A method and system for generating hyperspectral classification samples based on image simulation

Abstract

The present invention provides a method and system for generating hyperspectral classification samples based on image simulation. The scheme includes obtaining a surface reflectance image; setting the atmospheric conditions and observation geometry conditions of an atmospheric radiation transfer model; calculating atmospheric effect parameters using the atmospheric radiation transfer model according to model operation condition parameters; and simulating the surface reflectance image to generate an atmospheric top radiance image according to the atmospheric effect parameters. The scheme calculates atmospheric effect parameters under various atmospheric and observation geometry conditions based on the atmospheric radiation transfer model, combines the surface reflectance image, and simulates and calculates radiance images under various atmospheric and observation geometry conditions as samples for hyperspectral data classification, thereby enriching the classification sample data.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and more specifically, to a method and system for generating hyperspectral classification samples based on image simulation. Background Art

[0002] At present, the fine classification of hyperspectral images requires a large number of training samples. However, since it is difficult to see the difference between subdivided categories through visual images, the sample collection of hyperspectral images often needs to be obtained through field surveys, but this method of sample acquisition is costly, so the samples for fine classification of hyperspectral images are also very limited. Therefore, using data enhancement technology is one of the ways to increase the number of samples.

[0003] The existing methods of using data enhancement schemes to obtain data for classified samples mainly use spatial geometric transformations, such as flipping, cropping, translation, etc., or random superposition of noise, etc. However, these methods are based on signal and image processing algorithms, and do not consider the radiation properties of remote sensing images, which limits the data enhancement capabilities of remote sensing data. Summary of the invention

[0004] In view of the above problems, the present invention proposes a method and system for generating hyperspectral classification samples based on image simulation. The scheme calculates atmospheric effect parameters under various atmospheric and observation geometric conditions based on the atmospheric radiation transfer model, and combines the surface reflectance image to simulate and calculate the radiance images under various atmospheric and observation geometric conditions as samples for hyperspectral data classification, thereby increasing the amount of classification sample data.

[0005] According to a first aspect of an embodiment of the present invention, a method for generating hyperspectral classification samples based on image simulation is provided.

[0006] In one or more embodiments, preferably, the method for generating hyperspectral classification samples based on image simulation includes:

[0007] Obtain surface reflectance images;

[0008] Setting model operating condition parameters of the atmospheric radiation transfer model, wherein the model operating condition parameters include atmospheric conditions, observation geometry conditions, and calculation wavelengths of atmospheric effect parameters;

[0009] Calculating atmospheric effect parameters using the atmospheric radiation transfer model according to the model operation condition parameters;

[0010] According to the atmospheric effect parameters, the surface reflectivity image is simulated to generate an atmospheric top radiance image.

[0011] In one or more embodiments, preferably, the setting of the model operating condition parameters of the atmospheric radiation transfer model, wherein the model operating condition parameters include atmospheric conditions, observation geometry conditions, and calculation wavelengths of atmospheric effect parameters, specifically includes:

[0012] Setting a first surface reflectivity and a second surface reflectivity, wherein the first surface reflectivity and the second surface reflectivity are not equal;

[0013] Setting the atmospheric conditions according to a preset value library, wherein the atmospheric conditions include water vapor content and visibility;

[0014] The observation geometric conditions are set according to a preset value library, wherein the observation geometric conditions include a solar zenith angle and an observation zenith angle;

[0015] The atmospheric effect calculation interval is set to 1 nm.

[0016] In one or more embodiments, preferably, the preset value library specifically includes:

[0017] The water vapor content is selected from 0 to 4.2 g / m 3 one of the;

[0018] The visibility is selected from 10km to 50km;

[0019] The solar zenith angle is selected from 0 to 60 degrees;

[0020] The observation zenith angle is selected from 0 to 60 degrees.

[0021] In one or more embodiments, preferably, the calculating the atmospheric effect parameters using the atmospheric radiation transfer model according to the model operation condition parameters specifically includes:

[0022] Setting the atmospheric radiation transfer model according to the model operation condition parameters;

[0023] Obtaining a first intermediate variable and a second intermediate variable corresponding to the first surface reflectivity and the second surface reflectivity;

[0024] Obtaining a third intermediate variable and a fourth intermediate variable corresponding to the first surface reflectivity and the second surface reflectivity;

[0025] Calculate the albedo of the large balloon surface according to the first calculation formula by combining the first intermediate variable, the second intermediate variable, the first surface reflectivity and the second surface reflectivity, and save it as the first atmospheric effect parameter;

