Multi-source data processing method and system applied to territorial space planning

Abstract

The invention relates to the technical field of image processing, in particular to a multi-source data processing method and system applied to territorial space planning. For any scale, according to the morphological difference of the remote sensing image at different moments in the corresponding connected domain of each same area, the change stable area is screened out; forming a row vector of a morphological change feature matrix according to the morphological features of the connected domain corresponding to each change stable region at different moments; obtaining an augmented prediction matrix of the morphological change feature matrix, and obtaining the remote sensing attention of each change stability area according to the position distribution of pixel points in the connected domain corresponding to the change stability areas at different moments and the change stability corresponding to the augmented prediction matrix; obtaining a final fusion voting value of each variation stable region under each scale; and performing data fusion on the remote sensing image. According to the method, the accurate final fusion voting value of the remote sensing image on each scale is obtained, so that the remote sensing data fusion effect is improved.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a multi-source data processing method and system applied to national land space planning. Background Art

[0002] In national land space planning, multi-source data processing technology is used to better integrate, analyze and apply data from different sources, so as to provide a scientific basis for planning decisions. Specifically, multi-source refers to remote sensing data from different sensor platforms and different types of sensors. Different sensors can provide data with different fields of view, spatial resolution, temporal resolution and observation angles. Since the data obtained by different sensors in the same area may be redundant or contradictory, data fusion can fully utilize the advantages of different sensors, eliminate these inconsistencies, and obtain more accurate and reliable information. At the same time, it tries to retain and enhance important spatial, spectral, temporal, geometric and other key information to ensure that the final data has good performance in many aspects.

[0003] Multi-source remote sensing data need to be synthesized to accurately analyze the overall characteristics of the region. In the existing technology, mean coverage can be used to fuse multi-source remote sensing data. However, since remote sensing images show different characteristics at different scales as the time dimension changes, ignoring the multi-scale characteristics of the time dimension will result in direct mean fusion being unable to fully capture these changes, and the effect of remote sensing image data fusion is poor. Summary of the invention

[0004] In order to solve the technical problem that the multi-scale characteristics of the time dimension are ignored and the remote sensing image data fusion effect is poor, the purpose of the present invention is to provide a multi-source data processing method and system for national land space planning. The technical scheme adopted is as follows: The present invention proposes a multi-source data processing method for national land space planning, the method comprising: Obtain remote sensing images of different preset scales at each moment in the area to be integrated during national land space planning; According to the grayscale distribution of pixels in the remote sensing image, multiple connected domains of each remote sensing image are obtained; for any scale, according to the morphological differences of the connected domains corresponding to each same area of ​​the remote sensing image at different times, the change effect of the connected domain corresponding to each area is obtained; according to the change effect of the connected domains corresponding to all areas, the change area is screened out; according to the morphological differences of the connected domains corresponding to each change area between adjacent times, the change stability of the connected domain corresponding to each change area is obtained; according to the change stability of the connected domains corresponding to all change areas, the change stable area is screened out; The row vectors of the morphological change feature matrix are formed by the morphological features of the connected domain corresponding to each stable change region at different times; the augmented prediction matrix of the morphological change feature matrix is ​​obtained, and the remote sensing attention of each stable change region is obtained according to the position distribution of the pixel points in the connected domain corresponding to the stable change region at different times and the change stability corresponding to the augmented prediction matrix; According to the morphological characteristics of the connected domain corresponding to each stable region at different times at each scale and the remote sensing attention, the final fusion voting value of each stable region at each scale is obtained; based on the final fusion voting value of each stable region at different scales, data fusion is performed on the remote sensing images used in national land space planning.

[0005] Furthermore, the method for obtaining the change effect includes: Obtain the number of pixels in each connected domain as a morphological feature; For any scale, the differences in morphological features of the connected domains corresponding to the same area of ​​the remote sensing images between different adjacent moments are obtained, and the differences in morphological features between all adjacent moments are accumulated and normalized as the change effect of the connected domain corresponding to each area.

[0006] Furthermore, the method for obtaining the change region includes: If the change effect of the connected domain corresponding to the region is included in the preset change range, the corresponding region is used as the change region.

