CN114581793A - Cloud identification method and device for remote sensing image, electronic equipment and readable storage medium - Google Patents

Cloud identification method and device for remote sensing image, electronic equipment and readable storage medium Download PDF

Info

Publication number
CN114581793A
CN114581793A CN202210242573.XA CN202210242573A CN114581793A CN 114581793 A CN114581793 A CN 114581793A CN 202210242573 A CN202210242573 A CN 202210242573A CN 114581793 A CN114581793 A CN 114581793A
Authority
CN
China
Prior art keywords
cloud
hyperspectral
remote sensing
data
identification
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202210242573.XA
Other languages
Chinese (zh)
Inventor
张�浩
张舒宁
张兵
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Aerospace Information Research Institute of CAS
Original Assignee
Aerospace Information Research Institute of CAS
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Aerospace Information Research Institute of CAS filed Critical Aerospace Information Research Institute of CAS
Priority to CN202210242573.XA priority Critical patent/CN114581793A/en
Publication of CN114581793A publication Critical patent/CN114581793A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
    • Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Physics & Mathematics (AREA)
  • Evolutionary Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Medical Informatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Image Processing (AREA)

Abstract

The application discloses a cloud identification method and device for remote sensing images, electronic equipment and a readable storage medium. The method comprises the steps of updating an Fmark cloud identification algorithm suitable for the multispectral sensor in advance according to a data preprocessing mode and wave band information of the hyperspectral sensor to obtain a hyperspectral cloud identification method; and responding to the data input mode selection instruction, and acquiring target data to be identified from the hyperspectral remote sensing data. Based on a hyperspectral cloud identification method, image preprocessing is carried out on target data to be identified to obtain apparent reflectivity information; based on the apparent reflectivity information, calling a hyperspectral cloud identification method to respectively perform cloud identification and cloud shadow identification to obtain a cloud identification result and a cloud shadow identification result; and determining cloud information of the hyperspectral remote sensing data according to the cloud identification result and the cloud shadow identification result, so that the cloud identification precision of the hyperspectral image of the domestic hyperspectral satellite can be effectively improved.

