Method for dynamic monitoring of land use change based on unmanned aerial vehicle remote sensing
By combining the overlap analysis and standard segmentation of UAV remote sensing images with the calculation of the degree of homogeneity and heterogeneity between neighbors, adjacent merging is performed, which solves the problems of insufficient accuracy in type classification and computational complexity in the dynamic analysis of land use change by UAV remote sensing, and ensures the accuracy and real-time nature of the monitoring results.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHANDONG RHEIN TECH EQUIP
- Filing Date
- 2026-03-26
- Publication Date
- 2026-06-09
AI Technical Summary
Existing technologies for dynamic analysis of land use change based on UAV remote sensing suffer from insufficient accuracy in type classification or computational complexity due to the inability to select the optimal segmentation scale, which affects the accuracy and real-time performance of monitoring results.
Multiple remote sensing images are acquired through UAV remote sensing monitoring. Overlap analysis and cropping are performed, basic segmentation is carried out according to the preset standard segmentation scale, the degree of homogeneity and heterogeneity between adjacent targets in the images is calculated, and adjacent images are merged until they cannot be merged. Then, the images are stitched together and classified and labeled to obtain the land use change monitoring results.
It enables precise classification of different land use types, improves the accuracy and real-time nature of monitoring results, and solves the problems of insufficient classification accuracy and computational complexity.
Smart Images
Figure CN122176579A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of unmanned aerial vehicle (UAV) remote sensing technology, and particularly relates to a method for dynamic monitoring of land use change based on UAV remote sensing. Background Technology
[0002] Dynamic monitoring of land use change through UAV remote sensing involves acquiring images of land cover and land use in a specific area using UAV remote sensing technology. Then, through image processing, change detection, and information extraction, the UAV identifies and analyzes land use types and their changes. This process can identify changes such as the addition, reduction, transformation, or expansion of land use types, thus providing important data support for land resource planning, ecological protection, and land management decisions.
[0003] In existing technologies, dynamic analysis of land use change based on UAV remote sensing usually relies on segmenting remote sensing images to classify different land use types and then identify changes in different land use types. However, in practical applications, the inability to select the optimal segmentation scale can easily lead to insufficient accuracy in type classification or computational complexity, thereby affecting the accuracy or real-time performance of dynamic monitoring results of land use change. Summary of the Invention
[0004] The purpose of this invention is to provide a method for dynamic monitoring of land use change based on UAV remote sensing, which aims to solve the technical problems existing in the prior art mentioned in the background.
[0005] The embodiments of the present invention are implemented as follows: A method for dynamic monitoring of land use change based on UAV remote sensing, the method specifically includes the following steps: S101. Multiple remote sensing images are acquired through UAV remote sensing monitoring. Overlap analysis and cropping are performed on the multiple remote sensing images to obtain multiple partial monitoring images. S102. Perform basic segmentation on the multiple monitored images according to the preset standard segmentation scale to obtain multiple image targets; S103. Perform an adjacent comparison analysis on multiple image targets to calculate the degree of homogeneity and heterogeneity between adjacent image targets. S104. Perform a comprehensive analysis on the homogeneity and heterogeneity of the multiple neighboring units to determine whether adjacent units can be merged. If adjacent units can be merged, perform adjacent unit merging and then proceed to S103. S105. When adjacent images cannot be merged, multiple fully merged images are obtained, and image stitching, classification and labeling, and land change analysis are performed to obtain land use change monitoring results.
[0006] As a further limitation of the technical solution of this embodiment of the invention, the step of acquiring multiple remote sensing images through UAV remote sensing monitoring, and performing overlap analysis and cropping on the multiple remote sensing images to obtain multiple partial monitoring images specifically includes the following steps: S1011. Acquire remote sensing monitoring parameters and perform UAV remote sensing monitoring and control; S1012, Receive multiple remote sensing images acquired by UAV remote sensing monitoring; S1013. Perform image overlap analysis on multiple remote sensing monitoring images to determine multiple overlap directions and overlap ratios; S1014. According to the multiple overlapping directions and multiple overlapping ratios, perform image cropping processing on the multiple remote sensing monitoring images to obtain multiple partial monitoring images.
