Land utilization change monitoring method and system

Abstract

The invention provides a land utilization change monitoring method and system. The method comprises the following steps: acquiring a remote sensing image of a to-be-monitored land and sampling data of a plurality of sampling points in the to-be-monitored land; wherein the sampling data comprises one or more of soil parameters, air parameters, hydrological parameters, ecological parameters and climate parameters; generating a data distribution image of the to-be-monitored land based on the coordinates of each sampling point and the sampling data; fusing the data distribution image with the remote sensing image to obtain a fused image of the to-be-monitored land; and segmenting the fused image based on the image segmentation model to obtain a land utilization boundary of the to-be-monitored land, and determining the land utilization change of the to-be-monitored land based on the land utilization boundary. Compared with a traditional method, the scheme not only improves the monitoring precision, but also enhances the adaptability to complex scenes.

Description

Technical Field

[0001] The present invention relates to the technical field of land use monitoring, and particularly relates to a method and system for monitoring land use changes. Background Art

[0002] Land use change monitoring refers to the process of long-term tracking and analysis of land use types, spatial distributions, and their dynamic changes in a certain area. By monitoring land use changes, it is possible to provide a scientific basis for relevant departments, optimize the allocation of land resources, and formulate reasonable land use plans.

[0003] Traditional land use change monitoring methods are usually single remote sensing image analysis or ground sampling surveys. However, although remote sensing images can provide information on ground objects over a large area, their sensitivity to certain environmental parameters (such as soil nutrients, air pollution index, etc.) is relatively low. Although ground sampling data can provide detailed environmental parameter information, due to the limited number of sampling points, it is difficult to directly generate a continuous spatial distribution map. In addition, the boundaries of land use types often have complex geometric shapes and fuzzy transition regions, and traditional segmentation methods are difficult to accurately capture these details, resulting in low accuracy of current land use change monitoring. Summary of the Invention

[0004] Embodiments of the present invention provide a method and system for monitoring land use changes to solve the problem of improving the accuracy of land use change monitoring.

[0005] In a first aspect, embodiments of the present invention provide a method for monitoring land use changes, including: Obtaining a remote sensing image of the land to be monitored and sampling data of multiple sampling points within the land to be monitored; wherein, the sampling data includes one or more of soil parameters, air parameters, hydrological parameters, ecological parameters, and climate parameters; Generating a data distribution image of the land to be monitored based on the coordinates and sampling data of each sampling point; Fusing the data distribution image with the remote sensing image to obtain a fused image of the land to be monitored; Segmenting the fused image based on an image segmentation model to obtain the land use boundary of the land to be monitored, so as to determine the land use changes of the land to be monitored based on the land use boundary.

[0006] In a possible implementation manner, generating a data distribution image of the land to be monitored based on the coordinates and sampling data of each sampling point includes: For each type of sampling data, performing spatial interpolation based on the sampling data of each sampling point of this type to obtain multiple continuous distribution data corresponding to this type of sampling data; wherein, each continuous distribution data includes coordinates and values; For each type of sampled data, convert the values of the continuous distribution data corresponding to this type of sampled data into RGB values, and obtain the data distribution image corresponding to this type of sampled data according to the coordinate combination; Overlay the data distribution images corresponding to each type of sampled data to obtain the data distribution image of the land to be monitored.

[0007] In a possible implementation, for each type of sampled data, converting the values of the continuous distribution data corresponding to this type of sampled data into RGB values includes: For each continuous distribution data of the first type of sampled data, calculate the proportion of the value of this continuous distribution data in the corresponding value range, and determine the RGB value corresponding to the value of this continuous distribution data based on the proportion of the value and the RGB value range corresponding to the first type of sampled data; where the first type of sampled data is any type of sampled data.

[0008] In a possible implementation, before obtaining the sampled data of multiple sampling points in the land to be monitored, it further includes: For each land use type, use the distinction between this land use type and other land use types as the target variable, perform sensitivity analysis on the sampled data of each type, and obtain the type of sampled data corresponding to this land use type; Correspondingly, obtaining the sampled data of multiple sampling points in the land to be monitored includes: Based on the type of sampled data corresponding to the land to be monitored, obtain the sampled data of multiple sampling points in the land to be monitored.

[0009] In a possible implementation, fusing the data distribution image with the remote sensing image to obtain the fused image of the land to be monitored includes: Overlay the data distribution image and the remote sensing image based on the weighted average method to obtain the fused image of the land to be monitored.

[0010] In a possible implementation, overlaying the data distribution image and the remote sensing image based on the weighted average method to obtain the fused image of the land to be monitored includes: For each pixel position, calculate the local correlation between the data distribution image and the remote sensing image at this pixel position; where the calculation formula for the local correlation is:

[0011] Where, is the local correlation between the data distribution image and the remote sensing image at the pixel position , is the pixel value of the remote sensing image at the pixel position , is the pixel value of the data distribution image at the pixel position . is a local window centered at the pixel position , and is a pixel position within the local window . is the pixel mean value of the remote sensing image within the local window , and is the pixel mean value of the data distribution image within the local window ; Based on the local correlation of each pixel position, determine the weight of the data distribution image and the remote sensing image at this pixel position; Based on the weights of the data distribution image and the remote sensing image at each pixel position, overlay the data distribution image and the remote sensing image to obtain a fused image of the land to be monitored.

