A multispectral image band fusion ground surface deformation monitoring method, device and medium

Abstract

The application discloses a multispectral image band fusion ground surface deformation monitoring method, equipment and medium, relates to the technical field of remote sensing geological disaster monitoring, and the method comprises the following steps: carrying out deformation field extraction and error correction on the multi-temporal multispectral remote sensing image of a monitoring area, obtaining the corrected ground surface deformation field of each band, calculating the background noise standard deviation and slope consistency standard deviation of each band; based on the background noise standard deviation and slope consistency standard deviation of each band, determining an effective band set from the multiple bands; for each effective band, the entropy weight method is adopted, the corresponding background noise standard deviation and slope consistency standard deviation are used to calculate the comprehensive weight of each band, and then weighted fusion is performed in combination with the corrected ground surface deformation field of each band to obtain a fused ground surface deformation field. The application can realize optimal weighted fusion of multi-band deformation fields and improve monitoring accuracy.

Description

Technical Field

[0001] This application relates to the field of remote sensing geological disaster monitoring technology, and in particular to a method, equipment and medium for monitoring surface deformation by multispectral image band fusion. Background Technology

[0002] Surface deformation monitoring is a crucial tool for geological disaster prevention, resource and environmental management, and geodynamics research. Multispectral remote sensing imagery, due to its wide coverage, short revisit period, and low acquisition cost, has become an important data source for large-scale surface deformation monitoring. Multispectral images typically contain multiple bands, including visible and near-infrared. Different bands exhibit varying spectral response characteristics and sensitivities to changes in land cover, leading to significant differences in accuracy during image-matched deformation monitoring. Currently, the use of multispectral imagery for deformation monitoring generally suffers from the following shortcomings: First, there is a lack of quantitative evaluation methods for the deformation monitoring accuracy of each band, resulting in arbitrary and subjective band selection; second, existing fusion methods rely solely on a single accuracy index for weighting, failing to comprehensively consider observation noise and the consistency of physical motion direction; and third, when the number of bands is not fixed (two or more), there is a lack of a highly adaptable adaptive weighting calculation framework, hindering the effective utilization of the complementary advantages of multispectral imagery across multiple bands. Summary of the Invention

[0003] The purpose of this application is to provide a method, equipment and medium for monitoring surface deformation by multispectral image band fusion, which can achieve optimal weighted fusion of multi-band deformation fields and improve monitoring accuracy.

[0004] To achieve the above objectives, this application provides the following solution: Firstly, this application provides a method for monitoring surface deformation by multispectral image band fusion, including: Acquire multi-temporal, multispectral remote sensing images of the monitored area; Deformation field extraction and error correction are performed on the multi-temporal multispectral remote sensing images to obtain the corrected surface deformation field of each band. Based on the corrected surface deformation field of each band, the standard deviation of background noise and the standard deviation of slope aspect consistency of each band are calculated. Based on the standard deviation of background noise and the standard deviation of aspect consistency for each band, a set of effective bands is determined from multiple bands; For each effective band in the set of effective bands, the entropy weight method is used to calculate the comprehensive weight of each band based on the corresponding background noise standard deviation and slope aspect consistency standard deviation. Then, the weighted fusion is performed in combination with the corrected surface deformation field of each band to obtain the fused surface deformation field.

[0005] Secondly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement a multispectral image band fusion method for monitoring surface deformation.

[0006] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for monitoring surface deformation by multispectral image band fusion.

[0007] According to the specific embodiments provided in this application, the following technical effects are disclosed: This application constructs a dual-index accuracy evaluation system based on the standard deviation of background noise and the standard deviation of slope aspect consistency for each band. This system can quantitatively characterize the differences in deformation monitoring performance of each band from two independent dimensions: the observation noise level and the direction of physical motion. This overcomes the one-sidedness of single-index evaluation and provides an objective basis for band selection and adaptive weighting of multispectral images. Based on the above dual-index evaluation, effective bands are selected, and then the comprehensive weight of each band is calculated. Weighted fusion is then performed in combination with the corrected surface deformation fields of each band to obtain the fused surface deformation field. This weighted fusion of this application supports adaptive fusion of two or more arbitrary numbers of bands. The weight calculation process is independent of the number of bands and can automatically adapt to different sensors and different band combinations, exhibiting broad versatility and scalability. Furthermore, this application does not rely on specific sensors or a fixed number of bands, is applicable to various types of multispectral remote sensing images, and can be widely applied to various surface deformation monitoring scenarios such as landslides and glacier movement, providing a systematic and universal methodological framework for long-term time-series surface deformation monitoring of multispectral images.