[0026] Calculate a second atmospheric effect parameter using a second calculation formula according to the first intermediate variable, the second intermediate variable, the third intermediate variable, the fourth intermediate variable, the first surface reflectivity, and the second surface reflectivity;

[0027] Calculate the atmospheric path radiation caused by solar radiation using a third calculation formula according to the first intermediate variable, the second intermediate variable, the third intermediate variable, the fourth intermediate variable, the first surface reflectivity, and the second surface reflectivity, and save it as a third atmospheric effect parameter;

[0028] Specifically, the first intermediate variable, the second intermediate variable, the third intermediate variable, the fourth intermediate variable and the first atmospheric effect parameter, the second atmospheric effect parameter and the third atmospheric effect parameter are all variables and parameters related to wavelength, that is, each of the calculated wavelengths has the corresponding first intermediate variable, the second intermediate variable, the third intermediate variable, the fourth intermediate variable and the first atmospheric effect parameter, the second atmospheric effect parameter and the third atmospheric effect parameter;

[0029] The first calculation formula:

[0030]

[0031] Wherein, S(λ) is the first atmospheric effect parameter of the band λ, L gr1 is the first intermediate variable of the band λ, L gr2 is the second intermediate variable of the band λ, ρ1 is the first surface reflectivity, and ρ2 is the second surface reflectivity;

[0032] The second calculation formula:

[0033]

[0034] Wherein, D(λ) is the second atmospheric effect parameter of the band λ, L gr1 (λ) is the first intermediate variable of the band λ, L gr2 (λ) is the second intermediate variable of the band λ, L ss1 (λ) is the third intermediate variable of the band λ, L ss2 (λ) is the fourth intermediate variable, ρ1 is the first surface reflectivity, and ρ2 is the second surface reflectivity;

[0035] The third calculation formula:

[0036]

[0037] Wherein, C(λ) is the third atmospheric effect parameter of the band λ, L gr1(λ) is the first intermediate variable of the band λ, L gr2 (λ) is the second intermediate variable of the band λ, L ss1 (λ) is the third intermediate variable of the band λ, L ss2 (λ) is the fourth intermediate variable of the band λ, ρ1 is the first surface reflectivity, and ρ2 is the second surface reflectivity.

[0038] In one or more embodiments, preferably, the calculating the top-of-atmosphere radiance using the atmospheric effect parameter specifically includes:

[0039] Acquire the first atmospheric effect parameter, the second atmospheric effect parameter, and the third atmospheric effect parameter;

[0040] Calculating the top-of-atmosphere radiance for the surface reflectance image according to the fourth calculation formula;

[0041] The fourth calculation formula is:

[0042]

[0043] Wherein, L(λ) is the top-of-atmosphere radiance of the band λ, S(λ) is the first atmospheric effect parameter of the band λ, D(λ) is the second atmospheric effect parameter of the band λ, C(λ) is the third atmospheric effect parameter of the band λ, and ρ is the pixel surface reflectance.

[0044] According to a second aspect of an embodiment of the present invention, a system for generating hyperspectral classification samples based on image simulation is provided, the system comprising:

[0045] A data acquisition unit, used for acquiring a surface reflectivity image;

[0046] A parameter setting unit, used to set model operation condition parameters of the atmospheric radiation transfer model, wherein the model operation condition parameters include atmospheric conditions, observation geometry conditions, and calculation wavelengths of atmospheric effect parameters;

[0047] An atmospheric effect calculation unit, used to calculate atmospheric effect parameters using the atmospheric radiation transfer model according to the model operation condition parameters;

[0048] The image generation unit is used to simulate the surface reflectance image and generate an atmospheric top radiance image according to the atmospheric effect parameters.

[0049] In one or more embodiments, preferably, the hyperspectral classification sample generation system based on image simulation further includes a database, a storage medium, and a data display background;

[0050] The database is used to store the first atmospheric effect parameter, the second atmospheric effect parameter and the third atmospheric effect parameter;

[0051] The storage medium includes a mobile hard disk;

[0052] The data display background is used to display the top-of-atmosphere radiance image.

[0053] According to a third aspect of an embodiment of the present invention, there is provided a computer-readable storage medium on which computer program instructions are stored. When the computer program instructions are executed by a processor, the method according to any one of the first aspect of the embodiment of the present invention is implemented.

[0054] According to a fourth aspect of an embodiment of the present invention, there is provided an electronic device, comprising a memory and a processor, wherein the memory is used to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the steps as described in any one of the first aspect of the embodiment of the present invention.

