Multi-spectral fusion flight simulation terrain generation method, system and equipment and storage medium

Abstract

The invention relates to the technical field of flight simulation, provides a multispectral fusion flight simulation terrain generation method, system and device, and a storage medium, and solves the problem of low accuracy of flight simulation terrain generation. The method comprises the following steps: screening a target multispectral image of which the cloud coverage rate is lower than a set threshold value and target synthetic aperture radar data of a corresponding time window; extracting terrain edge features, dynamically adjusting a radar beam incident angle, and generating point cloud data with non-uniform resolution; a double-branch processing model is constructed, a first branch processes the target multispectral image, a second branch processes the point cloud data, and fusion features are obtained through a weighting mechanism; dynamically associating the vegetation index distribution of the terrain area to be measured with the elevation information in the fusion features to generate a space-time dynamic mapping result; and based on a space-time dynamic mapping result, generating a three-dimensional flight simulation terrain model with a semantic tag for planning a flight simulation path. The accuracy of flight simulation terrain generation is improved.

Description

Technical Field

[0001] The present application relates to the field of flight simulation technology, and in particular to a method and system for generating flight simulation terrain using multi-spectral fusion. Background Art

[0002] Flight simulation terrain refers to a virtual 3D terrain model constructed through technical means that matches real-world terrain features. It is primarily used in scenarios such as flight training, drone path planning, and aviation mission simulation. These scenarios place extremely high demands on terrain model accuracy, detail, and dynamic adaptability. The model must accurately reflect features such as topography, vegetation cover, and crop types, while maintaining sufficient detail in complex terrain areas (such as steep slopes and canyons). Furthermore, it must be able to cope with diverse weather conditions to ensure the authenticity and reliability of flight simulations.

[0003] Currently, there is a processing solution based on single multispectral image data to meet the technical requirements for flight simulation terrain generation. This solution collects multispectral images of the terrain area to be measured, directly performs band splitting and feature extraction on the multispectral images, generates terrain texture information, and then combines it with preset elevation data to construct a three-dimensional terrain model. Specifically, it first screens image data without cloud obstruction, identifies the general outline of the terrain through the distribution pattern of pixel values ​​in the image, and then simply overlays the extracted texture information with the fixed-resolution elevation data to form a terrain model for flight simulation. The entire process does not introduce other types of remote sensing data for supplementary optimization.

[0004] However, this existing solution has the following obvious defects: First, it relies solely on multispectral image data and is easily affected by weather. When there is cloud cover, the terrain texture information extraction will be inaccurate, and high-quality image data cannot be effectively screened out; second, the generated elevation data uses a uniform resolution, and the details in complex terrain areas (such as steep slopes and canyons) are insufficient, making it difficult to accurately reflect the true undulations of the terrain; third, vegetation information is not well reflected in the terrain model, making it difficult to take into account the impact of vegetation cover on flight safety when planning flight simulation paths. Summary of the Invention

[0005] The present application provides a multi-spectral fusion flight simulation terrain generation method and system to solve the problem of low accuracy of flight simulation terrain generation caused by low image data quality, difficulty in accurately reflecting the real undulations of the terrain, and lack of vegetation information in the existing technology.

[0006] In a first aspect, the present application provides a method for generating terrain for flight simulation using multi-spectral fusion, comprising: Collecting original multispectral images and original synthetic aperture radar data of the terrain area to be measured, identifying cloud cover based on the pixel value distribution of the original multispectral images, and screening target multispectral images and target synthetic aperture radar data of the corresponding time window with cloud cover below a set threshold; Extracting terrain edge features from the target multispectral image, and dynamically adjusting the radar beam incident angle based on the terrain edge features to generate point cloud data with non-uniform resolution. During the generation process, increasing the inclination of the radar beam incident angle increases the point cloud density in areas with complex terrain, such as steep slopes and canyon areas. A dual-branch processing model is constructed, where the first branch processes the target multispectral image and the second branch processes the point cloud data, and the high-resolution terrain features output by the two branches are fused through a weighted mechanism to obtain a fused feature; Dynamically associating the vegetation index distribution of the terrain area to be measured with the elevation information in the fusion feature to generate a spatiotemporal dynamic mapping result of the terrain and crop types in the vegetation coverage area; Based on the spatiotemporal dynamic mapping results, a three-dimensional flight simulation terrain model with semantic labels is generated for flight simulation path planning.