Description

Cloud identification method and device for remote sensing image, electronic equipment and readable storage medium
Technical Field
The present application relates to the field of remote sensing technologies, and in particular, to a cloud identification method and apparatus for a remote sensing image, an electronic device, and a readable storage medium.
Background
Cloud covers more than 50% of the global surface as an important part of the atmosphere, and has strong absorption and scattering effects on solar radiation and infrared radiation, so that the reflection of ground objects in remote sensing images generates attenuation and covering effects, and the ground observation based on visible light and infrared bands is almost interfered. Generally speaking, the brightening effect of the cloud area and the darkening effect of the cloud shadow can greatly influence the acquisition quality of the remote sensing information.
In remote sensing application, the cloud layer can cause large errors of spectral information of ground objects in a satellite image, particularly images with more thick clouds, the underlying surface can be shielded by the cloud layer to a great extent, and important spectral information is lost. In order to eliminate the influence of cloud layers on ground objects to the maximum extent, the primary task is to perform cloud identification on remote sensing images. The essence of cloud identification is to find the difference between the spectral feature or the texture feature of the cloud pixel and other ground features, so that the cloud pixel can be separated. For cloud and cloud shadow, the cloud is generally identified by using the high-reflectivity and low-temperature characteristics of the cloud and the low-reflectivity characteristics of the cloud shadow, and the current cloud identification methods can be roughly classified into a threshold method, a statistical method and a machine learning method. The threshold method is to use the spectral characteristics of the Cloud to construct a threshold criterion for identification, and is simple and widely applicable, such as isccp (international software Cloud similarity project) algorithm, clavr (the NOAA Cloud Advanced virtual High Resolution radiometer) algorithm, and APOLLO (AVHRR Processing scheme Over Clouds, Land and Ocean) algorithm. The statistical analysis method applies the statistical principle to carry out cloud identification, such as a Bayesian method, a discriminant analysis method, an SVM method and the like. The machine learning method utilizes an artificial neural network as a classifier to train sample data and perform cloud identification, although the neural network does not have physical significance, the spectrum and space characteristics can be extracted through strong self-learning and data analysis capability, and therefore cloud areas on remote sensing images can be identified, such as an Otsu cloud detection method, a cloud detection method based on a deep residual error full convolution network and the like. However, these cloud recognition algorithms have been developed for multispectral sensors mounted on satellites such as Landsat satellites and MODIS. However, the time for the appearance of the hyperspectral sensor is not long, the cloud identification algorithm for the hyperspectral sensor is not as much as that for the multispectral sensor, and particularly for domestic hyperspectral satellites, the cloud identification algorithm may not be suitable and cannot obtain a good cloud identification result.
In view of this, how to improve the cloud identification precision of the hyperspectral image of the domestic hyperspectral satellite is a technical problem to be solved by technical personnel in the field.
Disclosure of Invention
The application provides a cloud identification method and device for a remote sensing image, electronic equipment and a readable storage medium, which can effectively improve the cloud identification precision of a hyperspectral image of a domestic hyperspectral satellite.
In order to solve the above technical problems, embodiments of the present invention provide the following technical solutions:
an embodiment of the present invention provides a cloud identification method for a remote sensing image, including:
updating an Fmak cloud identification algorithm suitable for the multispectral sensor in advance according to a data preprocessing mode and wave band information of the hyperspectral sensor to obtain a hyperspectral cloud identification method;
responding to a data input mode selection instruction, and acquiring target data to be identified from the hyperspectral remote sensing data;
based on the hyperspectral cloud identification method, image preprocessing is carried out on the target data to be identified to obtain apparent reflectivity information;
based on the apparent reflectivity information, calling the hyperspectral cloud identification method to respectively perform cloud identification and cloud shadow identification to obtain a cloud identification result and a cloud shadow identification result;
determining cloud information of the hyperspectral remote sensing data according to the cloud identification result and the cloud shadow identification result;
the data input mode is determined according to wave band information of the hyperspectral sensor and the multispectral sensor, and the target data to be identified is ground object image information corresponding to wave bands matched with the hyperspectral sensor and the multispectral sensor.
Optionally, the obtaining target data to be identified from the hyperspectral remote sensing data includes:
when the data input mode is detected to be the central wavelength input mode, reading the median value of a plurality of single-waveband wavelength ranges of the multispectral sensor, and corresponding to the image data at the corresponding waveband of the hyperspectral remote sensing data;
and reading sun azimuth angles, sun zenith angle information, image row and column numbers and annual accumulation days from the hyperspectral remote sensing data to serve as target data to be identified.
Optionally, the obtaining target data to be identified from the hyperspectral remote sensing data includes:
when the data input mode is detected to be a weighted fitting input mode, determining the radiance value of each wave band after radiometric calibration according to the hyperspectral remote sensing data;
for each wave band, fitting a spectral response curve of the multispectral sensor in the current wave band and a radiance value of the hyperspectral remote sensor in the current wave band to obtain approximate multispectral data of the current wave band;
and taking the approximate multispectral data of each wave band as target data to be identified.
Optionally, the image preprocessing is performed on the target data to be recognized based on the hyperspectral cloud recognition method to obtain apparent reflectivity information, and the method includes:
acquiring a gain parameter and a bias parameter from the target data to be identified, and calculating the radiation brightness value of each pixel according to the gain parameter and the bias parameter;
and determining apparent reflectivity information according to the radiance value of each pixel, the distance between the day and the ground, the average solar irradiance of the top of the atmospheric layer and the solar zenith angle.
Optionally, before determining the apparent reflectivity information according to the radiance value of each pixel, the distance between the sun and the ground, the average solar irradiance at the top of the atmospheric layer, and the solar zenith angle, the method further includes:
acquiring a central wavelength and a half-wave width from the hyperspectral remote sensing data, and carrying out Gaussian fitting according to the central wavelength and the half-wave width to obtain a spectral curve of each wave band of the hyperspectral satellite;
and calculating the average solar irradiance of the top of the atmospheric layer of each wave band according to the spectral curve and the solar spectral curve of each wave band.
Optionally, the determining the cloud information of the hyperspectral remote sensing data according to the cloud identification result and the cloud shadow identification result includes:
performing a water test on the target data to be recognized, and determining a non-water area in the target data to be recognized;
for the non-water area, based on the apparent reflectivity information, cloud shadow recognition is carried out by utilizing an ATCOR algorithm to obtain a corrected cloud shadow recognition result;
cloud matching is carried out on the cloud identification result and the cloud shadow correction result respectively to obtain a shadow area and a corrected shadow area;
if the difference between the shadow area and the corrected shadow area is larger than a preset area threshold value, determining cloud information of the hyperspectral remote sensing data according to the cloud identification result and the cloud shadow identification result;
and if the difference between the shadow area and the corrected shadow area is smaller than or equal to a preset area threshold value, determining cloud information of the hyperspectral remote sensing data according to the cloud identification result and the corrected cloud shadow identification result.
Optionally, after the water test is performed on the target data to be identified, the method further includes:
and if the target data to be identified does not contain a non-water area, determining cloud information of the hyperspectral remote sensing data according to the cloud identification result and the cloud shadow identification result.
Another aspect of the embodiments of the present invention provides a cloud identification apparatus for remote sensing images, including:
the method processing module is used for updating an Fmark cloud identification algorithm suitable for the multispectral sensor in advance according to a data preprocessing mode and waveband information of the hyperspectral sensor to obtain a hyperspectral cloud identification method;
the data reading module is used for responding to a data input mode selection instruction and acquiring target data to be identified from the hyperspectral remote sensing data; the data input mode is determined according to wave band information of the hyperspectral sensor and the multispectral sensor, and the target data to be identified is ground object image information corresponding to wave bands matched with the hyperspectral sensor and the multispectral sensor;
the data preprocessing module is used for preprocessing images of the target data to be recognized based on the hyperspectral cloud recognition method to obtain apparent reflectivity information;
the identification module is used for calling the hyperspectral cloud identification method to respectively carry out cloud identification and cloud shadow identification based on the apparent reflectivity information to obtain a cloud identification result and a cloud shadow identification result;
and the result determining module is used for determining the cloud information of the hyperspectral remote sensing data according to the cloud identification result and the cloud shadow identification result.
An embodiment of the present invention further provides an electronic device, which includes a processor, and the processor is configured to implement the steps of the cloud identification method for remote sensing images according to any one of the foregoing embodiments when executing a computer program stored in a memory.
Finally, an embodiment of the present invention provides a readable storage medium, where a computer program is stored on the readable storage medium, and when the computer program is executed by a processor, the steps of the cloud identification method for remote sensing images according to any of the foregoing embodiments are implemented.