[0055] The technical solution provided by the embodiments of the present invention may have the following beneficial effects:

[0056] This technical solution uses image simulation to generate radiance images under different observation geometric conditions as fine classification samples of hyperspectral data. Compared with the existing data enhancement method, it further considers the radiation characteristics of remote sensing images. In actual situations, the images acquired by satellites in the same area will be different due to different atmospheric and observation geometric conditions. Therefore, the classification sample generation scheme provided by the present invention not only expands the samples, but also ensures that the generated classification samples are closer to the characteristics of real images and are more suitable for training classification models.

[0057] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings.

[0058] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0060] Figure 1The present invention is a flowchart of a method for generating hyperspectral classification samples based on image simulation according to an embodiment of the present invention.

[0061] Figure 2 The present invention is a flowchart of obtaining a surface reflectance image in a method for generating hyperspectral classification samples based on image simulation according to an embodiment of the present invention.

[0062] Figure 3 The present invention is a flowchart of setting model operation condition parameters of an atmospheric radiation transfer model in a method for generating hyperspectral classification samples based on image simulation according to an embodiment of the present invention.

[0063] Figure 4 It is a flowchart of calculating atmospheric effect parameters using the atmospheric radiation transfer model according to the model operating condition parameters in a method for generating hyperspectral classification samples based on image simulation in one embodiment of the present invention.

[0064] Figure 5 It is a flowchart of calculating the top-of-atmosphere radiance using the atmospheric effect parameters in a method for generating hyperspectral classification samples based on image simulation according to an embodiment of the present invention.

[0065] Figure 6 It is a structural diagram of a hyperspectral classification sample generation system based on image simulation according to an embodiment of the present invention.

[0066] Figure 7 It is a structural diagram of an electronic device in one embodiment of the present invention. DETAILED DESCRIPTION

[0067] In some of the processes described in the specification and claims of the present invention and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0068] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0069] At present, the fine classification of hyperspectral images requires a large number of training samples. However, since it is difficult to see the difference between subdivided categories through visual images, the sample collection of hyperspectral images often needs to be obtained through field surveys, but this method of sample acquisition is costly, so the samples for fine classification of hyperspectral images are also very limited. Therefore, the existing method is to use data enhancement technology to increase the number of samples.

[0070] However, the existing methods of using data enhancement schemes to obtain data for classified samples mainly use spatial geometric transformations, such as flipping, cropping, translation, etc., or random superposition of noise, etc. However, these methods are based on signal and image processing algorithms and do not consider the radiation properties of remote sensing images.

[0071] In an embodiment of the present invention, a method and system for generating hyperspectral classification samples based on image simulation are provided. The scheme calculates atmospheric effect parameters under various atmospheric and observation geometric conditions based on the atmospheric radiation transfer model, combines the surface reflectance image, and simulates and calculates the radiance images under various atmospheric and observation geometric conditions as samples for satellite hyperspectral data classification, thereby enriching the classification sample data.

[0072] According to a first aspect of an embodiment of the present invention, a method for generating hyperspectral classification samples based on image simulation is provided.

[0073] Figure 1 The present invention is a flowchart of a method for generating hyperspectral classification samples based on image simulation according to an embodiment of the present invention.

[0074] like Figure 1 As shown, in one or more embodiments, preferably, the method for generating hyperspectral classification samples based on image simulation includes:

[0075] S101, obtaining a surface reflectance image;

[0076] S102, setting model operation condition parameters of the atmospheric radiation transfer model, wherein the model operation condition parameters include atmospheric conditions, observation geometry conditions, and atmospheric effect calculation interval;

[0077] S103, calculating atmospheric effect parameters using the atmospheric radiation transfer model according to the model operation condition parameters;

[0078] S104: According to the atmospheric effect parameters, the surface reflectance image is simulated to generate an atmospheric top radiance image.

[0079] In the embodiment of the present invention, since the hyperspectral image is formed by the process of solar radiation passing through the atmosphere to reach the surface, interacting with the ground objects, and then passing through the atmosphere to enter the sensor, the surface reflectance image corrected by the atmosphere is used as input data, combined with the atmospheric effect parameters calculated by the atmospheric radiation transmission model at different observation geometric angles, the radiance image under different observation geometries is simulated, and the simulated data is used as a new sample for the training of the classification model. This solves the problem that the remote sensing data itself does not consider the radiation characteristics, improves the accuracy of the model classification training on the one hand, and increases the means of data enhancement on the other hand.

[0080] Figure 2 The present invention is a flowchart of obtaining a surface reflectance image in a method for generating hyperspectral classification samples based on image simulation according to an embodiment of the present invention.

