A farmland classification method and device, electronic equipment and storage medium

Abstract

The application provides a farmland classification method and device, electronic equipment and storage medium. The method comprises the following steps: determining mask data corresponding to a target region according to position information corresponding to the target region; determining a flight region and flight parameters according to the mask data and device information of a flight device; covering the target region by flight; controlling the flight device to perform a flight task in the flight region based on the flight parameters, and acquiring image data and point cloud data collected by the flight device; identifying a tree land in the target region according to the image data and the point cloud data, and removing the tree land from a farmland land corresponding to the target region to obtain a farmland land; and classifying the farmland land based on a preset farmland classification standard and the image data. Through multi-source data fusion, the application improves the accuracy of tree land identification, realizes the precision and standardization of farmland classification, saves labor cost, and improves the efficiency and accuracy of farmland classification.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a method, apparatus, electronic device and storage medium for classifying farmland. Background Technology

[0002] High-standard farmland, as high-quality arable land characterized by drought and flood resistance and high and stable yields, requires precise classification as a key step in optimizing the allocation of agricultural resources and carrying out scientific production planning. Traditional classification methods mainly rely on manual field surveys and marking, which have drawbacks such as low efficiency, high time consumption, and poor accuracy, making them difficult to meet the requirements of large-scale, refined farmland management. Summary of the Invention

[0003] This application provides a method, apparatus, electronic device, and storage medium for farmland classification, in order to solve the problems of low efficiency and long time consumption in the prior art of farmland classification by manual field survey and marking.

[0004] In a first aspect, embodiments of this application provide a method for classifying farmland, the method comprising: Based on the location information corresponding to the target region to be classified, determine the mask data corresponding to the target region; Based on the mask data and the equipment information of the flight equipment, the flight area and flight parameters are determined; the flight area covers the target area. The flight equipment is controlled to perform flight missions in the flight area based on the flight parameters, and to acquire image data and point cloud data collected by the flight equipment. Based on the image data and point cloud data, tree plots in the target area are identified, and the tree plots are removed from the corresponding cultivated land plots in the target area to obtain farmland plots; Based on the preset farmland classification standards and the image data, the farmland plots are classified.

[0005] Secondly, embodiments of this application also provide a farmland sorting device, the device comprising: The first determining module is used to determine the mask data corresponding to the target region based on the location information corresponding to the target region to be classified. The second determining module is used to determine the flight area and flight parameters based on the mask data and the equipment information of the flight equipment; the flight area covers the target area; The control module is used to control the flight equipment to perform flight missions in the flight area based on the flight parameters, and to acquire image data and point cloud data collected by the flight equipment. The first processing module is used to identify tree plots in the target area based on the image data and point cloud data, and remove the tree plots from the cultivated land plots corresponding to the target area to obtain farmland plots; The classification module is used to classify the farmland plots based on preset farmland classification standards and the image data.

[0006] Thirdly, embodiments of this application also provide an electronic device, which includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the above-described farmland classification method.

[0007] Fourthly, embodiments of this application also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described farmland classification method.

[0008] The embodiments of this application include at least the following technical effects: The technical solution of this application, through mask data of the target area and combined with flight equipment information, determines the flight area and flight parameters, thereby controlling the flight equipment to perform flight tasks, acquiring image data and point cloud data. Using data fusion technology, tree plots within the target area are identified and removed, thus obtaining farmland plots. These farmland plots are then classified according to preset farmland classification standards. This application utilizes mask data and flight planning to ensure the accuracy of data acquisition; through multi-source data fusion, it improves the accuracy of tree plot identification, achieving precise and standardized farmland classification, saving labor costs, and improving the efficiency and accuracy of farmland classification. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below.

[0010] Figure 1 This is a flowchart illustrating the farmland classification method provided in the embodiments of this application; Figure 2 This is a schematic diagram of the farmland sorting device provided in the embodiments of this application; Figure 3 A block diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0011] 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, 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.

[0012] It should be understood that the phrase "one embodiment" or "an embodiment" throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of this application. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments.

[0013] In the various embodiments of this application, it should be understood that the sequence number of each process described below does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0014] In related technologies, high-standard farmland, as high-quality arable land with characteristics of drought and flood resistance, high and stable yields, requires precise classification as a key step in optimizing the allocation of agricultural resources and carrying out scientific production planning. Traditional classification methods mainly rely on manual field surveys and marking, which have drawbacks such as low efficiency, high time consumption, and poor accuracy, making it difficult to meet the requirements of large-scale refined farmland management.

[0015] Based on this, in order to solve the problems of low efficiency and long time consumption in the existing technology of farmland classification by manual survey and marking, this application provides a farmland classification method, device, electronic device and storage medium. By fusing multi-source data, the accuracy of tree plot identification and farmland classification is improved, the precision and standardization of farmland classification are realized, labor costs are saved and the efficiency and accuracy of farmland classification are improved.

[0016] like Figure 1 As shown in the embodiment of this application, a method for classifying farmland is provided, the method comprising: Step 101: Determine the mask data corresponding to the target region based on the location information of the target region to be classified.