[0007] Optionally, dynamically adjusting the radar beam incident angle based on the terrain edge features to generate point cloud data with non-uniform resolution, and increasing the point cloud density in areas with complex terrain by increasing the inclination of the radar beam incident angle during the generation process, includes: By analyzing the terrain edge features, the contour of the terrain area to be measured is generated, the curvature change rate of the contour is calculated, and the terrain edge characteristic line is generated based on the curvature change rate; Based on the spatial distribution pattern of the terrain edge characteristic lines, the terrain area to be measured is divided into a plurality of sub-areas with similar terrain characteristics, and the corresponding dominant slope direction of each sub-area is determined; Calculating a complexity value of each sub-area based on a density value of the terrain edge characteristic line and a corresponding dominant slope aspect of each sub-area, and adjusting a radar beam incident angle according to the complexity value; Controlling the radar to scan each of the sub-areas according to the adjusted radar beam incident angle, obtaining a reflection signal from each of the sub-areas, and analyzing the return time and strength of the reflection signal to determine the position of the first reflection point; An initial point cloud data set is generated based on the position of the first reflection point. In the initial point cloud data, for areas with complex terrain, the point cloud density of the areas with complex terrain is increased by increasing the inclination of the radar beam incident angle to generate point cloud data with non-uniform resolution.

[0008] Optionally, the calculating the complexity value of each sub-region based on the density value of the terrain edge characteristic line and the corresponding dominant slope aspect of each sub-region includes: Counting the total length of all the terrain edge feature lines in each sub-region, and dividing the total length by the area of ​​the corresponding sub-region to obtain a density value of the terrain edge feature lines in the unit sub-region; Analyze the angle range between the corresponding dominant slope direction of each sub-area and the true north direction, and divide the angle range into multiple angle intervals; Determine the corresponding aspect coefficient according to the angle interval of the dominant aspect of each sub-area; Calculating an average curvature of a terrain edge characteristic line in each of the sub-areas, and dividing the average curvature into multiple levels, each level of the average curvature corresponding to a different complexity correction factor; The product of the density value of the terrain edge characteristic line in the unit sub-region, the corresponding slope aspect coefficient, and the corresponding complexity correction factor is used as the complexity value of each sub-region.

[0009] Optionally, generating an initial point cloud data set based on the position of the first reflection point, and increasing the point cloud density of a complex terrain area in the initial point cloud data by increasing the inclination of the radar beam incident angle to generate point cloud data with non-uniform resolution, includes: Arranging the position information of all the first reflection points according to three-dimensional space coordinates to generate an initial point cloud data set; Allocating each initial point cloud data in the initial point cloud data set to a corresponding sub-region; extracting a subregion as a complex terrain region from all subregions allocated with initial point cloud data, controlling the target synthetic aperture radar to scan the complex terrain region multiple times according to a target radar beam incident angle inclination to obtain position information of a second reflection point, wherein the target radar beam incident angle inclination is greater than the adjusted radar beam incident angle; Adding the position information of the second reflection point to the initial point cloud data of the corresponding complex terrain area to obtain target point cloud data of the complex terrain area, so as to increase the point cloud density of the complex terrain area; By comparing the point cloud densities corresponding to each of the sub-areas, it is verified that the point cloud density of the complex terrain area is higher than the point cloud density of the non-complex terrain area in the terrain area to be measured. After the verification is passed, point cloud data with non-uniform resolution is generated based on the target point cloud data of the complex terrain area and the initial point cloud data of the non-complex terrain area.