[0007] As a further limitation of the technical solution of this invention embodiment, the step of performing basic segmentation on multiple monitoring images according to a preset standard segmentation scale to obtain multiple image targets specifically includes the following steps: S1021. Perform radiometric and geometric corrections on multiple partial monitoring images to obtain multiple corrected monitoring images; S1022. Denoise the multiple corrected monitoring images to obtain multiple denoised monitoring images; S1023. Perform grayscale normalization processing on the multiple noise-reduced monitoring images to obtain multiple standard monitoring images; S1024. Load the preset standard segmentation scale; S1025. According to the standard segmentation scale, perform basic segmentation on multiple standard monitoring images to obtain multiple image targets.
[0008] As a further limitation of the technical solution of this embodiment of the invention, the step of performing adjacent comparison analysis on multiple image targets and calculating the degree of homogeneity and heterogeneity between multiple adjacent image targets specifically includes the following steps: S1031. Identify the adjacency relationships between multiple image targets; S1032. Based on the adjacency relationship between multiple image targets, perform homogeneity comparison analysis on multiple adjacent image targets, and calculate the degree of homogeneity between multiple adjacent image targets. S1033. Based on the adjacency relationship between multiple image targets, perform heterogeneous comparison analysis on multiple adjacent image targets and calculate the degree of inter-neighbor heterogeneity between multiple adjacent image targets.
[0009] As a further limitation of the technical solution of this embodiment of the invention, the calculation formula for the degree of homogeneity between the plurality of neighbors is as follows: ; in, For the adjacent first The image target and the first The degree of homogeneity between neighboring targets in an image For the first The target area of each image target. For the first The pixel standard deviation of an image target For the first The target area of each image target. For the first The standard deviation of pixels for each image target.
[0010] As a further limitation of the technical solution of this embodiment of the invention, the step of performing heterogeneous comparison analysis on multiple adjacent image targets and calculating the degree of inter-neighbor heterogeneity among multiple adjacent image targets specifically includes the following steps: S10331. Perform neighbor-to-neighbor smoothing comparison analysis on multiple adjacent image targets, and calculate the degree of neighbor-to-neighbor smoothing between multiple adjacent image targets. S10332. Perform a neighbor-to-neighbor compactness comparison analysis on multiple adjacent image targets and calculate the neighbor-to-neighbor compactness between multiple adjacent image targets. S10333. Based on the multiple neighbor smoothness degrees and multiple neighbor compactness degrees, calculate the neighbor heterogeneity degree between multiple adjacent image targets.
[0011] As a further limitation of the technical solution of this embodiment of the invention, the calculation formulas for the plurality of adjacent smoothness degrees are as follows: ; in, For the first The image target and the first The degree of neighbor-to-neighbor smoothing between image targets For the first The first image target and the first The total area after merging the image targets. For the first The first image target and the first Perimeter of the boundary after merging image targets. For the first The first image target and the first The perimeter of the bounding rectangle after merging the image targets. For the first The perimeter of the boundary of an image target. For the first The perimeter of the bounding rectangle of the image target. For the first The perimeter of the boundary of an image target. For the first Perimeter of the bounding rectangle of an image target; The formulas for calculating the compactness between the multiple neighbors are as follows: ; in, For the first The image target and the first The degree of neighbor compactness between image targets; The formulas for calculating the degree of heterogeneity between the aforementioned neighbors are as follows: ; ; ; in, For the first The image target and the first The degree of neighbor heterogeneity between image targets , , and The preset weighting factors, Represents the image layer. For the first Image layer weights, For the first The first image layer The first image target and the first The spectral standard deviation of the merged image targets For the first The first image layer The spectral standard deviation of each image target. For the first The first image layer The spectral standard deviation of an image target.