[0008] In summary, this application solves the core problems of unclear differences in deformation monitoring accuracy across different bands in multispectral images, insufficient robustness of single index weighting, and difficulty in adapting the fusion framework to any number of bands, ultimately improving the accuracy of deformation monitoring. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is an application environment diagram of the multispectral image band fusion surface deformation monitoring method in one embodiment of this application.

[0011] Figure 2 This is a flowchart illustrating a method for monitoring surface deformation by multispectral image band fusion in one embodiment of this application.

[0012] Figure 3 This is a band-fusion result of temporal deformation monitoring of surface landslides in one embodiment of this application.

[0013] Figure 4 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0014] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0015] This application achieves surface deformation monitoring through multispectral image band fusion based on background noise and slope aspect differences. It is particularly applicable to the field of multispectral remote sensing image surface deformation monitoring, especially involving band accuracy evaluation and adaptive weighted fusion of multispectral images with two or more bands to achieve high-precision monitoring of various surface deformations (including landslides, glacier movement, etc.).

[0016] This application uses the standard deviation of background noise and the standard deviation of deformation vector aspect consistency as dual evaluation indicators to comprehensively weight each band and adaptively select effective bands for fusion, thereby achieving optimal weighted fusion of multi-band deformation fields and improving monitoring accuracy. This application is applicable to any multispectral remote sensing image with two or more bands and can be widely used in various surface deformation monitoring scenarios such as landslides and glaciers.

[0017] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0018] The multispectral image band fusion method for monitoring surface deformation provided in this application can be applied to, for example... Figure 1In the application environment shown, terminal 101 communicates with server 102 via a network. A data storage system can store the data that server 102 needs to process. The data storage system can be set up independently, integrated into server 102, or placed in the cloud or on another server. Terminal 101 can send multi-temporal multispectral remote sensing images of the monitored area to server 102. After receiving the images, server 102 extracts the deformation field and corrects errors to obtain the corrected surface deformation field for each band. It calculates the background noise standard deviation and aspect consistency standard deviation for each band, and based on this, determines the effective band set from multiple bands. For each effective band, it calculates the comprehensive weight of each band based on the corresponding background noise standard deviation and aspect consistency standard deviation, and then performs weighted fusion with the corrected surface deformation fields of each band to obtain the fused surface deformation field. Server 102 can feed back the obtained fused surface deformation field to terminal 101. Furthermore, in some embodiments, the multispectral image band fusion surface deformation monitoring method can also be implemented independently by server 102 or terminal 101.

[0019] The terminal 101 can be, but is not limited to, various desktop computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. The server 102 can be implemented using a standalone server or a server cluster consisting of multiple servers, or it can be a cloud server.

[0020] In one exemplary embodiment, such as Figure 2 As shown, a method for monitoring surface deformation by multispectral image band fusion is provided. This method is executed by computer equipment, specifically by a terminal or server alone, or by both a terminal and a server. In this embodiment, the method is applied to... Figure 1 Taking server 102 as an example, the explanation includes the following steps 201 to 205.

[0021] Step 201: Acquire multi-temporal multispectral remote sensing images of the monitoring area; specifically, acquire multi-temporal multispectral remote sensing images covering the monitoring area and having the same spatial resolution.

[0022] Step 202: Extract deformation fields and correct errors in the multi-temporal multispectral remote sensing images to obtain the corrected surface deformation fields for each band.

[0023] In a specific application, step 202 includes the following steps (21)-(23).

[0024] (21) Using the digital elevation model (DEM), orthorectification is performed on the multispectral remote sensing images of each time phase to obtain the corresponding registered multispectral remote sensing images; specifically, the main image is used as the spatial reference to complete the fine geometric registration of each time phase and each band to achieve strict pixel-level alignment.