[0010] Optionally, the high-resolution terrain features include texture information and geometric information; The dual-branch processing model is constructed, where the first branch processes the target multispectral image and the second branch processes the point cloud data. The high-resolution terrain features output by the two branches are fused through a weighted mechanism to obtain fused features, including: Decomposing the target multispectral image into single-band images corresponding to the plurality of bands based on the first branch, determining a spatial distribution pattern of pixel values ​​in each of the single-band images, and extracting texture information corresponding to the terrain features based on the spatial distribution pattern; Based on the second branch, the point cloud data is organized into multiple data groups according to spatial position relationships, and the distribution density and relative position relationship of the point cloud data of each data group are analyzed to extract geometric information corresponding to the terrain features; Determining the credibility weight of the texture information based on the shooting parameters of the target multispectral image, and determining the credibility weight of the geometric information based on the acquisition parameters of the point cloud data; The texture information and the geometric information are weighted and merged according to corresponding credibility weights to obtain a fusion feature.

[0011] Optionally, decomposing the target multispectral image into single-band images corresponding to a plurality of bands based on the first branch, determining a spatial distribution pattern of pixel values ​​in each of the single-band images, and extracting texture information corresponding to terrain features based on the spatial distribution pattern includes: Splitting the target multispectral image into multiple independent single-band images according to a preset light wavelength range; Based on the values ​​of adjacent pixels in each of the single-band images, determining positions where a difference in the values ​​of adjacent pixels in each of the single-band images is greater than a preset difference threshold, dividing the corresponding single-band image according to the positions where the difference in the values ​​is greater than the preset difference threshold, and recording the distribution characteristics of each area in the single-band image; Determining a spatial distribution pattern of pixel values ​​in each of the single-band images according to the distribution characteristics of each region; According to the shape, terrain relief and terrain morphology of each area in the spatial distribution pattern, texture information corresponding to the terrain features is extracted from the single-band image.

[0012] Optionally, dynamically associating the vegetation index distribution of the terrain area to be measured with the elevation information in the fusion feature to generate a spatiotemporal dynamic mapping result of the vegetation coverage area terrain and crop type includes: Selecting an image of a specific band from the target multispectral image, and obtaining a vegetation index distribution of the terrain area to be measured by calculating a proportional relationship of pixel values ​​corresponding to vegetation in a single-band image of the specific band; Extracting elevation information from the fused features, and determining a variation range of elevation values ​​in the terrain area to be measured based on the elevation information; Associating the vegetation index distribution with the elevation information according to spatial positions to obtain a correlation degree between the vegetation index distribution value and the elevation value at each spatial position; According to the variation range of the elevation values, the degree of association is adjusted, and based on the adjusted degree of association, combined with the growth pattern of the crop type under the elevation information and the vegetation index distribution, a spatiotemporal dynamic mapping result is generated.

[0013] In a second aspect, the present application provides a multi-spectral fusion flight simulation terrain generation system, comprising: An acquisition module is configured to acquire original multispectral images and original synthetic aperture radar data of the terrain area to be measured, identify cloud coverage based on the pixel value distribution of the original multispectral images, and screen target multispectral images and target synthetic aperture radar data of the corresponding time window whose cloud coverage is lower than a set threshold; an extraction module for extracting terrain edge features from the target multispectral image, dynamically adjusting the radar beam incident angle based on the terrain edge features to generate point cloud data with non-uniform resolution, and increasing the point cloud density in areas with complex terrain, such as steep slopes and canyons, by increasing the inclination of the radar beam incident angle during the generation process; A construction module is used to construct a dual-branch processing model, wherein the first branch processes the target multispectral image and the second branch processes the point cloud data, and the high-resolution terrain features output by the two branches are fused through a weighted mechanism to obtain a fused feature; An association module is used to dynamically associate the vegetation index distribution of the terrain area to be measured with the elevation information in the fusion feature to generate a spatiotemporal dynamic mapping result of the terrain and crop types in the vegetation coverage area; A generation module is used to generate a three-dimensional flight simulation terrain model with semantic labels based on the spatiotemporal dynamic mapping result, which is used for planning the flight simulation path.

[0014] In a third aspect, the present application provides a computing device comprising a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a multi-spectral fusion flight simulation terrain generation method as described in any one of the first aspects.