[0017] The farmland classification method provided in this application first determines the target area to be classified and obtains the location information of the target area. The location information of the target area to be classified can come from standardized data sources, such as the Third National Land Survey data (i.e., the third survey data), land spatial planning vector data, etc. These data undergo full-process quality control and include precise corner coordinates of the target area (such as the latitude and longitude of the four vertices of a rectangular area, and the coordinates of the key boundary inflection points of an irregular area), regional boundary type, land ownership, and other related information.

[0018] Furthermore, based on the location information of the target area, boundary drawing and data transformation are performed using Geographic Information System (GIS) tools (such as ArcGIS and QGIS) to generate mask data. Mask data is a type of binary or vector boundary data that can mark valid areas and mask invalid areas, clearly defining the scope of subsequent flight and data processing operations. In addition, by defining valid areas in advance using mask data, invalid data can be directly eliminated later, reducing the hardware resource consumption for data storage and computation, and improving overall process efficiency.

[0019] For example, City A plans to classify 1,000 mu (approximately 67 hectares) of high-standard farmland within its jurisdiction. First, it obtains the third national land survey data for the area from the local natural resources department. This data clearly marks the latitude and longitude coordinates of the eight corner points of the farmland area. Then, using ArcGIS software, it generates vector mask data that perfectly matches the target area, clearly marking the boundary of the 1,000 mu of farmland.

[0020] Step 102: Determine the flight area and flight parameters based on the mask data and the equipment information of the flight equipment; the flight area covers the target area.

[0021] This application's embodiments classify farmland based on multi-source data collected by aerial equipment. Therefore, it is also necessary to obtain the equipment information of this aerial equipment, which includes at least the equipment type, onboard sensor parameters, and flight performance limitations. The equipment type specifically refers to the type of aerial equipment. The sensors onboard the aerial equipment include image acquisition devices and radar devices. The onboard sensor parameters include at least the pixel resolution and lens focal length of the image acquisition device, and the detection range and elevation accuracy of the radar device. Flight performance limitations include at least maximum speed, endurance, and maximum flight altitude.

[0022] Based on the mask data corresponding to the target area and the equipment information of the flight equipment, the parameters during the flight process are determined. These parameters include the flight area and flight parameters. The flight area must be defined within the target area as shown in the mask data, with a buffer zone reserved. That is, the flight area must completely cover the target area and be larger than it. During flight, there may be slight flight deviations and turning radius limitations; setting a buffer zone ensures that subsequent image and point cloud data are acquired without blind spots. Flight parameters include at least flight altitude, airspeed, image resolution, and overlap.

[0023] Specifically, flight altitude and image resolution are negatively correlated; lower altitude results in higher resolution but smaller coverage. A balance must be struck based on classification requirements. For example, to identify minute irrigation facilities within farmland, resolution needs to be increased, and flight altitude lowered accordingly. Conversely, to differentiate large-scale farmland types, resolution can be appropriately reduced, and flight altitude increased to minimize flight time. When determining flight speed and overlap, terrain adaptation is crucial. Plains with gentle terrain can be configured with higher speeds and lower overlaps; hilly and mountainous areas with significant terrain undulations require lower speeds and higher overlaps to ensure seamless image stitching and complete point cloud data. If the flight equipment has a short flight time, flight tasks should be split according to the flight area, and takeoff and landing points should be planned rationally to avoid data acquisition interruptions due to insufficient flight time. Flight altitude must match the elevation accuracy of the airborne radar to ensure that elevation data meets the requirements for tree and plot identification. Reasonable flight parameter planning can reduce unnecessary flight mileage, shorten operation time, and lower equipment energy consumption and labor costs, making it particularly suitable for large-area farmland classification projects.

[0024] Step 103: Control the flight equipment to perform flight missions in the flight area based on the flight parameters, and acquire image data and point cloud data collected by the flight equipment.

[0025] After determining the flight parameters and flight area, the flight control equipment executes flight missions within that area based on these parameters. Specifically, the flight parameters and flight area are imported into the software accompanying the flight control equipment to generate flight paths, typically in a grid pattern to ensure complete coverage without overlap or omissions. After takeoff, the flight control equipment automatically flies along the generated flight path.

[0026] Image acquisition equipment (such as orthophoto cameras) on the flight equipment continuously captures images of the flight area according to the set overlap and resolution; radar equipment on the flight equipment emits laser pulses, receives ground reflection signals, and generates point cloud data. The point cloud data includes the planar coordinates and elevation information of each point, which can reflect the terrain undulations and vegetation height.

[0027] After the flight of the aerial equipment is completed, the image data and point cloud data collected by the aerial equipment are acquired.

[0028] Step 104: Based on the image data and point cloud data, identify tree plots in the target area and remove the tree plots from the corresponding cultivated land plots in the target area to obtain farmland plots.

[0029] This application embodiment identifies tree plots in a target area by collecting multi-source data, namely image data and point cloud data.