[0012] In a possible implementation, before segmenting the fused image based on the image segmentation model to obtain the land use boundary of the land to be monitored, it further includes: Train the U-Net network with the labeled fused image to obtain an image segmentation model.

[0013] In a second aspect, an embodiment of the present invention provides a land use change monitoring system, including: An acquisition module, configured to acquire a remote sensing image of the land to be monitored and sampling data of multiple sampling points within the land to be monitored; wherein, the sampling data includes one or more of soil parameters, air parameters, hydrological parameters, ecological parameters, and climate parameters; A generation module, configured to generate a data distribution image of the land to be monitored based on the coordinates and sampling data of each sampling point; A fusion module, configured to fuse the data distribution image and the remote sensing image to obtain a fused image of the land to be monitored; A segmentation module, configured to segment the fused image based on the image segmentation model to obtain the land use boundary of the land to be monitored, so as to determine the land use change of the land to be monitored based on the land use boundary.

[0014] In a third aspect, an embodiment of the present invention provides a terminal, 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 the steps of the method described in the first aspect or any possible implementation manner of the first aspect above.

[0015] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, where the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, it implements the steps of the method described in the first aspect or any possible implementation manner of the first aspect above.

[0016] An embodiment of the present invention provides a method and system for monitoring land use changes. By combining remote sensing images and ground sampling data, the deficiencies of a single data source are made up for, and the characteristics of land use types can be reflected more comprehensively. An effective data fusion method is proposed to solve the problems of differences in resolution and dimension between remote sensing images and ground sampling data. Compared with traditional methods, this solution not only improves the monitoring accuracy but also enhances the adaptability to complex scenarios. By comparing the segmentation results at different times, the areas of land use changes can be quickly located, and their areas and influence ranges can be quantified. This method is applicable to a variety of application scenarios, such as urban expansion monitoring, deforestation detection, wetland degradation assessment, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0018] Figure 1 is a flowchart of the implementation of a method for monitoring land use changes provided by an embodiment of the present invention; Figure 2 is a schematic structural diagram of a system for monitoring land use changes provided by an embodiment of the present invention; Figure 3 is a schematic diagram of a terminal provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] In the following description, specific details such as specific system structures and technologies are proposed for the purpose of illustration rather than limitation, so as to thoroughly understand the embodiments of the present invention. However, those skilled in the art should clearly understand that the present invention can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid unnecessary details from interfering with the description of the present invention.

[0020] To make the objectives, technical solutions, and advantages of the present invention clearer, the following will be described through specific embodiments in conjunction with the drawings.

[0021] Refer to Figure 1 , which shows a flowchart of the implementation of a method for monitoring land use changes provided by an embodiment of the present invention, and is described in detail as follows: Step 101, obtain a remote sensing image of the land to be monitored and sampling data of multiple sampling points within the land to be monitored; wherein, the sampling data includes one or more of soil parameters, air parameters, hydrological parameters, ecological parameters, and climate parameters.

[0022] In this embodiment, the remote sensing image is the image data of the Earth's surface obtained through platforms such as satellites, drones, or airplanes. These images can be data in visible light, near-infrared, thermal infrared, or other bands, and can reflect the spatial distribution characteristics of surface cover types (such as vegetation, water bodies, buildings, etc.).

[0023] The sampling point refers to a specific location selected within the land to be monitored, which is a data point for collecting relevant parameters such as soil, air, and hydrology. The selection of sampling points is usually based on geographical distribution uniformity or the requirements of specific target areas.

[0024] Soil parameters can include soil texture, organic matter content, pH value, nutrient content (such as nitrogen, phosphorus, potassium), water content, etc., which are used to evaluate soil fertility and health status. Air parameters can include air quality index (AQI), particulate matter concentration (PM2.5, PM10), sulfur dioxide (SO2), nitrogen oxides (NOx), ozone (O3), etc., which are used to evaluate air pollution levels. Hydrological parameters can include groundwater level, river flow, water quality indicators (such as dissolved oxygen, turbidity, conductivity, etc.), which are used to monitor water resource conditions and water environment quality. Ecological parameters can include biodiversity index, vegetation coverage rate, ecosystem service value, etc., which are used to evaluate the health status and functions of ecosystems. Climate parameters can include temperature, humidity, precipitation, wind speed, solar radiation, etc., which are used to analyze regional climate conditions and their impacts on land use.

[0025] Remote sensing images can provide large-scale spatial data and are suitable for land use analysis at a macro scale. Sampling data reflects the specific characteristics of the local environment. Combining the two can more comprehensively describe the comprehensive characteristics of the land, and further tasks such as land use classification, boundary extraction, and change monitoring can be carried out, providing a scientific basis for natural resource management, ecological environment protection, etc.