[0025] (22) For the registered multispectral remote sensing images corresponding to each time, subpixel image matching is performed in each band to obtain the surface deformation field of the corresponding band; in particular, the surface deformation field of each band in two or more periods can be extracted.

[0026] (23) Model and correct the orbital error, strip / attitude error and topographic relief error of the surface deformation field of each band to obtain the corrected surface deformation field of each band. That is, model and correct the systematic error of the deformation field of each band, mainly including the orbital error, strip / attitude error and registration error caused by topographic relief, to obtain the error corrected surface deformation field of each band.

[0027] Step 203: Based on the corrected surface deformation field of each band, calculate the standard deviation of background noise and the standard deviation of slope aspect consistency for each band; specifically, for the corrected surface deformation field of N bands, construct the following dual-index accuracy evaluation system, and calculate the two accuracy indices for each band respectively.

[0028] The calculation process for the background noise standard deviation of each band includes: delineating a stable reference zone outside the monitoring area with no obvious surface movement; based on the zero deformation assumption, the matching offset within the stable reference zone is regarded as observation noise, and the standard deviation of the offset of all pixels of the surface deformation field of each band within the stable reference zone after correction is calculated as the background noise standard deviation; the smaller this value, the lower the matching noise and the higher the observation quality of the band under the current image conditions.

[0029] For gravity-driven surface deformation (such as landslides and glaciers), the deformation direction should be consistent with the topographic slope aspect. The calculation process for the standard deviation of slope aspect consistency for each band includes: within the deformation area, calculating the difference between the deformation vector direction extracted from the corrected surface deformation field of each band and the DEM slope aspect for each pixel, marking it as the difference value for each pixel, and then calculating the standard deviation as the standard deviation of slope aspect consistency. The smaller this value, the more closely the deformation direction extracted for that band matches the topographic driving direction, and the higher the physical rationality.

[0030] Step 204: Determine the effective band set from multiple bands based on the background noise standard deviation and aspect consistency standard deviation of each band.

[0031] In a specific application, based on the above two indicators, abnormal band screening is performed on N bands: when the standard deviation of background noise or the standard deviation of aspect consistency of a certain band exceeds the mean of the corresponding indicator of N bands plus 1.5 times the standard deviation between bands, i.e. μ+1.5σ, where μ and σ are the mean and standard deviation of the indicator of N bands, respectively, it is determined that the band is seriously affected by radiated noise or spectral interference, and it is removed and not included in the subsequent fusion.

[0032] The selection of the 1.5σ threshold is based on the following: Statistically, μ±1.5σ corresponds to approximately 86.6% confidence coverage, falling between μ±σ (68.3%, too strict, prone to incorrectly removing effective bands) and μ±2σ (95.4%, too lenient, difficult to exclude low-quality bands), achieving a reasonable balance between robustness and retention of effective bands. Experimentally, this application compared three thresholds: 1.0σ, 1.5σ, and 2.0σ, using the fusion deformation field closure error RMSE as the evaluation index. The results show that 1.5σ achieves the best fusion accuracy in all test scenarios; therefore, this value was selected as the threshold for removing abnormal bands. Finally, M (2<=M<=N) effective bands were determined to participate in the weighted fusion.

[0033] Step 205: For each effective band in the effective band set, calculate the comprehensive weight of each band based on the corresponding background noise standard deviation and slope aspect consistency standard deviation, and then perform weighted fusion with the corrected surface deformation field of each band to obtain the fused surface deformation field.

[0034] In a specific application, for the selected M (M>=2) effective bands, the comprehensive weight of each band is calculated and weighted fusion is performed, including the following steps (51)-(53).

[0035] (51) Single-index normalized weights: For any valid band, take the reciprocal of the corresponding background noise standard deviation and aspect consistency standard deviation, and normalize them within M valid bands to obtain two sets of single-index weights; the two sets of single-index weights are: .

[0036] .

[0037] in, The background noise weight for the i-th effective band is... Let M be the standard deviation of the background noise in the i-th effective band, and M be the number of effective bands. The slope consistency weight for the i-th effective band is... Let be the standard deviation of the slope aspect consistency of the i-th effective band.