The technical scheme provided by the application has the advantages that a data input mode and a data processing mode are improved through an Fmak cloud identification algorithm, so that the Fmak algorithm based on multispectral remote sensing satellite image cloud identification is applied to a domestic hyperspectral satellite, and as the data input and pretreatment part is arranged aiming at a hyperspectral sensor of the domestic hyperspectral satellite, accurate surface feature apparent reflectivity can be obtained, the device is suitable for setting the waveband of the hyperspectral sensor and can also obtain a better identification result, and the cloud identification precision of the hyperspectral image of the domestic hyperspectral satellite can be effectively improved.
In addition, the embodiment of the invention also provides a corresponding implementation device, electronic equipment and a readable storage medium for the cloud identification method of the remote sensing image, so that the method has higher practicability, and the device, the electronic equipment and the readable storage medium have corresponding advantages.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the related arts, the drawings used in the description of the embodiments or the related arts will be briefly described below, it is obvious that the drawings in the description below are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a schematic flow chart of a cloud identification method for a remote sensing image according to an embodiment of the present invention;
fig. 2 is a schematic flow chart of a cloud identification method for a remote sensing image according to an embodiment of the present invention;
fig. 3 is an unweighted recognition result of a high-score five-number image in 2019, 7 months, 7 days in an exemplary application scenario according to an embodiment of the present invention;
fig. 4 is an unweighted recognition result of a top-scoring fifth image of 11 months and 1 days in 2018 in an exemplary application scenario according to the embodiment of the present invention;
fig. 5 is an unweighted recognition result of a top-scoring fifth image at 11 days 3 and 2020 in an exemplary application scenario according to an embodiment of the present invention;
fig. 6 is an unweighted recognition result of a resource number one 02D image in 6/month and 1/month 2020 in an exemplary application scenario according to an embodiment of the present invention;
fig. 7 is an unweighted recognition result of a resource number one 02D image in 9/7/2020 in an exemplary application scenario according to an embodiment of the present invention;
fig. 8 is a weighted identification result of a high-score five-numbered image of 7 months and 7 days in 2019 in an exemplary application scenario according to an embodiment of the present invention;
fig. 9 is a weighted identification result of a top-scoring fifth image of 11 months and 1 days in 2018 in an exemplary application scenario according to an embodiment of the present invention;
fig. 10 shows a weighted recognition result of a top-scoring fifth image of 11 days 3/month 2020 in an exemplary application scenario according to an embodiment of the present invention;
fig. 11 is a result of weighted identification of resource number one 02D image in year 2020, 6 months, 1 days in an exemplary application scenario according to an embodiment of the present invention;
fig. 12 is a result of weighted identification of resource number one 02D image in 9/7/2020 in an exemplary application scenario according to an embodiment of the present invention;
fig. 13 is a structural diagram of a specific embodiment of a cloud identification apparatus for remote sensing images according to an embodiment of the present invention;
fig. 14 is a block diagram of a specific implementation manner of an electronic device according to an embodiment of the present invention.
Detailed Description
In order that those skilled in the art will better understand the disclosure, the invention will be described in further detail with reference to the accompanying drawings and specific embodiments. It is to be understood that the described embodiments are merely exemplary of the invention, and not restrictive of the full scope of the 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.
The terms "first," "second," "third," "fourth," and the like in the description and claims of this application and in the foregoing drawings are used for distinguishing between different objects and not necessarily for describing a particular sequential order. Furthermore, the terms "comprising" and "having," as well as any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, article, or apparatus that comprises a list of steps or elements is not limited to only those steps or elements but may include other steps or elements not expressly listed.
The inventor of the present application found through research that Zhe Zhu et al published in 2012 in the Remote Sensing of Environment journal entitled Object-based closed and closed shadow detection in the Landsat image discloses a famsk (function of mask) cloud identification algorithm, which is based on TM and ETM + sensors carried on Landsat series satellites, and then extends to Landsat8 and sentinel satellites, and achieves a better cloud identification effect. The Fmak cloud identification algorithm is a cloud identification algorithm for multispectral satellite remote sensing images, and can identify cloud, cloud shadow, water and snow pixels in the images with high precision. The algorithm uses the atmospheric Top (TOA) reflectivity and Brightness Temperature (BT) Of a multispectral sensor as inputs and identifies according to the spectral feature construction criteria Of ground objects such as clouds. In addition, the possible cloud shadow position is predicted by utilizing the visual angle and the illumination angle of the satellite sensor, so that the cloud and the cloud shadow are more accurately identified. Compared with the classic Cloud identification algorithm ACCA (automatic Cloud-Cover Assessment), the Cloud identification precision of the Famsk algorithm has great advantages, for example, the overall precision of the Famsk algorithm is 96.41%, and the overall precision of the ACCA algorithm is 84.8%; the producer precision of the Famsk algorithm is 92.1%, while the producer precision of the ACCA algorithm is 72.1%; the user precision of the Famsk algorithm is 89.4%, while the user precision of the ACCA algorithm is 91.8%. The Fmak cloud identification algorithm is divided into two parts, wherein the first part is used for determining potential cloud areas, and possible cloud pixels are determined through a series of threshold criteria; and the second part respectively calculates the probabilities of clouds on the land and the water body according to the properties of the clouds, and if the probabilities are greater than a certain value, the clouds are determined.
Wherein, (1) the detection of potential cloud areas, namely the PCPs test, comprises a basic test, a whiteness test, an HOT test and wave band5 and6 tests. The basic test is shown in equation (1). BandSWIR2 is the apparent reflectance value of the Landsat8 satellite in the seventh band, which has a wavelength in the range of 2.11 μm to 2.29 μm; BT is the pixel temperature calculated from the thermal infrared band, and the unit is centigrade; in the formula, the NDSI (normalized snow cover index) and NDVI (normalized vegetation index) are defined by the formulas (2) and (3). Wherein, the Band3, the Band4, the Band5 and the Band6 are respectively the apparent reflectance values of green, red, near infrared and SWIR1(1.57-1.65 μm) wave bands of Landsat8 satellite, and the wavelength range of SWIR1 wave Band is 1.57 μm-1.65 μm.
BandSWIR2＞0.03 and BT＜27 and NDSI＜0.8 and NDVI＜0.8 (1)
Figure BDA0003543120750000081
Figure BDA0003543120750000082
The whiteness test defines a new whiteness index as shown in formula (4), and then the whiteness test is performed using the index as shown in formula (5).
Figure BDA0003543120750000083
Figure BDA0003543120750000084
The HOT test definition is shown in equation (6).
HOT＝(Band2-0.5)×(Band4-0.08)＞0 (6)
The band5, 6 test definition is shown in formula (7).
Band5/Band6＞0.75 (7)
All the pixel points which pass the 4 tests are marked as potential cloud area pixels, and at least one pixel which does not pass the tests is marked as an absolute clear sky pixel.
(2) And (3) calculating the cloud probability of the air above the water body, wherein the temperature probability wTemperature _ Prob is calculated and shown in a formula (8), and the Brightness probability Brightness _ Prob is shown in a formula (9).
wTemperature_Prob＝(82.5％BTClear water body-BT)/4 (8)
Brightness_Prob＝min(Band6,0.11)/0.11 (9)
After the temperature probability and the brightness probability of the cloud above the water body are obtained, the cloud probability wCloud _ Prob above the water body can be calculated according to a formula (10), and in the Fmak algorithm, if the value is more than 0.5, the cloud probability wCloud _ Prob can be determined as a cloud pixel.
wCloud_Prob＝wTemperture_Prob×Brightness_Prob (10)
(3) Calculating the cloud probability over land, and calculating the temperature probability as shown in formula (11-13), wherein BTClear sky landIs the temperature of the land under clear sky conditions, BT is the temperature of the potential cloud pixel over the land, Tlow、ThighThe temperatures of clear air and land are respectively 17.5% and 82.5%, and lTemp _ Prob is the temperature probability over land. The Variability probability variabilty _ Prob is calculated as shown in equation (14).
Thigh＝82.5％×BTClear sky land (11)
Tlow＝17.5％×BTClear sky land (12)
lTemperature_Prob＝(Thigh+4-BT)/(Thigh+4-(Tlow-4)) (13)
Variability_Prob＝1-max(abs(NDVI,NDSI,Whiteness)) (14)
After obtaining the temperature probability and the variability probability, the probability of the terrestrial aerial cloud, lroud _ Prob, can be calculated according to equation (15).
lCloud_Prob＝lTemperature_Prob×Variability_Prob (15)
If the cloud probability is greater than Land _ threshold (equation (16)), it is determined as a cloud pel.
Land_threshold＝82.5％lCloud_ProbClear sky+0.2 (16)
Although the Famsk cloud identification algorithm has a good identification effect on the cloud of the multispectral satellite remote sensing image, the algorithm cannot preprocess data of the domestic hyperspectral sensor because related parameters such as head file formats of the sensor carried by the domestic hyperspectral satellite and the multispectral sensor aimed by the Famsk cloud identification algorithm are completely different. Further, in the Famsk cloud identification algorithm, a thermal infrared band is very important for distinguishing ground objects, but a thermal infrared band is not set in a hyperspectral satellite such as a hyperspectral sensor carried on a high-resolution five-number resource 02D, so that the Famsk cloud identification algorithm cannot be used by the domestic high-resolution satellite. In addition, the Famsk cloud identification algorithm cannot utilize the advantages of high spectral resolution and multiple wave bands of the hyperspectral sensor.
Because the method has the advantage of high precision in cloud identification based on the Famsk cloud identification algorithm, the Famsk cloud identification algorithm is improved, so that the method is suitable for a hyperspectral sensor of a domestic hyperspectral satellite, and the cloud identification precision of a hyperspectral image of the domestic hyperspectral satellite is improved.
Having described the technical solutions of the embodiments of the present invention, various non-limiting embodiments of the present application are described in detail below.
Referring to fig. 1, fig. 1 is a schematic flow chart of a cloud identification method for a remote sensing image according to an embodiment of the present invention, where the embodiment of the present invention may include the following:
s101: and updating an Fmak cloud identification algorithm suitable for the multispectral sensor in advance according to the data preprocessing mode and the wave band information of the hyperspectral sensor to obtain the hyperspectral cloud identification method.