[0015] In a fourth aspect, the present application provides a computer storage medium storing a computer program, which, when executed by a computer, implements a multi-spectral fusion flight simulation terrain generation method as described in any one of the first aspects.

[0016] The present application provides a method for generating flight simulation terrain using multispectral fusion, comprising: collecting original multispectral images and original synthetic aperture radar data of a terrain area to be measured; identifying cloud cover based on the pixel value distribution of the original multispectral images; and screening target multispectral images and target synthetic aperture radar data of corresponding time windows whose cloud cover is below a set threshold; extracting terrain edge features from the target multispectral images; dynamically adjusting the radar beam incidence angle based on the terrain edge features to generate point cloud data with non-uniform resolution. During the generation process, the point cloud density in complex terrain areas, such as steep slopes and canyon areas, is increased by increasing the inclination of the radar beam incidence angle; constructing a dual-branch processing model, wherein a first branch processes the target multispectral image and a second branch processes the point cloud data; and fusing the high-resolution terrain features output by the two branches through a weighted mechanism to obtain a fused feature; dynamically correlating the vegetation index distribution of the terrain area to be measured with the elevation information in the fused feature to generate a spatiotemporal dynamic mapping result of the terrain and crop types in the vegetation-covered area; and generating a semantically labeled three-dimensional flight simulation terrain model based on the spatiotemporal dynamic mapping result for flight simulation path planning.

[0017] This application has the following advantages: By collecting raw multispectral imagery and raw synthetic aperture radar data and screening target data with cloud coverage below a set threshold, cloud interference with image information can be reduced, ensuring the validity of subsequent data processing. By extracting terrain edge features from the target multispectral image, dynamically adjusting the radar beam incidence angle, and increasing the point cloud density in complex terrain areas, point cloud data with non-uniform resolution is generated. This allows the point cloud data to retain more details in complex terrain areas and adapt to the characteristic representation requirements of different terrains. By constructing a dual-branch processing model and weightedly fusing the terrain features output by the two branches to obtain fused features, the advantages of multispectral imagery and point cloud data can be combined to improve the integrity and accuracy of terrain features. By dynamically correlating the vegetation index distribution with elevation information to generate spatiotemporal dynamic mapping results, the relationship between terrain and vegetation and crop types can be established, reflecting their changing characteristics over time. By generating a three-dimensional flight simulation terrain model with semantic labels based on the spatiotemporal dynamic mapping results, a virtual environment support that conforms to actual terrain characteristics can be provided for flight simulation path planning.

[0018] Furthermore, by analyzing the terrain edge features, terrain edge feature lines are generated, sub-areas are divided and the dominant slope direction is determined. The complexity value of the sub-area is calculated by combining the terrain edge feature line density, slope coefficient and complexity correction factor. Based on this, the radar beam incident angle is adjusted, the reflection point position is scanned and an initial point cloud is generated. By increasing the inclination of the radar beam incident angle in the complex terrain area, its point cloud density is increased, and point cloud data with non-uniform resolution is generated. The above method can accurately quantify the complexity of the terrain, so that the adjustment of the radar beam incident angle is more in line with the terrain characteristics, so that the point cloud data forms a differentiated distribution in terrain areas of different complexity, while ensuring data processing efficiency, the detail richness of the point cloud data in the complex terrain area is improved.

[0019] These and other aspects of the present application will become more readily apparent from the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0021] Figure 1 A flowchart of a method for generating terrain for flight simulation using multi-spectral fusion provided in an embodiment of the present application; Figure 2 A schematic diagram of the structure of a multi-spectral fusion flight simulation terrain generation system provided in an embodiment of the present application; Figure 3 A schematic diagram of the structure of a computing device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0022] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0023] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this document or may be executed in parallel. The serial numbers of the operations, such as 11, 12, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any order of execution. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this document are used to distinguish different messages, devices, modules, etc., and do not represent a sequential order, nor do they limit "first" and "second" to being different types.

[0024] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of this application.