[0026] Step 102: Generate a data distribution image of the land to be monitored based on the coordinates and sampling data of each sampling point.

[0027] In this embodiment, the sampling point refers to a specific location selected within the land to be monitored, which is a data point for collecting relevant parameters such as soil, air, and hydrology. Each sampling point has clear geographical coordinates (such as longitude and latitude or projected coordinates) and corresponding environmental parameter values.

[0028] The data distribution image is to expand the discrete sampling point data into a continuous spatial distribution map to visually display the variation law of environmental parameters on the land to be monitored. Common data distribution images include soil organic matter distribution maps, air quality distribution maps, etc.

[0029] Discrete sampled point data can be extended into a data distribution image through an interpolation algorithm. Common interpolation algorithms include Kriging, Inverse Distance Weighting (IDW), and Nearest Neighbor Interpolation, etc. For example, the specific steps for interpolation can include: 1. Data preparation: Collect the coordinates (x, y) of the sampling points and the corresponding sampled data values (such as soil organic matter content, air quality index, etc.). Organize the data into a structured format, such as a matrix or table form.

[0030] 2. Select an interpolation algorithm: Select a suitable interpolation algorithm according to the characteristics of the sampled data and research requirements. For example: Kriging interpolation is suitable for data with spatial autocorrelation.

[0031] Inverse Distance Weighting is suitable for data with relatively smooth local variations.

[0032] 3. Perform interpolation: Use the interpolation algorithm to calculate the value of each grid cell (pixel) to generate regular grid data.

[0033] 4. Generate an image: Visualize the interpolation result as an image, using color coding to represent the range of different parameter values. For example, green represents a high vegetation coverage rate, and red represents low air quality.

[0034] Interpolation algorithms can estimate the parameter values in unsampled areas, making up for the problem of limited number of sampling points. The data distribution image can intuitively display the spatial distribution characteristics of environmental parameters on the land to be monitored, facilitating analysis and decision-making.

[0035] Step 103: Fuse the data distribution image with the remote sensing image to obtain a fused image of the land to be monitored.

[0036] In this embodiment, by using image fusion technology, combining discrete data distribution information with high-resolution remote sensing images can generate a fused image containing more information. The specific steps can be as follows: 1. Data preparation: Prepare the data distribution image (such as a soil organic matter distribution map) and the remote sensing image (such as satellite imagery or images taken by drones). Ensure that the geographic coordinate systems of the two are consistent and align the spatial resolutions of the images.

[0037] 2. Image preprocessing: Perform radiometric correction, atmospheric correction, and geometric correction on the remote sensing image to eliminate possible errors introduced during the imaging process. Perform normalization processing on the data distribution image to ensure that its numerical range is consistent with that of the remote sensing image.

[0038] 3. Select a fusion method: Select a suitable image fusion method according to application requirements. For example: Weighted average method: Simply superimpose the two images by assigning different weights to them.

[0039] Principal Component Analysis (PCA): Extract the main components of the remote sensing image and incorporate the information of the data distribution image into it.

[0040] Wavelet transform method: Use wavelet transform to decompose the multi-scale features of the image, and then fuse the features of the two images.

[0041] 4. Perform fusion: Combine the data distribution image and the remote sensing image using the selected fusion method to generate a fused image.

[0042] 5. Result verification: Evaluate the quality of the fused image to ensure that it retains the main features of the original image and enhances the information expression ability.

[0043] The data distribution image provides the spatial distribution characteristics of environmental parameters, while the remote sensing image shows the actual form or state of the surface. The fusion of the two can achieve information complementarity and provide more comprehensive monitoring results.

[0044] Step 104, segment the fused image based on the image segmentation model to obtain the land use boundary of the land to be monitored, and determine the land use change of the land to be monitored based on the land use boundary.

[0045] In this embodiment, the image segmentation model is an algorithm model that divides an image into multiple regions or objects through computer vision technology. Common image segmentation methods include threshold-based methods, edge detection, region growing, watershed algorithms, and semantic segmentation and instance segmentation models in deep learning (such as U-Net, Mask R-CNN, etc.).

[0046] The land use boundary refers to the boundary line between different land use types, such as the boundary between farmland and forest, urban and rural areas, water body and land. These boundaries reflect the spatial distribution and usage of land resources.

[0047] Land use change refers to the change in land use over time in a certain area, such as the conversion from farmland to construction land, or from forest to grassland. This change is usually caused by natural factors or human activities.

[0048] Use the image segmentation model to process the fused image, divide the image into different regions, and each region corresponds to a specific land use type (such as farmland, forest, water body, etc.). The model will automatically identify and label different types of regions according to the characteristics of pixels (such as color, texture, shape, etc.). Compare the segmentation results of the current time period with those of the historical time period, identify which regions have changed in land use type, and by calculating indicators such as change area and change ratio, the degree and trend of land use change can be quantitatively evaluated.