In this embodiment, the hyperspectral cloud identification method is a method for identifying a hyperspectral image obtained by improving an Fmask cloud identification algorithm, the Fmask cloud identification algorithm is an identification algorithm suitable for multispectral images, the multispectral images are collected by a multispectral sensor, a multispectral remote sensing satellite carried by the multispectral sensor is a Landsat8 satellite, and the hyperspectral image refers to a spectral resolution of 10-2Spectral images in the range of the order of λ.
In the embodiment, as the formats and the wave bands of the read-in head files of the multispectral sensor and the hyperspectral sensor are different, in order to enable the cloud identification algorithm applicable to the multispectral sensor to be applicable to the hyperspectral sensor and to be applicable to domestic hyperspectral satellites such as a hyperspectral fifth satellite and a resource first 02D satellite, the data input part, the cloud and the cloud shadow identification part of the Fmask cloud identification algorithm are improved, and a better cloud identification result is obtained. For the data pre-processing part: in the step, codes of a read-in head file part of the Fmask cloud identification algorithm are modified, and basic information of the hyperspectral sensor corresponding to a target waveband of the multispectral sensor is read, wherein the basic information comprises but is not limited to ground feature information, a solar azimuth angle and a solar zenith angle. Taking the Landsat8 satellite as an example, this embodiment reads in a DN value (Digital Number, remote sensing image pixel brightness value) image file of a high-resolution five-Number satellite or resource one 02D satellite corresponding to 2, 3, 4, 5, 6, 7, 9 bands of an OLI sensor in the Landsat8 satellite, and basic information such as a solar azimuth angle and a solar zenith angle.
For the cloud identification part: the wave band range of a visible short wave infrared hyperspectral camera of a hyperspectral sensor such as a high-resolution five-number and resource one 02D satellite is 400nm to 2500nm, and a thermal infrared wave band is not set, so that a source code of an Fmark cloud identification algorithm on a cloud identification part is modified, the part using the thermal infrared wave band is removed, the judgment of all use temperatures such as PCPs (physical control services) test and cloud probability calculation, threshold value calculation and the like are included, namely, the thermal infrared wave band is not set in a hyperspectral cloud identification method, and therefore the Fmark cloud identification algorithm is adapted to the wave band setting of the hyperspectral satellite such as the high-resolution five-number satellite to the maximum extent and is suitable for the hyperspectral sensor of the high-resolution five-number and resource one 02D satellite. For the cloud shadow recognition part: likewise, since the thermal infrared band is not set, the Fmask algorithm cloud shadow detection code is modified, removing all parts thereof with respect to temperature. The step modifies a data input and preprocessing part aiming at the characteristics of a hyperspectral sensor of a domestic hyperspectral satellite, and can obtain accurate apparent reflectivity of the ground object. The cloud and cloud shadow identification part is adjusted and improved according to the sensor, so that the hyperspectral sensor is suitable for the wave band setting of the hyperspectral sensor and can obtain a better identification result.
S102: and responding to the data input mode selection instruction, and acquiring target data to be identified from the hyperspectral remote sensing data.
In order to improve the data processing efficiency and the identification effect, some data matched with the type of the multispectral remote sensing data need to be read from the hyperspectral remote sensing data, that is, data similar to the multispectral remote sensing data used by the original Fmask algorithm needs to be read from the hyperspectral remote sensing data, and the data are determined based on the waveband information of the hyperspectral sensor and the multispectral sensor. The target data to be identified is surface feature image information corresponding to the wave band matched with the hyperspectral sensor and the multispectral sensor, the surface feature image information is used for recording surface feature data, such as a surface feature gray value image file, and the pixel brightness value of the remote sensing image can be obtained from the remote sensing data to be used as data representing the surface feature.
S103: based on a hyperspectral cloud identification method, image preprocessing is carried out on target data to be identified, and apparent reflectivity information is obtained.
In the step, data preprocessing is carried out on target data to be recognized by adopting an Fmask cloud recognition algorithm for modifying the format of a read-in header file, so that surface feature apparent reflectivity data is obtained.
S104: based on the apparent reflectivity information, calling a hyperspectral cloud identification method to respectively perform cloud identification and cloud shadow identification to obtain a cloud identification result and a cloud shadow identification result.
The method comprises the steps of carrying out cloud identification processing by adopting an Fmask cloud identification algorithm for removing a thermal infrared band to obtain a cloud identification result, and carrying out cloud shadow processing by adopting an Fmask cloud identification algorithm for removing a temperature part to obtain a cloud shadow identification result.
S105: and determining cloud information of the hyperspectral remote sensing data according to the cloud identification result and the cloud shadow identification result.
And matching the cloud identification result and the cloud shadow identification result in the last step, and determining a cloud area in the hyperspectral remote sensing data according to the matched result.
According to the technical scheme provided by the embodiment of the invention, a data input mode and a data processing mode are improved through an Fmak cloud recognition algorithm, so that the Fmak algorithm based on multispectral remote sensing satellite image cloud recognition is applied to a domestic hyperspectral satellite, and as the data input and preprocessing part is arranged aiming at a hyperspectral sensor of the domestic hyperspectral satellite, the accurate apparent reflectivity of a ground object can be obtained, the method is suitable for setting the waveband of the hyperspectral sensor and can also obtain a better recognition result, and the cloud recognition accuracy of the hyperspectral image of the domestic hyperspectral satellite can be effectively improved.
It should be noted that, in the present application, there is no strict sequential execution order among the steps, and as long as the logical order is met, the steps may be executed simultaneously or according to a certain preset order, and fig. 1 is only an exemplary manner, and does not represent that only the execution order is the order.
In the foregoing embodiment, how to execute step S102 is not limited, in this embodiment, two embodiments of obtaining target data to be identified from hyperspectral remote sensing data are provided, and correspondingly, the data input manner includes two manners, which may include the following steps:
as an optional implementation manner, when it is detected that the data input manner is a central wavelength input manner, reading a median value of a plurality of single-waveband wavelength ranges of the multispectral sensor, and corresponding to image data at a corresponding waveband of the hyperspectral remote sensing data; and reading sun azimuth angle, sun zenith angle information, image row number and year product day from the hyperspectral remote sensing data to serve as target data to be identified.
Taking the example that the multispectral sensor is the multispectral sensor OLI in Landsat8, in this embodiment, the hyperspectral single-band image at the median of the single-band wavelength range of the OLI is read in, for example, the wavelength range of the second band of Landsat8 is 0.45-0.51 μm, then the band corresponding to the high-numbered five of the bands read in is the 22 nd band with the wavelength of 0.48 μm, and the band read in the resource number one 02D satellite is the 11 th band with the wavelength of 0.48 μm, and so on, the data of seven bands in total, such as 3, 4, 5, 6, 7, 9, etc., are read in.
The application also provides another implementation method for acquiring target data to be identified from hyperspectral remote sensing data, which is parallel to the above implementation method, so that the implementation method can be used as another optional implementation method and can include:
when the data input mode is detected to be a weighted fitting input mode, determining the radiance value of each wave band after radiometric calibration according to the hyperspectral remote sensing data; for each wave band, fitting a spectral response curve of the multispectral sensor in the current wave band and a radiance value of the hyperspectral remote sensor in the current wave band to obtain approximate multispectral data of the current wave band; and taking the approximate multispectral data of each wave band as target data to be identified.
In this embodiment, the fine wave bands of the hyperspectral satellite such as the hyperspectral fifth satellite or the resource first 02D satellite are used to simulate the wave bands of the multispectral satellite such as the Landsat8 satellite to the maximum extent through weighted fitting, and fitting can be performed according to the following fitting relation (17):
Figure BDA0003543120750000131
wherein gi (λ) is a spectral response curve of each waveband of a multispectral sensor such as Landsat8, X (λ) is a radiance value of each hyperspectral waveband of hyperspectral sensor such as satellite data of high-resolution five or resource first 02D in the waveband range of λ 1 and λ 2 after radiometric calibration, and λ 1 and λ 2 are start-stop wavelengths of the waveband respectively. By means of weighted fitting, the high-spectrum data can be converted into the multi-spectrum data close to a multi-spectrum satellite such as a Landsat8 satellite, and the design initiatives of an Fmask cloud identification algorithm are better fitted. The matching can reduce the contingency of pixel values by utilizing the multiband characteristic of the hyperspectral data, thereby improving the defect of single hyperspectral waveband input in a multispectral algorithm.
According to the embodiment, through providing two input methods of center wavelength input and weighted fitting input, a user can flexibly select according to actual requirements, better results are obtained by focusing on different cloud layer types, and the user experience is improved.
The above embodiment does not limit the calculation process of the apparent reflectivity information, and the present application also provides an optional implementation manner for performing image preprocessing on target data to be recognized based on a hyperspectral cloud recognition method to obtain the apparent reflectivity information, which may include:
acquiring a gain parameter and a bias parameter from target data to be identified, and calculating the radiation brightness value of each pixel according to the gain parameter and the bias parameter; and determining apparent reflectivity information according to the radiance value of each pixel, the distance between the day and the ground, the average solar irradiance of the top of the atmospheric layer and the solar zenith angle.