[0030] Specifically, when identifying tree plots, the image data and point cloud data are first aligned to establish a one-to-one correspondence between image pixels and point cloud data points. For each pixel in the image data, a point with the same coordinates is found in the point cloud data, its elevation value is extracted and assigned to that pixel, generating fused image data that combines visual and elevation features. Further, the fused image data is preliminarily screened to identify areas where the elevation dispersion is greater than a preset dispersion threshold and the vegetation index is greater than a preset index threshold, thus identifying these as candidate tree plots. Finally, the elevation contours of the candidate areas are extracted. Based on the irregular convex shape of the tree canopy contour, which differs from the regular geometric shape of buildings, candidate tree plots that satisfy the tree canopy contour are identified as tree plots.

[0031] After obtaining the tree plots, the boundaries of the tree plots are overlaid with the boundaries of the cultivated land plots in the target area. The range of the tree plots is removed, and the remaining cultivated land area is retained to obtain the farmland plots. At the same time, the vector boundary data of the tree plots is output for subsequent traceability.

[0032] Step 105: Classify the farmland plots based on the preset farmland classification standards and the image data.

[0033] Based on the image data corresponding to farmland plots, feature extraction and matching are performed through image recognition algorithms to obtain various features of farmland plots, such as spectral features, texture features, and shape features. The extracted features are compared with the discrimination indicators in the preset farmland classification standards to achieve the classification of farmland plots.

[0034] In this embodiment, by combining mask data of the target area with flight equipment information, the flight area and flight parameters are determined, thereby controlling the flight equipment to perform flight missions, acquiring image data and point cloud data. Using data fusion technology, tree plots within the target area are identified and removed, resulting in farmland plots. These farmland plots are then classified according to preset farmland classification standards. This application utilizes mask data and flight planning to ensure the accuracy of data acquisition; through multi-source data fusion, the accuracy of tree plot identification is improved, achieving precise and standardized farmland classification, saving labor costs, and improving the efficiency and accuracy of farmland classification.

[0035] In an optional embodiment of this application, determining the flight area and flight parameters based on the mask data and the equipment information of the flight equipment includes: Based on the mask data and the device information corresponding to the image acquisition device mounted on the flight equipment, the first flight area and the first flight parameters are determined. Based on the mask data and the equipment information corresponding to the radar equipment mounted on the flight equipment, the second flight area and the second flight parameters are determined.

[0036] In determining the flight area and flight parameters, this application embodiment determines the flight area and flight parameters separately based on the differences in performance limitations and data acquisition requirements of the image acquisition equipment and radar equipment mounted on the flight equipment. This ensures that the data collected by the two types of equipment can not only fully cover the target area, but also give full play to their respective performance advantages, providing a high-quality data foundation for subsequent multi-source data fusion and tree and land plot identification.

[0037] Specifically, regarding the image acquisition equipment mounted on the flight equipment, this equipment can be an orthophoto camera or a multispectral camera. First, the equipment information of the image acquisition equipment is determined, including at least parameters such as sensor type, pixel resolution, lens focal length, field of view, imaging speed, and data storage capacity. Further, based on the mask data and the corresponding equipment information of the image acquisition equipment, the first flight area and the first flight parameters are determined.

[0038] In determining the initial flight area, the target area defined by the mask data is used as the core. A buffer area is reserved, taking into account the field of view and lens coverage of the image acquisition equipment, to avoid missed images or blurred images of the target area due to flight deviation or lens edge distortion. This ensures that the initial flight area completely covers the mask area (i.e., the target area), while avoiding excessively large areas that could cause data redundancy. When determining the initial flight parameters, the flight altitude must match the pixel resolution and imaging accuracy requirements, the flight speed must adapt to the imaging speed and overlap requirements, the image resolution must meet the detailed recognition requirements of subsequent classification, and the overlap must ensure the integrity of image stitching.

[0039] Specifically, regarding the radar equipment carried by the aircraft, this radar equipment can be lidar or synthetic aperture radar. First, the equipment information of the radar equipment is determined, including at least parameters such as radar type, detection range, elevation measurement accuracy, point cloud density, transmission frequency, and signal penetration capability. Further, based on the mask data and the radar equipment information, the second flight area and second flight parameters are determined.

[0040] In determining the second flight area, the target area defined by the mask data is used as the core. A buffer zone is reserved, taking into account the radar's detection range and signal propagation characteristics, to ensure complete radar signal coverage of the target area. When determining the second flight parameters, the flight altitude must match the radar's elevation measurement accuracy and point cloud density requirements; the flight speed must match the radar's transmission frequency to ensure uniform point cloud data coverage; the point cloud density threshold must meet the elevation dispersion analysis requirements for subsequent tree and land parcel identification; and the scanning angle must adapt to terrain undulations to avoid missed or repeated scans.

[0041] The above-mentioned implementation scheme of this application is based on mask data and combines the respective performance characteristics of the image acquisition equipment and radar equipment on the flight equipment to customize the flight area and flight parameters respectively, adapting to the data acquisition needs of the two types of equipment. This allows the image acquisition equipment to accurately capture the visual features of the target area and the radar equipment to accurately obtain elevation information, ensuring the high quality and integrity of the image data and point cloud data, which facilitates subsequent processes such as multi-source data fusion, tree and plot identification, and farmland classification.

