Geochemical and remote sensing image fusion method, system and device and storage medium

Abstract

The invention relates to the technical field of mathematical geology and geological big data, and discloses a geochemistry and remote sensing image fusion method, system and device and a storage medium, and the method comprises the steps: constructing a WVTNet model comprising a wavelet transformation module and a feature weight distribution module, and carrying out the training; wherein the wavelet transformation module is used for performing multi-scale image decomposition on geochemical data and a multispectral remote sensing image of a sample region to obtain a first image local feature and a second image local feature of different scales; the feature weight distribution module is used for acquiring key features in the plurality of first image local features and the plurality of second image local features based on a visual attention mechanism and fusing the key features to obtain a sample fusion image; and inputting the target geochemical data and the target multispectral remote sensing image of the target area into the trained WVTNet model to obtain a target fusion image. According to the method, the reliability and the information density of the fused data are improved, so that subsequent data analysis work is facilitated.

Description

Background Art

[0002] As a typical data-intensive discipline, geology includes remote sensing image data, geochemical anomaly data, geophysical anomaly data, drilling data, basic geological data and other data types with different structures and sources. These multi-source data have their own advantages in practical geological applications: remote sensing image data has the advantages of wide coverage, fast update and acquisition speed, and high spatial resolution; geochemical anomaly data has the advantage of higher sensitivity to shallow surface features and can more accurately reflect the mineral content information on the surface. In summary, different types of geological data have their own characteristics. In the face of practical problem applications, the judgment results based on a single data source are difficult to achieve the application goals. Therefore, in the face of massive multi-source geological big data, it is necessary to unify the multi-source data format, integrate the advantages of multi-source data, and fuse multiple geological data to deal with the complex problems in practical applications.

[0003] At present, in the field of multi-source geological data fusion, traditional wavelet transform methods focus on the fusion of different sensor data from a single data source. For example, high-spectral remote sensing data and high-spatial resolution remote sensing data are fused based on wavelet transform to obtain high-spectral and high-spatial resolution image data. On the other hand, although traditional wavelet transform can extract image features at multiple scales, the subsequent processing of features is still a simple superposition, resulting in the situation where the weights of some key features are consistent with the weights of non-key features, which is not conducive to subsequent data analysis. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a method, system, electronic device and storage medium for fusing geochemical and remote sensing images.

[0005] In a first aspect, the present invention provides a method for fusing geochemical and remote sensing images, and the technical solution of the method is as follows:

[0006] A WVTNet model including a wavelet transform module and a feature weight assignment module is constructed, and the WVTNet model is trained; wherein the wavelet transform module is used to perform multi-scale image decomposition on the target geochemical data and the target multispectral remote sensing image of the sample area, respectively, to obtain the first image local features of different scales corresponding to the target geochemical data and the second image local features of different scales corresponding to the target multispectral remote sensing image; the feature weight assignment module is used to obtain the first key feature among the multiple first image local features and the second key feature among the multiple second image local features based on the visual attention mechanism, and fuse the first key feature with the second key feature by using the convolution layer upsampling method to obtain the sample fused image;

[0007] The target geochemical data and target multispectral remote sensing images of the target area are input into the trained WVTNet model to obtain the target fused image.

[0008] The beneficial effects of the geochemical and remote sensing image fusion method of the present invention are as follows:

[0009] The method of the present invention realizes the fusion of geochemical data and remote sensing image data by constructing the WVTNet model, improves the reliability and information density of the fused data, and is helpful for subsequent data analysis.

[0010] Based on the above solution, the method for fusing geochemical and remote sensing images of the present invention can be further improved as follows.

[0011] In an optional manner, the method further includes:

[0012] Acquiring original geochemical data and original multispectral remote sensing images of the target area;

[0013] The original geochemical data of the target area are successively subjected to outlier elimination and Kriging difference processing to obtain the target geochemical data of the target area, and the original multispectral remote sensing image of the target area is successively subjected to radiation calibration, atmospheric correction and geometric correction to obtain the target multispectral remote sensing image of the target area.

