Infrared remote sensing image generation method and device and electronic equipment

Abstract

The invention relates to the technical field of image generation, and provides an infrared remote sensing image generation method and device and electronic equipment, and a plurality of temperature prediction functions are constructed according to a temperature prediction parameter set of a plurality of types of ground feature materials contained in a target shooting area at a preset moment, a ground surface temperature value and ground feature material multi-spectral reflectance. And calculating radiation brightness values of the plurality of types of ground feature materials at the plurality of moments according to the ground feature material multi-spectral reflectivity corresponding to the plurality of types of ground feature materials, the temperature prediction function, the temperature to-be-predicted parameter set at the plurality of moments, the infrared radiation parameter set and a preset radiation brightness value calculation function. And according to the plurality of radiation brightness values, obtaining initial infrared remote sensing images at the plurality of moments. And constructing a data training set of an infrared remote sensing image generation model according to the original visible light remote sensing image and the initial infrared remote sensing images at the multiple moments. And training the infrared remote sensing image generation model by using the data training set to obtain a trained target infrared remote sensing image generation model.

Description

Technical Field

[0001] The present invention relates to the technical field of image generation, and in particular to an infrared remote sensing image generation method, device and electronic equipment. Background Art

[0002] At present, infrared remote sensing technology has unique advantages over visible light remote sensing technology in penetrating haze, clouds and other atmospheric interference, so it is widely used in many fields such as environmental monitoring, disaster warning, agricultural monitoring, military reconnaissance, etc. With the development of artificial intelligence technology, infrared remote sensing images have shown great advantages in using network models to perform target detection and recognition tasks through infrared remote sensing images. Therefore, the demand for high-quality infrared remote sensing images is increasing day by day.

[0003] Based on this, in the prior art, infrared remote sensing images can be generated through physical models and deep learning network methods. Among them, the physical model method generates infrared remote sensing images by simulating the physical phenomena in the real infrared radiation transmission process, and the deep learning network generates infrared remote sensing images by automatically learning the complex mapping relationship between visible light images and infrared images.

[0004] However, using existing technologies, the physical model method is easily affected by external environmental factors, and the temperature prediction model in the physical model method is highly complex and time-consuming, while the deep learning network relies on a large amount of data and has problems of overfitting and poor model generalization ability, resulting in poor image quality and low efficiency in the generated infrared remote sensing images. Summary of the invention

[0005] Based on this, it is necessary to provide an infrared remote sensing image generation method, device and electronic equipment to address the above technical problems.

[0006] In a first aspect, an embodiment of the present invention provides a method for generating an infrared remote sensing image, the method comprising:

[0007] Constructing multiple temperature prediction functions according to a set of temperature prediction parameters, surface temperature values, and multi-spectral reflectivity of the surface material of multiple categories of surface material contained in the target shooting area at a preset time, wherein the set of temperature prediction parameters at least includes: the atmospheric temperature, atmospheric humidity, and wind speed of multiple categories of surface material at a preset time, and the surface temperature values ​​and multi-spectral reflectivity of the surface material are obtained according to an original visible light remote sensing image taken at the preset time of the target shooting area, and a corresponding original infrared remote sensing image;

[0008] According to the multi-spectral reflectivity of the ground object material corresponding to the multiple categories of ground object materials, the temperature prediction function, the temperature parameter set to be predicted at multiple times, the infrared radiation parameter set and the preset radiation brightness value calculation function, the radiation brightness values ​​of the multiple categories of ground object materials at multiple times are calculated, wherein the infrared radiation parameter set at least includes: solar radiation brightness value, sky radiation brightness value, atmospheric transmittance, and atmospheric path radiation brightness value;

[0009] According to multiple radiation brightness values, initial infrared remote sensing images at multiple moments are obtained through sensor imaging technology;

[0010] Constructing a data training set of an infrared remote sensing image generation model based on the original visible light remote sensing image and initial infrared remote sensing images at multiple moments;

[0011] An infrared remote sensing image generation model is trained using a data training set to obtain a trained target infrared remote sensing image generation model, wherein the target infrared remote sensing image generation model is used to generate a target infrared remote sensing image based on a visible light remote sensing image. The infrared remote sensing image generation model includes a generator and a discriminator. The generator includes at least: a downsampling layer, an intermediate layer, an upsampling layer, and a multi-head attention layer. The downsampling layer, the intermediate layer, and the upsampling layer include a hole residual block. Multi-head attention feature extraction is performed between the downsampling layer and the upsampling layer through a jump connection.

[0012] In one embodiment, the surface temperature value and the multi-spectral reflectivity of the ground object material are obtained based on the original visible light remote sensing image taken at a preset time of the target shooting area and the corresponding original infrared remote sensing image, including:

[0013] Acquire the original visible light remote sensing image of the target shooting area captured at a preset time, and the original infrared remote sensing image corresponding to the original visible light remote sensing image;

[0014] Determine multiple categories of ground object materials contained in the original visible light remote sensing image, and mark different categories of ground object materials;

[0015] According to the original visible light remote sensing image and the original infrared remote sensing image, the surface temperature values ​​of multiple categories of ground material at a preset time and the initial ground material multi-spectral reflectance curve are obtained, wherein the initial ground material multi-spectral reflectance curve is the first infrared band;

[0016] The multi-spectral reflectivity of the ground object material is determined according to the multi-spectral reflectance curve of the initial ground object material and the preset multi-spectral reflectance curves of multiple preset unknown categories of ground object materials in the preset software spectral data, wherein the multi-spectral reflectance curve of the ground object material is a second infrared band, and the second infrared band is larger than the first infrared band.

[0017] In one embodiment, determining the multi-spectral reflectivity of the ground object material according to the initial ground object material multi-spectral reflectance curve and a plurality of preset multi-spectral reflectance curves of preset unknown ground object materials in the preset software spectral data includes:

[0018] Matching the initial ground object material multi-spectral reflectance curve with a plurality of preset multi-spectral reflectance curves of a plurality of preset unknown ground object materials, determining that the preset unknown ground object material corresponding to the preset multi-spectral reflectance curve with the highest matching degree is the ground object material, and the preset multi-spectral reflectance curve with the highest matching degree is the target multi-spectral reflectance curve, wherein the preset multi-spectral reflectance curve is a third infrared band, and the third infrared band is greater than the second infrared band;

[0019] The multi-spectral reflectivity of the ground feature material is determined according to the target multi-spectral reflectivity curve.

[0020] In one embodiment, the method constructs multiple temperature prediction functions based on a set of temperature prediction parameters of multiple types of ground material contained in the target shooting area at a preset time, a ground surface temperature value, and a multi-spectral reflectivity of the ground material, including:

[0021] By using the least squares method, the temperature prediction parameter set, the surface temperature value and the multi-spectral reflectivity of the ground material of each category of ground material at a preset time are fitted to obtain multiple temperature prediction factors corresponding to each category of ground material;

[0022] A plurality of temperature prediction functions are constructed according to the temperature prediction parameter sets corresponding to the plurality of categories of ground material, the multi-spectral reflectivity of the ground material and the plurality of temperature prediction factors.

[0023] In one embodiment, the radiation brightness values ​​of multiple categories of ground material at multiple times are calculated based on the multi-spectral reflectivity of the ground material corresponding to multiple categories of ground material, the temperature prediction function, the temperature parameter set to be predicted at multiple times, the infrared radiation parameter set and the preset radiation brightness value calculation function, including:

[0024] According to the multi-spectral reflectivity of the ground object material corresponding to the multiple categories of ground object materials and the set of temperature parameters to be predicted at multiple moments, the predicted temperature values ​​of the multiple categories of ground object materials at multiple moments are obtained through the temperature prediction function;

[0025] According to the predicted temperature values ​​of multiple categories of ground object materials at multiple moments, the infrared radiation parameter set and the preset radiation brightness value calculation function, the radiation brightness values ​​of multiple categories of ground object materials at multiple moments are calculated.

[0026] In one embodiment, the method of calculating the radiation brightness values ​​of the multiple categories of ground material at multiple times according to the predicted temperature values ​​of the multiple categories of ground material at multiple times, the infrared radiation parameter set and the preset radiation brightness value calculation function includes:

[0027] According to the predicted temperature values ​​of multiple types of ground materials at multiple times, multiple blackbody radiation brightnesses are calculated using Planck's formula;

[0028] According to a plurality of blackbody radiation brightnesses, the infrared radiation parameter set and the preset radiation brightness value calculation function, the radiation brightness values ​​of a plurality of categories of ground object materials at a plurality of moments are calculated.

[0029] In one embodiment, the step of constructing a data training set of an infrared remote sensing image generation model based on the original visible light remote sensing image and the initial infrared remote sensing images at multiple moments includes:

[0030] Determine the original visible light remote sensing image and the initial infrared remote sensing image corresponding to each moment as a data training subset;

[0031] A data training set for an infrared remote sensing image generation model is constructed based on the data training subsets corresponding to a plurality of time moments.

[0032] In one embodiment, the downsampling layer and the upsampling layer are connected via a skip connection to perform multi-head attention feature extraction, including:

[0033] Determine the target upsampling sublayer corresponding to each downsampling sublayer;

[0034] The downsampled feature maps output by each downsampling sublayer are input to the target upsampling sublayer through the multi-head attention layer.

[0035] In a second aspect, an embodiment of the present invention provides an infrared remote sensing image generation device, comprising:

[0036] A function construction module, for constructing multiple temperature prediction functions according to a set of temperature prediction parameters of multiple types of ground material contained in a target shooting area at a preset time, a surface temperature value, and a multi-spectral reflectivity of the ground material, wherein the set of temperature prediction parameters at least includes: the atmospheric temperature, atmospheric humidity, and wind speed of multiple types of ground material at a preset time, and the surface temperature value and the multi-spectral reflectivity of the ground material are obtained according to an original visible light remote sensing image taken at the preset time of the target shooting area, and a corresponding original infrared remote sensing image;

[0037] A radiation brightness value acquisition module is used to calculate the radiation brightness values ​​of multiple categories of ground material at multiple times according to the multi-spectral reflectivity of the ground material corresponding to multiple categories of ground material, the temperature prediction function, the temperature parameter set to be predicted at multiple times, the infrared radiation parameter set and the preset radiation brightness value calculation function, wherein the infrared radiation parameter set includes at least: solar radiation brightness value, sky radiation brightness value, atmospheric transmittance, and atmospheric path radiation brightness value;

[0038] An initial infrared remote sensing image acquisition module is used to acquire initial infrared remote sensing images at multiple times through sensor imaging technology according to multiple radiation brightness values;

[0039] A training set acquisition module, used to construct a data training set of an infrared remote sensing image generation model based on the original visible light remote sensing image and initial infrared remote sensing images at multiple moments;

[0040] A training module is used to train an infrared remote sensing image generation model using a data training set to obtain a trained target infrared remote sensing image generation model, wherein the target infrared remote sensing image generation model is used to generate a target infrared remote sensing image based on a visible light remote sensing image, and the infrared remote sensing image generation model includes a generator and a discriminator, and the generator includes at least: a downsampling layer, an intermediate layer, an upsampling layer, and a multi-head attention layer, the downsampling layer, the intermediate layer, and the upsampling layer include a hole residual block, and multi-head attention feature extraction is performed between the downsampling layer and the upsampling layer through a jump connection.

[0041] In a third aspect, an embodiment of the present invention provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the infrared remote sensing image generation method described in the first aspect are implemented.

[0042] The technical solution provided by the embodiment of the present invention has the following advantages compared with the prior art:

[0043] An infrared remote sensing image generation method, device and electronic device provided by the embodiment of the present invention adopts this method to construct multiple temperature prediction functions through the temperature prediction parameter set of multiple categories of ground material contained in the target shooting area at preset time, the surface temperature value and the multi-spectral reflectivity of the ground material. According to the multi-spectral reflectivity of the ground material corresponding to the multiple categories of ground material, the temperature prediction function, the temperature parameter set to be predicted at multiple times, the infrared radiation parameter set and the preset radiation brightness value calculation function, the radiation brightness values ​​of the multiple categories of ground material at multiple times are calculated, so as to obtain the initial infrared remote sensing images at multiple times, which can reduce the parameters for obtaining the initial infrared remote sensing image and improve the calculation efficiency. Further, the original visible light remote sensing image and the initial infrared remote sensing images at multiple times are used to construct a data training set for training the infrared remote sensing image generation model, enhance the data training set, and improve the image quality of the target infrared remote sensing image generated by the trained target infrared remote sensing image generation model. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0046] Figure 1 A schematic diagram of a flow chart of a method for generating an infrared remote sensing image provided by an embodiment of the present invention;

[0047] Figure 2 A schematic diagram of a network model of a generator provided by an embodiment of the present invention;

[0048] Figure 3 A schematic structural diagram of an infrared remote sensing image generating device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0049] In order to more clearly understand the above-mentioned objectives, features and advantages of the present invention, the scheme of the present invention will be further described below. It should be noted that the embodiments of the present invention and the features in the embodiments can be combined with each other without conflict.

[0050] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present invention, rather than all of the embodiments.

[0051] At present, in order to obtain infrared remote sensing images, physical model methods and deep learning network methods are mainly used. Among them, the physical model method generates infrared remote sensing images by simulating the physical phenomena in the real infrared radiation transmission process, and the deep learning network method generates infrared remote sensing images by automatically learning the complex mapping relationship between visible light images and infrared images.

[0052] However, using existing technologies, the physical model method is easily affected by external environmental factors, and the temperature prediction model in the physical model method is highly complex and time-consuming, while the deep learning network relies on a large amount of data and has problems of overfitting and poor model generalization ability, resulting in poor image quality and low efficiency in the generated infrared remote sensing images.

[0053] Therefore, the present invention provides an infrared remote sensing image generation method, which constructs multiple temperature prediction functions through the temperature prediction parameter set of multiple categories of ground material contained in the target shooting area at preset moments, the surface temperature value and the multi-spectral reflectivity of the ground material. According to the multi-spectral reflectivity of the ground material corresponding to the multiple categories of ground material, the temperature prediction function, the temperature parameter set to be predicted at multiple moments, the infrared radiation parameter set and the preset radiation brightness value calculation function, the radiation brightness values ​​of the multiple categories of ground material at multiple moments are calculated, so as to obtain the initial infrared remote sensing images at multiple moments, which can reduce the parameters for obtaining the initial infrared remote sensing images, improve the calculation efficiency, further use the original visible light remote sensing images and the initial infrared remote sensing images at multiple moments to construct the data training set for training the infrared remote sensing image generation model, enhance the data training set, and improve the image quality of the target infrared remote sensing image generated by the trained target infrared remote sensing image generation model.

[0054] In one embodiment, Figure 1 As shown, Figure 1 A schematic flow chart of a method for generating an infrared remote sensing image provided by an embodiment of the present invention specifically includes the following steps:

[0055] S10: constructing multiple temperature prediction functions according to a set of temperature prediction parameters of multiple categories of ground material included in the target shooting area at a preset time, a ground surface temperature value, and a multi-spectral reflectivity of the ground material.

[0056] The target shooting area refers to an area corresponding to the target object being shot, such as grassland or city, but is not limited thereto. The present invention is not specifically limited thereto, and those skilled in the art may set it according to actual conditions.

[0057] The above-mentioned ground features refer to solid objects on the ground surface, and the ground feature material refers to the material corresponding to the solid object. The ground feature material can be, for example, water bodies, buildings, roads, soil and grass, but is not limited to this. The present invention does not specifically limit it, and those skilled in the art can set it according to actual conditions.

[0058] The above-mentioned preset time refers to a specific time, such as 8 o'clock, but is not limited to this. The present invention is not specifically limited, and those skilled in the art can set it according to actual conditions.

[0059] The above-mentioned temperature prediction parameter set includes at least: atmospheric temperature, atmospheric humidity, and wind speed of multiple categories of ground material at a preset time, wherein the atmospheric temperature, atmospheric humidity, and wind speed can be obtained from a publicly available meteorological data website.

[0060] The above-mentioned ground surface temperature value and multi-spectral reflectivity of ground material are obtained based on the original visible light remote sensing image taken at a preset time of the target shooting area and the corresponding original infrared remote sensing image.

[0061] Among them, the multi-spectral reflectivity of ground material refers to the reflectivity information of ground objects in different spectral bands, which can realize the fine identification and analysis of ground objects and reflect the reflectivity of ground objects to light. The multi-spectral reflectivity of ground material can be determined by the ratio of the radiant brightness in a specific spectral band to the solar radiation incident on the band.

[0062] Specifically, a set of temperature prediction parameters corresponding to multiple categories of ground material contained in the target shooting area at a preset time is obtained, that is, the atmospheric temperature, atmospheric humidity, wind speed, surface temperature value and multi-spectral reflectivity of the multiple categories of ground material at the preset time; and temperature prediction functions corresponding to the multiple categories of ground material are constructed according to the atmospheric temperature, atmospheric humidity, wind speed, surface temperature value and multi-spectral reflectivity of the multiple categories of ground material at the preset time.

[0063] S11: Calculate the radiation brightness values ​​of multiple categories of ground material at multiple moments according to the multi-spectral reflectivity of the ground material corresponding to multiple categories of ground material, the temperature prediction function, the set of temperature parameters to be predicted at multiple moments, the infrared radiation parameter set and the preset radiation brightness value calculation function.

[0064] The temperature parameter set to be predicted refers to parameter information used to obtain predicted temperature values ​​corresponding to multiple moments, and the temperature parameter set to be predicted includes at least: atmospheric temperature, atmospheric humidity, and wind speed of multiple categories of ground material at multiple moments.

[0065] The above-mentioned infrared radiation parameter set at least includes: solar radiation brightness value, sky radiation brightness value, atmospheric transmittance, atmospheric path radiation brightness value. The solar radiation brightness value, sky radiation brightness value, atmospheric transmittance, and atmospheric path radiation brightness value can be calculated by using preset software such as MODTRAN software. Specifically, the imaging time of image data at multiple times at Landsat8 data, the longitude and latitude of the shooting area, the solar altitude angle, the solar zenith angle, the observation zenith angle and the band information are obtained, and the solar radiation brightness value, the sky radiation brightness value, the atmospheric transmittance, and the atmospheric path radiation brightness value are calculated by preset software such as MODTRAN software, but it is not limited to this. The present invention is not specifically limited, and those skilled in the art can set it according to actual conditions.

[0066] The above-mentioned preset radiation brightness value calculation function refers to a function defined for calculating the radiation brightness value. The preset radiation brightness value calculation function can be defined as:

[0067] L a =[L b ×α+(L sun +L sky )×ρ]×τ+L p

[0068] Where, L b is the blackbody radiation brightness; α is the emissivity of the ground material; L sun The solar radiation brightness received by multiple types of ground materials; L sky is the sky background radiation brightness received by multiple types of ground material; ρ is the reflectivity of the ground material; τ is the atmospheric transmittance; L p is the atmospheric path radiation brightness, wherein α and ρ are not specifically limited in the present invention and those skilled in the art can set them according to actual conditions.

[0069] The above-mentioned radiation brightness value refers to the infrared radiation energy used to characterize the object, and the infrared remote sensing image can be obtained based on the radiation brightness value.

[0070] S12: Acquire initial infrared remote sensing images at multiple moments through sensor imaging technology according to the multiple radiation brightness values.

[0071] Specifically, after obtaining the radiation brightness values ​​corresponding to multiple categories of ground object materials at multiple times, the sensor imaging technology obtains the initial infrared remote sensing images at multiple times.

[0072] S13: Construct a data training set for an infrared remote sensing image generation model based on the original visible light remote sensing image and the initial infrared remote sensing images at multiple moments.

[0073] The data training set is used to train an initial infrared remote sensing image generation model for generating infrared remote sensing images, and the infrared remote sensing image generation model is a neural network model such as an adversarial network.

[0074] Optionally, based on the above embodiment, in some embodiments of the present invention, an implementation manner of S13 may be:

[0075] S131: Determine that the original visible light remote sensing image and the initial infrared remote sensing image corresponding to each moment are a data training subset.

[0076] S132: Constructing a data training set for an infrared remote sensing image generation model based on data training subsets corresponding to a plurality of time instants.

[0077] Specifically, for an original visible light remote sensing image, and initial infrared remote sensing images corresponding to the original visible light remote sensing image at multiple moments, an original visible light remote sensing image and an initial infrared remote sensing image corresponding to each moment are determined as a data training subset. Then, for an original visible light remote sensing image and an initial infrared remote sensing image corresponding to the original visible light remote sensing image at multiple moments, data training subsets corresponding to the multiple moments are obtained, thereby obtaining a data training set for training an infrared remote sensing image generation model.

[0078] In this way, this embodiment enhances the data training set by constructing a data training set including the original visible light remote sensing image and the initial infrared remote sensing images corresponding to the original visible light remote sensing image at multiple times, thereby improving the training effect of the infrared remote sensing image generation model, thereby improving the image quality of the infrared remote sensing image obtained by the infrared remote sensing image generation model.

[0079] S14: Use the data training set to train the infrared remote sensing image generation model to obtain a trained target infrared remote sensing image generation model.

[0080] Among them, the target infrared remote sensing image generation model is used to generate a target infrared remote sensing image based on a visible light remote sensing image. Specifically, the visible light remote sensing image is input into a trained target infrared remote sensing image generation model, and the target infrared remote sensing image is used to generate a target infrared remote sensing image.

[0081] The above-mentioned infrared remote sensing image generation model includes a generator and a discriminator. The generator includes at least: a downsampling layer, an intermediate upsampling layer and a multi-head attention layer, wherein the downsampling layer, the intermediate layer and the upsampling layer contain a hole residual block, and the residual block of the original infrared remote sensing image generation model is replaced with the hole residual block, so that the target infrared remote sensing image generation model can better extract the spatial information of the remote sensing image at different scales, improve the feature extraction capability, and thus improve the image quality of the generated infrared remote sensing image.

[0082] The above-mentioned downsampling layer and upsampling layer are connected through skip connections to perform multi-head attention feature extraction.

[0083] Among them, multi-head attention features refer to feature extraction performed through multi-head attention layers. In this way, the deep semantic features and shallow features of the input image can be superimposed from bottom to top to obtain more sufficient feature information, thereby improving the image quality of generated infrared remote sensing images.

[0084] Optionally, based on the above embodiments, in some embodiments of the present invention, Figure 2 As shown, the downsampling layer 21 includes multiple downsampling sublayers 211, and the upsampling layer 23 includes multiple upsampling sublayers 233. It should be noted that the number of downsampling sublayers is the same as that of upsampling sublayers. One implementation method of performing multi-head attention feature extraction through jump connections between the downsampling layer and the upsampling layer may be:

[0085] S131: Determine a target upsampling sublayer corresponding to each downsampling sublayer.

[0086] S132: The downsampled feature maps output by each downsampling sublayer are input to the target upsampling sublayer through the multi-head attention layer.

[0087] Specifically, the downsampling layer includes multiple downsampling sublayers, and the corresponding target upsampling sublayer is determined among the multiple upsampling sublayers. The downsampling feature maps output by each downsampling sublayer are extracted through the multi-head attention layer and then input into the target upsampling sublayer.

[0088] For example, continue to refer to Figure 2 As shown, each downsampling sublayer 211 included in the downsampling layer 21 determines the corresponding target upsampling sublayer 233 among the multiple upsampling sublayers 233 included in the upsampling layer 23, and the downsampling feature map output by each downsampling sublayer 211 is input into the target upsampling sublayer 233 after the attention feature is extracted through the multi-head attention layer 24.

[0089] In this way, this embodiment performs multi-head attention feature extraction through jump connections between the downsampling layer and the upsampling layer, so that the target infrared remote sensing image generation model can better extract the spatial information of the remote sensing image at different scales, improve the feature extraction capability, and thus improve the image quality of the generated infrared remote sensing image.

[0090] In this way, the infrared remote sensing image generation method provided by this embodiment constructs multiple temperature prediction functions through the temperature prediction parameter set, surface temperature value and multi-spectral reflectivity of multiple categories of ground material contained in the target shooting area at preset times. According to the multi-spectral reflectivity of ground material corresponding to multiple categories of ground material, temperature prediction function, temperature parameter set to be predicted at multiple times, infrared radiation parameter set and preset radiation brightness value calculation function, the radiation brightness values ​​of multiple categories of ground material at multiple times are calculated, so as to obtain the initial infrared remote sensing images at multiple times, which can reduce the parameters for obtaining the initial infrared remote sensing images and improve the calculation efficiency. Further, the original visible light remote sensing images and the initial infrared remote sensing images at multiple times are used to construct a data training set for training the infrared remote sensing image generation model, enhance the data training set, and improve the image quality of the target infrared remote sensing image generated by the trained target infrared remote sensing image generation model.

[0091] Optionally, based on the above embodiments, in some embodiments of the present invention, the ground surface temperature value and the multi-spectral reflectivity of the ground object material are obtained according to the original visible light remote sensing image taken at a preset time of the target shooting area, and the corresponding original infrared remote sensing image. One way of obtaining the image may be:

[0092] S201: Acquire an original visible light remote sensing image of a target shooting area captured at a preset time, and an original infrared remote sensing image corresponding to the original visible light remote sensing image.

[0093] Specifically, for the target shooting area, an original visible light remote sensing image taken at a preset time and an original infrared remote sensing image taken at the same angle as the original visible light remote sensing image at the preset time are obtained.

[0094] Exemplarily, it can be obtained from a publicly available data set such as the Landsat8 data set, but is not limited thereto. The present invention is not specifically limited thereto, and those skilled in the art can set it according to actual conditions.

[0095] S202: Determine multiple categories of ground object materials contained in the original visible light remote sensing image, and mark different categories of ground object materials.

[0096] Among them, marking processing refers to marking multiple categories of ground material by preset numbers and preset characters, so that multiple categories of ground material are represented by the preset numbers and preset characters. For example, for the original visible light remote sensing image, the five categories of ground material including water body, building, road, soil and grassland are determined, and the markings of water body, building, road, soil and grassland are determined to be 1, 2, 3, 4, 5, but not limited to this. The present invention is not specifically limited, and those skilled in the art can set it according to actual conditions.

[0097] Specifically, after obtaining the original visible light remote sensing image of the target shooting area captured at a preset time, multiple categories of ground object materials contained in the original visible light remote sensing image are determined, and different categories of ground object materials are marked.

[0098] Optionally, based on the above embodiments, in some embodiments of the present invention, the original visible light remote sensing image is classified by a semi-supervised grayscale threshold segmentation method to determine multiple categories of ground material contained in the original visible light remote sensing image.

[0099] S203: According to the original visible light remote sensing image and the original infrared remote sensing image, the surface temperature values ​​of multiple categories of ground material at a preset time and the multi-spectral reflectance curves of the initial ground material are obtained.

[0100] Among them, the initial multi-spectral reflectance curve of the ground object material is the first infrared band. Exemplarily, the first infrared band can be 1.57um-1.65um, but is not limited to this. The present invention is not specifically limited, and those skilled in the art can set it according to actual conditions.

[0101] Specifically, after obtaining the original visible light remote sensing image and the original infrared remote sensing image, the surface temperature values ​​and multi-spectral reflectance curves of multiple categories of ground materials at preset times are obtained based on the original visible light remote sensing image and the original infrared remote sensing image.

[0102] Optionally, based on the above embodiments, in some embodiments of the present invention, for the original visible light remote sensing images and the original infrared remote sensing images, the surface temperature values ​​of multiple categories of ground materials at preset times and the multi-spectral reflectance curves of the initial ground material are obtained through inversion technology.

[0103] S204: Determine the multi-spectral reflectivity of the ground object material according to the initial multi-spectral reflectance curve of the ground object material and the preset multi-spectral reflectance curves of multiple preset unknown categories of ground object materials in the preset software spectral data.

[0104] Among them, the multi-spectral reflectance curve of the ground object material is the second infrared band, and the second infrared band is larger than the first infrared band. Exemplarily, following the above embodiment, the first infrared band can be 1.57um-1.65um, for example, and the second infrared band is 1um-3um. Since the second infrared band is larger than the first infrared band, the multi-spectral reflectance curve of the ground object material corresponding to the second infrared band can provide more image information when the infrared remote sensing image is subsequently acquired, thereby improving the quality of the generated infrared remote sensing image.

[0105] The plurality of preset unknown-category ground object materials are a plurality of unknown-category ground object materials.

[0106] Optionally, based on the above embodiment, in some embodiments of the present invention, an implementation manner of S204 may be:

[0107] The initial ground object material multi-spectral reflectance curve is matched with the preset multi-spectral reflectance curves of multiple preset unknown category ground object materials, and the preset unknown category ground object material corresponding to the preset multi-spectral reflectance curve with the highest matching degree is determined as the ground object material, and the preset multi-spectral reflectance curve with the highest matching degree is determined as the target multi-spectral reflectance curve.

[0108] The preset multi-spectral reflectance curve is the third infrared band, and the third infrared band is greater than the second infrared band. Exemplarily, following the above embodiment, the first infrared band may be, for example, 1.57um-1.65um, the second infrared band is 1um-3um, and the third infrared band is 1um-14um, but is not limited thereto. The present invention is not specifically limited thereto, and those skilled in the art may set it according to actual conditions.

[0109] Specifically, the multi-spectral reflectance curve of the ground object material is matched with the preset multi-spectral reflectance curves of multiple preset unknown ground object materials, and the preset multi-spectral reflectance curve of the preset unknown ground object material with the highest matching degree is determined as the target multi-spectral reflectance curve. That is, it can be understood that the preset unknown ground object material with the highest matching degree is the ground object material currently being matched.

[0110] Exemplarily, assuming that the current ground object material is a water body, the multi-spectral reflection curve corresponding to the water body is curve 1, and the preset multi-spectral reflection curves corresponding to multiple preset unknown category ground object materials in the preset software spectral data are curve 2, curve 3, curve 4, curve 5, curve 6...curve n, and the third infrared bands corresponding to curve 2, curve 3, curve 4, curve 5, curve 6...curves are all greater than the first infrared band corresponding to curve 1. Curve 1 is matched with curve 2, curve 3, curve 4, curve 5, curve 6...curve n, and it is determined that the curve with the highest matching degree is curve 4. It is considered that the preset unknown category ground object material of the unknown category corresponding to curve 4 is a water body, and curve 4 corresponding to the water body is the target multi-spectral reflection curve, but is not limited to this. The present invention is not specifically limited, and those skilled in the art can set it according to actual conditions.

[0111] Determine the multi-spectral reflectivity of the ground material based on the target multi-spectral reflectance curve.

[0112] Specifically, after obtaining the target multi-spectral reflectance curve, the multi-spectral reflectance of the ground object material is determined according to the target multi-spectral reflectance curve.

[0113] Optionally, based on the above embodiment, in some embodiments of the present invention, a method for determining the multi-spectral reflectivity of the ground object material according to the target multi-spectral reflectivity curve may be:

[0114] The target multi-spectral reflectance curve is sampled and the multiple reflectivities obtained are averaged, and the average value obtained is the multi-spectral reflectance of the ground object material.

[0115] In this way, the infrared remote sensing image generation method provided by the present invention matches the initial multi-spectral reflectance curve of the ground object material corresponding to the ground object material, obtains the preset multi-spectral reflectance curve of the preset unknown category ground object material with a high matching degree as the target multi-spectral reflectance curve, and further determines the multi-spectral reflectivity of the ground object material according to the target multi-spectral reflectance curve. Since the second infrared band corresponding to the target multi-spectral reflectance curve is larger than the first infrared band of the initial multi-spectral reflectance curve of the ground object material, the obtained multi-spectral reflectivity of the ground object material can provide more image information when the infrared remote sensing image is subsequently acquired, thereby improving the quality of the generated infrared remote sensing image.

[0116] Optionally, based on the above embodiment, in some embodiments of the present invention, an implementation manner of S10 may be:

[0117] S101: Fitting the temperature prediction parameter set, the surface temperature value and the multi-spectral reflectivity of the ground material of each category at a preset time by the least square method to obtain multiple temperature prediction factors corresponding to each category of ground material.

[0118] S102: Constructing multiple temperature prediction functions according to temperature prediction parameter sets corresponding to multiple categories of ground material, multi-spectral reflectivity of ground material, and multiple temperature prediction factors.

[0119] Specifically, a set of temperature prediction parameters for each category of ground material at a preset time is obtained, such as the atmospheric temperature, atmospheric humidity, and wind speed of multiple categories of ground material at the preset time. The atmospheric temperature, atmospheric humidity, wind speed, surface temperature value, and multi-spectral reflectivity of the ground material of multiple categories of ground material at the preset time are fitted by the least squares method to obtain multiple temperature prediction factors of the multiple categories of ground material at the preset time. According to the temperature prediction parameter sets corresponding to the multiple categories of ground material, the multi-spectral reflectivity of the ground material, and the multiple temperature prediction factors, the temperature prediction functions corresponding to each category of ground material are constructed.

[0120] Optionally, based on the above embodiment, in some embodiments of the present invention, the temperature prediction function may be defined by the following expression:

[0121] T a =a0+a1T air +a2H+a3V+a4ρ

[0122] Among them, T airrepresents atmospheric temperature, H represents humidity, V represents wind speed, ρ represents shortwave reflectivity of ground material, a0, a1, a2, a3, a4 represent temperature prediction factors, and a0, a1, a2, a3, a4 are obtained by fitting.

[0123] In this way, this embodiment constructs a temperature prediction function corresponding to each type of ground object material, and can obtain the predicted temperature of each type of ground object material at multiple times through a small number of parameters and temperature prediction functions, thereby improving calculation efficiency.

[0124] Optionally, based on the above embodiment, in some embodiments of the present invention, an implementation manner of S11 may be:

[0125] S111: According to the multi-spectral reflectivity of the ground object materials corresponding to the multiple categories of ground object materials and the set of temperature parameters to be predicted at multiple times, the predicted temperature values ​​of the multiple categories of ground object materials at multiple times are obtained through the temperature prediction function.

[0126] S112: Calculate the radiation brightness values ​​of the multiple categories of ground object materials at multiple moments according to the predicted temperature values ​​of the multiple categories of ground object materials at multiple moments, the infrared radiation parameter set, and the preset radiation brightness value calculation function.

[0127] Specifically, obtain the multi-spectral reflectivity of the ground object material corresponding to multiple categories of ground object materials, and the set of temperature parameters to be predicted at multiple times, such as the atmospheric temperature, atmospheric humidity, and wind speed of multiple categories of ground object materials at multiple times, substitute the multi-spectral reflectivity of the ground object material, the atmospheric temperature, atmospheric humidity, and wind speed of multiple categories of ground object materials at multiple times into the temperature prediction function, and obtain the predicted temperature values ​​of multiple categories of ground object materials at multiple times. After obtaining the predicted temperature values ​​of multiple categories of ground object materials at multiple times, substitute the predicted temperature values ​​of multiple categories of ground object materials at multiple times and the infrared radiation parameter set into the preset radiation brightness value calculation function, and calculate the radiation brightness values ​​of multiple categories of ground object materials at multiple times.

[0128] Optionally, based on the above embodiment, in some embodiments of the present invention, an implementation manner of S112 may be:

[0129] According to the predicted temperature values ​​of multiple types of ground materials at multiple times, multiple blackbody radiation brightnesses are calculated using Planck's formula.

[0130] According to multiple blackbody radiation brightness, infrared radiation parameter sets and a preset radiation brightness value calculation function, the radiation brightness values ​​of multiple categories of ground material at multiple times are calculated.

[0131] Specifically, the predicted temperature values ​​of multiple categories of ground material at multiple times are substituted into the Planck formula to calculate the blackbody radiation brightness of each category of ground material at multiple times; multiple blackbody radiation brightness and infrared radiation parameter sets are substituted into a preset radiation brightness value calculation function to calculate the radiation brightness values ​​of multiple categories of ground material at multiple times.

[0132] It should be understood that although Figure 1 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 1 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0133] In one embodiment, Figure 3 As shown, an infrared remote sensing image generation device is provided, including: a function construction module 10, a radiation brightness value acquisition module 11, an initial infrared remote sensing image acquisition module 12, a training set acquisition module 13 and a training module 14.

[0134] Among them, the function construction module 10 is used to construct multiple temperature prediction functions according to the temperature prediction parameter set of multiple categories of ground materials contained in the target shooting area at a preset time, the surface temperature value and the multi-spectral reflectivity of the ground material, wherein the temperature prediction parameter set at least includes: the atmospheric temperature, atmospheric humidity, and wind speed of multiple categories of ground materials at a preset time, and the surface temperature value and the multi-spectral reflectivity of the ground material are obtained based on the original visible light remote sensing image taken at the preset time of the target shooting area, and the corresponding original infrared remote sensing image.

[0135] The radiation brightness value acquisition module 11 is used to calculate the radiation brightness values ​​of multiple categories of ground material at multiple times according to the multi-spectral reflectivity of the ground material corresponding to multiple categories of ground material, the temperature prediction function, the set of temperature parameters to be predicted at multiple times, the infrared radiation parameter set and the preset radiation brightness value calculation function, wherein the infrared radiation parameter set includes at least: solar radiation brightness value, sky radiation brightness value, atmospheric transmittance, and atmospheric path radiation brightness value.

[0136] The initial infrared remote sensing image acquisition module 12 is used to acquire initial infrared remote sensing images at multiple moments through sensor imaging technology according to multiple radiation brightness values.

[0137] The training set acquisition module 13 is used to construct a data training set of an infrared remote sensing image generation model based on the original visible light remote sensing image and the initial infrared remote sensing images at multiple moments.

[0138] The training module 14 is used to train the infrared remote sensing image generation model using the data training set to obtain a trained target infrared remote sensing image generation model, wherein the target infrared remote sensing image generation model is used to generate the target infrared remote sensing image based on the visible light remote sensing image, and the infrared remote sensing image generation model includes a generator and a discriminator. The generator includes at least: a downsampling layer, an intermediate layer, an upsampling layer, and a multi-head attention layer. The downsampling layer, the intermediate layer, and the upsampling layer include a hole residual block, and multi-head attention feature extraction is performed between the downsampling layer and the upsampling layer through a jump connection.

[0139] In the above embodiment, a plurality of temperature prediction functions are constructed by a function construction module according to a set of temperature prediction parameters, surface temperature values ​​and multi-spectral reflectivity of a plurality of categories of ground material contained in a target shooting area at a preset time. A radiation brightness value acquisition module calculates the radiation brightness values ​​of a plurality of categories of ground material at a plurality of times according to the multi-spectral reflectivity of the ground material corresponding to the plurality of categories of ground material, the temperature prediction function, the set of temperature parameters to be predicted at a plurality of times, the infrared radiation parameter set and the preset radiation brightness value calculation function, so that the initial infrared remote sensing images at a plurality of times are acquired by the initial infrared remote sensing image acquisition module, which can reduce the parameters for acquiring the initial infrared remote sensing images and improve the calculation efficiency. Furthermore, a data training set for training the infrared remote sensing image generation model is constructed by using the original visible light remote sensing images and the initial infrared remote sensing images at a plurality of times through the training set acquisition module, and the data training set is enhanced, so as to improve the image quality of the target infrared remote sensing image generated by the trained target infrared remote sensing image generation model.

[0140] For the specific definition of the infrared remote sensing image generation device, please refer to the definition of the infrared remote sensing image generation method above, which will not be repeated here. Each module in the above-mentioned server can be implemented in whole or in part by software, hardware and a combination thereof. Each of the above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0141] An embodiment of the present invention provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the infrared remote sensing image generation method provided in the embodiment of the present invention can be implemented. For example, when the processor executes the computer program, the infrared remote sensing image generation method provided in the embodiment of the present invention can be implemented. Figure 1The technical solution of any of the method embodiments shown has similar implementation principles and technical effects, which will not be repeated here.

[0142] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided by the present invention can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static random access memory (SRAM) and dynamic random access memory (DRAM).

[0143] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0144] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A method for generating an infrared remote sensing image, characterized in that: include: Constructing multiple temperature prediction functions according to a set of temperature prediction parameters, surface temperature values, and multi-spectral reflectivity of the surface material of multiple categories of surface material contained in the target shooting area at a preset time, wherein the set of temperature prediction parameters at least includes: the atmospheric temperature, atmospheric humidity, and wind speed of multiple categories of surface material at a preset time, and the surface temperature values ​​and multi-spectral reflectivity of the surface material are obtained according to an original visible light remote sensing image taken at the preset time of the target shooting area, and a corresponding original infrared remote sensing image; According to the multi-spectral reflectivity of the ground object material corresponding to the multiple categories of ground object materials, the temperature prediction function, the temperature parameter set to be predicted at multiple times, the infrared radiation parameter set and the preset radiation brightness value calculation function, the radiation brightness values ​​of the multiple categories of ground object materials at multiple times are calculated, wherein the infrared radiation parameter set at least includes: solar radiation brightness value, sky radiation brightness value, atmospheric transmittance, and atmospheric path radiation brightness value; According to multiple radiation brightness values, initial infrared remote sensing images at multiple moments are obtained through sensor imaging technology; Constructing a data training set of an infrared remote sensing image generation model based on the original visible light remote sensing image and initial infrared remote sensing images at multiple moments; An infrared remote sensing image generation model is trained using a data training set to obtain a trained target infrared remote sensing image generation model, wherein the target infrared remote sensing image generation model is used to generate a target infrared remote sensing image based on a visible light remote sensing image. The infrared remote sensing image generation model includes a generator and a discriminator. The generator includes at least: a downsampling layer, an intermediate layer, an upsampling layer, and a multi-head attention layer. The downsampling layer, the intermediate layer, and the upsampling layer include a hole residual block. Multi-head attention feature extraction is performed between the downsampling layer and the upsampling layer through a jump connection.

2. The method according to claim 1, characterized in that The surface temperature value and the multi-spectral reflectivity of the ground object material are obtained based on the original visible light remote sensing image taken at a preset time of the target shooting area and the corresponding original infrared remote sensing image, including: Acquire the original visible light remote sensing image of the target shooting area captured at a preset time, and the original infrared remote sensing image corresponding to the original visible light remote sensing image; Determine multiple categories of ground object materials contained in the original visible light remote sensing image, and mark different categories of ground object materials; According to the original visible light remote sensing image and the original infrared remote sensing image, the surface temperature values ​​of multiple categories of ground material at a preset time and the initial ground material multi-spectral reflectance curve are obtained, wherein the initial ground material multi-spectral reflectance curve is the first infrared band; The multi-spectral reflectivity of the ground object material is determined according to the multi-spectral reflectance curve of the initial ground object material and the preset multi-spectral reflectance curves of multiple preset unknown categories of ground object materials in the preset software spectral data, wherein the multi-spectral reflectance curve of the ground object material is a second infrared band, and the second infrared band is larger than the first infrared band.

3. The method according to claim 2, characterized in that The determining of the multi-spectral reflectivity of the ground object material according to the initial multi-spectral reflectance curve of the ground object material and the preset multi-spectral reflectance curves of a plurality of preset unknown ground object materials in the preset software spectral data comprises: Matching the initial ground object material multi-spectral reflectance curve with a plurality of preset multi-spectral reflectance curves of a plurality of preset unknown ground object materials, determining that the preset unknown ground object material corresponding to the preset multi-spectral reflectance curve with the highest matching degree is the ground object material, and the preset multi-spectral reflectance curve with the highest matching degree is the target multi-spectral reflectance curve, wherein the preset multi-spectral reflectance curve is a third infrared band, and the third infrared band is greater than the second infrared band; The multi-spectral reflectivity of the ground feature material is determined according to the target multi-spectral reflectivity curve.

4. The method according to claim 1, characterized in that: The method constructs multiple temperature prediction functions according to a set of temperature prediction parameters of multiple types of ground material contained in the target shooting area at a preset time, a surface temperature value, and a multi-spectral reflectivity of the ground material, including: By using the least squares method, the temperature prediction parameter set, the surface temperature value and the multi-spectral reflectivity of the ground material of each category of ground material at a preset time are fitted to obtain multiple temperature prediction factors corresponding to each category of ground material; A plurality of temperature prediction functions are constructed according to the temperature prediction parameter sets corresponding to the plurality of categories of ground material, the multi-spectral reflectivity of the ground material and the plurality of temperature prediction factors.

5. The method according to claim 1, characterized in that The method of calculating the radiation brightness values ​​of the plurality of types of ground material at a plurality of times according to the multi-spectral reflectivity of the ground material corresponding to the plurality of types of ground material, the temperature prediction function, the set of temperature parameters to be predicted at a plurality of times, the set of infrared radiation parameters and the preset radiation brightness value calculation function comprises: According to the multi-spectral reflectivity of the ground object material corresponding to the multiple categories of ground object materials and the set of temperature parameters to be predicted at multiple moments, the predicted temperature values ​​of the multiple categories of ground object materials at multiple moments are obtained through the temperature prediction function; According to the predicted temperature values ​​of multiple categories of ground object materials at multiple moments, the infrared radiation parameter set and the preset radiation brightness value calculation function, the radiation brightness values ​​of multiple categories of ground object materials at multiple moments are calculated.

6. The method according to claim 5, characterized in that The method of calculating the radiation brightness values ​​of the multiple categories of ground object materials at multiple times according to the predicted temperature values ​​of the multiple categories of ground object materials at multiple times, the infrared radiation parameter set and the preset radiation brightness value calculation function includes: According to the predicted temperature values ​​of multiple types of ground materials at multiple times, multiple blackbody radiation brightnesses are calculated using Planck's formula; According to a plurality of blackbody radiation brightnesses, the infrared radiation parameter set and the preset radiation brightness value calculation function, the radiation brightness values ​​of a plurality of categories of ground object materials at a plurality of moments are calculated.

7. The method according to claim 1, characterized in that The step of constructing a data training set of an infrared remote sensing image generation model based on the original visible light remote sensing image and the initial infrared remote sensing images at multiple moments includes: Determine the original visible light remote sensing image and the initial infrared remote sensing image corresponding to each moment as a data training subset; A data training set for an infrared remote sensing image generation model is constructed based on the data training subsets corresponding to a plurality of time moments.

8. The method according to claim 1, characterized in that The downsampling layer and the upsampling layer are connected via a jump connection to perform multi-head attention feature extraction, including: Determine the target upsampling sublayer corresponding to each downsampling sublayer; The downsampled feature maps output by each downsampling sublayer are input to the target upsampling sublayer through the multi-head attention layer.

9. An infrared remote sensing image generating device, characterized in that: include: A function construction module, for constructing multiple temperature prediction functions according to a set of temperature prediction parameters of multiple types of ground material contained in a target shooting area at a preset time, a surface temperature value, and a multi-spectral reflectivity of the ground material, wherein the set of temperature prediction parameters at least includes: the atmospheric temperature, atmospheric humidity, and wind speed of multiple types of ground material at a preset time, and the surface temperature value and the multi-spectral reflectivity of the ground material are obtained according to an original visible light remote sensing image taken at the preset time of the target shooting area, and a corresponding original infrared remote sensing image; A radiation brightness value acquisition module is used to calculate the radiation brightness values ​​of multiple categories of ground material at multiple times according to the multi-spectral reflectivity of the ground material corresponding to multiple categories of ground material, the temperature prediction function, the temperature parameter set to be predicted at multiple times, the infrared radiation parameter set and the preset radiation brightness value calculation function, wherein the infrared radiation parameter set includes at least: solar radiation brightness value, sky radiation brightness value, atmospheric transmittance, and atmospheric path radiation brightness value; An initial infrared remote sensing image acquisition module is used to acquire initial infrared remote sensing images at multiple times through sensor imaging technology according to multiple radiation brightness values; A training set acquisition module, used to construct a data training set of an infrared remote sensing image generation model based on the original visible light remote sensing image and initial infrared remote sensing images at multiple moments; A training module is used to train an infrared remote sensing image generation model using a data training set to obtain a trained target infrared remote sensing image generation model, wherein the target infrared remote sensing image generation model is used to generate a target infrared remote sensing image based on a visible light remote sensing image, and the infrared remote sensing image generation model includes a generator and a discriminator, and the generator includes at least: a downsampling layer, an intermediate layer, an upsampling layer, and a multi-head attention layer, the downsampling layer, the intermediate layer, and the upsampling layer include a hole residual block, and multi-head attention feature extraction is performed between the downsampling layer and the upsampling layer through a jump connection.

10. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 8 are implemented.

Images