[0042] In an optional embodiment of this application, controlling the flight equipment to perform flight missions in the flight area based on the flight parameters, and acquiring image data and point cloud data collected by the flight equipment, includes: The flight equipment is controlled to perform a first flight mission in the first flight area based on the first flight parameters, and to acquire image data collected by the image acquisition device mounted on the flight equipment. The flight equipment is controlled to perform a second flight mission in the second flight area based on the second flight parameters, and point cloud data collected by the radar equipment mounted on the flight equipment is acquired.

[0043] This application embodiment customizes different flight areas and flight parameters for image acquisition equipment and radar equipment, and breaks down the flight mission into targeted independent operations to ensure that the two types of equipment collect data under their respective optimal operating conditions, avoid mutual interference, maximize the performance of the equipment, and provide high-quality and highly adaptable basic data for subsequent multi-source data fusion, tree and plot identification and farmland separation.

[0044] For image acquisition equipment, the ground control system of the flight equipment imports the determined boundary coordinates of the first flight area and the first flight parameters to generate a dedicated flight path. The flight equipment automatically flies according to the flight path to perform the first flight mission. During the flight, it strictly follows the first flight parameters. The image acquisition equipment continuously takes pictures at the set imaging interval and resolution, records the visual information of the target area in real time, and generates continuous image data.

[0045] For radar equipment, the ground control system of the flight equipment imports the determined boundary coordinates of the second flight area and the second flight parameters to generate a dedicated flight path. The flight equipment flies automatically according to the flight path to perform the second flight mission. During the flight, it strictly follows the second flight parameters. The radar equipment continuously transmits detection signals and receives ground reflected waves to generate point cloud data containing plane coordinates and elevation information in real time.

[0046] The above-described implementation scheme of this application, by splitting flight missions, allows flight equipment to perform dedicated flight missions in corresponding flight areas based on the first and second flight parameters. This enables image acquisition equipment to focus on capturing visual features of the target area and radar equipment to focus on collecting elevation information, avoiding mutual interference between equipment, giving full play to the performance advantages of various equipment, and ensuring the high quality and integrity of image data and point cloud data. This provides reliable data support for subsequent processes such as multi-source data fusion, tree and plot identification, and farmland classification.

[0047] In an optional embodiment of this application, before identifying tree patches in the target area based on the image data and point cloud data, the method includes: Based on the mask data, the image data and the point cloud data are respectively processed to extract the target image data and target point cloud data corresponding to the target area.

[0048] Since the coverage area of ​​the raw data collected by aerial flight is larger than the target area and includes surrounding redundant areas, this embodiment of the application performs extraction processing on the image data and point cloud data based on mask data before fusing the image data and point cloud data. That is, by aligning the spatial boundary of the mask data with the coordinate system of the image data and point cloud data, the data part that completely overlaps with the mask area is filtered and retained, and irrelevant redundant data is removed, thereby obtaining the target image data and target point cloud data corresponding to the target area.

[0049] Specifically, during the extraction process, the first step is to use the same coordinate system for the mask data, image data, and point cloud data. This ensures a consistent spatial reference and prevents the extracted data from deviating from the target area due to coordinate misalignment. Next, the unified coordinate mask data is used as a spatial filtering template and overlaid on the image data. Using the spatial cropping function of a geographic information system (GIS) tool, only image pixels within the mask boundaries are retained, while all redundant pixels outside the boundaries are removed. During extraction, the visual characteristics of the image data must be preserved, ensuring that the color and texture information of the pixels are not distorted. The final result is target image data containing only the visual information of the target area. Again, using the unified coordinate mask data as a reference, the spatial filtering function of the point cloud processing software is used to retrieve all point cloud data points whose planar coordinates fall within the mask boundaries. The planar coordinates and elevation information of these points are retained, while invalid point clouds outside the boundaries are removed. After extraction, the three-dimensional structural integrity of the point cloud data must be preserved, ensuring that the elevation information and point cloud density do not change abnormally. The final result is target point cloud data covering only the target area.

[0050] The above-described implementation scheme of this application uses mask data as a spatial boundary reference to extract and process the image data and point cloud data collected by flight, respectively, and removes redundant data outside the target area to obtain target image data and target point cloud data that focus only on the target area. This avoids interference from irrelevant data in subsequent processes and reduces data processing costs.

[0051] In an optional embodiment of this application, identifying tree patches in the target area based on the image data and point cloud data includes: For each pixel in the target image data, obtain the elevation value of the corresponding planar coordinates of the pixel in the point cloud data; The elevation value is assigned to the pixel to obtain the target pixel. Based on the target pixel corresponding to each pixel in the target image data, the fused image data is determined; Based on the fused image data, tree patches in the target area are identified.

[0052] In this embodiment of the application, when identifying tree plots within a target area based on image data and point cloud data, the image data and point cloud data are first processed separately to remove redundant data outside the target area, resulting in target image data and target point cloud data focused on the target area. Further, a correspondence is established between pixels in the target image data and elevation values ​​in the target point cloud data, assigning the elevation information from the target point cloud data to each pixel in the target image data, generating fused image data that combines visual and elevation features. Finally, given that tree plots have greater elevation dispersion and more pronounced canopy outlines compared to ordinary farmland, tree plots are identified from the target area based on the fused image data.

