to a land blue light channel reflectivity cloud identification result, a land apparent pressure cloud identification result, a land polarization rainbow cloud identification result, a land reflectivity ratio clear-air identification result, a land two-channel standard deviation and average value difference cloud identification result or a land near-infrared reflectivity and minimum value difference cloud identification result corresponding to each layer angle of the multi-angle polarization imager pixel;

the corresponding ocean cloud identification result of the multi-angle polarization imager pixel at each layer angle is determined according to the corresponding ocean near infrared channel reflectivity cloud identification result, ocean atmospheric molecule optical thickness cloud identification result, ocean polarization rainbow cloud identification result, ocean reflectivity ratio clear air identification result, ocean near infrared standard deviation cloud identification result or ocean near infrared reflectivity and minimum value difference cloud identification result of the multi-angle polarization imager pixel at each layer angle;

the mixed pixel identification result corresponding to the multi-angle polarization imager pixel at each layer angle is determined according to the mixed pixel spatial dual-channel standard deviation cloud identification result corresponding to the multi-angle polarization imager pixel at each layer angle;

and the corresponding dazzling ocean cloud identification result of the multi-angle polarization imager pixel at each layer angle is determined according to the corresponding dazzling linear polarization degree cloud identification result or dazzling apparent pressure cloud identification result of the multi-angle polarization imager pixel at each layer angle.

In a second aspect, the present invention further provides a multi-angle polarized satellite cloud detection apparatus integrating spatial and spectral characteristics, including:

the acquisition module is used for acquiring observation data of the multi-angle polarization imager pixels under N layers of angles;

the cloud detection module is used for acquiring a land cloud identification result, a sea cloud identification result, a dazzling sea cloud identification result, a mixed pixel identification result or an uncertain pixel identification result of a multi-angle polarization imager pixel at each layer angle according to observation data of the multi-angle polarization imager pixel at each layer angle, and acquiring a cloud detection result corresponding to the multi-angle polarization imager pixel at each layer angle according to the land cloud identification result, the sea cloud identification result, the dazzling sea cloud identification result, the mixed pixel identification result or the uncertain pixel identification result of the multi-angle polarization imager pixel at each layer angle, wherein the cloud detection result comprises a cloud, a clear sky and an uncertain pixel;

the calculation module is used for determining the times that the multi-angle polarization imager pixel is identified as a cloud, the times that the multi-angle polarization imager pixel is identified as a clear sky and the times that the multi-angle polarization imager pixel is identified as an uncertain pixel under the angle of N layers according to the corresponding cloud detection result of the multi-angle polarization imager pixel under each layer of angle, and determining whether the multi-angle polarization imager pixel is a cloud or not according to the times that the multi-angle polarization imager pixel is identified as a cloud, the times that the multi-angle polarization imager pixel is identified as a clear sky and the times that the multi-angle polarization imager pixel is identified as an uncertain pixel under the angle of N layers;

In a third aspect, the present invention further provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the program, the method for detecting the multi-angle polarized satellite cloud with integrated spatial and spectral features according to the first aspect is implemented.

In a fourth aspect, the present invention further provides a non-transitory computer readable storage medium, on which a computer program is stored, which when executed by a processor, implements the steps of the multi-angle polarized satellite cloud detection method integrating spatial and spectral features as described in the first aspect.

According to the multi-angle polarization satellite cloud detection method integrating the spatial and spectral characteristics, the spectral characteristics and the spatial characteristics are utilized to realize high-precision detection of global cloud pixels, the problem that detection of cloud pixels above land and sea is missed due to the fact that spectrum information is only relied on is solved, the problem of dazzling cloud detection above sea is solved based on the polarization characteristics of the spectrum, the influence of sea dazzling high-reflectivity information on cloud detection is overcome, and detection precision is effectively improved.

Drawings

In order to more clearly illustrate the technical solutions of the present invention or the prior art, the drawings needed for the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.

FIG. 1 is a schematic flow chart of a multi-angle polarized satellite cloud detection method with comprehensive spatial and spectral characteristics according to the present invention;

FIG. 2 is a second flowchart of the multi-angle polarized satellite cloud detection method with integrated spatial and spectral features according to the present invention;

FIG. 3 is a schematic diagram illustrating a multi-angle detection result fusion strategy of a multi-angle polarized satellite cloud detection method with comprehensive spatial and spectral characteristics according to the present invention;

FIG. 4 is a schematic structural diagram of a multi-angle polarized satellite cloud detection device integrating spatial and spectral characteristics provided by the present invention;

fig. 5 is a schematic structural diagram of an electronic device provided in the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is obvious that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

Fig. 1 is a schematic flow chart of a multi-angle polarized satellite cloud detection method for integrating spatial and spectral characteristics provided by the present invention, and as shown in fig. 1, the multi-angle polarized satellite cloud detection method for integrating spatial and spectral characteristics provided by the present invention includes

steps

110, 120, and 130.

And 110, acquiring observation data of the multi-angle polarization imager pixel at N layers of angles, wherein N is a positive integer greater than 1.

In the embodiment of the present invention, the multi-angle polarization imager may be a high-resolution five-satellite multi-angle polarization imager, or may be a multi-angle polarization imager similar to the high-resolution five-satellite multi-angle polarization imager, which is not limited in the present invention.

The observation data of the multi-angle polarization imager is acquired based on the observation data of the domestic multi-angle polarization satellite sensor DPC/GF-5, and the observation data of each pixel at different angles in the multi-angle polarization imager can be acquired according to the observation data of the multi-angle polarization imager because the multi-angle polarization imager can observe the same pixel at multiple angles, namely the observation data of the pixel of the multi-angle polarization imager at N layers of angles is acquired.

It should be noted that, in the range of the number of observation angle layers that the multi-angle polarization imager can realize, the specific numerical value of the number of layers N of the observation angle of the multi-angle polarization imager can be set according to specific needs.

Step 120, acquiring a land cloud identification result, a sea cloud identification result, a dazzling sea cloud identification result, a mixed pixel identification result or an uncertain pixel identification result of the multi-angle polarization imager pixel at each layer angle according to observation data of the multi-angle polarization imager pixel at each layer angle, and acquiring a cloud detection result corresponding to the multi-angle polarization imager pixel at each layer angle according to the land cloud identification result, the sea cloud identification result, the dazzling sea cloud identification result, the mixed pixel identification result or the uncertain pixel identification result of the multi-angle polarization imager pixel at each layer angle, wherein the cloud detection result comprises a cloud, a clear sky and an uncertain pixel.

Optionally, the land cloud identification result corresponding to each layer angle of the multi-angle polarization imager pixel is determined according to a land blue light channel reflectivity cloud identification result, a land apparent pressure cloud identification result, a land polarization rainbow cloud identification result, a land reflectivity ratio clear-sky identification result, a land two-channel standard deviation and average value difference cloud identification result, or a land near-infrared reflectivity and minimum value difference cloud identification result corresponding to each layer angle of the multi-angle polarization imager pixel.

Optionally, the ocean cloud identification result corresponding to each layer of angle of the multi-angle polarization imager pixel is determined according to the ocean near infrared channel reflectivity cloud identification result, the ocean atmospheric molecule optical thickness cloud identification result, the ocean polarization iris cloud identification result, the ocean reflectivity ratio clear sky identification result, the ocean near infrared standard deviation cloud identification result or the ocean near infrared reflectivity and minimum value difference cloud identification result corresponding to each layer of angle of the multi-angle polarization imager pixel.

Optionally, the mixed pixel judgment result corresponding to the multi-angle polarization imager pixel at each layer angle is determined according to the mixed pixel space two-channel standard deviation cloud judgment result corresponding to the multi-angle polarization imager pixel at each layer angle.

Optionally, the flare ocean cloud identification result corresponding to each layer angle of the multi-angle polarization imager pixel is determined according to the flare linear polarization degree cloud identification result or flare apparent pressure cloud identification result corresponding to each layer angle of the multi-angle polarization imager pixel.