[0007] Furthermore, the method for obtaining the change stability includes: Obtain the difference in morphological characteristics of the corresponding connected domain of each changing region at each moment and the previous adjacent moment as the change difference of the corresponding changing region at each moment; Obtaining the difference of each change region corresponding to the change difference between adjacent moments, and performing negative correlation mapping as the difference similarity between adjacent moments; The mean value of the difference similarity between each changing region at all adjacent moments is calculated as the change stability of the connected domain corresponding to each changing region.

[0008] Furthermore, the method for obtaining the change stable region includes: If the change stability of the connected domain corresponding to the change region is included in the preset stability range, the corresponding change region is used as the change stable region.

[0009] Furthermore, the method for obtaining remote sensing attention includes: For each change-stable region, obtain the difference pixel points in the corresponding connected domain between each moment and the previous adjacent moment, connect each difference pixel point with the center of the connected domain at the corresponding moment with a straight line, and obtain the angle between the straight line and the horizontal direction as the change angle of each difference pixel point at the corresponding moment; Obtain the mean value of the change angles of all difference pixel points at the corresponding moment as the local moment change angle; obtain the mean value of the moment change angles at all moments as the overall moment change angle; obtain the difference between the local moment change angle and the overall moment change angle at each moment as the change angle difference; According to the change angle difference of each change stable area and the corresponding change stability in the augmented prediction matrix, the remote sensing attention of each change stable area is obtained. The change angle difference is negatively correlated with the remote sensing attention, and the change stability is positively correlated with the remote sensing attention.

[0010] Furthermore, the method for obtaining the final fusion voting value includes: Obtain the mean value of the morphological features of the connected domain corresponding to each stable region at all times at each scale as the local morphological features of the connected domain corresponding to each stable region at each scale; The local morphological features and remote sensing attention of each stable region at each scale are fused and normalized and mapped as the final fused voting value of each stable region at each scale.

[0011] Furthermore, the method for obtaining the augmented prediction matrix includes: The ARIMA prediction algorithm is used to predict each row vector of the morphological change feature matrix to obtain the predicted value of each row vector. The predicted value and the elements of the corresponding row vector of the morphological change matrix constitute the row vector of the augmented prediction matrix.

[0012] Furthermore, the method for obtaining the connected domain includes: The CANNY edge detection is used to obtain the edge detection grayscale image of each remote sensing image; the connected domain detection algorithm is used to obtain multiple connected domains of the edge detection grayscale image.

[0013] The present invention also proposes a multi-source data processing system applied to national land space planning, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements any step of the multi-source data processing method applied to national land space planning.

[0014] The present invention has the following beneficial effects: The present invention takes into account that different types of objects and regions are distributed in remote sensing images, and obtains multiple connected domains of each remote sensing image according to the grayscale distribution of pixel points in the remote sensing image; for any scale, according to the morphological differences of the connected domains corresponding to the remote sensing images in each same area at different times, the stable change regions are screened out, and those regions with stable changes and clear trends are screened out; the row vectors of the morphological change feature matrix are formed by the morphological features of the connected domains corresponding to each stable change region at different times; the augmented prediction matrix of the morphological change feature matrix is ​​obtained, the prediction ability and stability of the feature matrix are enhanced, and it is helpful to more accurately describe the laws and trends of the changes of the objects, and according to the changes at different times The position distribution of pixels in the connected domain corresponding to the stable region and the change stability corresponding to the augmented prediction matrix are used to obtain the remote sensing attention of each change stable region, which reflects the importance and attention of different change stable regions in remote sensing data analysis, and provides a weight basis for subsequent fusion voting and data fusion; considering the time variation characteristics of different scales, the final fusion voting value of each change stable region is obtained according to the morphological characteristics of the connected domain corresponding to each change stable region at different scales and different times, and the remote sensing attention; the remote sensing image is subjected to data fusion, and by fusing data from different sources and scales, the deficiency of a single data source can be made up, and the accuracy and efficiency of information extraction can be improved. The present invention improves the effect of remote sensing data fusion by obtaining accurate final fusion voting values ​​of remote sensing images at each scale. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0016] Figure 1 A flowchart of a multi-source data processing method for national land space planning provided by one embodiment of the present invention; Figure 2 A flow chart of a method for obtaining variable stability provided by an embodiment of the present invention; Figure 3 A flow chart of a method for acquiring remote sensing attention provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0017] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the multi-source data processing method and system for national land space planning proposed in accordance with the present invention, in combination with the accompanying drawings and preferred embodiments, its specific implementation method, structure, features and effects are described as follows. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0018] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0019] The following is a detailed description of a multi-source data processing method and system for national land space planning provided by the present invention in conjunction with the accompanying drawings.