[0049] In the embodiments of the present invention, by combining remote sensing images and ground sampling data, the deficiencies of a single data source are made up for, the characteristics of land use types can be reflected more comprehensively, and an effective data fusion method is proposed to solve the problems of differences in resolution and dimension between remote sensing images and ground sampling data. Compared with traditional methods, this solution not only improves the monitoring accuracy but also enhances the adaptability to complex scenarios. By comparing the segmentation results of different periods, the areas of land use change can be quickly located, and their areas and influence ranges can be quantified. This method is applicable to a variety of application scenarios, such as urban expansion monitoring, deforestation detection, wetland degradation assessment, etc.

[0050] In a possible implementation manner, based on the coordinates and sampling data of each sampling point, a data distribution image of the land to be monitored is generated, including: For each type of sampling data, spatial interpolation is performed based on the sampling data of each sampling point to obtain multiple continuous distribution data corresponding to the sampling data; where each continuous distribution data includes coordinates and values; For each type of sampling data, the values of the continuous distribution data corresponding to the sampling data are converted into RGB values, and according to the coordinate combination, a data distribution image corresponding to the sampling data is obtained; The data distribution images corresponding to each type of sampling data are superimposed to obtain the data distribution image of the land to be monitored.

[0051] In this embodiment, the value range of each type of sampling data can be mapped to the RGB color space. For example: Soil organic matter content: low values are mapped to red, and high values are mapped to green.

[0052] Air quality index: excellent and good values are mapped to blue, and polluted values are mapped to yellow.

[0053] The coordinates of each continuous distribution data and the corresponding RGB values form a pixel point. According to the RGB values and coordinate information of each type of sampling data, a data distribution image corresponding to the sampling data is generated, and comprehensive evaluation is realized through superposition analysis.

[0054] Taking the soil fertility assessment of a certain farmland as an example, the specific steps for generating the spatial distribution images of soil organic matter content, pH value, and moisture content are as follows: 1. Sampling data collection: 50 sampling points are evenly set in the farmland to cover the entire area. The soil organic matter content, pH value, and moisture content data of each sampling point are collected, and their geographical coordinates (latitude and longitude) are recorded.

[0055] 2. Spatial interpolation: Kriging interpolation method is used to perform spatial interpolation on soil organic matter content, pH value, and moisture content respectively. The interpolation results generate three continuous distribution data sets, and each data set contains coordinates and corresponding values.

[0056] 3. Numerical conversion to RGB values: Map the numerical range of soil organic matter content (e.g., 0 - 10 g / kg) to the RGB color space: Low values (0 - 3 g / kg) are mapped to red (R = 255, G = 0, B = 0).

[0057] High values (7 - 10 g / kg) are mapped to green (R = 0, G = 255, B = 0).

[0058] Map the numerical range of pH value (e.g., 4 - 8) to the RGB color space: Acidic (4 - 6) is mapped to blue (R = 0, G = 0, B = 255).

[0059] Neutral to alkaline (7 - 8) is mapped to yellow (R = 255, G = 255, B = 0).

[0060] Map the numerical range of moisture content (e.g., 10% - 30%) to the RGB color space: Dry (10% - 15%) is mapped to gray (R = 128, G = 128, B = 128).

[0061] Humid (25% - 30%) is mapped to cyan (R = 0, G = 255, B = 255).

[0062] 4. Generate data distribution images of single sampling data: Generate distribution maps of soil organic matter content, pH value, and moisture content according to the RGB values and coordinate information of each sampling data.

[0063] 5. Overlay data distribution images: Overlay the three data distribution images together to generate a data distribution image comprehensively reflecting the soil fertility status.

[0064] In a possible implementation, for each sampling data, convert the numerical values of the continuous distribution data corresponding to the sampling data into RGB values, including: For each continuous distribution data of the first sampling data, calculate the proportion of the numerical value of the continuous distribution data in the corresponding numerical range, and determine the RGB value corresponding to the numerical value of the continuous distribution data based on the proportion and the RGB value range corresponding to the first sampling data; where the first sampling data is any sampling data.

[0065] In this embodiment, the continuous distribution data is grid data containing coordinates and values generated by spatial interpolation, representing the continuous distribution of a certain sampling data within the entire monitoring area. The numerical range refers to the interval within which a certain sampling data may take values. For example, the numerical range of soil organic matter content may be 0 - 10 g / kg, and the numerical range of pH value may be 4 - 8. The RGB value is a numerical combination of the three primary colors, Red, Green, and Blue, used to define the color of each pixel in an image. The RGB value is usually represented by three integers between 0 and 255. The numerical proportion refers to the relative position of the value of a certain continuous distribution data within its corresponding numerical range, usually expressed as a percentage. For example, if the soil organic matter content is 6 g / kg and its numerical range is 0 - 10 g / kg, then the numerical proportion is 60%. The RGB value range refers to the color interval used when mapping the numerical range of the sampling data to the RGB color space. For example, the low value of soil organic matter content may be mapped to red (R = 255, G = 0, B = 0), and the high value may be mapped to green (R = 0, G = 255, B = 0).