According to the method and the device, the cloud detection result corresponding to the multi-angle polarization imager pixel at each layer angle is obtained according to the corresponding land cloud identification result, ocean cloud identification result, dazzling ocean cloud identification result, mixed pixel identification result or uncertain pixel identification result of the multi-angle polarization imager pixel at each layer angle.

For example, if the land cloud identification result of the multi-angle polarization imager pixel at the first layer angle is cloud, the corresponding cloud detection result of the multi-angle polarization imager pixel at the first layer angle is cloud; and judging that the multi-angle polarization imager pixel is the uncertain pixel by the uncertain pixel judgment result of the multi-angle polarization imager pixel at the fourth layer angle, and judging that the cloud detection result corresponding to the multi-angle polarization imager pixel at the fourth layer angle is the uncertain pixel.

The cloud detection result comprises cloud, clear sky and uncertain pixels, the multi-angle polarization imager pixels are judged through various detection methods, the judgment result of the multi-angle polarization imager pixels at each layer of angle is determined to be the cloud, clear sky or uncertain pixels, and then the corresponding cloud detection result of the multi-angle polarization imager pixels at each layer of angle is determined.

Step 130, determining the times of the multi-angle polarization imager pixels being identified as clouds, the times of the multi-angle polarization imager pixels being identified as clear sky and the times of the multi-angle polarization imager pixels being identified as uncertain pixels under N layers of angles according to the corresponding cloud detection results of the multi-angle polarization imager pixels under each layer of angles, and determining whether the multi-angle polarization imager pixels are clouds or not according to the times of the multi-angle polarization imager pixels being identified as clouds, the times of the multi-angle polarization imager pixels being identified as clear sky and the times of the multi-angle polarization imager pixels being identified as uncertain pixels under the N layers of angles.

According to the cloud detection result corresponding to the multi-angle polarization imager pixel under each layer angle, the times that the multi-angle polarization imager pixel is identified as the cloud, the times that the multi-angle polarization imager pixel is identified as the clear sky and the times that the multi-angle polarization imager pixel is identified as the uncertain pixel under the N layer angle are determined.

And according to the times of the multi-angle polarization imager pixels being identified as clouds under the N layers of angles, the times of the multi-angle polarization imager pixels being identified as clear sky and the times of the multi-angle polarization imager pixels being identified as uncertain pixels, fusing the detection results of the multi-angle polarization imager pixels with the N layers of angles according to a preset rule, and determining whether the multi-angle polarization imager pixels are clouds or not. The sum of the times of being identified as clouds, the times of being identified as clear sky and the times of being identified as uncertain pixels of each multi-angle polarization imager pixel is the same as the number N of the angle layers.

For example, the preset rule may be that the weight of the multi-angle polarization imager pixel identified as a cloud is 2, the weight of the multi-angle polarization imager pixel identified as a clear sky is 1, the weight of the multi-angle polarization imager pixel identified as an uncertain pixel is 0, and the maximum value of the product of the number of times of cloud identification and the corresponding weight, the product of the number of times of clear sky identification and the corresponding weight, and the product of the number of times of uncertain pixel identification and the corresponding weight is taken to determine whether the multi-angle polarization imager pixel is a cloud.

Namely, whether the pixel of the multi-angle polarization imager is cloud or not is determined according to a pixel identification result calculation formula, wherein the pixel identification result calculation formula comprises the following steps:

L＝max(N_cloud×2,N_clear×1,N_uncertain×0)

wherein L is the pixel identification result, N_cloud、N_clearAnd N_uncertainThe number of the pixels of the multi-angle polarization imager, which are identified as cloud, clear sky and uncertain, in N layers of angles is determined.

It should be noted that different preset rules can be set according to different needs, and the preset rules are not specifically limited in the present invention.

The multi-angle polarization satellite cloud detection method integrating the spatial and spectral characteristics, provided by the invention, is used for carrying out multiple detections on multi-angle polarization imager pixels observed by a multi-angle polarization imager at each layer angle on the basis of acquired DPC L1-level data, and determining a cloud detection result corresponding to the multi-angle polarization imager pixels observed by each multi-angle polarization imager at the current layer angle.

Fig. 2 is a second schematic flow chart of the multi-angle polarized satellite cloud detection method with comprehensive spatial and spectral characteristics, as shown in fig. 2, according to the observation data of the multi-angle polarized imager pixel at each layer angle, a land cloud identification result, a marine cloud identification result, a dazzling marine cloud identification result, a mixed pixel identification result or an uncertain pixel identification result corresponding to the multi-angle polarized imager pixel at each layer angle is obtained, which includes: step 121 and step 122.

And step 121, sea and land identification data are obtained, and the pixel category of the multi-angle polarization imager pixels at each layer angle is determined according to the sea and land identification data and the observation data of the multi-angle polarization imager pixels at each layer angle.

According to the embodiment of the invention, sea and land identification data are obtained, and the pixel category of the multi-angle polarization imager pixel at each layer angle is determined according to the sea and land identification data and observation data of the multi-angle polarization imager pixel at each layer angle.

The pixel categories are ocean pixels, land pixels and glaring ocean pixels respectively.

It should be noted that when the pixel types of the multi-angle polarization imager pixels at each layer of angle are determined, the pixels of the multi-angle polarization imager are not interfered with each other, and the pixel types are determined only according to the pixels of the multi-angle polarization imager.

Step 122, according to the pixel category of the multi-angle polarization imager pixel at each layer of angle, executing the following steps on the multi-angle polarization imager pixel at each layer of angle, including: step 1221, step 1222, step 1223, and step 1224.

Step 1221, in a case that a pixel category of a multi-angle polarization imager pixel at a current angle is a land pixel, if at least one of a land blue light channel reflectivity cloud judgment result, a land apparent pressure cloud judgment result, a land polarization iridescence cloud judgment result and a land reflectivity ratio clear sky judgment result corresponding to the multi-angle polarization imager pixel at the current angle is judged to be cloud or clear sky according to the at least one judgment result, determining that the land cloud judgment result of the multi-angle polarization imager pixel at the current angle is cloud or clear sky, otherwise determining that the multi-angle polarization imager pixel at the current angle is an uncertain pixel.

As shown in fig. 2, blue light reflectivity difference detection, apparent pressure detection, polarization rainbow detection and reflectivity ratio detection are sequentially performed on the multi-angle polarization imager pixel at the current angle, so as to obtain a land blue light channel reflectivity cloud identification result, a land apparent pressure cloud identification result, a land polarization rainbow cloud identification result and a land reflectivity ratio clear sky identification result corresponding to the multi-angle polarization imager pixel at the current angle.

The method comprises the steps that different detection is carried out on multi-angle polarization imager pixels at the current angle in sequence to obtain judgment and identification results, and if a certain judgment result directly determines the land cloud judgment and identification result of the multi-angle polarization imager pixels at the current angle in the judgment and identification process, subsequent detection is not carried out.

For example, under the condition that the land blue light channel reflectivity cloud identification result of the multi-angle polarization imager pixel at the current angle is cloud, the land cloud identification result of the multi-angle polarization imager pixel at the current angle is directly determined to be cloud, and subsequent detection is not performed on the multi-angle polarization imager pixel at the current angle.

And under the condition that the land reflectivity ratio of the multi-angle polarization imager pixel at the current angle is clear sky, directly determining that the land cloud judgment result of the multi-angle polarization imager pixel at the current angle is clear sky, and not carrying out subsequent detection on the multi-angle polarization imager pixel at the current angle.

In other embodiments of the present invention, blue light reflectivity difference detection, apparent pressure detection, polarization rainbow detection, and reflectivity ratio detection may also be performed on the multi-angle polarization imager pixel at the current angle at the same time, and according to the identification results corresponding to the above four detections, it is determined that the multi-angle polarization imager pixel at the current angle is a cloud, clear sky, or uncertain pixel.