[0020] See also Figure 1 , which shows a flowchart of a multi-source data processing method for national land space planning provided by an embodiment of the present invention, and the specific method includes: Step S1: Obtain remote sensing images of different scales at each moment in the area to be integrated during national land space planning.

[0021] In the embodiments of the present invention, remote sensing technology can be applied to multiple fields such as national land space planning, construction land monitoring, land geological survey, environmental protection, earthquake disasters, etc., to timely discover key information such as changes in surface cover types, trends in urban expansion, and the scope of impact after disasters. Since multi-source remote sensing data have their own characteristics and advantages, in order to highlight the key feature areas of images at different scales and improve the accuracy and reliability of remote sensing information, multi-source remote sensing data are fused. First, analysis is required at a fixed collection location to obtain remote sensing images of different scales of the area to be fused at each moment.

[0022] It should be noted that, in one embodiment of the present invention, the time interval is one week, the scale is different sizes of resolution and field of view, and the resolution and field of view are increased in an arithmetic manner until the maximum resolution and field of view are reached, and remote sensing images with different fields of view at different resolutions are analyzed; wherein each increase is 1 / 10 of the maximum resolution or field of view; in one embodiment of the present invention, the time interval and scale size can be set according to the specific situation, and are not limited or elaborated here.

[0023] Step S2: obtain multiple connected domains of each remote sensing image according to the grayscale distribution of pixels in the remote sensing image; for any scale, obtain the change effect of the connected domain corresponding to each area according to the morphological differences of the connected domains corresponding to the remote sensing image in each same area between different times; filter out the change area according to the change effect of the connected domains corresponding to all areas; obtain the change stability of the connected domain corresponding to each change area according to the morphological differences of the connected domains corresponding to each change area between adjacent times; filter out the change stable area according to the change stability of the connected domains corresponding to all change areas.

[0024] Since remote sensing images usually contain complex surface information, such as buildings, roads, vegetation, water bodies, etc., which show pixels with different grayscale characteristics in the images, multiple connected domains of each remote sensing image are obtained according to the grayscale distribution of pixels in the remote sensing image, which reduces the interference of noise on the analysis and helps to process and analyze each area separately in the subsequent process, thereby improving the accuracy and specificity of the processing.

[0025] Preferably, in one embodiment of the present invention, the method for acquiring the connected domain includes: The CANNY edge detection algorithm is used to obtain the edge detection grayscale image of each remote sensing image; the connected domain detection algorithm is used to obtain multiple connected domains of the edge detection grayscale image.

[0026] It should be noted that the specific CANNY edge detection algorithm and connected domain detection algorithm are technical means well known to those skilled in the art and will not be described in detail here.

[0027] Since vegetation areas may be affected by factors such as seasonal withering or fire, and when karst collapse occurs in the geology, the objects in the corresponding area will also change. At different times, the objects in the corresponding area of ​​the image may expand or shrink, so that the area of ​​the corresponding connected domain may increase or decrease, resulting in changes in the morphological characteristics of the connected domain. The change effect of the corresponding connected domain is relatively large; therefore, for any scale, the change effect of the connected domain under each sequence number is obtained based on the morphological differences of the connected domain under each same sequence number between different times.

[0028] Preferably, in one embodiment of the present invention, the method for obtaining the change effect includes: Obtain the number of pixels in each connected domain as a morphological feature; For any scale, the differences in morphological features of the connected domains corresponding to the same area of ​​the remote sensing images between different adjacent moments are obtained, and the differences in morphological features between all adjacent moments are accumulated and normalized as the change effect of the connected domain corresponding to each area.