[0053] Specifically, for each pixel in the target image data, its corresponding planar coordinates are first extracted using Geographic Information System (GIS) tools. These coordinates directly correlate with the pixel's location in actual geographic space. Next, spatial retrieval is performed in the target point cloud data to find point cloud data points that perfectly match or are closest to the original planar coordinates. Once a matching point cloud data point is found, its corresponding elevation value is extracted and used as the associated elevation data for the current pixel. If a pixel cannot find a directly matching point cloud data point due to insufficient point cloud data density or edge coverage issues, neighborhood interpolation can be used to supplement elevation information, avoiding missing elevation data and ensuring the integrity of the subsequent fused image. Then, the extracted elevation values ​​are bound to the corresponding image pixels, enriching the pixel's feature dimensions. This adds elevation attributes to pixels that originally only contained visual features such as color and texture, resulting in target pixels with both visual and elevation features. Finally, data integration is performed, reorganizing all target pixels according to the spatial arrangement of the original image to construct fused image data with both visual and elevation features. The fused image data has dual characteristics. At the visual level, it fully preserves the visual information of the original image, such as color, texture, and plot boundaries, and can intuitively distinguish the visual differences of different land cover types. At the elevation level, each pixel is associated with a clear elevation value, and the terrain undulations and vegetation height differences of the target area can be intuitively presented through elevation rendering.

[0054] After acquiring the fused image data, tree plots were identified from the target area by performing elevation and visual feature analysis.

[0055] The above-described implementation scheme of this application establishes a coordinate association between the target image data pixels and the target point cloud data, assigns elevation values ​​to the pixels to generate target pixels, and then constructs fused image data that combines visual and elevation features. Based on the multi-feature identification of the fused image data, the accuracy of tree plot identification in the target area is improved, and precise data support is provided for subsequent processes such as tree plot removal and farmland classification.

[0056] In an optional embodiment of this application, identifying tree patches in the target area based on the fused image data includes: The target region is divided into multiple sub-regions; For each sub-region, the standard deviation of elevation value and vegetation index corresponding to the sub-region are determined based on the fused image data; The sub-regions whose elevation standard deviation is greater than a preset elevation threshold and whose vegetation index is greater than a preset index threshold are identified as candidate sub-regions; Based on the fused image data, the contour shape corresponding to the candidate sub-region is determined; The continuous plot of land whose area is greater than a preset area threshold is determined as the tree plot. The candidate sub-regions whose outline shape conforms to the characteristics of trees are identified as the tree plot.

[0057] In this embodiment of the application, when identifying tree plots in a target area based on fused image data, tree plots, compared to ordinary farmland, have the characteristics of high elevation dispersion, pure vegetation attributes, unique outline morphology, and continuous distribution. By dividing the target area into sub-regions to reduce the complexity of analysis, the standard deviation of elevation values ​​and vegetation index of each sub-region are extracted one by one. Candidate sub-regions are determined by double threshold screening, and misjudged areas are eliminated by outline morphology verification and continuous area verification, and finally tree plots are identified.

[0058] Specifically, when segmenting the target region, the basic unit size is first determined to ensure that the sub-region can completely contain the features of a single tree or a small patch of trees, and that the size is not so large that it contains both trees and farmland. Then, based on this basic unit size, the target region is segmented into multiple sub-regions, and a unique identifier is assigned to each sub-region. The spatial boundary coordinates of each sub-region are recorded to ensure that subsequent feature extraction, filtering, and verification can be accurately linked to the corresponding region, avoiding data misalignment.

[0059] Furthermore, for each sub-region, the standard deviation of elevation values ​​and the vegetation index are determined. The fused image data includes the elevation value of each pixel. For a single sub-region, the elevation values ​​of all pixels within that sub-region are extracted, and the standard deviation of the elevation values ​​for that sub-region is calculated through statistical analysis. A larger standard deviation indicates more dramatic elevation fluctuations within the sub-region, suggesting a higher likelihood of tree presence; a smaller standard deviation indicates gentler terrain, suggesting more likely ordinary farmland. Simultaneously, based on the visual characteristics of the fused image data, the vegetation index for each sub-region is calculated. The vegetation index reflects the degree of vegetation cover and vegetation type. For a single sub-region, the color channel values ​​of all pixels within that sub-region are extracted from the fused image data, and then the vegetation index for that sub-region is determined based on the vegetation index calculation formula.

[0060] After obtaining the standard deviation of elevation values ​​and vegetation index corresponding to each sub-region of the target area, it is judged whether each sub-region meets the preset candidate conditions. Based on the fact that tree plots must simultaneously meet the requirements of high elevation dispersion and pure vegetation attributes, both of which are indispensable. If only the standard deviation of elevation values ​​meets the standard, it may be a non-vegetated area such as a terrain protrusion or a building. If only the vegetation index meets the standard, it may be a herbaceous area such as a tall crop. Only when both conditions are met can it be initially determined as a candidate area. In the embodiments of this application, the preset candidate conditions are that the standard deviation of elevation values ​​is greater than the preset elevation threshold and the vegetation index is greater than the preset index threshold.