[0081] like Figure 2 As shown, in one or more embodiments, preferably, obtaining a surface reflectance image specifically includes:

[0082] S200, setting a first surface reflectivity and a second surface reflectivity, wherein the first surface reflectivity and the second surface reflectivity are not equal;

[0083] Wherein, preferably, the first surface reflectivity is set to 0.5, and the second surface reflectivity is set to 0.8;

[0084] S201, acquiring the top-of-atmosphere radiance image through satellite raw data, and saving it as first correction data;

[0085] S202: Perform atmospheric correction on the first correction data to generate a surface reflectance product.

[0086] Specifically, atmospheric correction refers to the total radiance of the ground target finally measured by the sensor, but it is not a reflection of the true reflectance of the ground surface. The total radiance includes the radiation error caused by atmospheric absorption, especially scattering. Therefore, atmospheric correction is the process of eliminating these radiation errors caused by atmospheric influence and inverting the true surface reflectance of the ground object.

[0087] In the embodiment of the present invention, the acquired raw data is corrected. The main reason is that the atmospheric top radiance simulation calculation requires surface reflectance data as an input parameter. Therefore, the remote sensing data is firstly corrected for the atmosphere before data processing.

[0088] Figure 3 The present invention is a flowchart of setting model operation condition parameters of an atmospheric radiation transfer model in a method for generating hyperspectral classification samples based on image simulation according to an embodiment of the present invention.

[0089] like Figure 3 As shown, in one or more embodiments, preferably, the model operating condition parameters of the atmospheric radiation transfer model are set, wherein the model operating condition parameters include atmospheric conditions, observation geometry conditions, and atmospheric effect calculation intervals, specifically including:

[0090] S301, setting the atmospheric conditions according to a preset value library, wherein the atmospheric conditions include water vapor content and visibility;

[0091] S302, setting the observation geometry conditions according to a preset value library, wherein the observation geometry conditions include a solar zenith angle and an observation zenith angle;

[0092] S303: Setting a calculation wavelength for the atmospheric effect parameter, wherein the calculation wavelength for the atmospheric effect parameter is consistent with a central wavelength of each band of the hyperspectral image.

[0093] In one or more embodiments, preferably, the preset value library specifically includes:

[0094] The water vapor content is selected from 0 to 4.2 g / m 3 one of the;

[0095] The visibility is selected from one of 10km, 25km and 50km;

[0096] The solar zenith angle is selected from one of 0, 15, 30, 45, and 60;

[0097] The observation zenith angle is selected from one of 0, 15, 15, 30, 45, and 60.

[0098] Specifically, atmospheric radiation transmission refers to the propagation and transmission process of electromagnetic waves in the atmospheric medium. In this process, absorption and scattering occur due to the interaction between radiation energy and the medium, and the atmosphere also radiates radiation. The main components in the atmosphere that absorb solar radiation are oxygen, ozone, water vapor, carbon dioxide, methane, etc., and the main absorption components for long-wave radiation are water vapor, carbon dioxide and ozone. LOWTRAN is a single-parameter, spectral band atmospheric transmission model developed by the American Geophysical Laboratory. It is a software package for calculating atmospheric transmittance and radiation. It is suitable for the range from ultraviolet, visible, infrared to microwave and even wider electromagnetic spectrum, including atmospheric transmittance and background radiation of various atmospheric conditions such as clouds, fog, rain, etc. MODTRAN is an improved model of LOWTRAN, and the basic structure and framework of the program remain the same. MODTRAN covers 0-22600cm -1 The spectral range is 2cm -1spectral resolution. MODTRAN uses a two-stream approximation model to consider the atmospheric multiple scattering effect. MODTRAN is a medium-resolution atmospheric radiation transfer model that approximates multi-layered paths as equivalent uniform paths and can calculate thermal infrared radiation brightness, irradiance, etc.

[0099] In the embodiment of the present invention, preferably, MODTRAN is used as the atmospheric radiation transfer model.

[0100] In the embodiment of the present invention, before performing the top of atmosphere radiance simulation, parameter design is first performed. Specifically, it includes the design of two types of parameters, one is the setting of atmospheric conditions, and the other is the setting of observation geometry conditions. The setting process adopts a preset value method to ensure that the result after each setting has a certain basis. At the same time, since the setting process is standardized, it can be easy to analyze data.

[0101] Figure 4 It is a flowchart of calculating atmospheric effect parameters using the atmospheric radiation transfer model according to the model operating condition parameters in a method for generating hyperspectral classification samples based on image simulation in one embodiment of the present invention.