[0029] In one embodiment of the present invention, for any scale, the formula for the change effect is expressed as: ; in, Indicates The change effect of the connected domain corresponding to each region; Indicates The remote sensing image at the moment The regions correspond to the morphological features of the connected domains; Indicates The remote sensing image at the moment The regions correspond to the morphological features of the connected domains; represents the number of all moments; Indicates taking the absolute value; Represents the normalization function.

[0030] In the formula for the change effect, It means calculating the differences in morphological features of the connected domains corresponding to the remote sensing images in each same area between different moments. The greater the difference in morphological features, the greater the difference in the number of pixels in the connected domain, and the greater the change in remote sensing information in the connected domain, the more attention needs to be paid.

[0031] The change effect can reflect which areas of the image have changed significantly at different times. By screening out the areas with significant changes, the dynamic changes in the corresponding scene of the image can be understood in real time; therefore, the change areas are screened out according to the change effects of the corresponding connected domains of all areas.

[0032] Preferably, in one embodiment of the present invention, the method for obtaining the variable connected domain includes: If the change effect of the connected domain corresponding to the region is included in the preset change range, the corresponding region is used as the change region.

[0033] It should be noted that, in one embodiment of the present invention, the preset variation range is 0.8-1, and in other embodiments of the present invention, the size of the preset variation range can be set according to specific circumstances, which is not limited or elaborated here.

[0034] The continuous changes of the morphological features of the connected domain at different times reflect the stability and trend of the changes in the corresponding area. The more consistent the trend of the difference change, the relatively stable the change in the changing area, and the more helpful it is for subsequent analysis. According to the morphological differences of the connected domains corresponding to each changing area between adjacent moments, the change stability of the connected domain corresponding to each changing area is obtained.

[0035] Preferably, in one embodiment of the present invention, the method for obtaining the change stability includes referring to Figure 2 , which shows a flow chart of a method for obtaining variable stability, including: Step S201: obtaining the difference in morphological features of the connected domain corresponding to each change region at each moment and the previous adjacent moment as the change difference corresponding to the change region at each moment.

[0036] In one embodiment of the present invention, the formula for variation difference is expressed as: ; in, Indicates The time corresponds to The variation of the change areas; Indicates The remote sensing image at the moment The morphological characteristics of the connected domains corresponding to the changing regions; Indicates The remote sensing image at the moment The morphological characteristics of the connected domains corresponding to the changing regions; Indicates taking the absolute value.

[0037] In the formula for change variance, Indicates Moment and The remote sensing image is at The change areas correspond to the differences in morphological characteristics of the connected domains. The greater the difference, the greater the difference in change.

[0038] Step S202: Obtain the difference of the corresponding change differences of each change region between adjacent moments, and perform negative correlation mapping as the difference similarity between adjacent moments.

[0039] The corresponding change difference of each changing area between adjacent moments reflects the dynamic change of the corresponding area in time. Calculating the difference in change differences can quantify the consistency of regional dynamic changes. The larger the difference, the smaller the change consistency. Therefore, a negative correlation mapping is performed on the difference in change differences. The larger the difference, the smaller the difference similarity.

[0040] Step S203: Calculate the average value of the difference similarity between each change region at all adjacent moments as the change stability of the connected domain corresponding to each change region.

[0041] The overall distribution of the difference similarity of each change region between all adjacent moments is quantified by taking the average value to obtain the change stability. The smaller the difference similarity, the greater the change stability.

[0042] In one embodiment of the present invention, the formula for obtaining the variation stability is expressed as: ; in, Indicates The stability of changes in connected domains corresponding to the change regions; Indicates The time corresponds to The variation of the change areas; Indicates The time corresponds to The variation of the change areas; represents an exponential function with a natural constant as base; Represents the number of all moments.

