Method and device for determining variation type of cultivated land, medium and product

Abstract

The invention discloses a cultivated land change type determination method and device, a medium and a product, and relates to the technical field of cultivated land monitoring, and the method comprises the steps: obtaining original images of a monitoring region at two different moments; respectively preprocessing the original images at the two different moments to obtain preprocessed images at the two different moments; performing boundary identification on the preprocessed images of the monitoring area at the two different moments by using a boundary extraction model to obtain images with predicted cultivated land boundaries of the monitoring area at the two different moments; inputting images with predicted cultivated land boundaries at two different moments of the monitoring area into a change detection model to obtain predicted cultivated land area variation of the monitoring area; the change detection model is obtained by training the Siamese U-Net; and based on the predicted cultivated land area variation, determining a cultivated land abnormal change type of the monitoring area. According to the invention, the determination precision of the abnormal change type of the cultivated land is improved.

Description

Technical Field

[0001] This application relates to the field of farmland monitoring technology, and in particular to a method, apparatus, medium and product for determining the type of farmland change. Background Technology

[0002] Against the backdrop of rapid urbanization and industrialization, farmland protection faces severe challenges, necessitating efficient and accurate monitoring methods to grasp dynamic changes in farmland. Traditional monitoring methods largely rely on manual surveys or satellite remote sensing. While these methods offer advantages in terms of wide coverage, they are often limited by long data update cycles, high costs, and insufficient adaptability to complex terrains. In particular, when faced with small-scale, fragmented changes in farmland, they struggle to capture subtle differences and real-time dynamics, significantly compromising the timeliness and accuracy of monitoring results.

[0003] In this field, the core challenge lies in how to achieve efficient, accurate, and continuous identification and tracking of changes in arable land. In particular, the acquisition and processing of image data presents a contradiction between data resolution and coverage. High resolution often means limited coverage, while large-scale monitoring can easily sacrifice detail accuracy. This contradiction further evolves into the difficulty of comparative analysis of image data from different periods, because inconsistencies in data quality directly affect the accuracy of change detection, making the identification of issues such as changes in arable land boundaries or illegal occupation complex and prone to errors.

[0004] Therefore, a method for determining the change type of arable land with high classification accuracy is needed. Summary of the Invention

[0005] The purpose of this application is to provide a method, apparatus, medium, and product for determining the type of changes in arable land, so as to solve the problem of low accuracy in determining the type of abnormal changes in arable land.

[0006] To achieve the above objectives, this application provides the following solution.

[0007] Firstly, this application provides a method for determining the type of change in arable land, including: Acquire raw images of the monitored area at two different times; the raw images include UAV images and satellite images; The original images of the monitoring area at two different times were preprocessed to obtain the preprocessed images of the monitoring area at two different times. Using a boundary extraction model, boundary identification was performed on the preprocessed images of the monitoring area at two different times, resulting in images of the monitoring area with predicted farmland boundaries at two different times. The boundary extraction model was trained using the DeepLabv3+ model. Images of the monitored area at two different times, with predicted farmland boundaries, are input into the change detection model to obtain the predicted change in farmland area of ​​the monitored area; the change detection model is trained on Siamese U-Net. Based on the predicted changes in cultivated land area, the types of abnormal changes in cultivated land in the monitoring area are determined.

[0008] Secondly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described method for determining the type of change in cultivated land.

[0009] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for determining the type of change in cultivated land as described above.

[0010] Fourthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the method for determining the type of change in cultivated land as described above.

[0011] According to the specific embodiments provided in this application, the following technical effects are disclosed: This application discloses a method, apparatus, medium, and product for determining the type of farmland change. By fusing detailed UAV imagery with wide-coverage satellite imagery, it resolves the contradiction between resolution and coverage in traditional monitoring, reducing the detection error of abnormal farmland change types. Furthermore, by using a change detection model trained on Siamese U-Net, it achieves precise capture of subtle area changes, significantly improving the accuracy of identifying abnormal farmland change types. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 A schematic flowchart of a method for determining the type of change in cultivated land provided in an embodiment of this application; Figure 2 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

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

[0015] The purpose of this application is to provide a method, apparatus, medium, and product for determining the type of changes in arable land, with the aim of improving the accuracy of determining the types of abnormal changes in arable land.

[0016] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0017] In one exemplary embodiment, such as Figure 1 As shown, a method for determining the type of change in arable land is provided, including the following steps.

[0018] Step 1: Acquire raw images of the monitoring area at two different times; the raw images include UAV images and satellite images.

[0019] Step 2: Preprocess the original images of the monitoring area at two different times to obtain the preprocessed images of the monitoring area at two different times.

[0020] As an optional implementation, when performing step 2, the preprocessing of the original image at any time includes the following steps.

[0021] Step 21: Perform standardization correction on the UAV imagery and satellite imagery respectively to obtain the corrected UAV imagery and corrected satellite imagery; standardization correction includes radiometric correction and atmospheric correction.