[0066] Through the numerical proportion and linear mapping method, the value of the sampling data can be accurately converted into RGB values, ensuring that the color change is consistent with the data change. Using the color - coding method to display the spatial distribution characteristics of the sampling data makes complex data easier to understand and analyze. Different sampling data can be mapped using different RGB value ranges, facilitating the superposition of data distribution images of multiple parameters in the same image.

[0067] In one possible implementation, before obtaining the sampling data of multiple sampling points within the land to be monitored, it further includes: For each land use type, taking the distinction between this land use type and other land use types as the target variable, performing a sensitivity analysis on the sampling data of each type to obtain the sampling data type corresponding to this land use type; Correspondingly, obtaining the sampling data of multiple sampling points within the land to be monitored includes: Based on the sampling data type corresponding to the land to be monitored, obtaining the sampling data of multiple sampling points within the land to be monitored.

[0068] In this embodiment, the land use type refers to different categories divided according to the use or coverage characteristics of the land, such as farmland, forest, urban land, water body, etc. The sampling data type refers to the key environmental parameters related to a certain land use type, such as soil organic matter content, pH value, vegetation coverage rate, etc. The sensitivity analysis is a method for evaluating the influence degree of different parameters on the target variable, used to screen out the key parameters with the strongest ability to distinguish land use types.

[0069] Taking the land use classification of an area and generating a land use distribution map as an example, in order to improve the classification accuracy, a sensitivity analysis is carried out before sampling to screen out the key sampling data types with the strongest ability to distinguish different land use types. The specific steps are as follows: 1. Collect preliminary data: Randomly select several sampling points within the study area and collect various sampling data types, including: Soil parameters: Organic matter content, pH value, moisture content.

[0070] Hydrological parameters: Groundwater level, river flow.

[0071] Ecological parameters: Vegetation coverage rate, biodiversity index.

[0072] Climatic parameters: Precipitation, temperature.

[0073] 2. Sensitivity analysis: For each land use type (such as farmland, forest, urban land), the distinction from other land use types is used as the target variable. Use the random forest model to conduct a sensitivity analysis and extract the feature importance scores of each sampling data type. For example: Farmland vs. other types: Soil organic matter content > Vegetation coverage rate > Precipitation.

[0074] Forest vs. other types: Vegetation coverage rate > Soil moisture content > Temperature.

[0075] Urban land vs. other types: Precipitation > Groundwater level > pH value.

[0076] 3. Determine the key sampling data types: According to the sensitivity analysis results, determine the key sampling data types corresponding to each land use type: Farmland: Soil organic matter content, vegetation coverage rate.

[0077] Forest: Vegetation coverage rate, soil moisture content.

[0078] Urban land: Precipitation, groundwater level.

[0079] 4. Collect key sampling data: Set multiple sampling points within the land to be monitored and only collect the key sampling data types related to each land use type. For example, focus on collecting soil organic matter content and vegetation coverage rate data in the farmland area; focus on collecting vegetation coverage rate and soil moisture content data in the forest area.

[0080] 5. Build a classification model: Use the collected key sampling data to train a classification model (such as a support vector machine, random forest, or deep learning model) to achieve automatic classification of land use types.

[0081] 6. Result analysis: The classification results show that the model constructed based on the key sampling data types has a high classification accuracy, and the classification accuracies of farmland, forest, and urban land reach 95%, 92%, and 90% respectively.

[0082] By performing sensitivity analysis before sampling and screening out the key sampling data types, the accuracy of land use classification can be successfully improved, and the costs of data collection and processing can be significantly reduced.

[0083] In a possible implementation, the data distribution image is fused with the remote sensing image to obtain a fused image of the land to be monitored, including: The data distribution image and the remote sensing image are superimposed based on the weighted average method to obtain a fused image of the land to be monitored.

[0084] In this embodiment, weighted average is a common technique for image fusion, and the specific implementation steps include: 1. Data preparation: Convert the remote sensing image and the data distribution image into the form of a numerical matrix to ensure that the two images have the same resolution and geographic reference. Assume that the remote sensing image is matrix R and the data distribution image is matrix D.

[0085] 2. Define the objective: Determine the objective of the fusion. For example: If you want to retain more ground object information of the remote sensing image, the weight of the remote sensing image should be larger. If you want to highlight the spatial distribution characteristics of environmental parameters, the weight of the data distribution image should be larger.

[0086] 3. Weight selection method: (1) Based on prior knowledge Set the weights according to the research objective and experience. For example: The weight of the remote sensing image wR = 0.7 The weight of the data distribution image wD = 0.3 (2) Based on correlation analysis Calculate the correlation coefficient matrix between the remote sensing image and the data distribution image. For each pixel position, calculate the correlation between the two images, and assign a larger weight to the image with a higher correlation.

[0087] 4. Use the following formula to fuse the remote sensing image and the data distribution image:

[0088] Where: is the pixel value of the fused image, are the weights of the remote sensing image and the data distribution image respectively.

[0089] 5. Post-processing Normalize the fused image to ensure that the pixel values are within a reasonable range (such as [0, 1] or [0, 255]).