[0025] In order to solve the problem of low accuracy of flight simulation terrain generation caused by low image data quality, difficulty in accurately reflecting the real undulations of the terrain, and lack of vegetation information in the existing technology, the embodiment of the present application provides a multi-spectral fusion flight simulation terrain generation method, which adopts the following design concept: first collect multi-band images and radar detection data of the terrain to be detected, select images with less cloud cover and corresponding radar data; identify the edge shape of the terrain from the image, adjust the radar detection angle according to these shapes, and make the three-dimensional point data in the complex terrain area more dense; process the multi-band image and three-dimensional point data respectively through two branches, and then combine the terrain features of the two; correspond the growth index distribution of surface plants with the height information of the terrain to form an association result that can reflect the change of plant type and terrain over time; finally, based on the result, generate a three-dimensional terrain model with detailed labels for flight simulation path planning. In this way, the impact of clouds can be reduced, the details of the complex terrain area can be clearer, and plant information can be combined to generate a more realistic flight simulation terrain model.

[0026] Figure 1 A flowchart of a multi-spectral fusion flight simulation terrain generation method provided in an embodiment of the present application is shown in FIG. Figure 1 As shown, the method includes: S11. Collect original multispectral images and original synthetic aperture radar data of the terrain area to be measured, identify cloud coverage based on the pixel value distribution of the original multispectral image, and screen target multispectral images and target synthetic aperture radar data of the corresponding time window whose cloud coverage is lower than a set threshold.

[0027] Among them, the original multispectral image can be a terrain image containing information on multiple different light wavelengths, which can reflect the reflection of different light rays on the terrain surface; the original synthetic aperture radar data is data obtained by terrain detection by radar equipment, which can reflect the three-dimensional structural information of the terrain; the cloud cover rate is the area ratio occupied by clouds in the original multispectral image; the target multispectral image is the original multispectral image with cloud coverage lower than the set ratio; the target synthetic aperture radar data is the original synthetic aperture radar data with a similar shooting time to the target multispectral image; the result obtained by processing these data is the screened target multispectral image and the corresponding target synthetic aperture radar data.

[0028] In an embodiment of the present application, raw multispectral images and raw synthetic aperture radar (SAR) data of the terrain area to be inspected are first collected. The distribution of pixel values ​​in the raw multispectral images is then observed. Since clouds typically appear as areas with higher pixel values ​​in images, the ratio of the total area of ​​these high-pixel-value areas to the total image area is calculated to determine the cloud cover. Finally, an upper limit for the cloud cover is set, and raw multispectral images with coverage below the upper limit are selected as target multispectral images. Raw SAR data captured close to these target images are then selected as target SAR data. For example, when processing data for a certain area, if the total area of ​​the raw multispectral image is 1000 square meters and the area of ​​high-pixel-value cloud areas is 200 square meters, then the cloud cover is calculated as 20% (200 divided by 1000). If the upper limit is set to 30%, this image can be used as the target multispectral image. Radar data captured within an hour of this image is then selected as the target radar data.

[0029] S12. Extract terrain edge features from the target multispectral image and dynamically adjust the radar beam incident angle based on the terrain edge features to generate point cloud data with non-uniform resolution. During the generation process, increase the inclination of the radar beam incident angle to increase the point cloud density in complex terrain areas, such as steep slopes and canyon areas.

[0030] Among them, the target multispectral image is the screened target multispectral image; the terrain edge feature is the feature at the junction of different areas in the terrain, such as the boundary between mountains and plains, the edges of both sides of the canyon, etc.; the radar beam incidence angle is the angle between the detection signal emitted by the radar equipment and the terrain surface; the point cloud data is a data set composed of a large number of three-dimensional points, each point represents a position on the terrain surface; the point cloud data with non-uniform resolution is the situation where the density of points in the point cloud of different terrain areas is different; the complex terrain area is an area with large terrain undulations and complex structure, such as steep slopes, canyons, etc.; the result generated after processing is point cloud data with non-uniform resolution.