In this embodiment, a data preprocessing method of the Fmask cloud identification method is modified, a data preprocessing part of the hyperspectral cloud identification method adopts a preprocessing method different from that of the multispectral remote sensing satellite Landsat8, a suffix provided by a hyperspectral remote sensing satellite such as a high-score five or resource first 02D satellite can be used as gain and offset parameters in a RadCal file, and the radiance of each pixel is calculated according to the following radiance relational expression (18):
L＝Gain×DN+Bias (18)
wherein DN represents the pixel value in the original image of the multispectral remote sensing satellite such as Landsat8 satellite, L is the radiance corresponding to the pixel, and Gain and Bias parameters provided by Gain and Bias for the multispectral remote sensing satellite respectively. And the bias parameters of all wave bands of the satellite with the high score of five and the resource of 02D are all 0, namely the radiance value can be obtained only by multiplying the DN value file by the gain.
Figure BDA0003543120750000141
After the radiance value of each pixel is calculated, the apparent reflectivity can be calculated according to a ground object reflectivity relational expression 19, wherein L is radiance, D is the distance between the day and the ground obtained by imaging time, ESUN is the average solar irradiance of the top of the atmospheric layer, and theta is the solar zenith angle.
Because the ESUN value of each wave band of the hyperspectral satellite such as the hyperspectral fifth satellite and the resource first 02D satellite can not be obtained, before the executing step determines the apparent reflectivity information according to the radiance value of each pixel, the distance between the day and the earth, the average solar irradiance at the top of the atmospheric layer and the solar zenith angle, the method further comprises the following steps:
acquiring a central wavelength and a half-wave width from the hyperspectral remote sensing data, and carrying out Gaussian fitting according to the central wavelength and the half-wave width to obtain a spectral curve of each wave band of the hyperspectral satellite; and calculating the average solar irradiance of the top of the atmospheric layer of each wave band according to the spectral curve and the solar spectral curve of each wave band.
In this embodiment, the spectral curve of each wavelength band of the hyperspectral satellite can be fitted by using the center wavelength and the half-wave width in the file with the suffix provided by the high-score five or resource one 02D satellite as the Spectralresponse, and then the ESUN value of each wavelength band of the high-score five or resource one 02D satellite can be calculated by using the following relational expression (20) in combination with the solar spectral curve.
Figure BDA0003543120750000151
Wherein, E (λ) is a band spectrum curve, S (λ) is a solar spectrum curve, λ 1, λ 2 are the start-stop wavelengths of the band, respectively, and the start-stop wavelengths used in this study are the standard deviation of the central wavelength minus 3 times and the standard deviation of the central wavelength plus 3 times. And substituting the obtained ESUN data into the relational expression (19) to obtain the apparent reflectivity data required by the Fmak cloud identification algorithm.
In order to partially compensate for errors caused by temperature removal and cloud shadow matching, based on the above embodiment, the application further adds a cloud shadow criterion in an ATCOR atmospheric correction model to the whole technical scheme, and accordingly, as shown in formula (21), in the introduced process, various conditions are considered as comprehensively as possible, and as shown in fig. 2, the implementation process of determining the cloud information of the hyperspectral remote sensing data according to the cloud identification result and the cloud shadow identification result in the above steps may include:
performing a water test on target data to be recognized, and determining a non-water area in the target data to be recognized; for the non-water area, based on the apparent reflectivity information, cloud shadow recognition is carried out by utilizing an ATCOR algorithm to obtain a corrected cloud shadow recognition result; cloud matching is carried out on the cloud identification result and the cloud shadow correction result respectively to obtain a shadow area and a corrected shadow area;
if the difference between the shadow area and the corrected shadow area is larger than a preset area threshold value, determining cloud information of the hyperspectral remote sensing data according to the cloud identification result and the cloud shadow identification result; and if the difference between the shadow area and the corrected shadow area is smaller than or equal to a preset area threshold value, determining cloud information of the hyperspectral remote sensing data according to the cloud identification result and the corrected cloud shadow identification result.
And if the target data to be recognized does not contain the non-water area, determining cloud information of the hyperspectral remote sensing data according to the cloud recognition result and the cloud shadow recognition result.
In this embodiment, since the ATCOR algorithm is assumed to be a non-water region, a water test is performed before the criterion is performed, and if the criterion is a non-water region, an operation is performed. Meanwhile, the principle of formation of the mountain shadow and the cloud shadow is similar, and in the practical process, the ATCOR cloud shadow criterion cannot distinguish the mountain shadow and the cloud shadow, so that a change test is additionally added after a water test, and a preset area threshold value can be flexibly selected according to the actual situation, for example, the preset area threshold value can be 5% of the area of the whole image, correspondingly, the difference between the area of the shadow matched with the cloud shadow and the area of the shadow added with the ATCOR algorithm is judged, whether the difference exceeds 5% of the area of the whole image is judged, if the difference exceeds, the ATCOR criterion is not executed, and if the difference does not exceed, the ATCOR criterion is executed. Compared with the method that a small part of shadow is recognized as small error of the water body, the method has the advantage that the large error caused by the existence of the shadow in the mountainous area is more important to correct.
0.04＜ρ(NIR)＜0.12andρ(SWIR1)＜0.20 (21)
Where ρ (NIR) is the reflectance of the near infrared band and ρ (SWIR1) is the reflectance of the first short-wave infrared band.
Therefore, in the embodiment, the identification precision of the Fmask cloud identification algorithm is improved to the greatest extent by adding the cloud and cloud shadow related identification criteria, and the hyperspectral remote sensing image is further improved
In order to verify that the technical scheme provided by the application can effectively improve the cloud identification precision of the hyperspectral remote sensing image, the application also carries out a verification test, and as shown in fig. 2, the verification test can comprise the following contents:
the improved Fmask cloud identification algorithm, namely the hyperspectral cloud identification method, is tested by using a three-scene high-resolution five-number image and a two-scene resource one-number 02D satellite image, and the identification results are shown in fig. 3-12, wherein the input method of the central wavelength is adopted in fig. 3-7, and the input method of the weighted fitting is adopted in fig. 8-12. Fig. 3 to 5 are cloud recognition results of the hyperspectral cloud recognition method without the weighted fitting operation, and fig. 6 to 7 are cloud recognition results of the hyperspectral cloud recognition method without the weighted fitting operation. Fig. 3 shows unweighted recognition results of the 7/2019 high-resolution five-numbered image, fig. 4 shows unweighted recognition results of the 11/1/2018 high-resolution five-numbered image, and fig. 5 shows unweighted recognition results of the 11/3/2020 high-resolution five-numbered image. Fig. 6 shows the unweighted recognition result of the first resource 02D image on day 1/6/2020. Fig. 7 shows unweighted recognition results of the first resource number 02D image on day 7/9/2020. Fig. 8 to 10 are cloud recognition results of the hyperspectral cloud recognition method performing the weighted fitting operation, and fig. 11 to 12 are cloud recognition results of the hyperspectral cloud recognition method performing the weighted fitting operation. Fig. 8 shows the result of weighted recognition of the top-ranked five-numbered image at 7/2019, fig. 9 shows the result of weighted recognition of the top-ranked five-numbered image at 11/1/2018, and fig. 10 shows the result of weighted recognition of the top-ranked five-numbered image at 11/3/2020; fig. 11 shows the result of weighted identification of the first-day resource 02D video on day 1 in 2020, and fig. 12 shows the result of weighted identification of the first-day resource 02D video on day 7 in month 9 in 2020.
By comparing fig. 3 to fig. 12, and comparing the cloud identification results with cloud identification results of other cloud identification algorithms, it is verified that the hyperspectral cloud identification method provided by the application can effectively improve the cloud identification effect of the hyperspectral image.
The embodiment of the invention also provides a corresponding device for the cloud identification method of the remote sensing image, so that the method has higher practicability. Wherein the means can be described separately from the functional module point of view and the hardware point of view. In the following, the cloud identification device for remote sensing images provided by the embodiment of the present invention is introduced, and the cloud identification device for remote sensing images described below and the cloud identification method for remote sensing images described above may be referred to each other correspondingly.
Based on the angle of the functional module, referring to fig. 13, fig. 13 is a structural diagram of a cloud identification apparatus for remote sensing images according to an embodiment of the present invention in a specific implementation, where the apparatus may include:
the method processing module 131 is used for updating an Fmask cloud identification algorithm suitable for the multispectral sensor in advance according to a data preprocessing mode and waveband information of the hyperspectral sensor to obtain a hyperspectral cloud identification method;
the data reading module 132 is used for responding to a data input mode selection instruction and acquiring target data to be identified from the hyperspectral remote sensing data; the data input mode is determined according to the wave band information of the hyperspectral sensor and the multispectral sensor, and the target data to be identified is the ground object image information corresponding to the wave band matched with the hyperspectral sensor and the multispectral sensor;
the data preprocessing module 133 is configured to perform image preprocessing on target data to be recognized based on a hyperspectral cloud recognition method to obtain apparent reflectivity information;
the identification module 134 is used for calling a hyperspectral cloud identification method to respectively perform cloud identification and cloud shadow identification based on the apparent reflectivity information to obtain a cloud identification result and a cloud shadow identification result;
and a result determining module 135 for determining cloud information of the hyperspectral remote sensing data according to the cloud recognition result and the cloud shadow recognition result.
Optionally, in some embodiments of the present embodiment, the data reading module 132 may be configured to: when the data input mode is detected to be the central wavelength input mode, reading the median value of a plurality of single-waveband wavelength ranges of the multispectral sensor, and corresponding to the image data at the corresponding waveband of the hyperspectral remote sensing data; and reading sun azimuth angle, sun zenith angle information, image row number and year product day from the hyperspectral remote sensing data to serve as target data to be identified.