[0022] Step 22: Using the corrected UAV imagery as a reference, perform image registration on the corrected satellite imagery to obtain the registered satellite imagery.

[0023] Step 23: Determine the absolute value of the resolution difference between the corrected UAV image and the registered satellite image, and obtain the absolute value.

[0024] Step 24: Determine whether the absolute value is greater than the preset difference.

[0025] Step 25: If so, perform super-resolution reconstruction on the registered satellite imagery to obtain the reconstructed satellite imagery, and identify the reconstructed satellite imagery as the image to be fused.

[0026] Specifically, the super-resolution reconstruction employs the Efficient Sub-Pixel Convolutional Neural Network (ESPCN) model. This model directly generates high-resolution images from low-resolution feature maps through sub-pixel convolutional layers, featuring high computational efficiency and good reconstruction quality. It is suitable for real-time or near-real-time super-resolution reconstruction of wide-area satellite imagery to improve its spatial resolution, facilitating subsequent fusion and analysis with high-resolution UAV imagery.

[0027] Step 26: If not, determine the image to be fused from the registered satellite image.

[0028] Step 27: Perform weighted fusion of the image to be fused and the corrected UAV image to obtain the fused image.

[0029] As an optional implementation, step 27 includes the following steps.

[0030] Step 271: Using the corrected UAV imagery as a reference, spatially align the images to be fused to obtain the aligned images to be fused.

[0031] Step 272: Using the image to be fused as a reference, spatially align the corrected UAV image to obtain the aligned UAV image.

[0032] Step 273: Calculate the quality parameters of the aligned image to be fused and the aligned UAV image respectively; the quality parameters include: local structure similarity index and signal-to-noise ratio.

[0033] Step 274: Based on the quality parameters of the aligned image to be fused and the quality parameters of the aligned UAV image, determine the weights of the aligned image to be fused and the aligned UAV image.

[0034] Specifically, for regions with complex textures and dense edges, the aligned UAV images are given higher weights to preserve their rich details; while for regions with uniform color and large coverage, the weights of the aligned images to be fused are appropriately increased to maintain spectral consistency and regional integrity.

[0035] Step 275: Based on the weights of the aligned image to be fused and the aligned UAV image, the aligned image to be fused and the aligned UAV image are fused by a pixel-by-pixel weighted average to obtain the fused image.

[0036] Step 28: Perform geometric correction on the fused image to obtain the preprocessed image.

[0037] Step 3: Using the boundary extraction model, perform boundary identification on the preprocessed images of the monitoring area at two different times to obtain images of the monitoring area with predicted farmland boundaries at two different times.

[0038] The boundary extraction model is obtained by training the DeepLabv3+ model.

[0039] As an optional implementation, the process of determining the boundary extraction model in step 3 includes: Step 31: Acquire sample images of multiple first sample regions and sample images with actual farmland boundaries; Step 32: Build the DeepLabv3+ model; Step 33: Using the sample images of each first sample region as input and the sample images of the corresponding first sample region with real farmland boundaries as output, train the DeepLabv3+ model to obtain the boundary extraction model.

[0040] Step 4: Input the images of the monitoring area at two different times with the predicted farmland boundary into the change detection model to obtain the predicted change in farmland area of ​​the monitoring area.

[0041] The change detection model was trained using Siamese U-Net.

[0042] As an optional implementation, the process of determining the change detection model in step 4 includes: Step 411: Obtain images of multiple second sample regions at two different times with real farmland boundaries and the actual change in farmland area between the two different times; Step 412: Construct Siamese U-Net; Step 413: Using images with real farmland boundaries at two different times for each second sample region as input, and the actual farmland area change between the two different times for the corresponding second sample region as output, train the Siamese U-Net to obtain the change detection model.

[0043] Step 5: Based on the predicted changes in cultivated land area, determine the types of abnormal changes in cultivated land in the monitoring area.

[0044] As an optional implementation, step 5 includes the following: When the predicted change in cultivated land area is less than -5%, the abnormal change type of cultivated land in the monitoring area is determined to be an abnormal reduction in cultivated land. When the predicted change in cultivated land area is within the range of -5% to +5%, the type of abnormal change in cultivated land in the monitoring area is determined to be cultivated land stability. When the predicted change in cultivated land area is greater than +5%, the abnormal change type of cultivated land in the monitoring area is determined to be an abnormal increase in cultivated land.

[0045] Specifically, an abnormal decrease in arable land may be caused by non-agricultural construction, abandonment of farmland, or ecological conversion of farmland. Stable arable land falls within the scope of normal agricultural cultivation or minor adjustments. An abnormal increase in arable land may be caused by illegal reclamation, adjustments to planting structure, or data anomalies.