If the pixel of the multi-angle polarization imager pixel at the current angle is judged to be cloud or clear sky according to at least one judgment result of the land blue light channel reflectivity cloud judgment result, the land apparent pressure intensity cloud judgment result, the land polarization rainbow cloud judgment result and the land reflectivity ratio clear sky judgment result corresponding to the multi-angle polarization imager pixel at the current angle, determining that the land cloud judgment result of the multi-angle polarization imager pixel at the current angle is cloud or clear sky, and otherwise determining that the multi-angle polarization imager pixel at the current angle is uncertain.

And under the condition that the multi-angle polarization imager pixels at the current angle are uncertain pixels, classifying the multi-angle polarization imager pixels at the current angle again through the space window, and performing subsequent detection.

Step 1222, under the condition that the pixel category of the multi-angle polarization imager pixel at the current angle is an ocean pixel, if at least one of the ocean near infrared channel reflectivity cloud judgment result, the ocean atmospheric molecular optical thickness cloud judgment result, the ocean polarization rainbow cloud judgment result and the ocean reflectivity ratio clear sky judgment result corresponding to the multi-angle polarization imager pixel at the current angle is judged to be cloud or clear sky according to the at least one judgment result of the ocean near infrared channel reflectivity cloud judgment result, the ocean atmospheric molecular optical thickness cloud judgment result, the ocean polarization rainbow cloud judgment result and the ocean reflectivity ratio clear sky judgment result corresponding to the multi-angle polarization imager pixel at the current angle, determining that the ocean cloud judgment result of the multi-angle polarization imager pixel at the current angle is cloud or clear sky, otherwise determining that the multi-angle polarization imager pixel at the current angle is an uncertain pixel.

As shown in fig. 2, near-infrared reflectivity difference detection, atmospheric molecular optical thickness detection, polarization rainbow detection and reflectivity ratio detection are sequentially performed on the multi-angle polarization imager pixel at the current angle, so as to obtain an ocean near-infrared channel reflectivity cloud identification result, an ocean atmospheric molecular optical thickness cloud identification result, an ocean polarization rainbow cloud identification result and an ocean reflectivity ratio clear sky identification result corresponding to the multi-angle polarization imager pixel at the current angle.

The method comprises the steps of sequentially carrying out different detections on multi-angle polarization imager pixels at the current angle to obtain judgment and identification results, and if a certain judgment and identification result directly determines the ocean cloud judgment and identification result of the multi-angle polarization imager pixels at the current angle, not carrying out subsequent detections.

For example, when the ocean near infrared channel reflectivity cloud identification result of the multi-angle polarization imager pixel at the current angle is cloud, the ocean cloud identification result of the multi-angle polarization imager pixel at the current angle is directly determined to be cloud, and subsequent detection is not performed on the multi-angle polarization imager pixel at the current angle.

And under the condition that the ocean reflectivity ratio of the multi-angle polarization imager pixel at the current angle is clear sky, directly determining that the ocean cloud judgment result of the multi-angle polarization imager pixel at the current angle is clear sky, and not carrying out subsequent detection on the multi-angle polarization imager pixel at the current angle.

In other embodiments of the present invention, near infrared reflectivity difference detection, atmospheric molecule optical thickness detection, polarization rainbow detection, and reflectivity ratio detection may also be performed on the multi-angle polarization imager pixel at the current angle at the same time, and according to the identification results corresponding to the above four detections, it is determined that the multi-angle polarization imager pixel at the current angle is a cloud, clear sky, or uncertain pixel.

Step 1223, under the condition that the pixel category of the multi-angle polarization imager pixel at the current angle is a dazzling ocean pixel, if the pixel of the multi-angle polarization imager pixel at the current angle is judged to be a cloud according to at least one judgment result of a dazzling light linear polarization degree cloud judgment result and a dazzling apparent pressure intensity cloud judgment result corresponding to the multi-angle polarization imager pixel at the current angle, determining that the dazzling ocean cloud judgment result of the multi-angle polarization imager pixel at the current angle is a cloud, and otherwise determining an uncertain pixel judgment result of the multi-angle polarization imager pixel at the current angle;

as shown in fig. 2, line polarization degree detection and apparent pressure intensity detection are sequentially performed on the multi-angle polarization imager pixel at the current angle, so as to obtain a flare line polarization degree cloud identification result and a flare apparent pressure intensity cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle.

The method comprises the steps of sequentially carrying out different detections on multi-angle polarization imager pixels at the current angle to obtain judgment results, and if a certain judgment result directly determines the dazzling ocean cloud judgment result of the multi-angle polarization imager pixels at the current angle in the judgment process, not carrying out subsequent detections.

For example, under the condition that the judgment result of the cloud of the flare linear polarization degree of the multi-angle polarization imager pixel at the current angle is cloud, the judgment result of the ocean cloud of the multi-angle polarization imager pixel at the current angle is directly determined to be cloud, and subsequent detection is not performed on the multi-angle polarization imager pixel at the current angle.

In other embodiments of the present invention, the line polarization degree detection and the apparent pressure detection may also be performed on the multi-angle polarization imager pixel at the current angle at the same time, and the multi-angle polarization imager pixel at the current angle is determined to be a cloud or an uncertain pixel according to the identification result corresponding to the above detection. Determining that the ocean cloud identification result of the multi-angle polarization imager pixel at the current angle is cloud under the condition that the multi-angle polarization imager pixel at the current angle is cloud, and determining the uncertain pixel identification result of the multi-angle polarization imager pixel at the current angle under the condition that the multi-angle polarization imager pixel at the current angle is uncertain pixel.

Step 1224, in a case that the multi-angle polarization imager pixel at the current angle is an uncertain pixel, establishing a spatial window centered on the multi-angle polarization imager pixel at the current angle according to a preset window size, determining a plurality of identification pixels included in the spatial window, and performing the following steps: step 12241, step 12242, step 12243.

As shown in fig. 2, the step 1224 includes

steps

12241, 12242, and 12243, after performing blue light reflectance difference detection, apparent pressure detection, polarization rainbow detection, reflectance ratio detection, near-infrared reflectance difference detection, and atmospheric molecular optical thickness detection, obtaining that the multi-angle polarization imager pixels at the current angle are indeterminate pixels, and for the indeterminate pixels, with the indeterminate pixels as a center, establishing a spatial window of M × M size with the multi-angle polarization imager pixels at the current angle as a center, determining a plurality of identification pixels included in the spatial window, and re-determining the pixel categories of the indeterminate pixels by using pixel categories corresponding to the plurality of identification pixels in the spatial window.

The preset window size may be determined according to actual requirements, for example, the preset window size may be set to 3 × 3, or 5 × 5. That is, the size of the spatial window may be set to 3 × 3 or 5 × 5, and the present invention is not limited thereto.

It should be noted that the plurality of identification pixels included in the spatial window also includes the uncertain pixel. I.e. the uncertain picture element is one of a plurality of identification picture elements.

In the embodiment of the invention, in the identification process of a single multi-angle polarization imager pixel, the cloud detection result of the multi-angle polarization imager pixel in the cloud edge area is not easily obtained through blue light reflectivity difference detection, apparent pressure detection, polarization rainbow detection, reflectivity ratio detection, near infrared reflectivity difference detection and atmospheric molecule optical thickness detection.

Step 12241, in a case that a pixel category of each identification pixel of the space window is a land pixel, determining that the category of the multi-angle polarization imager pixel at the current angle is a land pixel, if the pixel of the multi-angle polarization imager pixel at the current angle is determined to be a cloud according to at least one of a land dual-channel standard deviation and average difference cloud determination result corresponding to the multi-angle polarization imager pixel at the current angle and a land near-infrared reflectivity and minimum difference cloud determination result, determining that the pixel of the multi-angle polarization imager pixel at the current angle is a cloud, otherwise determining that the multi-angle polarization imager pixel at the current angle is an uncertain pixel.