As another alternative implementation manner, in parallel with the above embodiment, the data reading module 132 may further be configured to: when the data input mode is detected to be a weighted fitting input mode, determining the radiance value of each wave band after radiometric calibration according to the hyperspectral remote sensing data; for each wave band, fitting a spectral response curve of the multispectral sensor in the current wave band and a radiance value of the hyperspectral remote sensor in the current wave band to obtain approximate multispectral data of the current wave band; and taking the approximate multispectral data of each wave band as target data to be identified.
Optionally, in another embodiment of this embodiment, the data preprocessing module 133 may further be configured to: acquiring a gain parameter and a bias parameter from target data to be identified, and calculating the radiation brightness value of each pixel according to the gain parameter and the bias parameter; and determining apparent reflectivity information according to the radiance value of each pixel, the distance between the day and the ground, the average solar irradiance of the top of the atmospheric layer and the solar zenith angle.
As an optional implementation manner of the foregoing embodiment, the data preprocessing module 133 may further be configured to: acquiring a central wavelength and a half-wave width from the hyperspectral remote sensing data, and carrying out Gaussian fitting according to the central wavelength and the half-wave width to obtain a spectral curve of each wave band of the hyperspectral satellite; and calculating the average solar irradiance of the top of the atmospheric layer of each wave band according to the spectral curve and the solar spectral curve of each wave band.
Optionally, in other embodiments of this embodiment, the result determining module 135 may further be configured to: performing water test on target data to be recognized, and determining a non-water area in the target data to be recognized; for the non-water area, based on the apparent reflectivity information, cloud shadow recognition is carried out by utilizing an ATCOR algorithm to obtain a corrected cloud shadow recognition result; cloud matching is carried out on the cloud identification result and the cloud shadow correction result respectively to obtain a shadow area and a corrected shadow area; if the difference between the shadow area and the corrected shadow area is larger than a preset area threshold value, determining cloud information of the hyperspectral remote sensing data according to the cloud identification result and the cloud shadow identification result; and if the difference between the shadow area and the corrected shadow area is smaller than or equal to a preset area threshold value, determining cloud information of the hyperspectral remote sensing data according to the cloud identification result and the corrected cloud shadow identification result.
As an optional implementation manner of the foregoing embodiment, the result determination module 135 may be further configured to: and if the target data to be recognized does not contain the non-water area, determining cloud information of the hyperspectral remote sensing data according to the cloud recognition result and the cloud shadow recognition result.
The functions of the functional modules of the cloud identification device for remote sensing images according to the embodiments of the present invention can be specifically implemented according to the method in the above method embodiments, and the specific implementation process may refer to the relevant description of the above method embodiments, which is not described herein again.
Therefore, the cloud identification precision of the hyperspectral image of the domestic hyperspectral satellite can be effectively improved.
The cloud recognition device for remote sensing images mentioned above is described from the perspective of functional modules, and further, the present application also provides an electronic device described from the perspective of hardware. Fig. 14 is a schematic structural diagram of an electronic device according to an embodiment of the present application. As shown in fig. 14, the electronic device includes a memory 140 for storing a computer program; the processor 141 is configured to implement the steps of the cloud identification method for remote sensing images according to any of the above embodiments when executing the computer program.
The processor 141 may include one or more processing cores, such as a 4-core processor, an 8-core processor, a controller, a microcontroller, a microprocessor, or other data processing chip, and the like. The processor 141 may be implemented in at least one hardware form of a DSP (Digital Signal Processing), an FPGA (Field-Programmable Gate Array), and a PLA (Programmable Logic Array). The processor 141 may also include a main processor and a coprocessor, where the main processor is a processor for Processing data in an awake state, and is also called a Central Processing Unit (CPU); a coprocessor is a low power processor for processing data in a standby state. In some embodiments, the processor 141 may be integrated with a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content required to be displayed by the display screen. In some embodiments, the processor 141 may further include an AI (Artificial Intelligence) processor for processing a computing operation related to machine learning.
Memory 140 may include one or more computer-readable storage media, which may be non-transitory. Memory 140 may also include high speed random access memory as well as non-volatile memory, such as one or more magnetic disk storage devices, flash memory storage devices. The memory 140 may be an internal storage unit of the electronic device, such as a hard disk of a server, in some embodiments. The memory 140 may also be an external storage device of the electronic device in other embodiments, such as a plug-in hard disk provided on a server, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), and the like. Further, the memory 140 may also include both an internal storage unit and an external storage device of the electronic device. The memory 140 can be used for storing various data and application software installed in the electronic device, such as: the code of the program that executes the vulnerability handling method, etc. may also be used to temporarily store data that has been output or is to be output. In this embodiment, the memory 140 is at least used for storing the following computer program 1401, wherein after the computer program is loaded and executed by the processor 141, the relevant steps of the cloud identification method for remote sensing images disclosed in any of the foregoing embodiments can be implemented. In addition, the resources stored by the memory 140 may also include an operating system 1402, data 1403, and the like, which may be stored in a transient or persistent manner. Operating system 1402 may include Windows, Unix, Linux, etc. The data 1403 may include, but is not limited to, data corresponding to a cloud recognition result of the remote sensing image, and the like.
In some embodiments, the electronic device may further include a display 142, an input/output interface 143, a communication interface 144 or network interface, a power supply 145, and a communication bus 146. The display 142 and the input/output interface 143, such as a Keyboard (Keyboard), belong to the user interface, and the optional user interface may further include a standard wired interface, a wireless interface, and the like. Alternatively, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch device, or the like. The display, which may also be referred to as a display screen or display unit, as appropriate, is used for displaying information processed in the electronic device and for displaying a visualized user interface. The communication interface 144 may optionally include a wired interface and/or a wireless interface, such as a WI-FI interface, a bluetooth interface, etc., typically used to establish a communication connection between an electronic device and other electronic devices. The communication bus 146 may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is shown in FIG. 14, but this is not intended to represent only one bus or type of bus.
Those skilled in the art will appreciate that the configuration shown in fig. 14 is not intended to be limiting of the electronic device and may include more or fewer components than those shown, such as a sensor 147 to perform various functions.
The functions of the functional modules of the electronic device according to the embodiments of the present invention may be specifically implemented according to the method in the above method embodiments, and the specific implementation process may refer to the description related to the above method embodiments, which is not described herein again.
Therefore, the cloud identification precision of the hyperspectral image of the domestic hyperspectral satellite can be effectively improved.
It is to be understood that, if the cloud identification method of the remote sensing image in the above embodiment is implemented in the form of a software functional unit and sold or used as a separate product, it may be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application may be substantially or partially implemented in the form of a software product, which is stored in a storage medium and executes all or part of the steps of the methods of the embodiments of the present application, or all or part of the technical solutions. And the aforementioned storage medium includes: a U disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), an electrically erasable programmable ROM, a register, a hard disk, a multimedia card, a card type Memory (e.g., SD or DX Memory, etc.), a magnetic Memory, a removable magnetic disk, a CD-ROM, a magnetic or optical disk, and other various media capable of storing program codes.
Based on this, the embodiment of the present invention further provides a readable storage medium, which stores a computer program, and the computer program is executed by a processor, and the steps of the cloud identification method for remote sensing images according to any one of the above embodiments are provided.
The embodiments are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same or similar parts among the embodiments are referred to each other. For hardware including devices and electronic equipment disclosed by the embodiment, the description is relatively simple because the hardware includes the devices and the electronic equipment correspond to the method disclosed by the embodiment, and the relevant points can be obtained by referring to the description of the method.
Those of skill would further appreciate that the various illustrative components and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the components and steps of the various examples have been described above generally in terms of their functionality in order to clearly illustrate this interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
The cloud identification method and device for the remote sensing image, the electronic device and the readable storage medium provided by the application are described in detail above. The principles and embodiments of the present invention are explained herein using specific examples, which are presented only to assist in understanding the method and its core concepts. It should be noted that, for those skilled in the art, it is possible to make various improvements and modifications to the present invention without departing from the principle of the present invention, and those improvements and modifications also fall within the scope of the claims of the present application.