[0031] In an embodiment of the present application, first, terrain edge features are extracted from the target multispectral image. Specifically, the locations where pixel values ​​suddenly change in the image are observed, such as the junction from low-pixel-valued flat land to high-pixel-valued mountains, so as to determine the edge of the terrain. Secondly, the incident angle of the radar beam is adjusted according to the extracted terrain edge features. For areas with complex terrain, the inclination of the incident angle of the radar beam is increased so that the radar signal can detect these areas more comprehensively. Then, the terrain is detected by the adjusted radar beam, and the reflected signals are collected and converted into point cloud data. Finally, since the incident angle of the radar beam in areas with complex terrain is more inclined, Obtaining more reflected signals makes the number of points in the point cloud data of these areas more and denser, forming point cloud data with non-uniform resolution. For example, in the processing of a certain area, the edge of a steep slope is found from the image, and the inclination of the radar beam incident angle in the area is adjusted from 30 degrees to 50 degrees. After detection, the number of point clouds per square meter in the steep slope area is 70, while the number per square meter of the surrounding flat land is 30. This is because the radar signal reflection points increase with the increase in inclination. The calculation method is the total number of points in the area divided by the area of ​​the area. For example, if the area of ​​the steep slope area is 10 square meters and the total number of points is 700, 700 divided by 10 equals 70 per square meter.

[0032] S13. Construct a dual-branch processing model. The first branch processes the target multispectral image, and the second branch processes the point cloud data. The high-resolution terrain features output by the two branches are fused through a weighted mechanism to obtain fused features.

[0033] The dual-branch processing model consists of two processing paths, each processing a different type of data. The first branch is dedicated to processing the target multispectral imagery within the dual-branch processing model, while the second branch is dedicated to processing point cloud data. The weighting mechanism assigns different weights to different data sets based on their reliability and then merges them. The fused feature is the combined feature of the terrain features obtained by processing the two branches. The resulting result is the fused feature. For example, the dual-branch processing model can be constructed based on a dual-stream optimized U-Net model.

[0034] In an embodiment of the present application, a dual-branch processing model is first constructed, wherein the first branch processes the target multispectral image, specifically analyzes the pixel distribution, color changes, etc. in the image, and extracts features related to the terrain, such as the texture features of the vegetation-covered area; the second branch processes the point cloud data, analyzes the distribution and distance of the three-dimensional points in the point cloud, and extracts the ups and downs features of the terrain; and then determines their corresponding weights according to their respective reliabilities; for example, the image is more reliable in vegetation recognition, and the weight is set to 0.6; the point cloud is more reliable in terrain undulation, and the weight is set to 0.4; finally, according to the determined weights, the features extracted by the first branch and the features extracted by the second branch are merged to obtain fused features; for example, in the processing of a certain area, the texture feature value extracted by the first branch is 0.8, and the terrain undulation feature value extracted by the second branch is 0.6, with weights of 0.5 and 0.5 respectively, then the fused feature value is 0.8 multiplied by 0.5 plus 0.6 multiplied by 0.5, which equals 0.7, thereby integrating the information of the two features.

[0035] S14. Dynamically associate the vegetation index distribution of the terrain area to be measured with the elevation information in the fusion feature to generate a spatiotemporal dynamic mapping result of the terrain and crop types in the vegetation coverage area.

[0036] Among them, the vegetation index distribution is a numerical distribution that reflects the vegetation growth conditions within the terrain area. Higher values ​​generally indicate more vigorous vegetation growth. The elevation information is the height information of each location within the terrain area. The spatiotemporal dynamic mapping result reflects the corresponding relationship between the terrain and crop types in the vegetation coverage area over time. The result generated after processing is the spatiotemporal dynamic mapping result.