[0038] (52) Comprehensive weight calculation: The entropy weight method is used to determine the weight coefficients of the background noise index and the slope aspect consistency index. , The weights are then linearly combined with the weights of two sets of single indicators to obtain the comprehensive weights for the bands.

[0039] The normalized weights obtained in the previous step (51) are used as the probability values ​​of each band under the two indicators, let subscript i Band numbering ( i =1,...,M), subscript j The index number is j=1, which corresponds to background noise, and j=2, which corresponds to slope aspect consistency.

[0040] Calculate the information entropy of the two indicators separately: , stipulates that hour, .

[0041] Calculate the information utility value of each indicator: .

[0042] The weighting coefficients of the two indicators are obtained using the entropy weighting method: .

[0043] in, + =1, , Both are greater than 0. Combining the two calculated weight coefficients with the two sets of single-index weights in a linear combination, the comprehensive weight of the i-th band is obtained. for: .

[0044] .

[0045] (53) Weighted fusion: The corrected surface deformation field of each band and the band comprehensive weight corresponding to the effective band are weighted and superimposed to obtain the fused surface deformation field.

[0046] The fused surface deformation field is : ;in, This represents the corrected surface deformation field of the i-th band.

[0047] The above framework has no limit on the number of bands (M>=2 is sufficient) and has good adaptability.

[0048] In a specific application example, the method further includes: reconstructing the temporal sequence of surface deformation in the monitoring area based on the fused surface deformation field, extracting the deformation rate and cumulative deformation, and also extracting spatial distribution characteristics.

[0049] Among them, for T acquisition times < < < ,by Using a reference epoch, the fused deformation field between epochs is accumulated pixel by pixel to obtain the result of pixel p at the epoch. Cumulative deformation of an epoch; the formula for calculating cumulative deformation is: ;in, For pixel p in the th Cumulative deformation of an epoch, , , The values ​​are the 0th epoch and the (0+1)th epoch, respectively; the cumulative deformation is linearly fitted to obtain the deformation rate v(p).

[0050] In addition, the internal consistency accuracy is evaluated by using a multi-temporal closed loop, and the external consistency is verified by combining high-resolution reference image interpretation or field measurement data, thus quantitatively evaluating the monitoring accuracy of the fusion results.

[0051] The specific verification method is as follows: In terms of internal consistency verification, several temporal images within the same time period as the integrated weighted band fusion step are selected to construct a multi-temporal closed loop (i.e., select three or more phases of images, take the first phase as a reference, calculate the deformation field between adjacent epochs in sequence, and finally evaluate the internal consistency accuracy of the deformation field of each epoch through the closure error). The root mean square error (RMSE) of the closure error is used as the internal consistency accuracy index and compared with the single-band results before fusion to verify the accuracy improvement effect of the fusion method.

[0052] For external compliance verification, high-resolution optical reference images or field GNSS measured displacement data are collected within the monitoring area. The displacement extracted from the fused deformation field is compared with the reference value point by point, and the RMSE and correlation coefficient are calculated to quantitatively evaluate the absolute accuracy of the fusion result.

[0053] In summary, this application includes four steps: image preprocessing and multi-band deformation field extraction and systematic error correction, dual-index band accuracy evaluation, comprehensive weighted band fusion, and monitoring and verification of temporal surface deformation. Taking the monitoring of surface deformation in a certain area using a multispectral remote sensing satellite image (containing four bands: blue, green, red, and near-infrared) as an example, the specific implementation process of this application is illustrated. In this embodiment, the bands are sequentially denoted as Band1 (blue), Band2 (green), Band3 (red), and Band4 (near-infrared), with an initial total number of bands participating in the evaluation, N=4.

[0054] (1) Corresponding to steps 201 and 202, acquire multi-temporal multispectral remote sensing images of the monitoring area, and use DEM to perform orthorectification on each temporal image. Perform subpixel image matching on N=4 bands respectively, extract the surface deformation field of each band, and then perform modeling and correction of orbital error, strip / attitude error and terrain undulation error in sequence to obtain the deformation fields D1, D2, D3 and D4 of each band after system error correction.