Claims (10)

1. A cloud identification method for remote sensing images is characterized by comprising the following steps:
updating an Fmak cloud identification algorithm suitable for the multispectral sensor in advance according to a data preprocessing mode and wave band information of the hyperspectral sensor to obtain a hyperspectral cloud identification method;
responding to a data input mode selection instruction, and acquiring target data to be identified from the hyperspectral remote sensing data;
based on the hyperspectral cloud identification method, image preprocessing is carried out on the target data to be identified to obtain apparent reflectivity information;
based on the apparent reflectivity information, calling the hyperspectral cloud identification method to respectively perform cloud identification and cloud shadow identification to obtain a cloud identification result and a cloud shadow identification result;
determining cloud information of the hyperspectral remote sensing data according to the cloud identification result and the cloud shadow identification result;
the data input mode is determined according to wave band information of the hyperspectral sensor and the multispectral sensor, and the target data to be identified is ground object image information corresponding to wave bands matched with the hyperspectral sensor and the multispectral sensor.
2. The cloud identification method for remote sensing images according to claim 1, wherein the obtaining target data to be identified from the hyperspectral remote sensing data comprises:
when the data input mode is detected to be the central wavelength input mode, reading the median value of a plurality of single-waveband wavelength ranges of the multispectral sensor, and corresponding to the image data at the corresponding waveband of the hyperspectral remote sensing data;
and reading sun azimuth angles, sun zenith angle information, image row and column numbers and annual accumulation days from the hyperspectral remote sensing data to serve as target data to be identified.
3. The cloud identification method for remote sensing images according to claim 1, wherein the obtaining target data to be identified from the hyperspectral remote sensing data comprises:
when the data input mode is detected to be a weighted fitting input mode, determining the radiance value of each wave band after radiometric calibration according to the hyperspectral remote sensing data;
for each wave band, fitting a spectral response curve of the multispectral sensor in the current wave band and a radiance value of the hyperspectral remote sensor in the current wave band to obtain approximate multispectral data of the current wave band;
and taking the approximate multispectral data of each wave band as target data to be identified.
4. The cloud identification method for remote sensing images according to claim 1, wherein the image preprocessing is performed on the target data to be identified based on the hyperspectral cloud identification method to obtain apparent reflectivity information, and the method comprises the following steps:
acquiring a gain parameter and a bias parameter from the target data to be identified, and calculating the radiation brightness value of each pixel according to the gain parameter and the bias parameter;
and determining apparent reflectivity information according to the radiance value of each pixel, the distance between the day and the ground, the average solar irradiance of the top of the atmospheric layer and the solar zenith angle.
5. The cloud identification method for remote sensing images according to claim 4, wherein before determining the apparent reflectivity information according to the radiance value of each pixel, the distance between the day and the earth, the average solar irradiance at the top of the atmospheric layer and the solar zenith angle, the method further comprises:
acquiring a central wavelength and a half-wave width from the hyperspectral remote sensing data, and carrying out Gaussian fitting according to the central wavelength and the half-wave width to obtain a spectral curve of each wave band of the hyperspectral satellite;
and calculating the average solar irradiance of the top of the atmospheric layer of each wave band according to the spectral curve and the solar spectral curve of each wave band.
6. The cloud identification method for remote sensing images according to any one of claims 1 to 5, wherein the determining the cloud information of the hyperspectral remote sensing data according to the cloud identification result and the cloud shadow identification result comprises:
performing a water test on the target data to be recognized, and determining a non-water area in the target data to be recognized;
for the non-water area, based on the apparent reflectivity information, cloud shadow recognition is carried out by utilizing an ATCOR algorithm to obtain a corrected cloud shadow recognition result;
cloud matching is carried out on the cloud identification result and the cloud shadow correction identification result respectively to obtain a shadow area and a corrected shadow area;
if the difference between the shadow area and the corrected shadow area is larger than a preset area threshold value, determining cloud information of the hyperspectral remote sensing data according to the cloud identification result and the cloud shadow identification result;
and if the difference between the shadow area and the corrected shadow area is smaller than or equal to a preset area threshold value, determining cloud information of the hyperspectral remote sensing data according to the cloud identification result and the corrected cloud shadow identification result.
7. The cloud identification method for remote sensing images according to claim 6, wherein after the water test is performed on the target data to be identified, the method further comprises:
and if the target data to be identified does not contain a non-water area, determining cloud information of the hyperspectral remote sensing data according to the cloud identification result and the cloud shadow identification result.
8. A cloud recognition device for a remote sensing image, comprising:
the method processing module is used for updating an Fmark cloud identification algorithm suitable for the multispectral sensor in advance according to a data preprocessing mode and waveband information of the hyperspectral sensor to obtain a hyperspectral cloud identification method;
the data reading module is used for responding to a data input mode selection instruction and acquiring target data to be identified from the hyperspectral remote sensing data; the data input mode is determined according to wave band information of the hyperspectral sensor and the multispectral sensor, and the target data to be identified is ground object image information corresponding to wave bands matched with the hyperspectral sensor and the multispectral sensor;
the data preprocessing module is used for preprocessing images of the target data to be recognized based on the hyperspectral cloud recognition method to obtain apparent reflectivity information;
the identification module is used for calling the hyperspectral cloud identification method to respectively carry out cloud identification and cloud shadow identification based on the apparent reflectivity information to obtain a cloud identification result and a cloud shadow identification result;
and the result determining module is used for determining the cloud information of the hyperspectral remote sensing data according to the cloud identification result and the cloud shadow identification result.
9. An electronic device, comprising a processor and a memory, the processor being configured to carry out the steps of the method of cloud recognition of remotely sensed images as claimed in any of claims 1 to 7 when executing a computer program stored in the memory.
10. A readable storage medium, characterized in that the readable storage medium has stored thereon a computer program which, when being executed by a processor, carries out the steps of a method for cloud recognition of remote sensing images according to any of claims 1 to 7.
CN202210242573.XA 2022-03-11 2022-03-11 Cloud identification method and device for remote sensing image, electronic equipment and readable storage medium Pending CN114581793A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210242573.XA CN114581793A (en) 2022-03-11 2022-03-11 Cloud identification method and device for remote sensing image, electronic equipment and readable storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210242573.XA CN114581793A (en) 2022-03-11 2022-03-11 Cloud identification method and device for remote sensing image, electronic equipment and readable storage medium