[0061] After identifying candidate sub-regions from the target region, the boundary contours of the candidate sub-regions are extracted from the fused image data. The inflection points and curve directions of the contours are determined by the edge detection algorithm. The geometric features of the contours, such as the degree of irregularity, edge smoothness, and number of protrusions, are analyzed. Finally, the extracted contour shapes are compared with the preset tree contour features to determine whether they conform to the tree morphological features.

[0062] Finally, after obtaining the outline morphology of each candidate sub-region, the spatial distribution of candidate sub-regions in the target area whose outline morphology matches the characteristics of trees is analyzed. Adjacent and connected areas are identified as continuous plots, and the area of ​​each continuous plot is calculated. The area of ​​the continuous plot is compared with a preset area threshold. This preset area threshold is set according to the farmland classification requirements to ensure that single trees or small, scattered trees are excluded, while large areas of trees are retained. When the area of ​​a continuous plot is greater than the preset area threshold, the continuous plot is identified as a tree plot.

[0063] The above-mentioned implementation scheme of this application is based on fused image data. Through a progressive process of segmenting the target area, extracting the standard deviation of the elevation value and vegetation index of the sub-region, screening candidate sub-regions with dual thresholds, verifying the contour morphology and verifying the area of ​​continuous plots, it realizes multi-dimensional feature cross-validation. It can effectively eliminate interference factors such as terrain protrusions, buildings, and scattered trees, and improve the accuracy and consistency of tree plot identification.

[0064] In an optional embodiment of this application, after acquiring the image data collected by the flight equipment, the method further includes: Obtain the weather parameters during the execution of the flight mission; The image data is corrected based on the weather parameters.

[0065] In this embodiment of the application, after acquiring the image data collected by the flight equipment, it is considered that the quality of the image data is directly affected by the weather conditions during flight. For example, factors such as lighting, atmospheric humidity, and precipitation can cause problems such as abnormal exposure, fogging, and color distortion in the image. The real-time weather parameters during the flight mission can be acquired, and the image data can be corrected according to the weather parameters, such as adjusting the brightness, contrast, and color balance of the image, in order to restore the true visual characteristics of the image.

[0066] Specifically, when acquiring weather data to determine weather parameters, the data collection time must be completely synchronized with the aerial image capture time, and the collection location must be consistent with the area to be flew, to avoid mismatches in weather conditions due to time differences.

[0067] When correcting image data based on weather parameters, a mapping relationship needs to be established beforehand. This involves determining the corresponding image distortion type based on common weather parameter components and then matching the appropriate correction algorithm. For example, strong light weather corresponds to overexposed images, requiring exposure compensation and highlight suppression algorithms; hazy weather corresponds to low contrast and foggy blurring, requiring defogging algorithms; low light weather corresponds to underexposed images, requiring brightness enhancement and noise reduction algorithms; and precipitation weather corresponds to blurred images and dull colors, requiring sharpening and color enhancement algorithms. Then, based on this mapping relationship, a correction algorithm matching the currently acquired weather parameters is determined, and the image data is processed using this algorithm to achieve the corrected image data.

[0068] The above-mentioned implementation scheme of this application obtains weather parameters during the flight mission and performs targeted correction processing for image distortion caused by weather factors. It effectively eliminates image defects such as overexposure, fog blurring, and color distortion caused by strong light, haze and other weather conditions, restores the true visual features and key details of the image, and improves the efficiency and accuracy of subsequent multi-source data fusion, tree and plot identification and farmland classification.

[0069] In an optional embodiment of this application, determining the mask data corresponding to the target region based on the location information corresponding to the target region to be classified includes: Based on the location information, determine the corner coordinates of the target area; The mask data is determined based on the corner coordinate information.

[0070] In this embodiment of the application, the location information of the target area can be derived from standardized data sources, such as the data from the Third National Land Survey (i.e., the Third Survey data), land spatial planning vector data, etc. These data undergo full-process quality control and include the precise corner coordinates of the target area (such as the latitude and longitude of the four vertices of a rectangular area, the coordinates of the key boundary inflection points of an irregular area), the type of regional boundary, land ownership and other related information.

[0071] Based on the location information of the target area, the corner coordinates of the target area can be determined. These corner coordinates represent the key inflection points of the target area's boundary. When determining the corner coordinates, the inflection point coordinates can be extracted in a clockwise or counterclockwise order along the boundary of the target area to ensure that the inflection points can completely outline the true contour of the target area without any omissions or redundancies.

[0072] Furthermore, using Geographic Information System (GIS) tools (such as ArcGIS and QGIS), the coordinates of all corner points are sequentially connected according to their spatial order to form a closed polygon boundary, which represents the spatial extent of the target area, and mask data is generated. This mask data, in digital form, clearly defines the effective extent of the target area, providing a unified spatial constraint for subsequent flight area planning, data acquisition, and data processing, and avoiding interference from invalid areas.