[0043] In the formula for variable stability, It means that the exponential function with the natural constant as the base is used to convert Perform negative correlation mapping, that is, difference similarity, Indicates Moment and The time corresponding to The difference in the change differences of the change areas, the greater the difference, the greater the difference in change differences, and the smaller the difference similarity; It means calculating the mean of the difference similarity between all adjacent moments of each change area, that is, the change stability. The smaller the mean of the difference similarity, the smaller the difference similarity, the greater the difference in change difference, the more inconsistent the change situation, and the smaller the change stability.

[0044] It should be noted that, in one embodiment of the present invention, right Negative correlation mapping is performed to form a negative correlation relationship in which the smaller the difference, the more consistent the change, and the greater the change stability; in other embodiments of the present invention, The negative correlation is achieved by taking the inverse. If the negative correlation is achieved by taking the inverse, it is also necessary to add an artificially set parameter in the denominator, such as 0.01, to avoid the denominator being 0, which would make the formula meaningless. The specific means are well known to those skilled in the art and will not be elaborated here.

[0045] In remote sensing images, there are often a large number of changing regions, but not all regions have the same stability. By screening the changing stable regions, those regions with frequent changes, instability or large noise can be removed, thereby reducing the redundancy and complexity of the data and helping to obtain useful information faster. The changing stable regions are screened out according to the changing stability of the connected domains corresponding to all the changing regions.

[0046] Preferably, in one embodiment of the present invention, the method for obtaining the variable stable connected domain includes: If the change stability of the connected domain corresponding to the change region is included in the preset stability range, the corresponding change region is used as the change stable region.

[0047] It should be noted that, in one embodiment of the present invention, the size of the preset stable range is 0.9-1. In other embodiments of the present invention, the size of the preset stable range can be set according to specific circumstances, which is not limited or elaborated here.

[0048] Step S3: The row vectors of the morphological change feature matrix are constructed with the morphological features of the connected domain corresponding to each stable changing region at different times; the augmented prediction matrix of the morphological change feature matrix is ​​obtained, and the remote sensing attention of each stable changing region is obtained according to the position distribution of pixel points in the connected domain corresponding to the stable changing region at different times and the change stability of the corresponding connected domain in the augmented prediction matrix.

[0049] The morphological change feature matrix quantifies the morphological features of each changing and stable region at different times, making the morphological changes concrete and visual, which helps to understand the changes in the region more intuitively. The row vectors of the morphological change feature matrix are composed of the morphological features of the connected domain corresponding to each changing and stable region at different times. It should be noted that each element in the row vector is the number of pixels in the connected domain corresponding to each changing and stable region at each moment.

[0050] In order to more accurately analyze the changes at different times, the augmented prediction matrix of the morphological change feature matrix is ​​obtained to more comprehensively understand the future change trend of the region.

[0051] Preferably, in one embodiment of the present invention, the method for obtaining the augmented prediction matrix includes: The ARIMA prediction algorithm is used to predict each row vector of the morphological change feature matrix to obtain the predicted value of each row vector. The predicted value and the elements of the corresponding row vector of the morphological change matrix constitute the row vector of the augmented prediction matrix.

[0052] It should be noted that the specific ARIMA prediction algorithm is a technical means well known to those skilled in the art and will not be described in detail here.

[0053] The position distribution of pixels can reflect the complexity and dynamics of the internal structure of the region. By comparing the position distribution at different times, we can identify which areas remain stable in time and which areas have changed significantly, which helps to understand which areas need attention. The augmented prediction matrix can evaluate future change trends. The greater the corresponding change stability, the greater the remote sensing attention. Combining the two for comprehensive analysis can more accurately evaluate the remote sensing attention of each change-stable area. According to the position distribution of pixels in the connected domain corresponding to the change-stable area at different times and the change stability corresponding to the augmented prediction matrix, the remote sensing attention of each change-stable area is obtained.

[0054] Preferably, in one embodiment of the present invention, the method for obtaining remote sensing attention degree can be found in Figure 3 , which shows a flow chart of a method for obtaining remote sensing attention, including: Step S301: For each change stable area, obtain the difference pixel points in the corresponding connected domain between each moment and the previous adjacent moment, connect each difference pixel point with the center of the connected domain at the corresponding moment with a straight line, and obtain the angle between the straight line and the horizontal direction as the change angle of each difference pixel point at the corresponding moment.