[0014] In an optional manner, the steps of performing multi-scale image decomposition on the target geochemical data and the target multispectral remote sensing image of the sample area to obtain first image local features of different scales corresponding to the target geochemical data and second image local features of different scales corresponding to the target multispectral remote sensing image include:

[0015] Performing two-dimensional discrete wavelet transform on the target geochemical data and the target multispectral remote sensing image of the sample area respectively, to obtain frequency domain mapping results of the target geochemical data of the sample area and frequency domain mapping results of the target multispectral remote sensing image;

[0016] Based on the DB-4 wavelet function, multi-scale image decomposition is performed on the frequency domain mapping results of the target geochemical data and the frequency domain mapping results of the target multispectral remote sensing image, respectively, to obtain the first image local features of different scales corresponding to the target geochemical data and the second image local features of different scales corresponding to the target multispectral remote sensing image.

[0017] In an optional manner, based on a visual attention mechanism, the step of obtaining a first key feature from among a plurality of first image local features and a second key feature from among a plurality of second image local features includes:

[0018] Based on the visual attention mechanism, weights are assigned to multiple first image local features to obtain the first key features, and weights are assigned to multiple second image features to obtain the second key features.

[0019] In an optional manner, the step of fusing the first key feature with the second key feature by using a convolutional layer upsampling method to obtain a sample fused image includes:

[0020] The first key feature and the second key feature are combined to obtain a fused key feature, and all output results of the fused key feature in the frequency domain are subjected to DB-4 transformation to obtain an image domain key feature and upsampled to obtain the sample fused image.

[0021] In a second aspect, the present invention provides a fusion system of geochemistry and remote sensing images, and the technical solution of the system is as follows:

[0022] Includes: building units and fusion units;

[0023] The construction unit is used to: construct a WVTNet model including a wavelet transform module and a feature weight assignment module, and train the WVTNet model; wherein the wavelet transform module is used to: perform multi-scale image decomposition on the target geochemical data and the target multispectral remote sensing image of the sample area, respectively, to obtain the first image local features of different scales corresponding to the target geochemical data and the second image local features of different scales corresponding to the target multispectral remote sensing image; the feature weight assignment module is used to: based on the visual attention mechanism, obtain the first key feature among the multiple first image local features and the second key feature among the multiple second image local features, and fuse the first key feature with the second key feature by using the convolution layer upsampling method to obtain the sample fused image;

[0024] The fusion unit is used to input the target geochemical data and the target multispectral remote sensing image of the target area into the trained WVTNet model to obtain the target fused image.

[0025] The beneficial effects of the fusion system of geochemistry and remote sensing images of the present invention are as follows:

[0026] The system of the present invention realizes the fusion of geochemical data and remote sensing image data by constructing the WVTNet model, improves the reliability and information density of the fused data, and is helpful for subsequent data analysis.

[0027] Based on the above solution, the geochemical and remote sensing image fusion system of the present invention can also be improved as follows.

[0028] In an optional manner, the method further includes: a preprocessing unit; the preprocessing unit is used to:

[0029] Acquiring original geochemical data and original multispectral remote sensing images of the target area;

[0030] The original geochemical data of the target area are successively subjected to outlier elimination and Kriging difference processing to obtain the target geochemical data of the target area, and the original multispectral remote sensing image of the target area is successively subjected to radiation calibration, atmospheric correction and geometric correction to obtain the target multispectral remote sensing image of the target area.

[0031] In an optional manner, the wavelet transform module is specifically used for:

[0032] Performing two-dimensional discrete wavelet transform on the target geochemical data and the target multispectral remote sensing image of the sample area respectively, to obtain frequency domain mapping results of the target geochemical data of the sample area and frequency domain mapping results of the target multispectral remote sensing image;

[0033] Based on the DB-4 wavelet function, multi-scale image decomposition is performed on the frequency domain mapping results of the target geochemical data and the frequency domain mapping results of the target multispectral remote sensing image, respectively, to obtain the first image local features of different scales corresponding to the target geochemical data and the second image local features of different scales corresponding to the target multispectral remote sensing image.

[0034] In a third aspect, a technical solution of an electronic device of the present invention is as follows:

[0035] The invention comprises a memory, a processor and a program stored in the memory and running on the processor. When the processor executes the program, the steps of the method for fusing geochemical and remote sensing images of the invention are realized.

[0036] In a fourth aspect, the present invention provides a computer-readable storage medium having the following technical solution:

[0037] Instructions are stored in the computer-readable storage medium. When the computer-readable storage medium reads the instructions, the computer-readable storage medium executes the steps of the method for fusing geochemical and remote sensing images of the present invention.