[0012] As a further limitation of the technical solution of this embodiment of the invention, the step of comprehensively analyzing the homogeneity and heterogeneity of multiple neighboring properties to determine whether adjacent merging can be performed, and then performing adjacent merging processing when adjacent merging can be performed, before proceeding to S103, specifically includes the following steps: S1041. Calculate multiple comprehensive merging indices based on multiple neighbor homogeneity levels and multiple neighbor heterogeneity levels; S1042. Compare the multiple comprehensive merging indices with the preset standard merging indices to determine whether adjacent merging can be performed. S1043. If multiple composite merging indices are greater than the standard merging index, it is determined that adjacent merging cannot be performed. In this case, proceed to S105. S1044. If one or more composite consolidation indices are not greater than the standard consolidation index, it is determined that adjacent consolidation can be carried out. In this case, proceed to S103.
[0013] As a further limitation of the technical solution of this embodiment of the invention, the calculation formulas for the plurality of comprehensive consolidation indices are as follows: ; in, For the first The image target and the first A comprehensive merging index among image targets. and This is a preset calculation factor.
[0014] As a further limitation of the technical solution of this invention, when adjacent images cannot be merged, obtaining multiple fully merged images and performing image stitching, classification and labeling, and land change analysis to obtain land use change monitoring results specifically includes the following steps: S1051. When adjacent merging is not possible, multiple fully merged images are obtained; S1052. Perform related image stitching on multiple fully merged images to obtain multiple stitched merged images; S1053. According to multiple preset land use types, classify and label the multiple stitched and merged images to obtain multiple stitched and labeled images; S1054. Acquire multiple historically labeled images; S1055. Perform land change analysis on multiple stitched and labeled images and multiple historical labeled images to obtain land use change monitoring results.
[0015] Compared with the prior art, the beneficial effects of the present invention are: This invention, after basic image segmentation, performs adjacent comparison analysis on multiple image targets to calculate the degree of homogeneity and heterogeneity between neighbors, and then judges and processes adjacent merging until adjacent merging is no longer possible. After that, it performs classification labeling and land change analysis to obtain land use change monitoring results. Thus, the merging process automatically realizes the classification of different land use types, which can solve the problems of insufficient classification accuracy and computational complexity, and effectively ensure the accuracy and real-time nature of dynamic monitoring results of land use change. Attached Figure Description
[0016] Figure 1 A flowchart of the land use change dynamic monitoring method based on UAV remote sensing provided in an embodiment of the present invention is shown; Figure 2 The flowchart of image overlap analysis and cropping in the method provided by the embodiment of the present invention is shown; Figure 3 A flowchart illustrating image-based segmentation in the method provided by an embodiment of the present invention is shown; Figure 4 A flowchart illustrating the neighbor comparison of image targets in the method provided by an embodiment of the present invention is shown; Figure 5 A flowchart of heterogeneous comparison analysis is shown in the method provided in the embodiments of the present invention; Figure 6 The flowchart illustrating the method for determining whether adjacent elements can be merged is shown in the embodiment of the present invention. Figure 7 The flowchart illustrating image stitching, classification and annotation, and land change analysis in the method provided by the embodiments of the present invention is shown. Detailed Implementation
[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.
[0018] Understandably, in existing technologies, dynamic analysis of land use change based on UAV remote sensing usually relies on segmenting remote sensing images to classify different land use types and then identify changes in different land use types. However, in practical applications, the inability to select the optimal segmentation scale can easily lead to problems such as insufficient accuracy in type classification or computational complexity, thereby affecting the accuracy or real-time performance of dynamic monitoring results of land use change.
[0019] To address the aforementioned issues, this invention discloses a land use change dynamic monitoring method based on UAV remote sensing. This method acquires multiple remote sensing images through UAV remote sensing monitoring, performs overlap analysis and cropping on these images to obtain multiple partial monitoring images. Following a preset standard segmentation scale, the multiple partial monitoring images are segmented to obtain multiple image targets. Adjacent image targets are compared and analyzed to calculate the degree of homogeneity and heterogeneity between adjacent targets. A comprehensive analysis of the homogeneity and heterogeneity between adjacent targets is performed to determine whether adjacent targets can be merged. If merging is possible, it is performed. If merging is not possible, multiple fully merged images are obtained, and image stitching, classification labeling, and land change analysis are performed to obtain the land use change monitoring results. After basic image segmentation, the system can perform neighbor comparison analysis on multiple image targets, calculate the degree of homogeneity and heterogeneity between neighbors, and then judge and process neighbor merging until neighbor merging is no longer possible. After that, it performs classification labeling and land change analysis to obtain land use change monitoring results. Thus, the system can automatically classify different land use types through the merging process, which can solve the problems of insufficient accuracy in type classification and complex calculation, and ensure the accuracy and real-time nature of dynamic monitoring results of land use change.