[0046] In one exemplary embodiment, a computer device is provided, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement a method for determining the type of change in arable land.

[0047] In one exemplary embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements a method for determining the type of change in arable land.

[0048] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements a method for determining the type of change in arable land.

[0049] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 2 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When executed by the processor, the computer program implements a method for determining the type of farmland change.

[0050] Those skilled in the art will understand that Figure 2 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0051] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0052] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0053] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0054] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0055] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for determining the type of change in arable land, characterized in that, The method for determining the type of change in cultivated land includes: Acquire raw images of the monitored area at two different times; the raw images include UAV images and satellite images; The original images of the monitoring area at two different times were preprocessed to obtain the preprocessed images of the monitoring area at two different times. Using a boundary extraction model, boundary identification was performed on the preprocessed images of the monitoring area at two different times, resulting in images of the monitoring area with predicted farmland boundaries at two different times. The boundary extraction model was trained using the DeepLabv3+ model. Images of the monitored area at two different times, with predicted farmland boundaries, are input into the change detection model to obtain the predicted change in farmland area of ​​the monitored area; the change detection model is trained on Siamese U-Net. Based on the predicted changes in cultivated land area, the types of abnormal changes in cultivated land in the monitoring area are determined.

2. The method for determining the type of change in cultivated land according to claim 1, characterized in that, When preprocessing the original images of the monitoring area at two different times to obtain preprocessed images of the monitoring area at two different times, the preprocessing process of the original image at any time includes: Standardization corrections were performed on both UAV and satellite imagery to obtain corrected UAV and satellite imagery; standardization corrections included radiometric correction and atmospheric correction. Using the corrected UAV imagery as a reference, image registration is performed on the corrected satellite imagery to obtain the registered satellite imagery. The absolute value is obtained by determining the absolute value of the resolution difference between the corrected UAV imagery and the registered satellite imagery; Determine if the absolute value is greater than a preset difference; If so, super-resolution reconstruction is performed on the registered satellite imagery to obtain the reconstructed satellite imagery, and the reconstructed satellite imagery is identified as the imagery to be fused. If not, the registered satellite images will be used to determine the images to be fused. The image to be fused and the corrected UAV image are weighted and fused to obtain the fused image; Geometric correction is performed on the fused image to obtain the preprocessed image.

3. The method for determining the type of change in cultivated land according to claim 2, characterized in that, The image to be fused and the corrected UAV image are weighted and fused to obtain the fused image, which includes: Using the corrected UAV imagery as a reference, spatial alignment is performed on the images to be fused to obtain the aligned images to be fused. Using the image to be fused as a reference, the corrected UAV image is spatially aligned to obtain the aligned UAV image; Calculate the quality parameters of the aligned image to be fused and the aligned UAV image separately; the quality parameters include: local structure similarity index and signal-to-noise ratio; Based on the quality parameters of the aligned images to be fused and the quality parameters of the aligned UAV images, the weights of the aligned images to be fused and the weights of the aligned UAV images are determined. Based on the weights of the aligned image to be fused and the aligned UAV image, the aligned image to be fused and the aligned UAV image are fused by a pixel-by-pixel weighted average to obtain the fused image.

4. The method for determining the type of change in cultivated land according to claim 1, characterized in that, The process of determining the boundary extraction model includes: Acquire sample images of multiple first sample regions and sample images with actual farmland boundaries; Build a DeepLabv3+ model; The DeepLabv3+ model is trained by taking the sample images of each first sample region as input and the sample images of the corresponding first sample region with real farmland boundaries as output, to obtain the boundary extraction model.

5. The method for determining the type of change in cultivated land according to claim 1, characterized in that, The process of determining the change detection model includes: Acquire images of multiple second sample regions at two different times with actual farmland boundaries and the actual change in farmland area between the two different times; Constructing Siamese U-Net; Using images of actual farmland boundaries at two different times for each second sample region as input, and the actual change in farmland area between the two different times for the corresponding second sample region as output, Siamese U-Net is trained to obtain a change detection model.

6. The method for determining the type of change in cultivated land according to claim 1, characterized in that, Based on the predicted changes in cultivated land area, the types of abnormal changes in cultivated land in the monitoring area are determined, including: When the predicted change in cultivated land area is less than -5%, the abnormal change type of cultivated land in the monitoring area is determined to be an abnormal reduction in cultivated land. When the predicted change in cultivated land area is within the range of -5% to +5%, the type of abnormal change in cultivated land in the monitoring area is determined to be cultivated land stability. When the predicted change in cultivated land area is greater than +5%, the abnormal change type of cultivated land in the monitoring area is determined to be an abnormal increase in cultivated land.

7. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the method for determining the type of change of cultivated land as described in any one of claims 1-6.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the method for determining the type of change of cultivated land as described in any one of claims 1-6.

9. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the method for determining the type of change of cultivated land as described in any one of claims 1-6.

Images