[0037] In an embodiment of the present application, the vegetation index distribution of the terrain area is first calculated based on the pixel values ​​of different bands in the target multispectral image. For example, the pixel values ​​of the two bands are compared, and the difference between the pixel value of band 1 and the pixel value of band 2 is divided by the sum of the pixel value of band 1 and the pixel value of band 2 to obtain the vegetation index value of each position. Secondly, the elevation information is extracted from the fusion feature to determine the height value of each position. Then, the vegetation index distribution and the elevation information are matched according to the spatial position, so that the vegetation index value of each position is matched with the height value of the position. Then, the correlation between the two is adjusted according to the change of the height value. Places with large altitude changes make the correlation between vegetation index and altitude closer, while places with small altitude changes make the correlation slightly weaker. Finally, the correlation results at different time points are recorded. Combined with the growth of different crops at different altitudes and vegetation indices, the correspondence between the terrain of the vegetation coverage area and the crop type over time is generated, and the spatiotemporal dynamic mapping results are obtained. For example, at a certain location, the pixel value of band 1 is 80, and the pixel value of band 2 is 20. The vegetation index is the difference of 80 minus 20 divided by the sum of 80 plus 20, which is equal to 0.6. The elevation of this location is 200 meters. In spring, this correlation corresponds to crop E. In summer, the vegetation index changes to 0.7, which still corresponds to crop E.

[0038] S15. Based on the spatiotemporal dynamic mapping results, a three-dimensional flight simulation terrain model with semantic labels is generated for flight simulation path planning.

[0039] Among them, the three-dimensional flight simulation terrain model with semantic labels is a virtual three-dimensional terrain model that contains category labels such as terrain, vegetation, and crop types; flight simulation path planning is to design a route suitable for flight in the virtual terrain model; the result generated after processing is a three-dimensional flight simulation terrain model with semantic labels, which is used for flight simulation path planning.

[0040] In an embodiment of the present application, a three-dimensional terrain model is first constructed based on the spatiotemporal dynamic mapping results obtained previously. This model can reflect the ups and downs of the terrain. Secondly, semantic labels are added to the three-dimensional model, such as labeling steep slope areas as steep slopes and labeling areas where crop E is planted as crop E. Then, this three-dimensional model with semantic labels is used as a flight simulation terrain model. Finally, this model is used for flight simulation path planning. According to labels such as steep slopes and high-altitude crops E (high altitude) in the model, flight routes that avoid these areas are planned. For example, in a certain model, canyons (complex terrain), crop E areas (height 2 meters), and flat land are marked. When planning the route, the distance and safety factor of each area from the starting point to the end point are calculated, and the canyons and crop E areas with low safety factors are avoided, and flat land with high safety factors is selected.

[0041] For example, when generating terrain for flight simulation in area B, we first collect the original multispectral imagery and original synthetic aperture radar (SAR) data for that area. We calculate the cloud cover percentage of the original multispectral imagery. For example, if one image has a total area of ​​500 square meters and a cloud cover area of ​​100 square meters, the cloud cover percentage is 100 ÷ 500 = 20%. With a threshold of 30%, we select this imagery as the target multispectral imagery. We also select radar data acquired within an hour of each other as the target SAR data. Next, terrain edge features were extracted from the target multispectral imagery, revealing a 20-square-meter canyon and a 15-square-meter steep slope. Based on these edge features, the radar beam's incident angle was adjusted from 30 to 50 degrees in these two areas. The resulting detection yielded 1,600 points in the canyon (1,600 ÷ 20 = 80 points per square meter), 1,050 points on the steep slope (1,050 ÷ 15 = 70 points per square meter), and 3,000 points in the 100-square-meter flat area (3,000 ÷ 100 = 30 points per square meter). This generated point cloud data with non-uniform resolution. This was then processed using a two-branch model: the first branch extracted texture features (valued at 0.7) from the target multispectral imagery, and the second branch extracted terrain relief features (valued at 0.8) from the point cloud data. Each branch was assigned a weight of 0.5, resulting in a fused feature value of 0.7 × 0.5 + 0.8 × 0.5 = 0.75. The vegetation index was then calculated. For region C, where Band 1 has a pixel value of 100 and Band 2 has a pixel value of 20, the index is (100-20) ÷ (100+20) ≈ 0.67, corresponding to an elevation of 200 meters. For region D, where Band 1 has a pixel value of 40 and Band 2 has a pixel value of 30, the index is (40-30) ÷ (40+30) ≈ 0.14, corresponding to an elevation of 300 meters. Combined with temporal variation records, region C corresponds to crop E in spring and region D to crop F. Despite the index change in summer, the index still corresponds to the same crop, resulting in a spatiotemporal dynamic mapping result. Finally, based on this result, a 3D model was constructed, annotated with labels such as "Canyon," "Crop Area E (2-meter elevation)," and "Flat Land." Flight routes were planned to avoid canyons and crop areas E and focus on flat land, completing the entire process.