[0020] Specifically, Figure 1 A flowchart of a land use change dynamic monitoring method based on UAV remote sensing provided in an embodiment of the present invention is shown.
[0021] In a preferred embodiment of the present invention, a method for dynamic monitoring of land use change based on UAV remote sensing specifically includes the following steps: S101. Multiple remote sensing images are acquired through UAV remote sensing monitoring. Overlap analysis and cropping are performed on the multiple remote sensing images to obtain multiple partial monitoring images.
[0022] In this embodiment of the invention, pre-planned remote sensing monitoring parameters are obtained, and monitoring and control of the UAV are performed according to the remote sensing monitoring parameters. Then, multiple remote sensing monitoring images acquired by the UAV are received, and image overlap analysis is performed on the multiple remote sensing monitoring images to determine the overlap direction and overlap ratio of the multiple remote sensing monitoring images. After that, according to the multiple overlap directions and multiple overlap ratios, the overlapping images of the multiple corresponding remote sensing monitoring images are cropped and removed to obtain multiple retained monitoring images.
[0023] Specifically, Figure 2 The flowchart of image overlap analysis and cropping in the method provided by the embodiment of the present invention is shown.
[0024] In a preferred embodiment of the present invention, the step of acquiring multiple remote sensing images through UAV remote sensing monitoring, and performing overlap analysis and cropping on the multiple remote sensing images to obtain multiple partial monitoring images specifically includes the following steps: S1011. Acquire remote sensing monitoring parameters and perform UAV remote sensing monitoring and control; S1012, Receive multiple remote sensing images acquired by UAV remote sensing monitoring; S1013. Perform image overlap analysis on multiple remote sensing monitoring images to determine multiple overlap directions and overlap ratios; S1014. According to the multiple overlapping directions and multiple overlapping ratios, perform image cropping processing on the multiple remote sensing monitoring images to obtain multiple partial monitoring images.
[0025] Furthermore, the land use change dynamic monitoring method based on UAV remote sensing also includes the following steps: S102. According to the preset standard segmentation scale, perform basic segmentation on the multiple monitoring images to obtain multiple image targets.
[0026] In this embodiment of the invention, radiometric and geometric corrections are performed on multiple partial monitoring images to obtain multiple corrected monitoring images. Then, median filtering or Gaussian filtering is used to denoise the multiple corrected monitoring images to remove random noise and sensor interference, resulting in multiple denoised monitoring images. Subsequently, grayscale normalization is performed on the multiple denoised monitoring images to unify the image pixel values within a preset range, resulting in multiple standard monitoring images. Then, a preset standard segmentation scale is loaded, and basic segmentation is performed on the multiple standard monitoring images according to the standard segmentation scale to obtain multiple image targets.
[0027] It is understood that in the embodiments of the present invention, there is no need to calculate and determine the optimal segmentation scale. Instead, the basic segmentation process is directly performed according to the preset standard segmentation scale, and the standard segmentation scale can be adjusted accordingly according to the monitoring accuracy requirements of land use change.
[0028] Specifically, Figure 3 A flowchart illustrating image-based segmentation is shown in the method provided by an embodiment of the present invention.