[0090] In a possible implementation, the data distribution image is superimposed on the remote sensing image based on the weighted average method to obtain a fused image of the land to be monitored, including: For each pixel position, the local correlation between the data distribution image and the remote sensing image at this pixel position; wherein, the calculation formula of the local correlation is:

[0091] Wherein, is the local correlation between the data distribution image and the remote sensing image at the pixel position , is the pixel value of the remote sensing image at the pixel position , is the pixel value of the data distribution image at the pixel position , is the local window centered on the pixel position , is the local window inside a pixel position is the pixel mean value of the remote sensing image inside the local window , is the pixel mean value of the data distribution image inside the local window ; Based on the local correlation of each pixel position, determine the weight of the data distribution image and the remote sensing image at this pixel position; Based on the weights of the data distribution image and the remote sensing image at each pixel position, superimpose the data distribution image and the remote sensing image to obtain a fused image of the land to be monitored.

[0092] In this embodiment, the core of the weighted average method lies in determining a reasonable weight value. By calculating the local correlation and dynamically adjusting the weight, the adaptive fusion of the remote sensing image and the data distribution image can be realized. Each symbol in the formula directly corresponds to the pixel values and their statistical characteristics of the two images, ensuring the scientific and reasonable fusion process. This method can better retain the important information of the two images and reduce the influence of irrelevant noise at the same time.

[0093] In a possible implementation, before segmenting the fused image based on the image segmentation model to obtain the land use boundary of the land to be monitored, it further includes: Train the U-Net network with the labeled fused image to obtain an image segmentation model.

[0094] In this embodiment, U-Net is a convolutional neural network model commonly used in medical image segmentation and remote sensing image segmentation. Its structure is in the shape of a "U", consisting of two parts: downsampling (contraction path) and upsampling (expansion path). The core idea of the U-Net network is to extract high-level features of the image through the encoder (downsampling), and then gradually restore the spatial details of the image through the decoder (upsampling).

[0095] The encoder part uses convolutional layers and pooling layers to gradually reduce the image resolution and extract global features; the decoder part gradually increases the resolution through transposed convolutional layers to restore detailed features.

[0096] In each decoder stage, U-Net concatenates the feature maps of the corresponding stage of the encoder with the current feature map to enhance the retention of spatial information, effectively extract multi-scale features of the image, and perform accurate segmentation.

[0097] The specific steps for training and using the U-Net network can include: 1. Prepare annotated fusion images: Collect a certain number of fusion images, and have experts perform pixel-level annotation on these images to generate corresponding label images. For example, in the label image, the farmland area is represented by the number "1", the forest area is represented by the number "2", and the water area is represented by the number "3".

[0098] 2. Construct the U-Net network: Use the U-Net network as the basic architecture to design an image segmentation model suitable for the characteristics of the land to be monitored. The U-Net network extracts global features through downsampling and restores detailed features through upsampling, and finally outputs a segmentation result with the same size as the input image.

[0099] 3. Train the image segmentation model: Divide the annotated fusion images into a training set and a validation set.

[0100] Use the training set to train the U-Net network, and adjust the network weights through the backpropagation algorithm to make the prediction results of the model as close as possible to the true annotations.

[0101] Evaluate the model performance on the validation set and select the best model parameters.

[0102] 4. Apply the image segmentation model: Use the trained image segmentation model to segment the fusion images of the land to be monitored to obtain the land use boundary.

[0103] It should be understood that the sequence numbers of the steps in the above embodiments do not imply the order of execution. The execution order of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0104] The following is a system embodiment of the present invention. For details not described in detail, reference may be made to the corresponding method embodiments above.

[0105] Figure 2 The structural schematic diagram of a land use change monitoring system provided by an embodiment of the present invention is shown. For the convenience of description, only the parts related to the embodiment of the present invention are shown and are described in detail as follows: As Figure 2 shown, a land use change monitoring system 2 includes: An acquisition module 21, configured to acquire a remote sensing image of the land to be monitored and sampling data of a plurality of sampling points within the land to be monitored; wherein, the sampling data includes one or more of soil parameters, air parameters, hydrological parameters, ecological parameters, and climate parameters; A generation module 22, configured to generate a data distribution image of the land to be monitored based on the coordinates and sampling data of each sampling point; A fusion module 23, configured to fuse the data distribution image with the remote sensing image to obtain a fused image of the land to be monitored; A segmentation module 24, configured to segment the fused image based on an image segmentation model to obtain the land use boundary of the land to be monitored, so as to determine the land use change of the land to be monitored based on the land use boundary.

[0106] In a possible implementation manner, the generation module 22 is specifically configured to: For each type of sampling data, perform spatial interpolation based on the sampling data of each sampling point to obtain a plurality of continuous distribution data corresponding to the type of sampling data; wherein, each continuous distribution data includes coordinates and values; For each type of sampling data, convert the values of the continuous distribution data corresponding to the type of sampling data into RGB values, and combine them according to the coordinates to obtain a data distribution image corresponding to the type of sampling data; Overlay the data distribution images corresponding to each type of sampling data to obtain a data distribution image of the land to be monitored.