[0038] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented according to the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The accompanying drawings are only used to illustrate the embodiments and are not to be considered as limiting the present invention. In addition, the same reference symbols are used to represent the same components throughout the accompanying drawings. In the accompanying drawings:

[0040] Figure 1 A schematic diagram of a flow chart of an embodiment of a method for fusing geochemical and remote sensing images of the present invention;

[0041] Figure 2 It is a schematic diagram of the principle of two-dimensional discrete wavelet transform;

[0042] Figure 3 This is a schematic diagram of the structure of the WVTNet model;

[0043] Figure 4 This is a schematic diagram of the principle of the feature weight allocation module;

[0044] Figure 5 It is a structural schematic diagram of an embodiment of a geochemical and remote sensing image fusion system of the present invention;

[0045] Figure 6 The figure is a schematic structural diagram of an embodiment of an electronic device of the present invention. DETAILED DESCRIPTION

[0046] Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0047] Figure 1 A flow chart of an embodiment of a method for fusing geochemical and remote sensing images provided by the present invention is shown. The method for fusing geochemical and remote sensing images can be executed by an electronic device such as a terminal device or a server. The terminal device can be any fixed or mobile terminal such as a user equipment (UE), a mobile device, a user terminal, a terminal, a cellular phone, a cordless phone, a personal digital assistant (PDA), a handheld device, a computing device, a vehicle-mounted device, a wearable device, etc. The server can be a single server or a server cluster composed of multiple servers. Any electronic device can implement the method for fusing geochemical and remote sensing images by calling computer-readable instructions stored in a memory through a processor. Figure 1 As shown, the following steps are included:

[0048] S1. Construct a WVTNet model including a wavelet transform module and a feature weight allocation module, and train the WVTNet model.

[0049] Among them, the WVTNet model is a Wavelet Transform Vision Transformer Network model, which includes a wavelet transform module and a feature weight assignment module. The wavelet transform module is used to perform multi-scale image decomposition on the target geochemical data and the target multispectral remote sensing image of the sample area, respectively, to obtain the first image local features of different scales corresponding to the target geochemical data and the second image local features of different scales corresponding to the target multispectral remote sensing image; the feature weight assignment module is used to obtain the first key feature among multiple first image local features and the second key feature among multiple second image local features based on the visual attention mechanism, and fuse the first key feature with the second key feature by using the convolution layer upsampling method to obtain the sample fusion image.

[0050] It should be noted that the sample area is the area used for model training. Geochemical data can reveal the geological composition and characteristics of different regions of the earth, including the element content, distribution and composition of rocks, soils, water bodies, etc., which helps to understand the internal structure and evolution of the earth. On the other hand, geochemical data can be used to monitor pollutants in the environment, such as heavy metals, organic matter, etc., and evaluate environmental quality and ecosystem health. According to the 1:20w geochemical anomaly map standard, the SiO2, Al2O3, CaO, MgO, Fe2O3, Na2O and K2O content data of the target area are uniformly sampled at equal intervals. The multispectral remote sensing image is ASTER multispectral remote sensing data, including 3 visible near-infrared bands (0.52-0.86μm) with a spatial resolution of 15m, 6 short-wave infrared bands (1.60-2.43μm) with a spatial resolution of 30m, and 5 thermal infrared bands (8.13-11.65μm) with a spatial resolution of 90m. Each ASTER scene gets a 60km strip, subdivided into 60×60km 2 scene.

[0051] S2. Input the target geochemical data and target multispectral remote sensing image of the target area into the trained WVTNet model to obtain the target fused image.

[0052] The target area is the area where the fusion of geochemical data and multispectral remote sensing images is actually required. The target geochemical data is the preprocessed geochemical data, and the target multispectral remote sensing image is the preprocessed multispectral remote sensing image. The target fusion image is the image obtained by fusion of the target geochemical data and the target multispectral remote sensing image.

[0053] In an optional manner, the method further includes:

[0054] Obtain original geochemical data and original multispectral remote sensing images of the target area.

[0055] Among them, the original geochemical data refers to the unprocessed geochemical data, and the original multispectral remote sensing images refers to the unprocessed multispectral remote sensing images.

[0056] The original geochemical data of the target area are processed with outliers and Kriging difference to obtain the target geochemical data of the target area. The original multispectral remote sensing image of the target area is calibrated with radiation, corrected with atmosphere and corrected with geometry to obtain the target multispectral remote sensing image of the target area.

[0057] Among them, the original geochemical data is subjected to outlier removal to obtain usable geochemical sampling point data. The geochemical sampling point data is subjected to Kriging interpolation processing to obtain a geochemical anomaly map, and the size is unified to 384*384 as the target geochemical data.

[0058] Among them, the original multispectral remote sensing image is cropped to a size of 384*384, and radiation calibration is performed to obtain a radiation calibration result; the radiation calibration result is atmospherically corrected to obtain an atmospheric correction result; the atmospheric correction result is geometrically corrected to obtain a geometric correction result, that is, the target multispectral remote sensing image.

[0059] In an optional manner, the steps of performing multi-scale image decomposition on the target geochemical data and the target multispectral remote sensing image of the sample area to obtain first image local features of different scales corresponding to the target geochemical data and second image local features of different scales corresponding to the target multispectral remote sensing image include:

[0060] The target geochemical data and the target multispectral remote sensing image of the sample area are respectively subjected to two-dimensional discrete wavelet transform to obtain the frequency domain mapping results of the target geochemical data and the frequency domain mapping results of the target multispectral remote sensing image of the sample area.

[0061] Based on the DB-4 wavelet function, multi-scale image decomposition is performed on the frequency domain mapping results of the target geochemical data and the frequency domain mapping results of the target multispectral remote sensing image, respectively, to obtain the first image local features of different scales corresponding to the target geochemical data and the second image local features of different scales corresponding to the target multispectral remote sensing image.

[0062] Specifically, Figure 2As shown in the figure, two-dimensional discrete wavelet transform is performed on the target geochemical data and the target multispectral remote sensing image respectively, and the two images (target geochemical data and target multispectral remote sensing image) are converted from the spatial domain to the frequency domain to obtain the frequency domain mapping results of the two images. Based on the DB-4 (Daubechies-4) wavelet function, the frequency domain mapping results of the target geochemical data and the frequency domain mapping results of the target multispectral remote sensing image are respectively subjected to multi-scale image decomposition, and the expression is: Among them, a k is the kth wavelet coefficient, k refers to the relative position index at different scales, k=0,1,2...N-1; ψ(x) is the wavelet function, x refers to the original spatial position of the pixel, N refers to the number of decomposition layers, and N is 4 in this embodiment. The high-frequency and low-frequency feature information of the two images at different scales are obtained respectively, wherein each image includes 1 low-frequency component (representing the main structure and large-scale feature information of the image, including LL4) and 12 high-frequency components (representing the detail texture, edge or small-scale feature information in the image, including LH1, HL1, HH1, LH2, HL2, HH2, LH3, HL3, HH3, LH4, HL4, HH4).

[0063] It should be noted that the multi-scale image decomposition process of the target geochemical data is the same as that of the target multispectral remote sensing image. Both are input as image data in the form of rasters, so the processing methods are the same. The above formulas and steps are explained using any image.

[0064] In an optional manner, based on a visual attention mechanism, the step of obtaining a first key feature from among a plurality of first image local features and a second key feature from among a plurality of second image local features includes:

[0065] Based on the visual attention mechanism, multiple first image local features are weighted to obtain a first key feature, and multiple second image features are weighted to obtain a second key feature. Specifically:

[0066] ① Quantify multiple local features of the first image and multiple local features of the second image respectively, and obtain Dimensional token sequence;

[0067] ②An additional set of category embedding vectors is defined to record the weight of each feature finally included. The vector result is the weighted average of the features of each part. The weight is learned by the back propagation algorithm. At the same time, the transformer position encoding vector is defined, using sine and cosine position encoding. For each position pos and dimension i:

[0068] ③If Figure 3As shown in the figure, in the vector correlation measurement, the token sequence, position encoding vector and category embedding vector are input into the transformer encoder layer, and the weighted feature vector is calculated by the encoder. Among them, Q is the Query vector, whose input is the features extracted from the geochemical anomaly map and its corresponding position vector, K is the Key vector, whose input is the feature vector extracted from the remote sensing image and its corresponding position encoding vector, V is the Value vector, which corresponds to the category embedding vector, and d k is the dimension of the Key vector, and the correlation between vectors is calculated by multiplying Q and K. It should be noted that the token sequence is a combination of feature vectors extracted from geochemical anomaly maps and remote sensing images.

[0069] ④ Input the feature vector T1 output by the transformer encoder layer into the FFN layer (Feed Forward Neural Network), which defines two fully connected layers. The activation function of the first layer is LeakyReLU, and the activation function of the second layer is ReLU activation function.

[0070] ⑤ Such as Figure 3 As shown, in the back propagation algorithm, the constructed network is trained according to the back propagation algorithm, using the cross entropy loss function, which is defined as: Among them, y i represents the real image component, It represents the decomposed image components, minimizes the difference between the two, and uses the NesterovAdam (Nadam) optimizer to make the loss function tend to the global minimum. The initial learning rate is set to 0.0001, the decay rate is (0.99, 0.99), and the weight matrix of the output feature vector is obtained.

[0071] It should be noted that before training, the vector groups corresponding to the two images (the vector combination of the token sequence, the position encoding vector and the category embedding vector) are encapsulated into a module (transformer encoder module), and its weight matrix is ​​trained to assign the fusion weights of the feature vectors of the two images.

[0072] In an optional manner, the step of fusing the first key feature with the second key feature by upsampling the convolution layer to obtain a sample fused image includes:

[0073] The first key feature is combined with the second key feature to obtain a fused key feature, and all output results of the fused key feature in the frequency domain are subjected to DB-4 transformation to obtain an image domain key feature and upsampled to obtain a sample fused image.

[0074] like Figure 4 As shown, specifically: ① Combine the first key feature (key low-frequency feature) and the second key feature (key high-frequency feature) to obtain the fused key feature; ② Perform DB-4 transformation on all output results of the fused key feature in the frequency domain, transform the results from the frequency domain to the image domain, and obtain the image domain key feature. ③ Upsample the image domain key features and restore them to the same dimension as the original image to obtain a sample fused image.

[0075] The technical solution of this embodiment realizes the fusion of geochemical data and remote sensing image data by constructing the WVTNet model. The fused image reduces the dimension of features and increases the information density of the fused data. While retaining the accuracy of the geochemical data, it also improves the spatial resolution of the geochemical data in combination with remote sensing data, which is helpful for subsequent data analysis.

[0076] Figure 5 FIG. 2 is a schematic diagram showing a structural embodiment of a geochemical and remote sensing image fusion system 200 provided by the present invention. Figure 5 As shown, the system 200 includes: a construction unit 210 and a fusion unit 220;

[0077] The construction unit 210 is used to: construct a WVTNet model including a wavelet transform module and a feature weight assignment module, and train the WVTNet model; wherein the wavelet transform module is used to: perform multi-scale image decomposition on the target geochemical data and the target multispectral remote sensing image of the sample area, respectively, to obtain first image local features of different scales corresponding to the target geochemical data and second image local features of different scales corresponding to the target multispectral remote sensing image; the feature weight assignment module is used to: obtain a first key feature from a plurality of first image local features and a second key feature from a plurality of second image local features based on a visual attention mechanism, and fuse the first key feature with the second key feature by using a convolutional layer upsampling method to obtain a sample fused image;

[0078] The fusion unit 220 is used to input the target geochemical data and the target multispectral remote sensing image of the target area into the trained WVTNet model to obtain the target fused image.

[0079] In an optional manner, the method further includes: a preprocessing unit; the preprocessing unit is used to:

[0080] Obtain original geochemical data and original multispectral remote sensing images of the target area;

[0081] The original geochemical data of the target area are processed with outliers and Kriging difference to obtain the target geochemical data of the target area. The original multispectral remote sensing image of the target area is calibrated with radiation, corrected with atmosphere and corrected with geometry to obtain the target multispectral remote sensing image of the target area.

[0082] In an optional manner, the wavelet transform module is specifically used for:

[0083] Performing two-dimensional discrete wavelet transform on the target geochemical data and the target multispectral remote sensing image of the sample area respectively, and obtaining the frequency domain mapping results of the target geochemical data and the frequency domain mapping results of the target multispectral remote sensing image of the sample area;

[0084] Based on the DB-4 wavelet function, multi-scale image decomposition is performed on the frequency domain mapping results of the target geochemical data and the frequency domain mapping results of the target multispectral remote sensing image, respectively, to obtain the first image local features of different scales corresponding to the target geochemical data and the second image local features of different scales corresponding to the target multispectral remote sensing image.

[0085] In an optional manner, the feature weight assignment module is used to:

[0086] Based on the visual attention mechanism, weights are assigned to multiple first image local features to obtain first key features, and weights are assigned to multiple second image features to obtain second key features.

[0087] In an optional manner, the feature weight assignment module is used to:

[0088] The first key feature is combined with the second key feature to obtain a fused key feature, and all output results of the fused key feature in the frequency domain are subjected to DB-4 transformation to obtain an image domain key feature and upsampled to obtain a sample fused image.

[0089] The technical solution of this embodiment realizes the fusion of geochemical data and remote sensing image data by constructing the WVTNet model, improves the reliability and information density of the fused data, and facilitates subsequent data analysis.

[0090] The above parameters and steps for each module to implement corresponding functions in the geochemical and remote sensing image fusion system 200 of this embodiment can refer to the parameters and steps in the embodiment of the geochemical and remote sensing image fusion method above, and will not be repeated here.

[0091] like Figure 6As shown, an electronic device 300 of an embodiment of the present invention includes a processor 320, the processor 320 is coupled to a memory 310, and the memory 310 stores at least one computer program 330, and the at least one computer program 330 is loaded and executed by the processor 320, so that the electronic device 300 implements any of the above-mentioned geochemical and remote sensing image fusion methods, specifically:

[0092] The electronic device 300 may have relatively large differences due to different configurations or performances, and may include one or more processors 320 (Central Processing Units, CPU) and one or more memories 310, wherein the one or more memories 310 store at least one computer program 330, and the at least one computer program 330 is loaded and executed by the one or more processors 320, so that the electronic device 300 implements any of the geochemical and remote sensing image fusion methods provided in the above embodiments. Of course, the electronic device 300 may also have components such as a wired or wireless network interface, a keyboard, and an input and output interface for input and output, and the electronic device 300 may also include other components for implementing device functions, which will not be described in detail here.

[0093] A computer-readable storage medium according to an embodiment of the present invention stores at least one computer program, and the at least one computer program is loaded and executed by a processor so that a computer implements any of the above-mentioned methods for fusing geochemistry and remote sensing images.

[0094] Optionally, the computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a compact disc (CD-ROM), a magnetic tape, a floppy disk, an optical data storage device, and the like.

[0095] In an exemplary embodiment, a computer program product or a computer program is also provided, the computer program product or the computer program comprising computer instructions, the computer instructions being stored in a computer-readable storage medium. A processor of an electronic device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the electronic device performs any of the above-mentioned methods for fusing geochemical and remote sensing images.

[0096] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application are used to distinguish similar objects and represent the definition of a specific order or sequence. The order of use of similar objects can be interchanged where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than the order shown or described.

[0097] Those skilled in the art know that the present invention can be implemented as a system, method or computer program product. Therefore, the present disclosure can be specifically implemented in the following forms, namely: it can be complete hardware, it can be complete software (including firmware, resident software, microcode, etc.), or it can be a combination of hardware and software, which is generally referred to as "circuit", "module" or "system" herein. In addition, in some embodiments, the present invention can also be implemented in the form of a computer program product in one or more computer-readable media, and the computer-readable medium contains computer-readable program code.

[0098] Any combination of one or more computer-readable media can be used. Computer-readable media can be computer-readable signal media or computer-readable storage media. Computer-readable storage media can be, for example, but not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or devices, or any combination of the above. More specific examples of computer-readable storage media (a non-exhaustive list) include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by an instruction execution system, device or device or used in combination with it.

[0099] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A method for fusing geochemical and remote sensing images, characterized in that: include: A WVTNet model including a wavelet transform module and a feature weight assignment module is constructed, and the WVTNet model is trained; wherein the wavelet transform module is used to perform multi-scale image decomposition on the target geochemical data and the target multispectral remote sensing image of the sample area, respectively, to obtain the first image local features of different scales corresponding to the target geochemical data and the second image local features of different scales corresponding to the target multispectral remote sensing image; the feature weight assignment module is used to obtain the first key feature among the multiple first image local features and the second key feature among the multiple second image local features based on the visual attention mechanism, and fuse the first key feature with the second key feature by using the convolution layer upsampling method to obtain the sample fused image; The target geochemical data and target multispectral remote sensing images of the target area are input into the trained WVTNet model to obtain the target fused image.

2. The method for fusing geochemical and remote sensing images according to claim 1, characterized in that: Also includes: Acquiring original geochemical data and original multispectral remote sensing images of the target area; The original geochemical data of the target area are successively subjected to outlier elimination and Kriging difference processing to obtain the target geochemical data of the target area, and the original multispectral remote sensing image of the target area is successively subjected to radiation calibration, atmospheric correction and geometric correction to obtain the target multispectral remote sensing image of the target area.

3. The method for fusing geochemical and remote sensing images according to claim 1, characterized in that: The steps of performing multi-scale image decomposition on the target geochemical data and the target multispectral remote sensing image of the sample area to obtain first image local features of different scales corresponding to the target geochemical data and second image local features of different scales corresponding to the target multispectral remote sensing image include: Performing two-dimensional discrete wavelet transform on the target geochemical data and the target multispectral remote sensing image of the sample area respectively, to obtain frequency domain mapping results of the target geochemical data of the sample area and frequency domain mapping results of the target multispectral remote sensing image; Based on the DB-4 wavelet function, multi-scale image decomposition is performed on the frequency domain mapping results of the target geochemical data and the frequency domain mapping results of the target multispectral remote sensing image, respectively, to obtain the first image local features of different scales corresponding to the target geochemical data and the second image local features of different scales corresponding to the target multispectral remote sensing image.

4. The method for fusing geochemical and remote sensing images according to claim 1, characterized in that: The step of obtaining a first key feature from a plurality of first image local features and a second key feature from a plurality of second image local features based on a visual attention mechanism includes: Based on the visual attention mechanism, weights are assigned to multiple first image local features to obtain the first key features, and weights are assigned to multiple second image features to obtain the second key features.

5. The method for fusing geochemical and remote sensing images according to claim 4, characterized in that: The step of fusing the first key feature with the second key feature by using a convolutional layer upsampling method to obtain a sample fused image includes: The first key feature and the second key feature are combined to obtain a fused key feature, and all output results of the fused key feature in the frequency domain are subjected to DB-4 transformation to obtain an image domain key feature and upsampled to obtain the sample fused image.

6. A geochemical and remote sensing image fusion system, characterized in that: include: Building blocks and fusion blocks; The construction unit is used to: construct a WVTNet model including a wavelet transform module and a feature weight assignment module, and train the WVTNet model; wherein the wavelet transform module is used to: perform multi-scale image decomposition on the target geochemical data and the target multispectral remote sensing image of the sample area, respectively, to obtain the first image local features of different scales corresponding to the target geochemical data and the second image local features of different scales corresponding to the target multispectral remote sensing image; the feature weight assignment module is used to: based on the visual attention mechanism, obtain the first key feature among the multiple first image local features and the second key feature among the multiple second image local features, and fuse the first key feature with the second key feature by using the convolution layer upsampling method to obtain the sample fused image; The fusion unit is used to input the target geochemical data and the target multispectral remote sensing image of the target area into the trained WVTNet model to obtain the target fused image.

7. The fusion system of geochemistry and remote sensing images according to claim 6, characterized in that: Also includes: Pre-processing unit; The pre-processing unit is used for: Acquiring original geochemical data and original multispectral remote sensing images of the target area; The original geochemical data of the target area are successively subjected to outlier elimination and Kriging difference processing to obtain the target geochemical data of the target area, and the original multispectral remote sensing image of the target area is successively subjected to radiation calibration, atmospheric correction and geometric correction to obtain the target multispectral remote sensing image of the target area.

8. The geochemical and remote sensing image fusion system according to claim 6, characterized in that: The wavelet transform module is specifically used for: Performing two-dimensional discrete wavelet transform on the target geochemical data and the target multispectral remote sensing image of the sample area respectively, to obtain frequency domain mapping results of the target geochemical data of the sample area and frequency domain mapping results of the target multispectral remote sensing image; Based on the DB-4 wavelet function, multi-scale image decomposition is performed on the frequency domain mapping results of the target geochemical data and the frequency domain mapping results of the target multispectral remote sensing image, respectively, to obtain the first image local features of different scales corresponding to the target geochemical data and the second image local features of different scales corresponding to the target multispectral remote sensing image.

9. An electronic device, characterized in that: The electronic device includes a processor, the processor is coupled to a memory, at least one computer program is stored in the memory, and the at least one computer program is loaded and executed by the processor so that the electronic device implements the method for fusing geochemical and remote sensing images as described in any one of claims 1 to 5.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores at least one computer program, and the at least one computer program is loaded and executed by the processor so that the computer-readable storage medium implements the method for fusing geochemical and remote sensing images as described in any one of claims 1 to 5.

Images