[0029] In a preferred embodiment of the present invention, the step of performing basic segmentation on multiple monitored images according to a preset standard segmentation scale to obtain multiple image targets specifically includes the following steps: S1021. Perform radiometric and geometric corrections on multiple partial monitoring images to obtain multiple corrected monitoring images; S1022. Denoise the multiple corrected monitoring images to obtain multiple denoised monitoring images; S1023. Perform grayscale normalization processing on the multiple noise-reduced monitoring images to obtain multiple standard monitoring images; S1024. Load the preset standard segmentation scale; S1025. According to the standard segmentation scale, perform basic segmentation on multiple standard monitoring images to obtain multiple image targets.
[0030] Furthermore, the land use change dynamic monitoring method based on UAV remote sensing also includes the following steps: S103. Perform neighbor comparison analysis on multiple image targets to calculate the degree of homogeneity and heterogeneity between multiple adjacent image targets.
[0031] In this embodiment of the invention, by identifying the adjacency relationships between multiple image targets, and then performing a homogeneity comparison analysis on the multiple adjacent image targets according to the adjacency relationships, the degree of homogeneity between the multiple adjacent image targets is calculated. Next, according to the adjacency relationships between the multiple image targets, a smoothness comparison analysis is performed on the multiple adjacent image targets to calculate the smoothness between the multiple adjacent image targets. Furthermore, a compactness comparison analysis is performed on the multiple adjacent image targets to calculate the compactness between the multiple adjacent image targets. Finally, based on the smoothness and compactness of the multiple adjacent image targets, the degree of heterogeneity between the multiple adjacent image targets is calculated. Specifically, the formula for calculating the degree of homogeneity between multiple adjacent targets is: ; in, For the adjacent first The image target and the first The degree of homogeneity between neighboring targets in an image For the first The target area of each image target. For the first The pixel standard deviation of an image target For the first The target area of each image target. For the first The standard deviation of pixels for each image target; The formula for calculating the smoothness of multiple neighboring nodes is as follows: ; in, For the first The image target and the first The degree of neighbor-to-neighbor smoothing between image targets For the first The first image target and the first The total area after merging the image targets. For the first The first image target and the first Perimeter of the boundary after merging image targets. For the first The first image target and the first The perimeter of the bounding rectangle after merging the image targets. For the first The perimeter of the boundary of an image target. For the first The perimeter of the bounding rectangle of the image target. For the first The perimeter of the boundary of an image target. For the first Perimeter of the bounding rectangle of an image target; The formula for calculating the compactness of multiple adjacent nodes is as follows: ; in, For the first The image target and the first The degree of neighbor compactness between image targets; The formula for calculating the degree of heterogeneity among multiple neighbors is: ; ; ; in, For the first The image target and the first The degree of neighbor heterogeneity between image targets , , and The preset weighting factors, Represents the image layer. For the first Image layer weights, For the first The first image layer The first image target and the first The spectral standard deviation of the merged image targets For the first The first image layer The spectral standard deviation of each image target. For the first The first image layer The spectral standard deviation of an image target.
[0032] It is understandable that the lower the degree of homogeneity between neighboring targets, the higher the homogeneity between two adjacent image targets; the lower the degree of heterogeneity between neighboring targets, the lower the heterogeneity between two adjacent image targets.
[0033] Specifically, Figure 4 A flowchart illustrating the neighbor comparison of image targets in the method provided by an embodiment of the present invention is shown.
[0034] In a preferred embodiment of the present invention, the step of performing adjacent comparison analysis on multiple image targets and calculating the degree of homogeneity and heterogeneity between adjacent image targets specifically includes the following steps: S1031. Identify the adjacency relationships between multiple image targets; S1032. Based on the adjacency relationship between multiple image targets, perform homogeneity comparison analysis on multiple adjacent image targets, and calculate the degree of homogeneity between multiple adjacent image targets. S1033. Based on the adjacency relationship between multiple image targets, perform heterogeneous comparison analysis on multiple adjacent image targets and calculate the degree of inter-neighbor heterogeneity between multiple adjacent image targets.
[0035] Specifically, Figure 5 A flowchart illustrating heterogeneous comparison analysis is shown in the method provided in this embodiment of the invention.