[0055] (2) Corresponding to steps 203 and 204, calculate the standard deviation of background noise and the standard deviation of aspect consistency for each band to determine the effective bands. Examples of the dual-index calculation results for the four bands are as follows: Band 1 (Blue): =0.29 m, =55.31 degrees.

[0056] Band 2 (Green): =0.26 m, =47.15 degrees.

[0057] Band3 (Red): =0.30 m, =59.24 degrees.

[0058] Band4 (Near Infrared): =0.69 m, =60.77 degrees.

[0059] The standard deviation of background noise in Band 4 (0.69 m) exceeds that of 4 bands σ. noise The mean (μ=0.385 m) plus 1.5 times the inter-band standard deviation (σ=0.177 m) equals a threshold of 0.385 + 1.5 × 0.177 = 0.651 m, which is considered an anomalous band. The aspect consistency standard deviation of Band 4 (60.77 degrees) does not exceed σ for 4 bands. aspect The mean (μ=55.62 degrees) plus 1.5 times the standard deviation between bands (σ=5.28 degrees) equals the threshold of 55.62+1.5×5.28≈63.54 degrees, which is considered a normal band.

[0060] Considering both indicators, with background noise exceeding the standard deviation threshold being the primary cause, Band4 was still identified as an abnormal band and removed. Ultimately, the number of effective bands was M=3, meaning Band1, Band2, and Band3 were selected as effective bands for subsequent fusion.

[0061] (3) Corresponding to step 205, the reciprocals of the standard deviation of background noise and the standard deviation of aspect consistency are taken and normalized, and a weighting coefficient is introduced to obtain the comprehensive weight of each band through linear combination. The deformation field is weighted and superimposed with the comprehensive weight of each band to obtain the optimal fused deformation field, which adaptively supports any number of bands. For M=3 effective bands (Band1, Band2, Band3), the following processing is performed: Calculate the normalized weights of background noise for each band. : Substituting the data: 1 / 0.29=3.448, 1 / 0.26=3.846, 1 / 0.30=3.333, total 10.627.

[0062] =3.448 / 10.627=0.324.

[0063] =3.846 / 10.627=0.362.

[0064] =3.333 / 10.627=0.314.

[0065] Calculate the aspect uniformity normalization weights for each band. Similarly, substitute the data: 1 / 55.31 = 0.01808, 1 / 47.15 = 0.02120, 1 / 59.24 = 0.01688, total 0.05616.

[0066] =0.01808 / 0.05616=0.322.

[0067] =0.02120 / 0.05616=0.378.

[0068] =0.01688 / 0.05616=0.300.

[0069] The entropy weight method is used to calculate the index weight coefficients. α and β Let the normalized weights obtained above be probability values, and let: = =0.324, = =0.362, = =0.314.

[0070] = =0.322, = =0.378, = =0.300.

[0071] Calculate the information entropy of the two indicators separately: =-(1 / ln3)×(0.324×ln0.324+0.362×ln0.362+0.314×ln0.314)=0.9983.

[0072] =-(1 / ln3)×(0.322×ln0.322+0.378×ln0.378+0.300×ln0.300)=0.9956.

[0073] Information utility value: =1- =0.0017, =1- =0.0044.

[0074] The weighting coefficients, obtained from the entropy weighting method, are: α= / ( + )=0.2844, β= / ( + =0.7156.

[0075] The slope consistency index has lower information entropy and is given higher weight, reflecting its stronger ability to distinguish the quality of fusion.

[0076] Calculate the overall weight of each band : =0.2844×0.324+0.7156×0.322=0.3226.

[0077] =0.2844×0.362+0.7156×0.378=0.3734.

[0078] =0.2844×0.314+0.7156×0.300=0.3040.

[0079] Weighted fusion: .

[0080] (4) Reconstruct the surface deformation time series based on the fused deformation field, extract the deformation rate and cumulative deformation; use multi-temporal closed loop to evaluate the internal consistency accuracy, and combine reference images or field measured data to verify the external consistency.