[0073] The above-described implementation scheme of this application, based on the location information of the target area, determines the precise corner coordinate information by transformation, and then generates mask data by constructing a closed polygon boundary based on the corner coordinate information, realizes the precise digital definition of the target area, provides a unified and accurate spatial reference for subsequent flight area planning, data collection, data processing and other links, reduces the cost of invalid operations and redundant data processing, and ensures the consistency and traceability of data.

[0074] The above describes the farmland classification method provided in the embodiments of this application. The farmland classification device provided in the embodiments of this application will be described below with reference to the accompanying drawings.

[0075] like Figure 2 As shown, this embodiment of the invention also provides a farmland sorting device, the device comprising: The first determining module 210 is used to determine the mask data corresponding to the target region based on the location information corresponding to the target region to be classified. The second determining module 220 is used to determine the flight area and flight parameters based on the mask data and the equipment information of the flight equipment; the flight area covers the target area; The control module 230 is used to control the flight equipment to perform flight missions in the flight area based on the flight parameters, and to acquire image data and point cloud data collected by the flight equipment; The first processing module 240 is used to identify tree plots in the target area based on the image data and point cloud data, and remove the tree plots from the cultivated land plots corresponding to the target area to obtain farmland plots; The classification module 250 is used to classify the farmland plots based on the preset farmland classification standards and the image data.

[0076] Optionally, the second determining module includes: The first determining submodule is used to determine the first flight area and the first flight parameters based on the mask data and the device information corresponding to the image acquisition device mounted on the flight equipment. The second determining submodule is used to determine the second flight area and the second flight parameters based on the mask data and the equipment information corresponding to the radar equipment mounted on the flight equipment.

[0077] Optionally, the control module includes: The first control submodule is used to control the flight equipment to perform a first flight mission in the first flight area based on the first flight parameters, and to acquire image data collected by the image acquisition device mounted on the flight equipment. The second control submodule is used to control the flight equipment to perform a second flight mission in the second flight area based on the second flight parameters, and to acquire point cloud data collected by the radar equipment mounted on the flight equipment.

[0078] Optionally, before identifying tree patches in the target area based on the image data and point cloud data, the device includes: The second processing module is used to perform extraction processing on the image data and the point cloud data based on the mask data to obtain target image data and target point cloud data corresponding to the target area.

[0079] Optionally, the first processing module includes: The acquisition submodule is used to acquire the elevation value of the planar coordinates of each pixel in the target image data in the point cloud data. The processing submodule is used to assign the elevation value to the pixel point to obtain the target pixel point; The third determining submodule is used to determine the fused image data based on the target pixel corresponding to each pixel in the target image data; The identification submodule is used to identify tree plots in the target area based on the fused image data.

[0080] Optionally, the recognition submodule includes: A segmentation unit is used to segment the target region into multiple sub-regions; The first determining unit is used to determine the standard deviation of the elevation value and the vegetation index corresponding to each sub-region based on the fused image data. The second determining unit is used to determine the sub-regions where the standard deviation of the elevation value is greater than a preset elevation threshold and the vegetation index is greater than a preset index threshold as candidate sub-regions. The third determining unit is used to determine the contour shape corresponding to the candidate sub-region based on the fused image data; The fourth determining unit is used to determine the continuous plot of land whose area is greater than a preset area threshold, which is composed of candidate sub-regions whose outline shape conforms to the characteristics of trees, as the tree plot.

[0081] Optionally, after acquiring the image data collected by the aerial equipment, the device further includes: The acquisition module is used to acquire weather parameters when the flight mission is performed; The third processing module is used to correct the image data according to the weather parameters.

[0082] Optionally, the first determining module includes: The fourth determining submodule is used to determine the corner coordinates of the target area based on the location information. The fifth determining submodule is used to determine the mask data based on the corner coordinate information.

[0083] The farmland classification device provided in this application determines the flight area and flight parameters by combining mask data of the target area with flight equipment information, thereby controlling the flight equipment to perform flight tasks, acquiring image data and point cloud data. Using data fusion technology, it identifies and removes tree plots within the target area, thus obtaining farmland plots, which are then classified according to preset farmland classification standards. This application utilizes mask data and flight planning to ensure the accuracy of data acquisition; through multi-source data fusion, it improves the accuracy of tree plot identification, achieving precise and standardized farmland classification, saving labor costs, and improving the efficiency and accuracy of farmland classification.

[0084] As the device embodiment is basically similar to the method embodiment, the description is relatively simple, and relevant parts can be found in the description of the method embodiment.