[0036] In a preferred embodiment of the present invention, the step of performing heterogeneous comparison analysis on multiple adjacent image targets and calculating the degree of inter-neighbor heterogeneity among multiple adjacent image targets specifically includes the following steps: S10331. Perform neighbor-to-neighbor smoothing comparison analysis on multiple adjacent image targets, and calculate the degree of neighbor-to-neighbor smoothing between multiple adjacent image targets. S10332. Perform a neighbor-to-neighbor compactness comparison analysis on multiple adjacent image targets and calculate the neighbor-to-neighbor compactness between multiple adjacent image targets. S10333. Based on the multiple neighbor smoothness degrees and multiple neighbor compactness degrees, calculate the neighbor heterogeneity degree between multiple adjacent image targets.
[0037] Furthermore, the land use change dynamic monitoring method based on UAV remote sensing also includes the following steps: S104. Perform a comprehensive analysis on the homogeneity and heterogeneity of the multiple neighboring units to determine whether adjacent units can be merged. If adjacent units can be merged, perform adjacent unit merging processing and then proceed to S103.
[0038] In this embodiment of the invention, after calculating the degree of homogeneity and heterogeneity between multiple adjacent image targets, a comprehensive merging index is calculated based on the degree of homogeneity and heterogeneity between multiple adjacent image targets. The comprehensive merging index is compared with a preset standard merging index to determine whether adjacent merging can be performed. Specifically, if all comprehensive merging indices are greater than the standard merging index, it is determined that all adjacent image targets cannot be merged, and the process proceeds to S105. However, if one or more comprehensive merging indices are not greater than the standard merging index, it is determined that one or more pairs of adjacent image targets can be merged. In this case, adjacent merging processing is performed on the corresponding pair or more adjacent image targets. Then, the process proceeds to S103 to recalculate the degree of homogeneity and heterogeneity between multiple adjacent image targets. Specifically, the formula for calculating the multiple comprehensive merging indices is as follows: ; in, For the first The image target and the first A comprehensive merging index among image targets. and This is a preset calculation factor.
[0039] Specifically, Figure 6 The flowchart illustrating the method for determining whether adjacent elements can be merged is shown in the embodiment of the present invention.
[0040] In a preferred embodiment of the present invention, the step of comprehensively analyzing the homogeneity and heterogeneity of multiple neighboring properties to determine whether neighboring properties can be merged, and then performing neighboring merging processing when neighboring properties can be merged, before proceeding to S103, specifically includes the following steps: S1041. Calculate multiple comprehensive merging indices based on multiple neighbor homogeneity levels and multiple neighbor heterogeneity levels; S1042. Compare the multiple comprehensive merging indices with the preset standard merging indices to determine whether adjacent merging can be performed. S1043. If multiple composite merging indices are greater than the standard merging index, it is determined that adjacent merging cannot be performed. In this case, proceed to S105. S1044. If one or more composite consolidation indices are not greater than the standard consolidation index, it is determined that adjacent consolidation can be carried out. In this case, proceed to S103.
[0041] Furthermore, the land use change dynamic monitoring method based on UAV remote sensing also includes the following steps: S105. When adjacent images cannot be merged, multiple fully merged images are obtained, and image stitching, classification and labeling, and land change analysis are performed to obtain land use change monitoring results.
[0042] In this embodiment of the invention, when all adjacent image targets cannot be merged, multiple fully merged images are obtained. Then, according to the positional relationship of the multiple fully merged images, relevant image stitching processing is performed to obtain multiple stitched merged images. Then, according to multiple preset land use types, the multiple stitched merged images are classified and labeled to obtain multiple stitched labeled images. Multiple historical labeled images are also obtained. Then, land change analysis is performed on the multiple stitched labeled images and the multiple historical labeled images to obtain land use change monitoring results. The change monitoring results record data such as the change ratio and change location of different land use types.
[0043] Specifically, Figure 7 The flowchart illustrating image stitching, classification and annotation, and land change analysis in the method provided by the embodiments of the present invention is shown.