[0081] In this embodiment, internal consistency accuracy is evaluated using a multi-temporal closed loop, and external consistency verification is conducted in conjunction with reference data. Verification shows that the standard deviation of background noise and the standard deviation of slope aspect consistency in the fused deformation field are both superior to the results of any single band, indicating that the dual-index comprehensive weighted fusion method effectively improves the accuracy and robustness of surface deformation monitoring. The specific verification process is as follows: In terms of internal consistency verification, a closed loop was constructed using three effective bands (Band1, Band2, and Band3) at T=4 time phases (t1–t4). The closure error RMSE of each band and the fusion result was calculated. The closure error RMSE of the fusion result was 0.11 m, which is better than the monitoring results of single bands (Band1 (0.21 m), Band2 (0.19 m), and Band3 (0.22 m), indicating that the fusion method effectively reduces random errors.

[0082] In terms of external compliance verification, measured displacement data from 12 GNSS reference stations in the monitoring area were collected during the same period. The extracted values ​​of the corresponding points of the fused deformation field were compared with the measured values ​​of GNSS. The correlation coefficient R=0.97 and RMSE=0.13 m were both better than any single band, further verifying the effectiveness of the dual-index comprehensive weighted fusion method.

[0083] like Figure 3 As shown, this is the result of band-fusion surface landslide temporal deformation monitoring obtained using the above method.

[0084] Based on the same inventive concept, this application also provides a system. The solution provided by this system is similar to the solution described in the above method. Therefore, the specific limitations of one or more system embodiments provided below can be found in the limitations of the method above, and will not be repeated here.

[0085] In one exemplary embodiment, a multispectral image band fusion surface deformation monitoring system is provided, comprising the following modules.

[0086] The image acquisition module is used to acquire multi-temporal, multispectral remote sensing images of the monitored area.

[0087] The deformation field extraction and correction module is used to extract the deformation field and correct the error of the multi-temporal multispectral remote sensing images to obtain the corrected surface deformation field of each band.

[0088] The dual-index calculation module is used to calculate the standard deviation of background noise and the standard deviation of slope aspect consistency for each band based on the corrected surface deformation field of each band.

[0089] The band selection module is used to determine the effective band set from multiple bands based on the standard deviation of background noise and the standard deviation of aspect consistency for each band.

[0090] The weighted fusion module is used to calculate the comprehensive weight of each effective band in the effective band set according to the corresponding background noise standard deviation and slope aspect consistency standard deviation, and then perform weighted fusion with the corrected surface deformation field of each band to obtain the fused surface deformation field.

[0091] Compared with the prior art, this application has the following advantages: (1) A dual-index accuracy evaluation system combining background noise standard deviation and slope aspect consistency standard deviation was constructed. This system can quantitatively characterize the differences in deformation monitoring performance of each band from two independent dimensions: observation noise level and physical motion direction. It overcomes the one-sidedness of single-index evaluation and provides an objective basis for band selection and adaptive weighting of multispectral images.

[0092] (2) The proposed integrated weighted fusion framework naturally supports adaptive fusion of two or more arbitrary number of bands. By normalizing the inverse of the standard deviation of the two items respectively, the entropy weight method is used to adaptively determine the index weight coefficients and linearly combine them, taking into account the differentiated information of each index. The weight calculation process is independent of the number of bands and can automatically adapt to different sensors and different band combinations, with wide versatility and scalability.

[0093] (3) It does not rely on specific sensors or a fixed number of bands, and is applicable to various types of multispectral remote sensing images. It can be widely used in various surface deformation monitoring scenarios such as landslides and glacier movement, providing a systematic and universal methodological framework for long-term surface deformation monitoring of multispectral images.

[0094] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 4As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and databases. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media to run. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network. When the computer program is executed by the processor, it implements a multispectral image band fusion method for monitoring surface deformation.

[0095] Those skilled in the art will understand that Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0096] In one exemplary embodiment, a computer device is also provided, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps in the above-described method embodiments.

[0097] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0098] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0099] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0100] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application 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, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0101] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0102] The technical features of the above embodiments can be combined in any way. For the sake of brevity, 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.