[0055] The difference pixels represent the pixels whose connected domains have changed between adjacent moments, reflecting the dynamic changes in the area. By calculating the angle between the straight line between the difference pixel and the center of the connected domain and the horizontal direction, the change direction of each difference pixel can be quantified, which helps to understand the trend of changes.

[0056] It should be noted that the center of the connected domain refers to the centroid of the connected domain, that is, the average position of all pixels in the connected domain.

[0057] Step S302: obtain the average change angle of all difference pixel points at the corresponding moment as the local moment change angle; obtain the average change angle of all moments as the overall moment change angle; obtain the difference between the local moment change angle and the overall moment change angle at each moment as the change angle difference.

[0058] By calculating the mean of the change angles of all difference pixels, we can reflect the direction and trend of changes in the connected domain at the corresponding moment, as well as the overall direction and trend of changes in the connected domain at all moments, and make a global description. By comparing the difference between the local change angle and the overall change angle at each moment, we can quantify the degree of difference between the change at each moment and the overall change, reflecting the directional inconsistency between the change at each moment and the overall change, and reflecting the complexity or uncertainty of the change.

[0059] Step S303: According to the change angle difference of each change stable area and the change stability corresponding to the augmented prediction matrix, the remote sensing attention of each change stable area is obtained. The change angle difference is negatively correlated with the remote sensing attention, and the change stability is positively correlated with the remote sensing attention.

[0060] It should be noted that, in one embodiment of the present invention, the method for obtaining the change stability of each change stable region in the augmented prediction matrix is ​​the same as the change stability of the connected domain corresponding to each change region. The change stability of each change stable region in the augmented prediction matrix is ​​obtained by obtaining the difference between adjacent elements in the row vector of the augmented prediction matrix corresponding to each change stable region.

[0061] Among them, the positive correlation means that the dependent variable will increase as the independent variable increases, and the dependent variable will decrease as the independent variable decreases, that is, the greater the change stability, the more similar the change differences between adjacent moments are, the more it is necessary to fusion and retain features, and the greater the remote sensing attention; the negative correlation means that the dependent variable will decrease as the independent variable increases, and the dependent variable will increase as the independent variable decreases, that is, the larger the change angle, the more uncertain factors there are, which affect the efficiency of fusion and the smaller the remote sensing attention.

[0062] In one embodiment of the present invention, for any variable stable region, the remote sensing attention degree is expressed as: ; in, Indicates Remote sensing attention to stable regions; Indicates The change stability regions correspond to the change stability in the augmented prediction matrix; Indicates The average change angle of all difference pixels at the time, that is, The local moment change angle of the moment; It represents the mean of the moment-to-moment change angle at all moments, that is, the overall moment-to-moment change angle; represents the number of all moments; Indicates the adjustment parameter.

[0063] In the formula of remote sensing attention, Indicates The difference between the local moment change angle and the overall moment change angle at a certain moment, that is, the change angle difference, Add an adjustment parameter to the change angle difference to avoid the formula being meaningless when it is 0. The adjustment parameter can be 0.01; It means that after negatively correlating the change angle differences, the mean of all negative correlation results is calculated. The larger the mean, the smaller the change angle difference, the more consistent the change direction, the more necessary it is to perform fusion, the greater the remote sensing attention, the more inconsistent the change direction, the more random the change, the more likely it is that the change is caused by noise, avoid paying attention, and the smaller the remote sensing attention; the greater the change stability, the greater the remote sensing attention.

[0064] Step S4: According to the morphological characteristics of the connected domain corresponding to each stable region at different times at each scale, and the remote sensing attention, the final fusion voting value of each stable region at each scale is obtained; based on the final fusion voting value of each stable region at different scales, data fusion is performed on the remote sensing images used in national land space planning.

[0065] The morphological features at different times at each scale can reflect the overall trend of objects in the region and more comprehensively reflect the representativeness of each stable region to the actual situation. The greater the overall trend, the greater the representativeness, and the greater the final fusion voting value; remote sensing attention usually reflects the importance and attention of the research area or specific target. The more remote sensing information changes in the region, the more remote sensing attention. Comprehensive analysis can ensure that important and characteristic regions are given priority in the fusion process. The final fusion voting value of each stable region at each scale is obtained according to the morphological features of the connected domain corresponding to each stable region at different times at each scale and the remote sensing attention.