[0107] In a possible implementation manner, the generation module 22 is specifically configured to: For each continuous distribution data of the first type of sampling data, calculate the value proportion of the value of the continuous distribution data in the corresponding value range, and determine the RGB value corresponding to the value of the continuous distribution data based on the value proportion and the RGB value range corresponding to the first type of sampling data; wherein, the first type of sampling data is any type of sampling data.

[0108] In a possible implementation, the acquisition module 21 is further configured to: Before acquiring the sampling data of multiple sampling points in the land to be monitored, for each land use type, taking the distinction between this land use type and other land use types as the target variable, performing sensitivity analysis on the sampling data of each type, and obtaining the sampling data type corresponding to this land use type; Based on the sampling data type corresponding to the land to be monitored, acquire the sampling data of multiple sampling points in the land to be monitored.

[0109] In a possible implementation, the fusion module 23 is specifically configured to: Overlay the data distribution image and the remote sensing image based on the weighted average method to obtain the fused image of the land to be monitored.

[0110] In a possible implementation, the fusion module 23 is specifically configured to: For each pixel position, calculate the local correlation between the data distribution image and the remote sensing image at this pixel position; where the calculation formula for the local correlation is:

[0111] Where is the local correlation between the data distribution image and the remote sensing image at the pixel position is the local correlation, is the pixel value of the remote sensing image at the pixel position is the pixel value, is the pixel value of the data distribution image at the pixel position is the pixel value, is the local window centered on the pixel position is the local window, is the local window is a pixel position within the local window, is the pixel mean of the remote sensing image within the local window is the pixel mean, is the pixel mean of the data distribution image within the local window is the pixel mean; Based on the local correlation of each pixel position, determine the weight of the data distribution image and the remote sensing image at this pixel position; Based on the weights of the data distribution image and the remote sensing image at each pixel position, overlay the data distribution image and the remote sensing image to obtain the fused image of the land to be monitored.

[0112] In a possible implementation, the segmentation module 24 is further configured to: Before segmenting the fused image based on the image segmentation model to obtain the land use boundary of the land to be monitored, the U-Net network is trained with the labeled fused image to obtain the image segmentation model.

[0113] In the embodiment of the present invention, by combining remote sensing images and ground sampling data, the deficiencies of a single data source are made up for, and the characteristics of land use types can be reflected more comprehensively. An effective data fusion method is proposed to solve the problems of differences in resolution and dimension between remote sensing images and ground sampling data. Compared with traditional methods, this solution not only improves the monitoring accuracy but also enhances the adaptability to complex scenarios. By comparing the segmentation results of different periods, the areas of land use change can be quickly located, and their areas and influence ranges can be quantified. This method is applicable to a variety of application scenarios, such as urban expansion monitoring, deforestation detection, wetland degradation assessment, etc.

[0114] Figure 3 is a schematic diagram of the terminal provided by the embodiment of the present invention. As Figure 3 shown, the terminal 3 of this embodiment includes: a processor 30, a memory 31, and a computer program 32 stored in the memory 31 and operable on the processor 30. When the processor 30 executes the computer program 32, the steps in each of the above embodiments of a land use change monitoring method are implemented. Alternatively, when the processor 30 executes the computer program 32, the functions of each module / unit in each of the above system embodiments are implemented.

[0115] Exemplarily, the computer program 32 can be divided into one or more modules / units. The one or more modules / units are stored in the memory 31 and executed by the processor 30 to complete the present invention. The one or more modules / units can be a series of computer program instruction segments capable of performing specific functions, and these instruction segments are used to describe the execution process of the computer program 32 in the terminal 3.

[0116] The terminal 3 can be a computing device such as a desktop computer, a notebook, a palm computer, and a cloud server. The terminal 3 may include, but is not limited to, a processor 30 and a memory 31. Those skilled in the art can understand that Figure 3 merely examples of the terminal 3, which do not constitute a limitation to the terminal 3. It may include more or fewer components than shown in the figure, or combine certain components, or different components. For example, the terminal may further include input / output devices, network access devices, a bus, etc.

[0117] The so-called processor 30 may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0118] The memory 31 may be an internal storage unit of the terminal 3, such as the hard disk or memory of the terminal 3. The memory 31 may also be an external storage device of the terminal 3, such as a plug-in hard disk equipped on the terminal 3, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc. Further, the memory 31 may also include both the internal storage unit of the terminal 3 and the external storage device. The memory 31 is used to store the computer program and other programs and data required by the terminal. The memory 31 may also be used to temporarily store data that has been output or is to be output.

[0119] Those skilled in the art can clearly understand that, for the convenience and conciseness of description, only the above division of each functional unit and module is used as an example. In practical applications, the above functions can be allocated to different functional units and modules according to needs, that is, the internal structure of the system is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiment can be integrated into a processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit. In addition, the specific names of each functional unit and module are only for the convenience of mutual distinction and do not limit the protection scope of this application. The specific working processes of the units and modules in the above system can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0120] In the above embodiments, the descriptions of the respective embodiments have their own emphases. For parts not detailed or recorded in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0121] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. A professional technician can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0122] In the embodiments provided by the present invention, it should be understood that the disclosed device / terminal and method can be implemented in other ways. For example, the device / terminal embodiments described above are merely illustrative. For example, the division of the modules or units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the devices or units can be in electrical, mechanical or other forms.