[0085] This application also provides an electronic device, including: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the various processes of the above-described farmland classification method embodiments and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0086] For example, Figure 3 A schematic diagram of the physical structure of an electronic device is shown. (For example...) Figure 3As shown, the electronic device may include: a processor 310, a communication interface 320, a memory 330, and a communication bus 340, wherein the processor 310, the communication interface 320, and the memory 330 communicate with each other through the communication bus 340. The processor 310 can call logical instructions in the memory 330, and the processor 310 is used to perform the following steps: determining the mask data corresponding to the target area based on the location information corresponding to the target area to be classified; determining the flight area and flight parameters based on the mask data and the equipment information of the flight equipment; the flight area covers the target area; controlling the flight equipment to perform flight tasks in the flight area based on the flight parameters, and acquiring image data and point cloud data collected by the flight equipment; identifying tree plots in the target area based on the image data and point cloud data, and removing the tree plots from the cultivated land plots corresponding to the target area to obtain farmland plots; classifying the farmland plots based on a preset farmland classification standard and the image data. The processor 310 can also execute other schemes in the embodiments of this application, which will not be further described here.

[0087] Furthermore, the logical instructions in the aforementioned memory 330 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application.

[0088] This application also provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the various processes of the above-described farmland classification method embodiments and achieves the same technical effects. To avoid repetition, it will not be described again here. The computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0089] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0090] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0091] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

[0092] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this application can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0093] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0094] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0095] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0096] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0097] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0098] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for classifying farmland, characterized in that, The method includes: Based on the location information corresponding to the target region to be classified, determine the mask data corresponding to the target region; Based on the mask data and the equipment information of the flight equipment, the flight area and flight parameters are determined; the flight area covers the target area. The flight equipment is controlled to perform flight missions in the flight area based on the flight parameters, and to acquire image data and point cloud data collected by the flight equipment. Based on the image data and point cloud data, tree plots in the target area are identified, and the tree plots are removed from the corresponding cultivated land plots in the target area to obtain farmland plots; Based on the preset farmland classification standards and the image data, the farmland plots are classified.

2. The farmland classification method according to claim 1, characterized in that, Based on the mask data and the equipment information of the flight equipment, the flight area and flight parameters are determined, including: Based on the mask data and the device information corresponding to the image acquisition device mounted on the flight equipment, the first flight area and the first flight parameters are determined. Based on the mask data and the equipment information corresponding to the radar equipment mounted on the flight equipment, the second flight area and the second flight parameters are determined.

3. The farmland classification method according to claim 2, characterized in that, The flight control equipment is controlled to perform flight missions within the flight area based on the flight parameters, and to acquire image data and point cloud data collected by the flight control equipment, including: The flight equipment is controlled to perform a first flight mission in the first flight area based on the first flight parameters, and to acquire image data collected by the image acquisition device mounted on the flight equipment. The flight equipment is controlled to perform a second flight mission in the second flight area based on the second flight parameters, and point cloud data collected by the radar equipment mounted on the flight equipment is acquired.

4. The farmland classification method according to claim 1, characterized in that, Before identifying tree patches in the target area based on the image data and point cloud data, the method includes: Based on the mask data, the image data and the point cloud data are respectively processed to extract the target image data and target point cloud data corresponding to the target area.

5. The farmland classification method according to claim 4, characterized in that, Based on the image data and point cloud data, identify tree patches in the target area, including: For each pixel in the target image data, obtain the elevation value of the corresponding planar coordinates of the pixel in the point cloud data; The elevation value is assigned to the pixel to obtain the target pixel. Based on the target pixel corresponding to each pixel in the target image data, the fused image data is determined; Based on the fused image data, tree patches in the target area are identified.

6. The farmland classification method according to claim 5, characterized in that, Based on the fused image data, tree patches in the target area are identified, including: The target region is divided into multiple sub-regions; For each sub-region, the standard deviation of elevation value and vegetation index corresponding to the sub-region are determined based on the fused image data; The sub-regions whose elevation standard deviation is greater than a preset elevation threshold and whose vegetation index is greater than a preset index threshold are identified as candidate sub-regions; Based on the fused image data, the contour shape corresponding to the candidate sub-region is determined; The continuous plot of land whose area is greater than a preset area threshold is determined as the tree plot. The candidate sub-regions whose outline shape conforms to the characteristics of trees are identified as the tree plot.

7. The farmland classification method according to claim 1, characterized in that, After acquiring the image data collected by the flight equipment, the method further includes: Obtain the weather parameters during the execution of the flight mission; The image data is corrected based on the weather parameters.

8. The farmland classification method according to claim 1, characterized in that, Based on the location information corresponding to the target region to be classified, determine the mask data corresponding to the target region, including: Based on the location information, determine the corner coordinates of the target area; The mask data is determined based on the corner coordinate information.

9. A farmland sorting device, characterized in that, include: The first determining module is used to determine the mask data corresponding to the target region based on the location information corresponding to the target region to be classified. The second determining module is used to determine the flight area and flight parameters based on the mask data and the equipment information of the flight equipment; the flight area covers the target area; The control module is used to control the flight equipment to perform flight missions in the flight area based on the flight parameters, and to acquire image data and point cloud data collected by the flight equipment. The first processing module is used to identify tree plots in the target area based on the image data and point cloud data, and remove the tree plots from the cultivated land plots corresponding to the target area to obtain farmland plots; The classification module is used to classify the farmland plots based on preset farmland classification standards and the image data.

10. An electronic device, characterized in that, It includes a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the farmland classification method as described in any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the farmland classification method as described in any one of claims 1 to 8.

Images