[0044] In a preferred embodiment of the present invention, when adjacent images cannot be merged, obtaining multiple fully merged images and performing image stitching, classification and labeling, and land change analysis to obtain land use change monitoring results specifically includes the following steps: S1051. When adjacent merging is not possible, multiple fully merged images are obtained; S1052. Perform related image stitching on multiple fully merged images to obtain multiple stitched merged images; S1053. According to multiple preset land use types, classify and label the multiple stitched and merged images to obtain multiple stitched and labeled images; S1054. Acquire multiple historically labeled images; S1055. Perform land change analysis on multiple stitched and labeled images and multiple historical labeled images to obtain land use change monitoring results.
[0045] The above-described embodiments are merely examples of several implementations of the present invention, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of the present invention. For those skilled in the art, various modifications and improvements can be made without departing from the concept of the present invention, and these modifications and improvements all fall within the protection scope of the present invention.
Claims
1. A method for dynamic monitoring of land use change based on UAV remote sensing, characterized in that, The method specifically includes the following steps: S101. Multiple remote sensing images are acquired through UAV remote sensing monitoring. Overlap analysis and cropping are performed on the multiple remote sensing images to obtain multiple partial monitoring images. S102. Perform basic segmentation on the multiple monitored images according to the preset standard segmentation scale to obtain multiple image targets; S103. Perform an adjacent comparison analysis on multiple image targets to calculate the degree of homogeneity and heterogeneity between adjacent image targets. S104. Perform a comprehensive analysis on the homogeneity and heterogeneity of the multiple neighboring units to determine whether adjacent units can be merged. If adjacent units can be merged, perform adjacent unit merging and then proceed to S103. S105. When adjacent images cannot be merged, multiple fully merged images are obtained, and image stitching, classification and labeling, and land change analysis are performed to obtain land use change monitoring results.
2. The method for dynamic monitoring of land use change based on UAV remote sensing according to claim 1, characterized in that, The process of acquiring multiple remote sensing images through UAV remote sensing monitoring, and performing overlap analysis and cropping on these multiple images to obtain multiple partial monitoring images specifically includes the following steps: S1011. Acquire remote sensing monitoring parameters and perform UAV remote sensing monitoring and control; S1012, Receive multiple remote sensing images acquired by UAV remote sensing monitoring; S1013. Perform image overlap analysis on multiple remote sensing monitoring images to determine multiple overlap directions and overlap ratios; S1014. According to the multiple overlapping directions and multiple overlapping ratios, perform image cropping processing on the multiple remote sensing monitoring images to obtain multiple partial monitoring images.
3. The method for dynamic monitoring of land use change based on UAV remote sensing according to claim 1, characterized in that, The step of performing basic segmentation on multiple monitored images according to a preset standard segmentation scale to obtain multiple image targets specifically includes the following steps: S1021. Perform radiometric and geometric corrections on multiple partial monitoring images to obtain multiple corrected monitoring images; S1022. Denoise the multiple corrected monitoring images to obtain multiple denoised monitoring images; S1023. Perform grayscale normalization processing on the multiple noise-reduced monitoring images to obtain multiple standard monitoring images; S1024. Load the preset standard segmentation scale; S1025. According to the standard segmentation scale, perform basic segmentation on multiple standard monitoring images to obtain multiple image targets.
4. The method for dynamic monitoring of land use change based on UAV remote sensing according to claim 1, characterized in that, The step of performing neighbor comparison analysis on multiple image targets and calculating the degree of homogeneity and heterogeneity between multiple adjacent image targets specifically includes the following steps: S1031. Identify the adjacency relationships between multiple image targets; S1032. Based on the adjacency relationship between multiple image targets, perform homogeneity comparison analysis on multiple adjacent image targets, and calculate the degree of homogeneity between multiple adjacent image targets. S1033. Based on the adjacency relationship between multiple image targets, perform heterogeneous comparison analysis on multiple adjacent image targets and calculate the degree of inter-neighbor heterogeneity between multiple adjacent image targets.
5. The method for dynamic monitoring of land use change based on UAV remote sensing according to claim 4, characterized in that, The formulas for calculating the homogeneity between the aforementioned neighbors are as follows: ； in, For the adjacent first The image target and the first The degree of homogeneity between neighboring targets in an image For the first The target area of each image target. For the first The pixel standard deviation of an image target For the first The target area of each image target. For the first The standard deviation of pixels for each image target.
6. The method for dynamic monitoring of land use change based on UAV remote sensing according to claim 5, characterized in that, The heterogeneous comparison analysis of multiple adjacent image targets, and the calculation of the degree of inter-neighbor heterogeneity among multiple adjacent image targets, specifically includes the following steps: S10331. Perform neighbor-to-neighbor smoothing comparison analysis on multiple adjacent image targets, and calculate the degree of neighbor-to-neighbor smoothing between multiple adjacent image targets. S10332. Perform a neighbor-to-neighbor compactness comparison analysis on multiple adjacent image targets and calculate the neighbor-to-neighbor compactness between multiple adjacent image targets. S10333. Based on the multiple neighbor smoothness degrees and multiple neighbor compactness degrees, calculate the neighbor heterogeneity degree between multiple adjacent image targets.
7. The method for dynamic monitoring of land use change based on UAV remote sensing according to claim 6, characterized in that, The formulas for calculating the smoothness between multiple neighbors are as follows: ； in, For the first The image target and the first The degree of neighbor-to-neighbor smoothing between image targets For the first The first image target and the first The total area after merging the image targets. For the first The first image target and the first Perimeter of the boundary after merging image targets. For the first The first image target and the first The perimeter of the bounding rectangle after merging the image targets. For the first The perimeter of the boundary of an image target. For the first The perimeter of the bounding rectangle of the image target. For the first The perimeter of the boundary of an image target. For the first Perimeter of the bounding rectangle of an image target; The formulas for calculating the compactness between the multiple neighbors are as follows: ； in, For the first The image target and the first The degree of neighbor compactness between image targets; The formulas for calculating the degree of heterogeneity between the aforementioned neighbors are as follows: ； ； ； in, For the first The image target and the first The degree of neighbor heterogeneity between image targets , , and The preset weighting factors, Represents the image layer. For the first Image layer weights, For the first The first image layer The first image target and the first The spectral standard deviation of the merged image targets For the first The first image layer The spectral standard deviation of each image target. For the first The first image layer The spectral standard deviation of an image target.
8. The method for dynamic monitoring of land use change based on UAV remote sensing according to claim 7, characterized in that, The step of comprehensively analyzing the homogeneity and heterogeneity of multiple neighboring units to determine whether adjacent units can be merged, and then performing adjacent unit merging when it is possible, before proceeding to S103, specifically includes the following steps: S1041. Calculate multiple comprehensive merging indices based on multiple neighbor homogeneity levels and multiple neighbor heterogeneity levels; S1042. Compare the multiple comprehensive merging indices with the preset standard merging indices to determine whether adjacent merging can be performed. S1043. If multiple composite merging indices are greater than the standard merging index, it is determined that adjacent merging cannot be performed. In this case, proceed to S105. S1044. If one or more composite consolidation indices are not greater than the standard consolidation index, it is determined that adjacent consolidation can be carried out. In this case, proceed to S103.
9. The method for dynamic monitoring of land use change based on UAV remote sensing according to claim 8, characterized in that, The formulas for calculating the various composite consolidation indices are as follows: ； in, For the first The image target and the first A comprehensive merging index among image targets. and This is a preset calculation factor.
10. The method for dynamic monitoring of land use change based on UAV remote sensing according to claim 1, characterized in that, When adjacent images cannot be merged, multiple fully merged images are obtained, and image stitching, classification and labeling, and land change analysis are performed to obtain land use change monitoring results. Specifically, this includes the following steps: S1051. When adjacent merging is not possible, multiple fully merged images are obtained; S1052. Perform related image stitching on multiple fully merged images to obtain multiple stitched merged images; S1053. According to multiple preset land use types, classify and label the multiple stitched and merged images to obtain multiple stitched and labeled images; S1054. Acquire multiple historically labeled images; S1055. Perform land change analysis on multiple stitched and labeled images and multiple historical labeled images to obtain land use change monitoring results.