[0123] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place, or can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0124] In addition, the functional units in various embodiments of the present invention can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.

[0125] When the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, to implement all or part of the processes in the above-described embodiment methods of the present invention, it can also be completed by a computer program instructing relevant hardware. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above-described various embodiments of a land use change monitoring method can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the content included in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice within the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0126] The above-described embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A method for monitoring land use change, characterized in that, Including: Obtaining a remote sensing image of the land to be monitored and sampling data of multiple sampling points within the land to be monitored; wherein, the sampling data includes one or more of soil parameters, air parameters, hydrological parameters, ecological parameters, and climate parameters; Generating a data distribution image of the land to be monitored based on the coordinates and sampling data of each sampling point; Fusing the data distribution image with the remote sensing image to obtain a fused image of the land to be monitored; Segmenting the fused image based on an image segmentation model to obtain the land use boundary of the land to be monitored, so as to determine the land use change of the land to be monitored based on the land use boundary.

2. The land use change monitoring method according to claim 1, characterized in that The generating a data distribution image of the land to be monitored based on the coordinates and sampling data of each sampling point includes: For each type of sampling data, performing spatial interpolation based on the sampling data of each sampling point of this type to obtain a plurality of continuous distribution data corresponding to this type of sampling data; wherein, each continuous distribution data includes coordinates and values; For each type of sampling data, converting the values of the continuous distribution data corresponding to this type of sampling data into RGB values, and combining them according to the coordinates to obtain a data distribution image corresponding to this type of sampling data; Overlaying the data distribution images corresponding to each type of sampling data to obtain the data distribution image of the land to be monitored.

3. The land use change monitoring method according to claim 2, characterized in that, The converting the values of the continuous distribution data corresponding to each type of sampling data into RGB values includes: For each continuous distribution data of the first type of sampling data, calculating the numerical proportion of the value of this continuous distribution data within the corresponding numerical range, and determining the RGB value corresponding to the value of this continuous distribution data based on the numerical proportion and the RGB value range corresponding to the first type of sampling data; wherein, the first type of sampling data is any type of sampling data.

4. A method for monitoring land use change according to claim 1, characterized in that, Before obtaining the sampling data of multiple sampling points within the land to be monitored, it further includes: For each land use type, using the distinction between this land use type and other land use types as the target variable, performing sensitivity analysis on the sampling data of each type to obtain the sampling data type corresponding to this land use type; Correspondingly, obtaining the sampling data of multiple sampling points within the land to be monitored includes: Based on the sampling data type corresponding to the land to be monitored, obtaining the sampling data of multiple sampling points within the land to be monitored.

5. A method for monitoring land use change according to claim 1, characterized in that, The fusing the data distribution image with the remote sensing image to obtain a fused image of the land to be monitored includes: Overlaying the data distribution image and the remote sensing image based on the weighted average method to obtain a fused image of the land to be monitored.

6. A method for monitoring land use change according to claim 5, characterized in that, The overlaying the data distribution image and the remote sensing image based on the weighted average method to obtain a fused image of the land to be monitored includes: For each pixel position, calculating the local correlation between the data distribution image and the remote sensing image at this pixel position; wherein, the calculation formula of the local correlation is: wherein, is the local correlation between the data distribution image and the remote sensing image at the pixel position , is the pixel value of the remote sensing image at the pixel position , is the pixel value of the data distribution image at the pixel position , is the local window centered at the pixel position , is the local window is a pixel position within the local window is the pixel mean of the remote sensing image within the local window , is the pixel mean of the data distribution image within the local window ; Based on the local correlation of each pixel position, determining the weight between the data distribution image and the remote sensing image at this pixel position. Based on the weights of the data distribution image and the remote sensing image at each pixel position, the data distribution image and the remote sensing image are superimposed to obtain the fused image of the land to be monitored.

7. A method for monitoring land use change according to claim 1, characterized in that, Before segmenting the fused image based on the image segmentation model to obtain the land use boundary of the land to be monitored, it further includes: Training the U-Net network with the annotated fused image to obtain an image segmentation model.

8. A land use change monitoring system, characterized in that, It includes: An acquisition module for acquiring a remote sensing image of the land to be monitored and sampling data of multiple sampling points within the land to be monitored; wherein, the sampling data includes one or more of soil parameters, air parameters, hydrological parameters, ecological parameters, and climate parameters; A generation module for generating a data distribution image of the land to be monitored based on the coordinates and sampling data of each sampling point; A fusion module for fusing the data distribution image and the remote sensing image to obtain the fused image of the land to be monitored; A segmentation module for segmenting the fused image based on the image segmentation model to obtain the land use boundary of the land to be monitored, so as to determine the land use change of the land to be monitored based on the land use boundary.

9. A terminal, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7 above.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the method according to any one of claims 1 to 7 above.

Images