[0102] like Figure 4 As shown, in one or more embodiments, preferably, the calculating of the atmospheric effect parameters using the atmospheric radiation transfer model according to the model operating condition parameters specifically includes:

[0103] S401, setting the atmospheric radiation transfer model according to the model operation condition parameters;

[0104] S402, obtaining a first intermediate variable and a second intermediate variable corresponding to the first surface reflectivity and the second surface reflectivity;

[0105] S403, obtaining a third intermediate variable and a fourth intermediate variable corresponding to the first surface reflectivity and the second surface reflectivity;

[0106] S404, calculating the albedo of the giant balloon surface according to the first calculation formula by combining the first intermediate variable, the second intermediate variable, the first surface reflectivity, and the second surface reflectivity, and saving it as a first atmospheric effect parameter;

[0107] S405, calculating a second atmospheric effect parameter using a second calculation formula according to the first intermediate variable, the second intermediate variable, the third intermediate variable, the fourth intermediate variable, the first surface reflectivity, and the second surface reflectivity;

[0108] S406. Calculate the atmospheric path radiation caused by solar radiation using a third calculation formula according to the first intermediate variable, the second intermediate variable, the third intermediate variable, the fourth intermediate variable, the first surface reflectivity, and the second surface reflectivity, and save it as a third atmospheric effect parameter.

[0109] Specifically, the first intermediate variable, the second intermediate variable, the third intermediate variable, the fourth intermediate variable and the first atmospheric effect parameter, the second atmospheric effect parameter and the third atmospheric effect parameter are all variables and parameters related to wavelength, that is, each of the calculated wavelengths has the corresponding first intermediate variable, the second intermediate variable, the third intermediate variable, the fourth intermediate variable and the first atmospheric effect parameter, the second atmospheric effect parameter and the third atmospheric effect parameter;

[0110] The first calculation formula:

[0111]

[0112] Wherein, S(λ) is the first atmospheric effect parameter of the band λ, L gr1 is the first intermediate variable of the band λ, L gr2 is the second intermediate variable of the band λ, ρ1 is the first surface reflectivity, and ρ2 is the second surface reflectivity;

[0113] The second calculation formula:

[0114]

[0115] Wherein, D(λ) is the second atmospheric effect parameter of the band λ, L gr1 (λ) is the first intermediate variable of the band λ, L gr2 (λ) is the second intermediate variable of the band λ, L ss1 (λ) is the third intermediate variable of the band λ, L ss2 (λ) is the fourth intermediate variable, ρ1 is the first surface reflectivity, and ρ2 is the second surface reflectivity;

[0116] The third calculation formula:

[0117]

[0118] Wherein, C(λ) is the third atmospheric effect parameter of the band λ, L gr1 (λ) is the first intermediate variable of the band λ, L gr2 (λ) is the second intermediate variable of the band λ, L ss1 (λ) is the third intermediate variable of the band λ, L ss2(λ) is the fourth intermediate variable of the band λ, ρ1 is the first surface reflectivity, and ρ2 is the second surface reflectivity.

[0119] Specifically, the first intermediate variable and the second intermediate variable represent GRNDRFLT in the MODTRAN calculation results when the reflectivities are ρ1 and ρ2 respectively, and the third intermediate variable and the fourth intermediate variable represent SOL SCAT in the MODTRAN calculation results when the reflectivities are ρ1 and ρ2 respectively; wherein GRNDRFLT and SOL SCAT are both components of the top-of-atmosphere radiance simulated by MODTRAN.

[0120] In an embodiment of the present invention, when the first surface reflectivity and the second surface reflectivity are adopted, the atmospheric effect parameters under the corresponding model operating condition parameters are calculated respectively using the first calculation formula, the second calculation formula and the third calculation formula, and these atmospheric effect parameters will change with the change of the model operating condition parameters, and then different top-of-atmosphere radiances will be formed, generating corresponding different top-of-atmosphere radiance images.

[0121] Figure 5 It is a flowchart of calculating the top-of-atmosphere radiance using the atmospheric effect parameters in a method for generating hyperspectral classification samples based on image simulation according to an embodiment of the present invention.

[0122] like Figure 5 As shown, in one or more embodiments, preferably, the calculating the top of atmosphere radiance by using the atmospheric effect parameter specifically includes:

[0123] S501, obtaining the first atmospheric effect parameter, the second atmospheric effect parameter, and the third atmospheric effect parameter;

[0124] S502, calculating the top-of-atmosphere radiance for the surface reflectance image according to a fourth calculation formula;

[0125] The fourth calculation formula is:

[0126]

[0127] Wherein, L(λ) is the top-of-atmosphere radiance of the band λ, S(λ) is the first atmospheric effect parameter of the band λ, D(λ) is the second atmospheric effect parameter of the band λ, C(λ) is the third atmospheric effect parameter of the band λ, and ρ is the pixel surface reflectance.

[0128] The fourth calculation formula is obtained by simplifying the fifth calculation formula. Since the spatial resolution of satellite hyperspectral data is usually not high, for example, 30m, the simplified assumption is that the proximity effect is not considered. Specifically, for the visible light to short infrared spectrum, the top of atmosphere radiance is obtained using the following fifth calculation formula.

[0129]

[0130] Among them, ρ e is the average reflectance of the neighboring pixels, A λ , B λ are atmospheric effect parameters determined by atmospheric conditions and observation geometry, and A λ With B λ The sum of is D λ , S λ is the first atmospheric effect parameter, D λ is the second atmospheric effect parameter, C λ is the third atmospheric effect parameter.

[0131] In an embodiment of the present invention, the top-of-atmosphere radiance is calculated based on the first atmospheric effect parameter, the second atmospheric effect parameter, and the third atmospheric effect parameter. Since each new data input set by the preset value library will produce different atmospheric effect parameters, a different top-of-atmosphere radiance image will be generated.

[0132] According to a second aspect of an embodiment of the present invention, a hyperspectral classification sample generation system based on image simulation is provided.

[0133] Figure 6 It is a structural diagram of a hyperspectral classification sample generation system based on image simulation according to an embodiment of the present invention.

[0134] like Figure 6 As shown, the hyperspectral classification sample generation system based on image simulation includes:

[0135] A data acquisition unit 601 is used to acquire a surface reflectivity image;

[0136] A parameter setting unit 602 is used to set model operation condition parameters of the atmospheric radiation transfer model, wherein the model operation condition parameters include atmospheric conditions, observation geometry conditions, and atmospheric effect calculation intervals;

[0137] An atmospheric effect calculation unit 603, configured to calculate atmospheric effect parameters using the atmospheric radiation transfer model according to the model operation condition parameters;

[0138] The image generation unit 604 is used to simulate the surface reflectance image and generate an atmosphere top radiance image according to the atmospheric effect parameters.

[0139] In one or more embodiments, preferably, the hyperspectral classification sample generation system based on image simulation further includes a database 605, a storage medium 606, and a data display background 607;

[0140] The database 605 is used to store the first atmospheric effect parameter, the second atmospheric effect parameter and the third atmospheric effect parameter;

[0141] The storage medium 606 includes a mobile hard disk;

[0142] The data display background 607 is used to display the top-of-atmosphere radiance image.

[0143] This technical solution combines the atmospheric effect parameters calculated by the atmospheric radiation transfer model at different observation geometric angles to simulate the radiance images under different observation geometries, and then uses the simulated data as new samples to train the classification model to improve the accuracy of sample classification.

[0144] According to a third aspect of an embodiment of the present invention, there is provided a computer-readable storage medium on which computer program instructions are stored. When the computer program instructions are executed by a processor, the method as described in any one of the first aspect of the embodiment of the present invention is implemented.

[0145] According to a fourth aspect of an embodiment of the present invention, an electronic device is provided. Figure 7 It is a structural diagram of an electronic device in one embodiment of the present invention. Figure 7 The electronic device shown is a general hyperspectral classification sample generation device, which includes a general computer hardware structure, which at least includes a processor 701 and a memory 702. The processor 701 and the memory 702 are connected via a bus 703. The memory 702 is suitable for storing instructions or programs executable by the processor 701. The processor 701 can be an independent microprocessor or a collection of one or more microprocessors. Thus, the processor 701 executes the instructions stored in the memory 702 to execute the method flow of the embodiment of the present invention as described above to realize the processing of data and the control of other devices. The bus 703 connects the above-mentioned multiple components together, and at the same time connects the above-mentioned components to the display controller 704 and the display device and the input / output (I / O) device 705. The input / output (I / O) device 705 can be a mouse, a keyboard, a modem, a network interface, a touch input device, a somatosensory input device, a printer, and other devices known in the art. Typically, the input / output device 705 is connected to the system via an input / output (I / O) controller 706.

[0146] The technical solution provided by the embodiments of the present invention may have the following beneficial effects:

[0147] This technical solution uses image simulation to generate radiance images under different observation geometric conditions as fine classification samples of hyperspectral data. Compared with the existing data enhancement method, it takes into account the radiation characteristics of remote sensing images. In actual situations, the images acquired by satellites in the same area will be different due to different atmospheric and observation geometric conditions. Therefore, the classification sample generation scheme provided by the present invention not only expands the samples, but also ensures that the generated classification samples are closer to the characteristics of real images and are more suitable for training classification models.

[0148] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) containing computer-usable program code.

[0149] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0150] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0151] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0152] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for generating hyperspectral classification samples based on image simulation, characterized in that: include: Obtain surface reflectance images; Setting model operating condition parameters of the atmospheric radiation transfer model, wherein the model operating condition parameters include atmospheric conditions, observation geometry conditions, and calculation wavelengths of atmospheric effect parameters; Calculating atmospheric effect parameters using the atmospheric radiation transfer model according to the model operation condition parameters; According to the atmospheric effect parameters, the surface reflectance image is simulated to generate an atmospheric top radiance image; The step of calculating the atmospheric effect parameters by using the atmospheric radiation transfer model according to the model operation condition parameters specifically includes: Setting the atmospheric radiation transfer model according to the model operation condition parameters; Obtaining a first intermediate variable and a second intermediate variable corresponding to the first surface reflectivity and the second surface reflectivity; Obtaining a third intermediate variable and a fourth intermediate variable corresponding to the first surface reflectivity and the second surface reflectivity; Calculate the albedo of the large balloon surface according to the first calculation formula by combining the first intermediate variable, the second intermediate variable, the first surface reflectivity and the second surface reflectivity, and save it as the first atmospheric effect parameter; Calculate a second atmospheric effect parameter using a second calculation formula according to the first intermediate variable, the second intermediate variable, the third intermediate variable, the fourth intermediate variable, the first surface reflectivity, and the second surface reflectivity; Calculate the atmospheric path radiation caused by solar radiation using a third calculation formula according to the first intermediate variable, the second intermediate variable, the third intermediate variable, the fourth intermediate variable, the first surface reflectivity, and the second surface reflectivity, and save it as a third atmospheric effect parameter; Specifically, the first intermediate variable, the second intermediate variable, the third intermediate variable, the fourth intermediate variable and the first atmospheric effect parameter, the second atmospheric effect parameter and the third atmospheric effect parameter are all variables and parameters related to wavelength, that is, each of the calculated wavelengths has the corresponding first intermediate variable, the second intermediate variable, the third intermediate variable, the fourth intermediate variable and the first atmospheric effect parameter, the second atmospheric effect parameter and the third atmospheric effect parameter; The first calculation formula: Among them, S(λ) is the first atmospheric effect parameter of band λ, L gr1 is the first intermediate variable of the band λ, L gr2 is the second intermediate variable of the band λ, ρ1 is the first surface reflectivity, and ρ2 is the second surface reflectivity; The second calculation formula: Wherein, D(λ) is the second atmospheric effect parameter of the band λ, L gr1 (λ) is the first intermediate variable of the band λ, L gr2 (λ) is the second intermediate variable of the band λ, L ss1 (λ) is the third intermediate variable of the band λ, L ss2 (λ) is the fourth intermediate variable, ρ1 is the first surface reflectivity, and ρ2 is the second surface reflectivity; The third calculation formula: Wherein, C(λ) is the third atmospheric effect parameter of the band λ, L gr1 (λ) is the first intermediate variable of the band λ, L gr2 (λ) is the second intermediate variable of the band λ, L ss1 (λ) is the third intermediate variable of the band λ, L ss2 (λ) is the fourth intermediate variable of the band λ, ρ1 is the first surface reflectivity, and ρ2 is the second surface reflectivity.

2. The method for generating hyperspectral classification samples based on image simulation according to claim 1, characterized in that: The model operating condition parameters of the atmospheric radiation transfer model are set, wherein the model operating condition parameters include atmospheric conditions, observation geometry conditions, and calculation wavelengths of atmospheric effect parameters, specifically including: Setting a first surface reflectivity and a second surface reflectivity, wherein the first surface reflectivity and the second surface reflectivity are not equal; Setting the atmospheric conditions according to a preset value library, wherein the atmospheric conditions include water vapor content and visibility; The observation geometric conditions are set according to a preset value library, wherein the observation geometric conditions include a solar zenith angle and an observation zenith angle; The calculation wavelength is set so that the calculation wavelength of the atmospheric effect parameter is consistent with the central wavelength of each band of the hyperspectral image.

3. The method for generating hyperspectral classification samples based on image simulation according to claim 2, characterized in that: The preset value library specifically includes: The water vapor content is selected from 0 to 4.2 g / m 3 one of the; The visibility is selected from 10km to 50km; The solar zenith angle is selected from 0 to 60 degrees; The observation zenith angle is selected from 0 to 60 degrees.

4. The method for generating hyperspectral classification samples based on image simulation according to claim 1, characterized in that: The calculating the top-of-atmosphere radiance by using the atmospheric effect parameters specifically includes: Acquire the first atmospheric effect parameter, the second atmospheric effect parameter, and the third atmospheric effect parameter; calculate the top-of-atmosphere radiance for the surface reflectance image according to a fourth calculation formula; The fourth calculation formula is: Wherein, L(λ) is the top-of-atmosphere radiance of the band λ, S(λ) is the first atmospheric effect parameter of the band λ, D(λ) is the second atmospheric effect parameter of the band λ, C(λ) is the third atmospheric effect parameter of the band λ, and ρ is the pixel surface reflectance.

5. A hyperspectral classification sample generation system based on image simulation, characterized in that: The system includes: A data acquisition unit, used for acquiring a surface reflectivity image; A parameter setting unit, used to set model operation condition parameters of the atmospheric radiation transfer model, wherein the model operation condition parameters include atmospheric conditions, observation geometry conditions, and calculation wavelengths of atmospheric effect parameters; An atmospheric effect calculation unit, used to calculate atmospheric effect parameters using the atmospheric radiation transfer model according to the model operation condition parameters; An image generating unit, configured to simulate the surface reflectance image and generate an atmosphere top radiance image according to the atmospheric effect parameter; The step of calculating the atmospheric effect parameters by using the atmospheric radiation transfer model according to the model operation condition parameters specifically includes: Setting the atmospheric radiation transfer model according to the model operation condition parameters; Obtaining a first intermediate variable and a second intermediate variable corresponding to the first surface reflectivity and the second surface reflectivity; Obtaining a third intermediate variable and a fourth intermediate variable corresponding to the first surface reflectivity and the second surface reflectivity; Calculate the albedo of the large balloon surface according to the first calculation formula by combining the first intermediate variable, the second intermediate variable, the first surface reflectivity and the second surface reflectivity, and save it as the first atmospheric effect parameter; Calculate a second atmospheric effect parameter using a second calculation formula according to the first intermediate variable, the second intermediate variable, the third intermediate variable, the fourth intermediate variable, the first surface reflectivity, and the second surface reflectivity; Calculate the atmospheric path radiation caused by solar radiation using a third calculation formula according to the first intermediate variable, the second intermediate variable, the third intermediate variable, the fourth intermediate variable, the first surface reflectivity, and the second surface reflectivity, and save it as a third atmospheric effect parameter; Specifically, the first intermediate variable, the second intermediate variable, the third intermediate variable, the fourth intermediate variable and the first atmospheric effect parameter, the second atmospheric effect parameter and the third atmospheric effect parameter are all variables and parameters related to wavelength, that is, each of the calculated wavelengths has the corresponding first intermediate variable, the second intermediate variable, the third intermediate variable, the fourth intermediate variable and the first atmospheric effect parameter, the second atmospheric effect parameter and the third atmospheric effect parameter; The first calculation formula: Among them, S(λ) is the first atmospheric effect parameter of band λ, L gr1 is the first intermediate variable of the band λ, L gr2 is the second intermediate variable of the band λ, ρ1 is the first surface reflectivity, and ρ2 is the second surface reflectivity; The second calculation formula: Wherein, D(λ) is the second atmospheric effect parameter of the band λ, L gr1 (λ) is the first intermediate variable of the band λ, L gr2 (λ) is the second intermediate variable of the band λ, L ss1 (λ) is the third intermediate variable of the band λ, L ss2 (λ) is the fourth intermediate variable, ρ1 is the first surface reflectivity, and ρ2 is the second surface reflectivity; The third calculation formula: Wherein, C(λ) is the third atmospheric effect parameter of the band λ, L gr1 (λ) is the first intermediate variable of the band λ, L gr2 (λ) is the second intermediate variable of the band λ, L ss1 (λ) is the third intermediate variable of the band λ, L ss2 (λ) is the fourth intermediate variable of the band λ, ρ1 is the first surface reflectivity, and ρ2 is the second surface reflectivity.

6. The hyperspectral classification sample generation system based on image simulation according to claim 5, characterized in that: It also includes database, storage media, and data display background; The database is used to store the first atmospheric effect parameter, the second atmospheric effect parameter and the third atmospheric effect parameter; The storage medium includes a mobile hard disk; The data display background is used to display the top-of-atmosphere radiance image.

7. A computer-readable storage medium storing computer program instructions, characterized in that: The computer program instructions implement the method according to any one of claims 1 to 4 when executed by a processor.

8. An electronic device, comprising a memory and a processor, characterized in that: The memory is used to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the method according to any one of claims 1-4.

Images