[0103] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for monitoring surface deformation by multispectral image band fusion, characterized in that, The method includes: Acquire multi-temporal, multispectral remote sensing images of the monitored area; Deformation field extraction and error correction are performed on the multi-temporal multispectral remote sensing images to obtain the corrected surface deformation field of each band. Based on the corrected surface deformation field of each band, the standard deviation of background noise and the standard deviation of slope aspect consistency of each band are calculated. Based on the standard deviation of background noise and the standard deviation of aspect consistency for each band, a set of effective bands is determined from multiple bands; For each effective band in the set of effective bands, the entropy weight method is used to calculate the comprehensive weight of each band based on the corresponding background noise standard deviation and slope aspect consistency standard deviation. Then, the weighted fusion is performed in combination with the corrected surface deformation field of each band to obtain the fused surface deformation field.

2. The multispectral image band fusion method for monitoring surface deformation according to claim 1, characterized in that, Deformation field extraction and error correction are performed on the multi-temporal multispectral remote sensing images to obtain the corrected surface deformation fields for each band, including: Using a digital elevation model, orthorectification is performed on multispectral remote sensing images of different time phases to obtain corresponding registered multispectral remote sensing images. For the registered multispectral remote sensing images corresponding to each time, subpixel image matching is performed in each band to obtain the surface deformation field of the corresponding band. Modeling and correction of orbital error, strip / attitude error and topographic relief error are performed on the surface deformation field of each band to obtain the corrected surface deformation field of each band.

3. The multispectral image band fusion method for monitoring surface deformation according to claim 1, characterized in that, The calculation process for the standard deviation of background noise in each band includes: Outside the monitoring area, a stable reference zone is defined; the standard deviation of the offset of all pixels in the stable reference zone after correction of the surface deformation field of each band is calculated and used as the standard deviation of the background noise. The calculation process for the standard deviation of aspect consistency for each band includes: Within the deformation area, the difference between the direction of the deformation vector extracted from the corrected surface deformation field of each band and the slope aspect of the DEM is calculated pixel by pixel and marked as the difference value of each pixel. Then the standard deviation is calculated as the slope aspect consistency standard deviation.

4. The method for monitoring surface deformation by multispectral image band fusion according to claim 1, characterized in that, For each effective band in the effective band set, the entropy weight method is used to calculate the comprehensive weight of each band based on the corresponding background noise standard deviation and slope aspect consistency standard deviation. Then, a weighted fusion is performed on each band's surface deformation field after correction to obtain the fused surface deformation field, including: For any valid band, take the reciprocal of the corresponding background noise standard deviation and the slope consistency standard deviation and normalize them to obtain two sets of single index weights; The entropy weight method is used to determine the weight coefficients of the background noise index and the slope consistency index, and then linearly combine them with the weights of the two sets of single indices to obtain the comprehensive weight of the band. The corrected surface deformation fields of each band are weighted and superimposed with the band comprehensive weights corresponding to the effective bands to obtain the fused surface deformation fields.

5. The multispectral image band fusion method for monitoring surface deformation according to claim 4, characterized in that, The weights of the two sets of single indicators are: ； ； in, The background noise weight for the i-th effective band is... Let M be the standard deviation of the background noise in the i-th effective band, and M be the number of effective bands. The slope consistency weight for the i-th effective band is... Let be the standard deviation of the slope aspect consistency of the i-th effective band; Band composite weight for: ;in, , All are trade-off coefficients.

6. The multispectral image band fusion method for monitoring surface deformation according to claim 5, characterized in that, The fused surface deformation field is : ;in, This represents the corrected surface deformation field of the i-th band.

7. The method for monitoring surface deformation by multispectral image band fusion according to claim 1, characterized in that, The method further includes: Based on the fused surface deformation field, the surface deformation time series of the monitoring area is reconstructed, and the deformation rate and cumulative deformation are extracted.

8. The method for monitoring surface deformation by multispectral image band fusion according to claim 7, characterized in that, The formula for calculating the cumulative deformation is as follows: ;in, For pixel p in the th Cumulative deformation of an epoch, To integrate the surface deformation field, , They are the 0th epoch and the (0+1)th epoch, respectively. n =2,3,...,T, where T is the total number of time phases; The cumulative deformation is linearly fitted to obtain the deformation rate.

9. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the multispectral image band fusion method for monitoring surface deformation according to any one of claims 1-8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the multispectral image band fusion method for monitoring surface deformation as described in any one of claims 1-8.

Images