Publications (1)

Publication Number Publication Date
CN114581793A true CN114581793A (en) 2022-06-03

Family

ID=81781341

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210242573.XA Pending CN114581793A (en) 2022-03-11 2022-03-11 Cloud identification method and device for remote sensing image, electronic equipment and readable storage medium

Country Status (1)

Country Link
CN (1) CN114581793A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115984715A (en) * 2023-03-22 2023-04-18 武汉大学 Hyperspectral satellite adaptive atmosphere correction method based on prior ground object spectrum matching

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115984715A (en) * 2023-03-22 2023-04-18 武汉大学 Hyperspectral satellite adaptive atmosphere correction method based on prior ground object spectrum matching

Similar Documents

Publication Publication Date Title
US9230170B2 (en) Plant species identification apparatus and method
Kale et al. A research review on hyperspectral data processing and analysis algorithms
US8509476B2 (en) Automated system and method for optical cloud shadow detection over water
Chen et al. The application of the tasseled cap transformation and feature knowledge for the extraction of coastline information from remote sensing images
Marcinkowska-Ochtyra et al. Subalpine and alpine vegetation classification based on hyperspectral APEX and simulated EnMAP images
CN112001374B (en) Cloud detection method and device for hyperspectral image
US9383478B2 (en) System and method for atmospheric parameter enhancement
Oishi et al. A new Landsat 8 cloud discrimination algorithm using thresholding tests
CN114563378B (en) Method, device, medium and equipment for quantitatively describing space distribution of cyanobacterial bloom in lakes and reservoirs
CN110991430A (en) Ground feature identification and coverage rate calculation method and system based on remote sensing image
Zhou et al. Restoration of information obscured by mountainous shadows through Landsat TM/ETM+ images without the use of DEM data: a new method
CN115372282A (en) Farmland soil water content monitoring method based on hyperspectral image of unmanned aerial vehicle
CN114581793A (en) Cloud identification method and device for remote sensing image, electronic equipment and readable storage medium
CN113284066B (en) Automatic cloud detection method and device for remote sensing image
Kashtan et al. Deep learning technology for automatic burned area extraction using satellite high spatial resolution images
CN108229517B (en) Neural network training and hyperspectral image interpretation method and device and electronic equipment
US20200364835A1 (en) Image processing device, image processing method and storage medium
Wang et al. CG-CFPANet: a multi-task network for built-up area extraction from SDGSAT-1 and Sentinel-2 remote sensing images
Andreev et al. Cloud detection from the Himawari-8 satellite data using a convolutional neural network
Sah et al. Semi-automated cloud/shadow removal and land cover change detection using satellite imagery
Wu et al. A novel algorithm for differentiating cloud from snow sheets using Landsat 8 OLI imagery
Liu et al. Hyperspectral Real-time Online Processing Local Anomaly Detection via Multi-Line Multi-Band Progressing
CN113627292A (en) Remote sensing image identification method and device based on converged network
Jian et al. Transfer-learning-based cloud detection for Zhuhai-1 satellite hyperspectral imagery
Roy Hybrid algorithm for hyperspectral target detection

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination