An adaptive remote sensing geological lithology information enhancement method and terminal

Abstract

The present invention discloses an adaptive remote sensing geological lithology information enhancement method and terminal, which pre-processes a multi-spectral remote sensing image to obtain an ortho-fused multi-spectral remote sensing image, removes interference factor pixels, obtains a bedrock remote sensing pixel spectrum set, performs cluster analysis on the pixel spectrum set, establishes a spectrum subset of each spectral geological unit, establishes a linear discriminant analysis model based on the spectrum subset of each spectral geological unit, calculates a spectral conversion matrix, uses the spectral conversion matrix to perform dimensionality reduction processing on the multi-band remote sensing image, and then performs color synthesis to obtain an enhanced remote sensing image. The present application introduces spectral clustering and linear discriminant analysis technology, and proposes an adaptive remote sensing geological lithology information enhancement method based on machine statistical learning and not dependent on prior knowledge, obtains the optimal spectral projection conversion scheme for remote sensing geological lithology information enhancement within the scope of linear changes, and achieves the purpose of enhancing remote sensing geological lithology information as a whole.

Description

Technical Field

[0001] The invention relates to the technical field of remote sensing geological survey, and in particular to an adaptive remote sensing geological lithology information enhancement method and terminal. Background Art

[0002] Lithology mapping is an important part of geological survey. Using remote sensing technology to interpret and identify lithology information is one of the most widely used geological survey techniques and plays an important role in geological survey. At present, spectral feature enhancement methods are used to enhance remote sensing geological lithology information, including band combination method, ratio method, principal component transformation method, HLS transformation method and chromatic aberration enhancement method. These remote sensing lithology information enhancement methods have their own advantages and disadvantages. Some rely on expert experience, some only enhance local lithology information but cannot effectively utilize the full-band spectral information, and some are not targeted and the transformation results are unpredictable. Therefore, designing a global optimization method to enhance remote sensing geological lithology information as a whole is a problem that needs to be solved urgently. Summary of the invention

[0003] The purpose of the present invention is to provide an adaptive remote sensing geological lithology information enhancement method and terminal, introduce spectral clustering and linear discriminant analysis technology, propose an adaptive remote sensing geological lithology information enhancement method based on machine statistical learning and independent of prior knowledge, obtain the optimal spectral projection conversion scheme for remote sensing geological lithology information enhancement within the scope of linear change, and achieve the purpose of enhancing remote sensing geological lithology information as a whole.

[0004] In the first aspect, the above-mentioned invention object of the present invention is achieved through the following technical solutions:

[0005] An adaptive remote sensing geological lithology information enhancement method is provided. The method preprocesses multispectral remote sensing images to obtain orthophoto fused multispectral remote sensing images, removes interference factor pixels, obtains bedrock remote sensing pixel spectrum sets, performs cluster analysis on the pixel spectrum sets, establishes spectrum subsets of each spectral geological unit, establishes a linear discriminant analysis model based on the spectrum subsets of each spectral geological unit, calculates the spectrum conversion matrix, uses the spectrum conversion matrix to perform dimensionality reduction processing on multi-band remote sensing images, and then performs color synthesis to obtain enhanced remote sensing images.

[0006] The present invention is further configured to pre-process the multispectral remote sensing image, including radiation correction, atmospheric correction, orthorectification and fusion processing.

[0007] The present invention is further configured to: establish a bedrock outcrop area mask, eliminate interference factor pixels, use the high-end cutting of the blue light band of multispectral imaging to extract cloud and snow interference factor pixels, and use the low-end cutting of the short-wave infrared band / blue light band ratio to extract water body and shadow interference factor pixels; use the NDVI index method to extract vegetation interference factor pixels, and obtain the bedrock area remote sensing pixel spectrum set.

[0008] The present invention is further configured as follows: combining existing geological data, according to a spectral clustering algorithm, the spectral set of remote sensing pixels in the bedrock area is divided into K types of remote sensing spectral geological units, and spectral subsets of each type of remote sensing spectral geological unit are established respectively, wherein K is a positive integer greater than or equal to 4.

[0009] The present invention is further configured to calculate the spectrum subsets X of various remote sensing spectrum geological units i The covariance matrix S i , the overall covariance matrix S of all remote sensing spectral geological units, solve the matrix (∑S i ) -1 (S-∑S i ), and take the eigenvectors corresponding to the first three eigenvalues ​​to construct the spectral transformation matrix W.

[0010] The present invention is further configured as follows: the spectral subset of the i-th remote sensing spectral geological unit Centralize each spectral subset to obtain the centralized spectral set A i , calculate each sub-covariance matrix S i =A i T ×A i ,in n represents the number of bands, m represents the number of bands, i Indicates the number of remote sensing pixels corresponding to the i-th remote sensing spectral geological unit; μ i ∈R n represents the mean spectral vector of the spectral subset;

[0011] make represents the total spectrum set, where μ∈R n represents the mean spectral vector of the overall spectral set;

[0012] The overall covariance matrix S = A T A, Where T represents the matrix transpose operation.

[0013] The present invention is further configured to construct a linear discriminant analysis model for spectral subsets of various remote sensing spectral geological units based on minimizing spectral similarity within remote sensing spectral geological units and maximizing spectral differentiation between remote sensing spectral geological units.

[0014] The present invention is further configured as follows: the spectral similarity within the remote sensing spectral geological unit is defined as The spectral heterogeneity between each remote sensing spectral geological unit is defined as tr(W T (S-∑S i )W), the optimization of lithology information enhancement problem in the scope of linear transformation is to solve In the formula, tr(*) represents the matrix trace calculation, W represents the spectrum conversion matrix, S i represents the covariance matrix of the spectral subset of the i-th remote sensing spectral geological unit, and S represents the overall covariance matrix.

[0015] The present invention is further configured as follows: by solving (S-∑S i )W=λ∑S i The generalized eigenvalue corresponding to W is realized The solution of the square matrix (∑S i ) -1 (S-∑S i ) to perform eigenvalue decomposition and sort the obtained eigenvalues ​​into the order λ1>λ2>λ3…>λ K-1 , take the eigenvectors corresponding to the first three eigenvalues ​​to construct the spectral conversion matrix W = (ω1, ω2, ω3), where K is the number of remote sensing spectral geological unit categories.

[0016] In the first aspect, the above-mentioned invention object of the present invention is achieved through the following technical solutions:

[0017] An adaptive remote sensing geological lithology information enhancement terminal comprises 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 method described in the present application is implemented.

[0018] Compared with the prior art, the beneficial technical effects of this application are:

[0019] 1. This application obtains remote sensing spectral geological units by preprocessing multispectral remote sensing images and removing interference factors, establishes a linear discriminant analysis model of the spectral set of remote sensing spectral geological units, and realizes geological and lithological information enhancement based on multispectral remote sensing images;

[0020] 2. Furthermore, the present application solves the minimum spectral differentiation within the same lithological unit and the maximum spectral difference between different lithological units within the scope of linear transformation, thereby achieving global balanced optimization between the tonal purity within various lithological units and the contrast between units, thereby enhancing the geological lithological information. DETAILED DESCRIPTION

[0021] The present invention is described in further detail below. Specific embodiment 1

[0023] The present invention discloses an adaptive remote sensing geological lithology information enhancement method, which pre-processes a multispectral remote sensing image to obtain an orthophoto fusion multispectral remote sensing image with n bands, establishes a bedrock outcrop area mask, removes interference factors, obtains a remote sensing pixel image, divides the remote sensing pixel image, obtains K types of remote sensing spectral geological units, and establishes a spectral set X of various types of remote sensing spectral geological units. i , based on each spectrum set X i , establish a linear discriminant analysis model, solve the problem of minimizing the spectral differentiation within the unit and maximizing the spectral difference between the units, calculate the spectral conversion matrix, use the spectral conversion matrix to reduce the dimension of the multi-band remote sensing image, and then perform color synthesis to obtain enhanced remote sensing images.

[0024] The adaptive remote sensing geological lithology information enhancement method comprises the following steps:

[0025] S1. Acquire a multispectral remote sensing image, perform preprocessing, and obtain an orthorectified fused multispectral remote sensing image with n bands, wherein the preprocessing includes radiation correction, atmospheric correction, orthorectified correction, fusion, etc.

[0026] S2. Establish a bedrock outcrop area mask, remove pixels of geological and lithological information interference factors such as water bodies, vegetation, snow, fog, and shadows, and obtain a remote sensing pixel image of the bedrock area;

[0027] Specifically, the high-end cutting of the blue light band of multispectral images is used to extract the pixels of cloud and snow interference factors, and the low-end cutting of the short-wave infrared band / blue light band ratio is used to extract the pixels of water body and shadow interference factors; the NDVI index method is used to extract the pixels of vegetation interference factors.

[0028] S3. Based on the existing geological data and the spectral clustering algorithm, the remote sensing pixel images of the bedrock area are divided into K types of remote sensing spectral geological units; wherein K is a positive integer greater than or equal to 4;

[0029] S4. Calculate the spectral subset X of the i-th remote sensing spectral geological unit i , Among them, m i It represents the number of remote sensing pixels corresponding to the i-th remote sensing spectral geological unit, and n represents the number of bands;

[0030] For each spectral subset X i Perform centralized processing to obtain the centralized processing spectrum set A i ,

[0031]

[0032] in, μ i ∈R n (2), represents the mean spectral vector of the spectral subset;

[0033] Calculate the sub-covariance matrix S i :S i =A i T ×A i (3);

[0034] Where T represents the matrix transpose operation.

[0035] Let the total spectrum set be X,

[0036] in, μ∈R n (5), represents the mean spectral vector of the overall spectral set;

[0037] The overall covariance matrix S = A T ·A, (6);

[0038]

[0039] Let W∈R n×3 is the spectral conversion matrix, then the spectral similarity within each spectral geological unit after conversion can be expressed as:

[0040]

[0041] The spectral differentiation between each remote sensing spectral geological unit is characterized by:

[0042] tr(W T (S-∑S i )W) (9);

[0044] Among them, tr(*) represents the matrix trace calculation.

[0045] The core of remote sensing geological lithology information enhancement is to minimize the spectral differentiation within the same lithology unit and maximize the spectral difference between different lithology units. Therefore, within the scope of linear transformation, the lithology information enhancement problem can be optimized to solve:

[0046]

[0047] Where W represents the spectral conversion matrix.

[0048] By using the Lagrange multiplier method and calculating the generalized eigenvalues ​​corresponding to equation (10), the optimal solution of equation (9) can be achieved.

[0049] (S-∑S i )W=λ∑S i W (10);

[0050] Specifically, solve the eigenvalue λ of the square matrix (11):

[0051] (∑S i ) -1 (S-∑S i ) (11);

[0052] Sort the obtained eigenvalues ​​λ from large to small, λ1>λ2>λ3…>λ K-1 , where K represents the number of remote sensing spectral geological unit categories.

[0053] Take the eigenvectors corresponding to the first three eigenvalues ​​to construct the spectral conversion matrix W:

[0054] W = (ω1, ω2, ω3) (22).

[0055] S5. According to the spectral conversion matrix W, the multi-band remote sensing image is subjected to dimensionality reduction processing, and the multi-band remote sensing image is reduced from the spectral dimension n to the dimension 3. The 3D remote sensing image is subjected to RGB color synthesis, and the synthesized remote sensing image is subjected to linear stretching and saturation enhancement processing to obtain a remote sensing image with balanced and optimized tonal purity within the spectral geological unit and differentiation between units. Specific embodiment 2

[0057] An embodiment of the present invention provides an adaptive remote sensing geological lithology information enhancement terminal device. The terminal device of this embodiment includes: a processor, a memory, and a computer program stored in the memory and executable on the processor, such as a computer program for constructing a transformation matrix. When the processor executes the computer program, the method described in Example 1 is implemented.

[0058] Exemplarily, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of completing specific functions, which are used to describe the execution process of the computer program in the adaptive remote sensing geological lithology information enhancement terminal device. For example, the computer program may be divided into multiple modules, and the specific functions of each module are as follows:

[0059] 1. Covariance matrix calculation module, used to calculate the covariance matrix;

[0060] 2. Spectral conversion matrix calculation module, used to calculate the spectral conversion matrix.

[0061] 3. Dimensionality reduction module, used to reduce dimension.

[0062] The adaptive remote sensing geological lithology information enhancement terminal device can be a computing device such as a desktop computer, a notebook, a PDA, and a cloud server. The terminal device may include, but is not limited to, a processor and a memory. Those skilled in the art can understand that the above examples are only examples of adaptive remote sensing geological lithology information enhancement terminal devices, and do not constitute a limitation on the adaptive remote sensing geological lithology information enhancement terminal device, and may include more or fewer components, or a combination of certain components, or different components. For example, the adaptive remote sensing geological lithology information enhancement terminal device may also include input and output devices, network access devices, buses, etc.

[0063] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the adaptive remote sensing geological lithology information enhancement terminal device, and uses various interfaces and lines to connect the various parts of the entire adaptive remote sensing geological lithology information enhancement terminal device.

[0064] The memory can be used to store the computer program and / or module, and the processor realizes various functions of the adaptive remote sensing geological lithology information enhancement terminal device by running or executing the computer program and / or module stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0065] The embodiments of this specific implementation method are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An adaptive remote sensing geological lithology information enhancement method, characterized by: Preprocess the multispectral remote sensing image to obtain the orthophoto fusion multispectral remote sensing image, remove the interference factor pixels, obtain the bedrock remote sensing pixel spectrum set, perform cluster analysis on the pixel spectrum set, establish the spectrum subset of each spectral geological unit, establish a linear discriminant analysis model based on the spectrum subset of each spectral geological unit, calculate the spectrum conversion matrix, use the spectrum conversion matrix to reduce the dimension of the multi-band remote sensing image, and then perform color synthesis to obtain the enhanced remote sensing image; Combined with the existing geological data, according to the spectral clustering algorithm, the spectrum set of remote sensing pixels in the bedrock area is divided into K types of remote sensing spectrum geological units, and the spectrum subsets of each type of remote sensing spectrum geological unit are established respectively, where K is a positive integer greater than or equal to 4; Calculate the spectral subset X of various remote sensing spectral geological units i The covariance matrix S i , the overall covariance matrix S of all remote sensing spectral geological units, solve the matrix (∑S i ) -1 (S-∑S i ), take the eigenvalues ​​corresponding to the first three eigenvalues ​​to construct the spectrum conversion matrix W; Spectral subset of the i-th remote sensing spectral geological unit Centralize each spectral subset to obtain the centralized spectral set A i , calculate each sub-covariance matrix S i =A i T ×A i ,in n represents the number of bands, m represents the number of bands, i Indicates the number of remote sensing pixels corresponding to the i-th remote sensing spectral geological unit; μ i ∈R n represents the mean spectral vector of the spectral subset; make represents the total spectrum set, where μ∈R n represents the mean spectral vector of the overall spectral set; The overall covariance matrix S = A T A, In the formula, T represents the matrix transpose operation; The spectral similarity within a geological unit of remote sensing spectra is defined as The spectral heterogeneity between each remote sensing spectral geological unit is defined as tr(W T (S-∑S i )W), the optimization of lithology information enhancement problem in the scope of linear transformation is to solve In the formula, tr(*) represents the matrix trace calculation, W represents the spectrum conversion matrix, S i represents the covariance matrix of the spectral subset of the i-th remote sensing spectral geological unit, S represents the overall covariance matrix; by solving (S-∑S i )W=λ∑S i The generalized eigenvalue corresponding to W is realized The solution of the square matrix (∑S i ) -1 (S-∑S i ) to perform eigenvalue decomposition and sort the obtained eigenvalues ​​λ1>λ2>λ3…>λ K-1 , take the eigenvectors corresponding to the first three eigenvalues ​​to construct the spectral conversion matrix W = (ω1, ω2, ω3), where K is the number of remote sensing spectral geological unit categories.

2. The method for enhancing the adaptive remote sensing geological lithology information according to claim 1, characterized in that: Preprocess multispectral remote sensing images, including radiation correction, atmospheric correction, orthorectification and fusion processing.

3. The method for enhancing the adaptive remote sensing geological lithology information according to claim 1, characterized in that: A mask for the bedrock outcrop area is established to remove interfering pixels. The high-end cutting of the blue light band of multispectral images is used to extract the interfering pixels of clouds and snow. The low-end cutting of the shortwave infrared band / blue light band ratio is used to extract the interfering pixels of water bodies and shadows. The NDVI index method is used to extract the interfering pixels of vegetation, and the spectral set of remote sensing pixels in the bedrock area is obtained.

4. The method for enhancing the adaptive remote sensing geological lithology information according to claim 1, characterized in that: Based on minimizing the spectral similarity within remote sensing spectral geological units and maximizing the spectral differentiation between remote sensing spectral geological units, a linear discriminant analysis model for spectral subsets of various remote sensing spectral geological units was constructed.

5. An adaptive remote sensing geological lithology information enhancement terminal, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 4 is implemented.

Images