Step 12242, in a case that the pixel category of each identification pixel of the space window is an ocean pixel, determining that the category of the multi-angle polarization imager pixel at the current angle is an ocean pixel, if the ocean cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is at least one of the ocean near infrared standard deviation cloud identification result and the ocean near infrared reflectivity and minimum value difference cloud identification result, determining that the pixel of the multi-angle polarization imager pixel at the current angle is a cloud, and otherwise determining that the multi-angle polarization imager pixel at the current angle is an uncertain pixel.

Step 12243, under the condition that the pixel of the identification pixel in the space window includes both an ocean pixel and a land pixel, determining that the type of the multi-angle polarization imager pixel at the current angle is a mixed pixel, if the pixel of the multi-angle polarization imager pixel at the current angle is judged to be cloud according to the mixed pixel space dual-channel standard deviation cloud judgment result corresponding to the multi-angle polarization imager pixel at the current angle, determining that the mixed pixel judgment result of the multi-angle polarization imager pixel at the current angle is cloud, otherwise determining that the multi-angle polarization imager pixel at the current angle is an uncertain pixel.

In step 12241, step 12242, and step 12243, if it is determined that the multi-angle polarization imager pixel at the current angle is an uncertain pixel, determining an uncertain pixel judgment result corresponding to the multi-angle polarization imager pixel at the current angle.

Optionally, the land blue light channel reflectivity cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

obtaining the blue light wave band apparent reflectivity data of the minimum observation zenith angle of the pixel of the multi-angle polarization imager at the current angle, and performing atmospheric molecule scattering correction;

and carrying out cloud identification on the multi-angle polarization imager pixel at the current angle by a blue light channel reflectivity threshold method to obtain a land blue light channel reflectivity cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle.

It should be noted that, in the sky above the land, most of the ground features have a low reflectivity in the blue light band, and the cloud reflectivity is high, so that the cloud and clear sky pixels can be effectively distinguished by using the reflectivity difference in the blue light band.

Optionally, the land apparent pressure cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

acquiring 763nm and 765nm apparent reflectivity data and observation geometric data of the multi-angle polarization imager pixel at the current angle, and calculating the apparent pressure corresponding to the multi-angle polarization imager pixel at the current angle;

acquiring 865nm and 670nm apparent reflectivity data and observation geometric data of a multi-angle polarization imager pixel at a current angle, and calculating a normalized vegetation index corresponding to the multi-angle polarization imager pixel at the current angle;

and performing cloud judgment on the multi-angle polarization imager pixel at the current angle according to the apparent pressure and the normalized vegetation index corresponding to the multi-angle polarization imager pixel at the current angle to obtain a land apparent pressure cloud judgment result corresponding to the multi-angle polarization imager pixel at the current angle.

It should be noted that the apparent pressure detection is performed by using the height difference between the cloud and the clear sky ground object, the apparent pressure of the clear sky pixel is about the earth surface pressure (the pressure at the standard sea level height is about 1013hPa), the value is high, and the cloud is at a certain height, the value of the apparent pressure is low. Thus, a higher cloud layer is more advantageous for the detection of the apparent pressure criterion, even if the cloud layer is thinner.

Alternatively, the apparent pressure is calculated by the formula:

wherein, P₀For sea level pressure 1013.25hPa, C, A1, A2, A3 and A4 are polynomial fitting coefficients 192.33, -837.93, 1376.38, -1022.08, 288.35, R₇₆₃、R₇₆₅Is 763nm and 765nm in apparent reflectivity,

the scattering reflectivity of 763nm and 765nm atmospheric molecules, and m is the mass number of the atmosphere.

Optionally, the land polarization iris cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

acquiring 865nm polarization component data and observation geometric data of a multi-angle polarization imager pixel at a current angle, calculating a polarization reflectivity and correcting the observation geometric data;

and carrying out cloud judgment on the multi-angle polarization imager pixel at the current angle according to the polarization reflectivity of the scattering angle within the range of 135-150 degrees, and obtaining a land polarization rainbow cloud judgment result corresponding to the multi-angle polarization imager pixel at the current angle.

Alternatively, the calculation formula of the polarized reflectance is as follows:

R′_p865＝(μ_s+μ_v)×R_p865

wherein R is_Q865、R_U865Polarization Q component and U component of 865nm, R_p865Is 865nm polarized reflectivity, R'_p865Is the polarized reflectance corrected for the observation geometry.

Optionally, the clear-sky judgment result of the land reflectivity ratio corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

acquiring 865nm and 443nm apparent reflectivity data and observation geometric data of a multi-angle polarization imager pixel at a current angle, and calculating a reflectivity ratio;

and carrying out cloud judgment on the multi-angle polarization imager pixel at the current angle according to a reflectivity ratio method to obtain a clear and empty judgment result of the land reflectivity ratio corresponding to the multi-angle polarization imager pixel at the current angle.

It should be noted that the reflectivity of a general land pixel generally has a certain variation characteristic in visible light-near infrared, and the reflection of the near infrared band is significantly higher than that of a short band, while the reflectivity of a cloud pixel remains stable in visible light-near infrared, so that the cloud and the ground object can be distinguished by using the reflectivity ratio.

Optionally, the marine near-infrared channel reflectivity cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

acquiring near-infrared band apparent reflectivity data and observation geometric data of a multi-angle polarization imager pixel at a current angle, and correcting atmospheric molecular scattering;

carrying out cloud judgment on the multi-angle polarization imager pixel at the current angle according to a near-infrared channel reflectivity threshold method to obtain a marine near-infrared channel reflectivity cloud judgment result corresponding to the multi-angle polarization imager pixel at the current angle;

it should be noted that, in the air above the ocean, the water body has a lower reflectivity in the near-infrared band, and the cloud reflectivity is higher, so that the cloud and clear sky pixels can be effectively distinguished by using the reflectivity difference of the near-infrared band.

Optionally, the marine atmospheric molecular optical thickness cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

acquiring 490nm polarization component data and observation geometric data of a multi-angle polarization imager pixel at a current angle, and calculating the polarization reflectivity and the atmospheric molecular optical thickness;

and carrying out cloud identification on the multi-angle polarization imager pixel at the current angle according to the atmospheric molecular optical thickness with the scattering angle within the range of 80-120 degrees, and obtaining an ocean atmospheric molecular optical thickness cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle.

Alternatively, the atmospheric molecular optical thickness calculation formula is as follows:

wherein, tau₄₉₀Optical thickness of 490nm atmospheric molecule, R_p490Is 490nm polarization reflectance.

Optionally, the marine polarization iris cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

according to the polarized reflectivity of the scattering angle within the range of 135-150 degrees, cloud judgment is carried out on the multi-angle polarization imager pixel under the current angle, and an ocean polarization rainbow cloud judgment result corresponding to the multi-angle polarization imager pixel under the current angle is obtained;

optionally, the judgment result of the clear sky based on the ocean reflectivity ratio corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

and carrying out angle-by-angle cloud identification on the multi-angle polarization imager pixels at the current angle according to a reflectivity ratio method to obtain a clear sky identification result of the ocean reflectivity ratio corresponding to the multi-angle polarization imager pixels at the current angle.

It should be noted that the reflectivity of the general ocean pixel generally has a certain change characteristic in the visible light-near infrared, the reflection of the near infrared band is significantly lower than that of the short band, and the reflectivity of the cloud pixel remains stable in the visible light-near infrared, so that the cloud and the ground objects can be distinguished by using the reflectivity ratio.

Optionally, the flare linear polarization degree cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

acquiring 865nm polarization component data and observation geometric data of a multi-angle polarization imager pixel at a current angle, and calculating a flare angle and a linear polarization degree;

and performing cloud identification on the multi-angle polarization imager pixel at the current angle according to the linear polarization degree of the flare angle within the range of 0-30 degrees to obtain a flare linear polarization degree cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle.

Optionally, the flare apparent pressure cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

acquiring 763nm and 765nm apparent reflectivity data and observation geometric data of the multi-angle polarization imager pixel at the current angle, and calculating apparent pressure;

and carrying out cloud judgment on the multi-angle polarization imager pixel at the current angle according to an apparent pressure method to obtain a dazzling apparent pressure cloud judgment result corresponding to the multi-angle polarization imager pixel at the current angle.

Optionally, the land dual-channel standard deviation and average difference cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

obtaining 443nm and 763nm apparent reflectivity data of multi-angle polarization imager pixels at a front angle to obtain reflectivity data corresponding to each identification pixel, and calculating a standard deviation and an average value of reflectivity corresponding to the multi-angle polarization imager pixels at a current angle according to the reflectivity data corresponding to each identification pixel;

and obtaining a land dual-channel standard deviation and average value difference cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle according to the difference between the standard deviation and the average value of the reflectivity corresponding to the multi-angle polarization imager pixel at the current angle.

For example, the average value of the multi-angle polarization imager pixels at the current angle is obtained through an average value formula of the multi-angle polarization imager pixels at the current angle, and the average value formula of the multi-angle polarization imager pixels at the current angle is as follows:

wherein R is_1,1Is the reflectivity, R, of the 1 st row and 1 st column of picture elements in an MxM window_1,2Is the reflectivity, R, of the row 1 and column 2 pixels in an MxM window_m,mIs the reflectivity of the M row and M column picture elements in the M x M window.

The standard deviation of the multi-angle polarization imager pixel at the current angle is obtained through a standard deviation formula of the multi-angle polarization imager pixel at the current angle, and the standard deviation formula of the multi-angle polarization imager pixel at the current angle is as follows:

wherein R is_1,1Is the reflectivity, R, of the 1 st row and 1 st column element in an MxM window_1,2Is the reflectivity, R, of the row 1 and column 2 pixels in an MxM window_m,mIs the reflectivity of the M-th row and M-th column of pixels in the M × M window, and Mean is the average of M × M pixels in the M × M window.

Optionally, the land near-infrared reflectivity and minimum value difference cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

acquiring 763nm apparent reflectivity data of the multi-angle polarization imager pixel at the current angle to obtain reflectivity data corresponding to each identification pixel, and calculating the minimum value of the reflectivity corresponding to the multi-angle polarization imager pixel at the current angle according to the reflectivity data corresponding to each identification pixel;

and obtaining a land near-infrared reflectivity and minimum value difference cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle according to the difference between the reflectivity and the minimum value corresponding to the multi-angle polarization imager pixel at the current angle.

It should be noted that, according to the reflectivity data corresponding to each identification pixel, the minimum value of the reflectivity data corresponding to the identification pixel in the spatial window may be determined, so as to obtain the minimum value of the reflectivity corresponding to the pixel of the multi-angle polarization imager at the current angle.

Optionally, the ocean near infrared standard deviation cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

reading 763nm apparent reflectivity data of the multi-angle polarization imager pixels at the current angle to obtain reflectivity data corresponding to each identification pixel, and calculating the standard deviation of the reflectivity corresponding to the multi-angle polarization imager pixels at the current angle according to the reflectivity data corresponding to each identification pixel;

and obtaining a marine near-infrared standard deviation cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle according to the standard deviation of the reflectivity corresponding to the multi-angle polarization imager pixel at the current angle.

It should be noted that, according to the reflectivity data corresponding to each identification pixel, the standard deviation corresponding to a plurality of identification pixels in the spatial window is calculated, so as to obtain the standard deviation of the reflectivity corresponding to the multi-angle polarization imager pixel at the current angle.

Optionally, the ocean near-infrared reflectivity and minimum value difference cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

763nm apparent reflectivity data of the multi-angle polarization imager pixels at the current angle are obtained, reflectivity data corresponding to each identification pixel are obtained, and the minimum value of the reflectivity corresponding to the multi-angle polarization imager pixels at the current angle is calculated according to the reflectivity data corresponding to each identification pixel;

and obtaining a cloud identification result of the difference between the ocean near-infrared reflectivity and the minimum value corresponding to the multi-angle polarization imager pixel at the current angle according to the difference between the reflectivity and the minimum value corresponding to the multi-angle polarization imager pixel at the current angle.

It should be noted that according to the reflectivity data corresponding to each identification pixel, the minimum value of the reflectivity data corresponding to the identification pixel in the spatial window can be determined, so as to obtain the minimum value of the reflectivity corresponding to the multi-angle polarization imager pixel at the current angle.

Optionally, the mixed pixel spatial dual-channel standard deviation cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

obtaining 443nm and 763nm apparent reflectivity data and sea and land identification data of multi-angle polarization imager pixels at the current angle, obtaining reflectivity data and sea and land identification data of each identification pixel, and calculating a standard deviation of reflectivity corresponding to the multi-angle polarization imager pixels at the current angle according to the reflectivity data and the sea and land identification data of each identification pixel;

and obtaining a mixed pixel space dual-channel standard deviation cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle according to the standard deviation of the reflectivity corresponding to the multi-angle polarization imager pixel at the current angle.

It should be noted that, according to the reflectivity data of each identification pixel and the sea-land identification data, the standard deviation of the reflectivity corresponding to the identification pixel in the space window is calculated, so as to obtain the standard deviation of the reflectivity corresponding to the multi-angle polarization imager pixel at the current angle.

Fig. 3 is a schematic diagram of a multi-angle detection result fusion strategy of the multi-angle polarized satellite cloud detection method with comprehensive spatial and spectral characteristics, as shown in fig. 3, a gray square in the diagram indicates that a cloud detection result of a multi-angle polarized imager pixel at a current angle is cloud, a white square indicates that a cloud detection result of a multi-angle polarized imager pixel at a current angle is clear sky, a dotted square indicates that a cloud detection result of a multi-angle polarized imager pixel at a current angle is an uncertain pixel, and cloud detection results at multi-layer angles are fused to obtain a final cloud detection result.

Fig. 4 is a schematic structural diagram of a multi-angle polarized satellite cloud detection device with integrated spatial and spectral characteristics provided by the present invention, and as shown in fig. 4, the multi-angle polarized satellite cloud detection device with integrated spatial and spectral characteristics provided by the present invention includes:

the obtaining module 410 is used for obtaining observation data of the multi-angle polarization imager pixels under N layers of angles;

the cloud detection module 420 is configured to obtain a land cloud identification result, a sea cloud identification result, a dazzling sea cloud identification result, a mixed pixel identification result or an uncertain pixel identification result corresponding to a multi-angle polarization imager pixel at each layer angle according to observation data of the multi-angle polarization imager pixel at each layer angle, and obtain a cloud detection result corresponding to the multi-angle polarization imager pixel at each layer angle according to the land cloud identification result, the sea cloud identification result, the dazzling sea cloud identification result, the mixed pixel identification result or the uncertain pixel identification result corresponding to the multi-angle polarization imager pixel at each layer angle, where the cloud detection result includes a cloud, a clear sky and an uncertain pixel;

the calculating module 430 is used for determining the times that the multi-angle polarization imager pixel is identified as a cloud, the times that the multi-angle polarization imager pixel is identified as a clear sky and the times that the multi-angle polarization imager pixel is identified as an uncertain pixel under the angle of N layers according to the corresponding cloud detection result of the multi-angle polarization imager pixel under each layer of angle, and determining whether the multi-angle polarization imager pixel is a cloud or not according to the times that the multi-angle polarization imager pixel is identified as a cloud, the times that the multi-angle polarization imager pixel is identified as a clear sky and the times that the multi-angle polarization imager pixel is identified as an uncertain pixel under the angle of N layers;

Optionally, the obtaining, according to the observation data of the multi-angle polarization imager pixel at each layer angle, a land cloud identification result, a sea cloud identification result, a dazzling sea cloud identification result, a mixed pixel identification result, or an uncertain pixel identification result corresponding to the multi-angle polarization imager pixel at each layer angle includes:

obtaining sea and land identification data, and determining the pixel category of the multi-angle polarization imager pixel at each layer angle according to the sea and land identification data and the observation data of the multi-angle polarization imager pixel at each layer angle;

according to the pixel category of the multi-angle polarization imager pixel under each layer of angle, the following steps are executed to the multi-angle polarization imager pixel under each layer of angle:

under the condition that the pixel category of a multi-angle polarization imager pixel at the current angle is a land pixel, if the pixel of the multi-angle polarization imager pixel at the current angle is judged to be cloud or clear sky according to at least one judgment result in a land blue light channel reflectivity cloud judgment result, a land apparent pressure intensity cloud judgment result, a land polarization rainbow cloud judgment result and a land reflectivity ratio clear sky judgment result corresponding to the multi-angle polarization imager pixel at the current angle, determining that the land cloud judgment result of the multi-angle polarization imager pixel at the current angle is cloud or clear sky, and otherwise determining that the multi-angle polarization imager pixel at the current angle is an uncertain pixel;

under the condition that the pixel category of a multi-angle polarization imager pixel at the current angle is an ocean pixel, if at least one of an ocean near infrared channel reflectivity cloud judgment result, an ocean atmospheric molecular optical thickness cloud judgment result, an ocean polarization iridescence cloud judgment result and an ocean reflectivity ratio clear sky judgment result which correspond to the multi-angle polarization imager pixel at the current angle is judged to be cloud or clear sky, determining that the ocean cloud judgment result of the multi-angle polarization imager pixel at the current angle is cloud or clear sky, otherwise determining that the multi-angle polarization imager pixel at the current angle is uncertain;

under the condition that the pixel category of a multi-angle polarization imager pixel at the current angle is a dazzling ocean pixel, if the pixel of the multi-angle polarization imager pixel at the current angle is judged to be a cloud according to at least one judgment result of a dazzling linear polarization degree cloud judgment result and a dazzling apparent pressure intensity cloud judgment result corresponding to the multi-angle polarization imager pixel at the current angle, determining that the dazzling ocean cloud judgment result of the multi-angle polarization imager pixel at the current angle is a cloud, and otherwise determining the uncertain pixel judgment result of the multi-angle polarization imager pixel at the current angle;

under the condition that the multi-angle polarization imager pixel at the current angle is an uncertain pixel, establishing a space window taking the multi-angle polarization imager pixel at the current angle as a center according to the size of a preset window, determining a plurality of identification pixels contained in the space window, and executing the following steps:

determining that the type of the multi-angle polarization imager pixel at the current angle is a land pixel under the condition that the pixel type of each identification pixel of the space window is a land pixel, if the pixel of the multi-angle polarization imager pixel at the current angle is determined to be a cloud according to at least one of a land dual-channel standard deviation and average value difference cloud determination result corresponding to the multi-angle polarization imager pixel at the current angle and a land near-infrared reflectivity and minimum value difference cloud determination result, determining that the land cloud determination result of the multi-angle polarization imager pixel at the current angle is a cloud, otherwise determining that the multi-angle polarization imager pixel at the current angle is an uncertain pixel;

determining that the type of the multi-angle polarization imager pixel at the current angle is an ocean pixel under the condition that the pixel type of each identification pixel of the space window is an ocean pixel, if the pixel of the multi-angle polarization imager pixel at the current angle is determined to be a cloud according to at least one of an ocean near infrared standard deviation cloud determination result corresponding to the multi-angle polarization imager pixel at the current angle and an ocean near infrared reflectivity and minimum value difference cloud determination result, determining that the ocean cloud determination result of the multi-angle polarization imager pixel at the current angle is a cloud, and otherwise determining that the multi-angle polarization imager pixel at the current angle is an uncertain pixel;

and under the condition that the pixel of the identification pixel in the space window comprises both an ocean pixel and a land pixel, determining that the type of the multi-angle polarization imager pixel at the current angle is a mixed pixel, if the pixel of the multi-angle polarization imager pixel at the current angle is judged to be cloud according to a mixed pixel space double-channel standard deviation cloud judgment result corresponding to the multi-angle polarization imager pixel at the current angle, determining that the mixed pixel judgment result of the multi-angle polarization imager pixel at the current angle is cloud, and otherwise determining that the multi-angle polarization imager pixel at the current angle is an uncertain pixel.

Optionally, the cloud identification result of the reflectivity of the terrestrial blue light channel corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

carrying out cloud identification on the multi-angle polarization imager pixel at the current angle through a blue light channel reflectivity threshold value method to obtain a cloud identification result of the terrestrial blue light channel reflectivity corresponding to the multi-angle polarization imager pixel at the current angle;

the land apparent pressure cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

acquiring 763nm and 765nm apparent reflectivity data and observation geometric data of the multi-angle polarization imager pixel at the current angle, and calculating the corresponding apparent pressure of the multi-angle polarization imager pixel at the current angle;

performing cloud judgment on the multi-angle polarization imager pixel at the current angle according to the apparent pressure and the normalized vegetation index corresponding to the multi-angle polarization imager pixel at the current angle to obtain a land apparent pressure cloud judgment result corresponding to the multi-angle polarization imager pixel at the current angle;

the land polarization rainbow cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

according to the polarized reflectivity of the scattering angle within the range of 135-150 degrees, cloud identification is carried out on the multi-angle polarization imager pixel under the current angle, and a land polarization rainbow cloud identification result corresponding to the multi-angle polarization imager pixel under the current angle is obtained;

the clear-sky judgment result of the land reflectivity ratio corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

the marine atmospheric molecule optical thickness cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

carrying out cloud judgment on the multi-angle polarization imager pixel at the current angle according to the atmospheric molecular optical thickness with the scattering angle within the range of 80-120 degrees to obtain an ocean atmospheric molecular optical thickness cloud judgment result corresponding to the multi-angle polarization imager pixel at the current angle;

the marine polarization rainbow cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

the ocean reflectivity ratio clear sky judgment result corresponding to the multi-angle polarization imager pixel under the current angle is obtained through the following steps:

Optionally, the judgment result of the cloud of the polarization degree of the flare light corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

performing cloud judgment on the multi-angle polarization imager pixel at the current angle according to the linear polarization degree of the flare angle within the range of 0-30 degrees to obtain a flare linear polarization degree cloud judgment result corresponding to the multi-angle polarization imager pixel at the current angle;

the flare apparent pressure cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

obtaining 443nm and 763nm apparent reflectivity data of the multi-angle polarization imager pixel at the previous angle to obtain reflectivity data corresponding to each identification pixel, and calculating the standard deviation and the average value of the reflectivity corresponding to the multi-angle polarization imager pixel at the current angle according to the reflectivity data corresponding to each identification pixel;

according to the difference between the standard deviation and the average value of the reflectivity corresponding to the multi-angle polarization imager pixel at the current angle, obtaining a land dual-channel standard deviation and average value difference cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle;

the land near-infrared reflectivity and minimum value difference cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

obtaining a marine near-infrared standard deviation cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle according to the standard deviation of the reflectivity corresponding to the multi-angle polarization imager pixel at the current angle;

the ocean near infrared reflectivity and minimum value difference cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

obtaining 443nm and 763nm apparent reflectivity data and sea and land identification data of a multi-angle polarization imager pixel at a current angle to obtain reflectivity data and sea and land identification data of each identification pixel, and calculating a standard deviation of reflectivity corresponding to the multi-angle polarization imager pixel at the current angle according to the reflectivity data and the sea and land identification data of each identification pixel;

It should be noted that the multi-angle polarized satellite cloud detection device with integrated spatial and spectral characteristics provided in the embodiments of the present invention can implement all the method steps implemented by the multi-angle polarized satellite cloud detection method with integrated spatial and spectral characteristics, and can achieve the same technical effects, which are not described herein again.

The multi-angle polarization satellite cloud detection device integrating the spatial and spectral characteristics, provided by the invention, realizes high-precision detection of global cloud pixels by using the spectral characteristics and the spatial characteristics, solves the problem of detection leakage of the cloud pixels above land and sea caused by only depending on spectral information, solves the problem of dazzling cloud detection above sea based on the polarization characteristics of the spectrum, overcomes the influence of sea dazzling high-reflectivity information on cloud detection, and effectively improves the detection precision.

Fig. 5 illustrates a physical structure diagram of an electronic device, which may include, as shown in fig. 5: a processor (processor)510, a communication Interface (Communications Interface)520, a memory (memory)530 and a communication bus 540, wherein the processor 510, the communication Interface 520 and the memory 530 communicate with each other via the communication bus 540. Processor 510 may invoke logic instructions in memory 530 to perform a method of multi-angle polarized satellite cloud detection of integrated spatial and spectral features, the method comprising: acquiring observation data of a multi-angle polarization imager pixel at N-layer angles; according to observation data of multi-angle polarization imager pixels at each layer angle, acquiring corresponding land cloud identification results, ocean cloud identification results, dazzling ocean cloud identification results, mixed pixel identification results or uncertain pixel identification results of the multi-angle polarization imager pixels at each layer angle, and according to the corresponding land cloud identification results, ocean cloud identification results, dazzling ocean cloud identification results, mixed pixel identification results or uncertain pixel identification results of the multi-angle polarization imager pixels at each layer angle, acquiring corresponding cloud detection results of the multi-angle polarization imager pixels at each layer angle, wherein the cloud detection results comprise clouds, clear sky and uncertain pixels; determining the times that the multi-angle polarization imager pixel is identified as cloud, the times that the multi-angle polarization imager pixel is identified as clear sky and the times that the multi-angle polarization imager pixel is identified as uncertain pixel under N layers of angles according to the corresponding cloud detection result of the multi-angle polarization imager pixel under each layer of angles, and determining whether the multi-angle polarization imager pixel is cloud or not according to the times that the multi-angle polarization imager pixel is identified as cloud, the times that the multi-angle polarization imager pixel is identified as clear sky and the times that the multi-angle polarization imager pixel is identified as uncertain pixel under the N layers of angles; the land cloud identification result corresponding to each layer angle of the multi-angle polarization imager pixel is determined according to a land blue light channel reflectivity cloud identification result, a land apparent pressure cloud identification result, a land polarization rainbow cloud identification result, a land reflectivity ratio clear-air identification result, a land two-channel standard deviation and average value difference cloud identification result or a land near-infrared reflectivity and minimum value difference cloud identification result corresponding to each layer angle of the multi-angle polarization imager pixel; the corresponding ocean cloud identification result of the multi-angle polarization imager pixel at each layer angle is determined according to the corresponding ocean near infrared channel reflectivity cloud identification result, ocean atmospheric molecule optical thickness cloud identification result, ocean polarization rainbow cloud identification result, ocean reflectivity ratio clear air identification result, ocean near infrared standard deviation cloud identification result or ocean near infrared reflectivity and minimum value difference cloud identification result of the multi-angle polarization imager pixel at each layer angle; the mixed pixel identification result corresponding to the multi-angle polarization imager pixel at each layer angle is determined according to the mixed pixel spatial dual-channel standard deviation cloud identification result corresponding to the multi-angle polarization imager pixel at each layer angle; the flare ocean cloud identification result corresponding to each layer of angle of the multi-angle polarization imager pixel is determined according to the flare linear polarization degree cloud identification result or flare apparent pressure intensity cloud identification result corresponding to each layer of angle of the multi-angle polarization imager pixel.

Furthermore, the logic instructions in the memory 530 may be implemented in the form of software functional units and stored in a computer readable storage medium when the software functional units are sold or used as independent products. Based on such understanding, the technical solution of the present invention or a part thereof which substantially contributes to the prior art may be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.

In another aspect, the present invention also provides a computer program product, the computer program product comprising a computer program, the computer program being stored on a non-transitory computer readable storage medium, wherein when the computer program is executed by a processor, the computer is capable of executing the method for multi-angle polarized satellite cloud detection with integrated spatial and spectral features provided by the methods, the method comprising: acquiring observation data of a multi-angle polarization imager pixel at N-layer angles; according to observation data of multi-angle polarization imager pixels at each layer angle, acquiring corresponding land cloud identification results, ocean cloud identification results, dazzling ocean cloud identification results, mixed pixel identification results or uncertain pixel identification results of the multi-angle polarization imager pixels at each layer angle, and according to the corresponding land cloud identification results, ocean cloud identification results, dazzling ocean cloud identification results, mixed pixel identification results or uncertain pixel identification results of the multi-angle polarization imager pixels at each layer angle, acquiring corresponding cloud detection results of the multi-angle polarization imager pixels at each layer angle, wherein the cloud detection results comprise clouds, clear sky and uncertain pixels; determining the times of the multi-angle polarization imager pixels being identified as clouds, the times of the multi-angle polarization imager pixels being identified as clear sky and the times of the multi-angle polarization imager pixels being identified as uncertain pixels under N layers of angles according to the corresponding cloud detection results of the multi-angle polarization imager pixels under each layer of angles, and determining whether the multi-angle polarization imager pixels are clouds or not according to the times of the multi-angle polarization imager pixels being identified as clouds, the times of the multi-angle polarization imager pixels being identified as clear sky and the times of the multi-angle polarization imager pixels being identified as uncertain pixels under the N layers of angles; the land cloud identification result corresponding to each layer angle of the multi-angle polarization imager pixel is determined according to a land blue light channel reflectivity cloud identification result, a land apparent pressure cloud identification result, a land polarization rainbow cloud identification result, a land reflectivity ratio clear-air identification result, a land two-channel standard deviation and average value difference cloud identification result or a land near-infrared reflectivity and minimum value difference cloud identification result corresponding to each layer angle of the multi-angle polarization imager pixel; the corresponding ocean cloud identification result of the multi-angle polarization imager pixel at each layer angle is determined according to the corresponding ocean near infrared channel reflectivity cloud identification result, ocean atmospheric molecule optical thickness cloud identification result, ocean polarization rainbow cloud identification result, ocean reflectivity ratio clear air identification result, ocean near infrared standard deviation cloud identification result or ocean near infrared reflectivity and minimum value difference cloud identification result of the multi-angle polarization imager pixel at each layer angle; the mixed pixel identification result corresponding to the multi-angle polarization imager pixel at each layer angle is determined according to the mixed pixel spatial dual-channel standard deviation cloud identification result corresponding to the multi-angle polarization imager pixel at each layer angle; and the corresponding dazzling ocean cloud identification result of the multi-angle polarization imager pixel at each layer angle is determined according to the corresponding dazzling linear polarization degree cloud identification result or dazzling apparent pressure cloud identification result of the multi-angle polarization imager pixel at each layer angle.

In yet another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored, the computer program, when executed by a processor, implementing a multi-angle polarized satellite cloud detection method for integrating spatial and spectral features provided by the above methods, the method comprising: acquiring observation data of a multi-angle polarization imager pixel at N-layer angles; according to observation data of multi-angle polarization imager pixels at each layer angle, acquiring corresponding land cloud identification results, ocean cloud identification results, dazzling ocean cloud identification results, mixed pixel identification results or uncertain pixel identification results of the multi-angle polarization imager pixels at each layer angle, and according to the corresponding land cloud identification results, ocean cloud identification results, dazzling ocean cloud identification results, mixed pixel identification results or uncertain pixel identification results of the multi-angle polarization imager pixels at each layer angle, acquiring corresponding cloud detection results of the multi-angle polarization imager pixels at each layer angle, wherein the cloud detection results comprise clouds, clear sky and uncertain pixels; determining the times of the multi-angle polarization imager pixels being identified as clouds, the times of the multi-angle polarization imager pixels being identified as clear sky and the times of the multi-angle polarization imager pixels being identified as uncertain pixels under N layers of angles according to the corresponding cloud detection results of the multi-angle polarization imager pixels under each layer of angles, and determining whether the multi-angle polarization imager pixels are clouds or not according to the times of the multi-angle polarization imager pixels being identified as clouds, the times of the multi-angle polarization imager pixels being identified as clear sky and the times of the multi-angle polarization imager pixels being identified as uncertain pixels under the N layers of angles; the land cloud identification result corresponding to each layer angle of the multi-angle polarization imager pixel is determined according to a land blue light channel reflectivity cloud identification result, a land apparent pressure cloud identification result, a land polarization rainbow cloud identification result, a land reflectivity ratio clear-air identification result, a land two-channel standard deviation and average value difference cloud identification result or a land near-infrared reflectivity and minimum value difference cloud identification result corresponding to each layer angle of the multi-angle polarization imager pixel; the corresponding ocean cloud identification result of the multi-angle polarization imager pixel at each layer angle is determined according to the corresponding ocean near infrared channel reflectivity cloud identification result, ocean atmospheric molecule optical thickness cloud identification result, ocean polarization rainbow cloud identification result, ocean reflectivity ratio clear air identification result, ocean near infrared standard deviation cloud identification result or ocean near infrared reflectivity and minimum value difference cloud identification result of the multi-angle polarization imager pixel at each layer angle; the mixed pixel identification result corresponding to the multi-angle polarization imager pixel at each layer angle is determined according to the mixed pixel spatial dual-channel standard deviation cloud identification result corresponding to the multi-angle polarization imager pixel at each layer angle; the flare ocean cloud identification result corresponding to each layer of angle of the multi-angle polarization imager pixel is determined according to the flare linear polarization degree cloud identification result or flare apparent pressure intensity cloud identification result corresponding to each layer of angle of the multi-angle polarization imager pixel.

The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.

Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.

Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims

1. A multi-angle polarized satellite cloud detection method integrating spatial and spectral characteristics is characterized by comprising the following steps:

determining the times that the multi-angle polarization imager pixel is identified as cloud, the times that the multi-angle polarization imager pixel is identified as clear sky and the times that the multi-angle polarization imager pixel is identified as uncertain pixel under N layers of angles according to the corresponding cloud detection result of the multi-angle polarization imager pixel under each layer of angles, and determining whether the multi-angle polarization imager pixel is cloud or not according to the times that the multi-angle polarization imager pixel is identified as cloud, the times that the multi-angle polarization imager pixel is identified as clear sky and the times that the multi-angle polarization imager pixel is identified as uncertain pixel under the N layers of angles;

2. The method for detecting the multi-angle polarized satellite cloud with the integrated spatial and spectral characteristics according to claim 1, wherein the obtaining of the land cloud identification result, the ocean cloud identification result, the dazzling ocean cloud identification result, the mixed pixel identification result or the uncertain pixel identification result corresponding to the multi-angle polarized imager pixel at each layer angle according to the observation data of the multi-angle polarized imager pixel at each layer angle comprises:

under the condition that the pixel category of a multi-angle polarization imager pixel at the current angle is a land pixel, if the pixel of the multi-angle polarization imager pixel at the current angle is judged to be cloud or clear sky according to at least one judgment result of a land blue light channel reflectivity cloud judgment result, a land apparent pressure intensity cloud judgment result, a land polarization rainbow cloud judgment result and a land reflectivity ratio clear sky judgment result corresponding to the multi-angle polarization imager pixel at the current angle, determining that the land cloud judgment result of the multi-angle polarization imager pixel at the current angle is cloud or clear sky, and otherwise determining that the pixel of the multi-angle polarization imager pixel at the current angle is an uncertain pixel;

under the condition that the pixel category of a multi-angle polarization imager pixel at the current angle is an ocean pixel, if the pixel of the multi-angle polarization imager pixel at the current angle is judged to be cloud or clear sky according to at least one judgment result of an ocean near infrared channel reflectivity cloud judgment result, an ocean atmospheric molecular optical thickness cloud judgment result, an ocean polarization iridescence cloud judgment result and an ocean reflectivity ratio clear sky judgment result corresponding to the multi-angle polarization imager pixel at the current angle, determining that the ocean cloud judgment result of the multi-angle polarization imager pixel at the current angle is cloud or clear sky, and otherwise determining that the pixel of the multi-angle polarization imager pixel at the current angle is uncertain;

under the condition that the pixel category of a multi-angle polarization imager pixel at the current angle is a flare ocean pixel, if the pixel of the multi-angle polarization imager pixel at the current angle is judged to be a cloud according to at least one judgment result of a flare light linear polarization degree cloud judgment result and a flare apparent pressure intensity cloud judgment result corresponding to the multi-angle polarization imager pixel at the current angle, determining the flare ocean cloud judgment result of the multi-angle polarization imager pixel at the current angle to be a cloud, and otherwise, determining the uncertain pixel judgment result of the multi-angle polarization imager pixel at the current angle;

determining that the category of the multi-angle polarization imager pixel at the current angle is an ocean pixel under the condition that the pixel category of each identification pixel of the space window is an ocean pixel, if the pixel of the multi-angle polarization imager pixel at the current angle is determined to be a cloud according to at least one of an ocean near infrared standard deviation cloud determination result corresponding to the multi-angle polarization imager pixel at the current angle and an ocean near infrared reflectivity and minimum value difference cloud determination result, determining that the ocean cloud determination result of the multi-angle polarization imager pixel at the current angle is a cloud, and otherwise determining that the multi-angle polarization imager pixel at the current angle is an uncertain pixel;

3. The method for multi-angle polarized satellite cloud detection based on spatial and spectral characteristics according to claim 2,

the cloud identification result of the reflectivity of the terrestrial blue light channel corresponding to the pixel of the multi-angle polarization imager at the current angle is obtained through the following steps:

carrying out cloud judgment on the multi-angle polarization imager pixel at the current angle by a blue light channel reflectivity threshold method to obtain a land blue light channel reflectivity cloud judgment result corresponding to the multi-angle polarization imager pixel at the current angle;

according to the polarized reflectivity of the scattering angle within the range of 135-150 degrees, cloud judgment is carried out on the multi-angle polarization imager pixel under the current angle, and a land polarization rainbow cloud judgment result corresponding to the multi-angle polarization imager pixel under the current angle is obtained;

4. The method for multi-angle polarized satellite cloud detection based on spatial and spectral characteristics according to claim 2,

the ocean near infrared channel reflectivity cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

5. The method for multi-angle polarized satellite cloud detection based on spatial and spectral characteristics according to claim 1,

the flare line polarization degree cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

6. The method for multi-angle polarized satellite cloud detection based on spatial and spectral characteristics according to claim 2,

the land dual-channel standard deviation and average value difference cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

acquiring 763nm apparent reflectivity data of pixels of the multi-angle polarization imager at the current angle to obtain reflectivity data corresponding to each identification pixel, and according to the reflectivity data corresponding to each identification pixel;

7. The method for multi-angle polarized satellite cloud detection based on spatial and spectral characteristics according to claim 2,

the ocean near infrared standard deviation cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

8. The method for multi-angle polarized satellite cloud detection based on spatial and spectral characteristics according to claim 2,

the mixed pixel space dual-channel standard deviation cloud identification result corresponding to the multi-angle polarization imager pixel at the current angle is obtained through the following steps:

9. The utility model provides a multi-angle polarization satellite cloud detection device of comprehensive space and spectral feature which characterized in that includes:

the land cloud identification result corresponding to each layer angle of the multi-angle polarization imager pixels is determined according to a land blue light channel reflectivity cloud identification result, a land apparent pressure cloud identification result, a land polarization rainbow cloud identification result, a land reflectivity ratio clear air identification result, a land two-channel standard deviation and average value difference cloud identification result or a land near-infrared reflectivity and minimum value difference cloud identification result corresponding to each layer angle of the multi-angle polarization imager pixels;

the mixed pixel identification result corresponding to the multi-angle polarization imager pixel under each layer angle is determined according to the mixed pixel space two-channel standard deviation cloud identification result corresponding to the multi-angle polarization imager pixel under each layer angle;

10. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements the multi-angle polarized satellite cloud detection method integrating spatial and spectral features according to any one of claims 1 to 8.