[0066] Preferably, in one embodiment of the present invention, the method for obtaining the final fusion voting value includes: Obtain the mean value of the morphological features of the connected domain corresponding to each stable region at all times at each scale as the local morphological features of the connected domain corresponding to each stable region at each scale; The local morphological features and remote sensing attention of each stable region at each scale are fused and normalized and mapped as the final fused voting value of each stable region at each scale.

[0067] It should be noted that, in some embodiments of the present invention, the overall morphological characteristics and remote sensing attention of each variable stable area can be fused by addition or multiplication. The specific method is a technical means well known to those skilled in the art and will not be elaborated here.

[0068] In one embodiment of the present invention, the formula for the final fusion voting value is expressed as: ; in, Indicates The stable region of change is The final fusion voting value at each scale; Indicates Remote sensing attention to stable regions; Indicates The stable region of change is The scale The morphological characteristics of the connected domain corresponding to each moment; represents the number of all moments; Represents the normalization function.

[0069] In the formula for the final fusion voting value, It means calculating the mean morphological features of the connected domain corresponding to each stable region at all times at each scale, that is, the local morphological features. The larger the local morphological features, the larger the area corresponding to the connected domain, the greater the remote sensing attention, the more features are displayed, and the larger the final fusion voting value.

[0070] It should be noted that, in another embodiment of the present invention, after obtaining the final fusion voting value, the remote sensing image can be fused, and the fusion method is: for each variable stable area, the final fusion voting value at each scale is used as the weight of the image area at the corresponding scale, and the product of the final fusion voting value at each scale and the remote sensing data in the image area is calculated as the weighted remote sensing data of each scale; the mean of the weighted remote sensing data at all scales is calculated as the fused remote sensing data; for other areas, the fusion method of directly calculating the mean is used to obtain the fused remote sensing data; after obtaining the fused remote sensing data, the fusion result is further optimized through filtering, morphological operations and other methods to remove unreasonable details, ensure the continuity and consistency of the fused image in space, help improve the observability of the data, and more clearly understand the information representation of the remote sensing details in the region.

[0071] In summary, for any scale, the present invention selects the changing stable areas according to the morphological differences of the connected domains corresponding to the remote sensing images at different times in each same area; the row vectors of the morphological change feature matrix are formed by the morphological features of the connected domains corresponding to each changing stable area at different times; the augmented prediction matrix of the morphological change feature matrix is ​​obtained, and the remote sensing attention of each changing stable area is obtained according to the position distribution of the pixel points in the connected domains corresponding to the changing stable areas at different times and the corresponding change stability in the augmented prediction matrix; the final fusion voting value of each changing stable area is obtained by combining the morphological features of the connected domains corresponding to each changing stable area at different scales at different times; and the remote sensing images are subjected to data fusion. The present invention improves the effect of remote sensing data fusion by obtaining accurate final fusion voting values ​​of remote sensing images at each scale.

[0072] The present invention also proposes a multi-source data processing system applied to national land space planning, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements any step of a multi-source data processing method applied to national land space planning.

[0073] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0074] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A multi-source data processing method applied to national land space planning, characterized in that: The method comprises: Obtain remote sensing images of different preset scales at each moment in the area to be integrated during national land space planning; According to the grayscale distribution of pixels in the remote sensing image, multiple connected domains of each remote sensing image are obtained; for any scale, according to the morphological differences of the connected domains corresponding to each same area of ​​the remote sensing image at different times, the change effect of the connected domain corresponding to each area is obtained; according to the change effect of the connected domains corresponding to all areas, the change area is screened out; according to the morphological differences of the connected domains corresponding to each change area between adjacent times, the change stability of the connected domain corresponding to each change area is obtained; according to the change stability of the connected domains corresponding to all change areas, the change stable area is screened out; The row vectors of the morphological change feature matrix are formed by the morphological features of the connected domain corresponding to each stable change region at different times; the augmented prediction matrix of the morphological change feature matrix is ​​obtained, and the remote sensing attention of each stable change region is obtained according to the position distribution of the pixel points in the connected domain corresponding to the stable change region at different times and the change stability corresponding to the augmented prediction matrix; According to the morphological characteristics of the connected domain corresponding to each stable region at different times at each scale and the remote sensing attention, the final fusion voting value of each stable region at each scale is obtained; based on the final fusion voting value of each stable region at different scales, data fusion is performed on the remote sensing images used in national land space planning.

2. The multi-source data processing method for national land space planning according to claim 1 is characterized in that: The method for obtaining the change effect includes: Obtain the number of pixels in each connected domain as a morphological feature; For any scale, the differences in morphological features of the connected domains corresponding to the same area of ​​the remote sensing images between different adjacent moments are obtained, and the differences in morphological features between all adjacent moments are accumulated and normalized as the change effect of the connected domain corresponding to each area.

3. The multi-source data processing method for national land space planning according to claim 1 is characterized in that: The method for obtaining the change area includes: If the change effect of the connected domain corresponding to the region is included in the preset change range, the corresponding region is used as the change region.

4. The multi-source data processing method for national land space planning according to claim 2 is characterized in that: The method for obtaining the variation stability includes: Obtain the difference in morphological characteristics of the corresponding connected domain of each changing region at each moment and the previous adjacent moment as the change difference of the corresponding changing region at each moment; Obtaining the difference of each change region corresponding to the change difference between adjacent moments, and performing negative correlation mapping as the difference similarity between adjacent moments; The mean value of the difference similarity between each changing region at all adjacent moments is calculated as the change stability of the connected domain corresponding to each changing region.

5. The multi-source data processing method for national land space planning according to claim 1 is characterized in that: The method for obtaining the change stable region comprises: If the change stability of the connected domain corresponding to the change region is included in the preset stability range, the corresponding change region is used as the change stable region.

6. The multi-source data processing method for national land space planning according to claim 1 is characterized in that: The method for obtaining the remote sensing attention comprises: For each change-stable region, obtain the difference pixel points in the corresponding connected domain between each moment and the previous adjacent moment, connect each difference pixel point with the center of the connected domain at the corresponding moment with a straight line, and obtain the angle between the straight line and the horizontal direction as the change angle of each difference pixel point at the corresponding moment; Obtain the mean value of the change angles of all difference pixel points at the corresponding moment as the local moment change angle; obtain the mean value of the moment change angles at all moments as the overall moment change angle; obtain the difference between the local moment change angle and the overall moment change angle at each moment as the change angle difference; According to the change angle difference of each change stable area and the corresponding change stability in the augmented prediction matrix, the remote sensing attention of each change stable area is obtained. The change angle difference is negatively correlated with the remote sensing attention, and the change stability is positively correlated with the remote sensing attention.

7. The multi-source data processing method for national land space planning according to claim 2 is characterized in that: The method for obtaining the final fusion voting value includes: Obtain the mean value of the morphological features of the connected domain corresponding to each stable region at all times at each scale as the local morphological features of the connected domain corresponding to each stable region at each scale; The local morphological features and remote sensing attention of each stable region at each scale are fused and normalized and mapped as the final fused voting value of each stable region at each scale.

8. The multi-source data processing method for national land space planning according to claim 1 is characterized in that: The method for obtaining the augmented prediction matrix includes: The ARIMA prediction algorithm is used to predict each row vector of the morphological change feature matrix to obtain the predicted value of each row vector. The predicted value and the elements of the corresponding row vector of the morphological change matrix constitute the row vector of the augmented prediction matrix.

9. The multi-source data processing method for national land space planning according to claim 1 is characterized in that: The method for obtaining the connected domain includes: The CANNY edge detection is used to obtain the edge detection grayscale image of each remote sensing image; the connected domain detection algorithm is used to obtain multiple connected domains of the edge detection grayscale image.

10. A multi-source data processing system for national land space planning, the system 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, it implements the steps of a multi-source data processing method applied to national land space planning as described in any one of claims 1 to 9.

Images