[0042] By executing S11 to S15, the embodiment of the present application processes the entire process, screens out high-quality multispectral images and radar data from the data acquisition stage, reduces the interference of clouds on terrain information, and ensures the temporal consistency of data; in the point cloud data generation stage, by dynamically adjusting the radar detection angle, the point cloud in the complex terrain area is made denser, which not only retains the details of the key area but also avoids data redundancy; the fusion processing of the dual-branch model integrates the texture information of the multispectral image and the three-dimensional structure information of the point cloud data, making the terrain features more comprehensive and accurate; the dynamic correlation between the vegetation index and elevation reflects the temporal and spatial change relationship between the terrain and vegetation and crop types, and enriches the semantic information of the model; the final generated three-dimensional model with semantic labels provides a reliable reference for flight simulation path planning, takes into account the terrain authenticity and flight safety, and overall improves the accuracy and practicality of flight simulation terrain generation.

[0043] In one possible embodiment, point cloud data with uniform resolution tends to generate data redundancy in areas with simple terrain and lack details in complex areas. Generating point cloud data with non-uniform resolution can make points in complex areas denser to preserve details and points in simple areas fewer to reduce redundancy, thus balancing accuracy and efficiency. Therefore, in S12, the radar beam incident angle is dynamically adjusted based on terrain edge features to generate point cloud data with non-uniform resolution. During the generation process, the point cloud density in areas with complex terrain is increased by increasing the inclination of the radar beam incident angle, including: Step 121 : Generate the outline of the terrain area to be measured by analyzing the terrain edge features, calculate the curvature change rate of the outline, and generate the terrain edge feature line based on the curvature change rate.

[0044] Among them, the terrain edge features are the features of the junction of different areas in the terrain (such as plains and mountains, and both sides of canyons); the contour of the terrain area to be measured is the approximate outer boundary of the terrain in that area; the curvature change rate is the speed of change of the curvature degree of each point on the contour; the terrain edge feature line is a line generated based on the curvature change rate that can reflect the specific position and direction of the terrain edge. The result generated after processing is the terrain edge feature line.

[0045] In the embodiment of the present application, the terrain edge features are first analyzed, and the contour of the terrain area to be measured is outlined by identifying the position and shape of the intersection of different areas; then the curvature change rate of each point on the contour is calculated, specifically by measuring the change in the tangent direction of adjacent points on the contour to determine the change in the degree of curvature of each point; finally, based on the magnitude of the curvature change rate, the positions with obvious changes are connected to generate the terrain edge characteristic line. For example, in area A, the boundary features between the flat land and the steep slope are first analyzed to outline the contour. The tangent directions of two adjacent points of a certain contour are calculated to be 30 degrees and 60 degrees respectively, with an angle difference of 30 degrees. The distance between the two points is 2 meters, and the curvature change rate is 30 degrees ÷ 2 meters = 15 degrees / meter. The positions with large curvature change rates are connected to generate the terrain edge characteristic line of the steep slope.

[0046] Step 122: Based on the spatial distribution pattern of the terrain edge characteristic lines, the terrain area to be measured is divided into multiple sub-areas with similar terrain characteristics, and the corresponding dominant slope direction of each sub-area is determined.

[0047] Among them, the spatial distribution pattern of the terrain edge characteristic lines refers to the distribution density, direction, etc. of the characteristic lines in the terrain area to be measured; the sub-area is a small area with similar terrain characteristics divided from the terrain area to be measured; the dominant slope direction is the main inclination direction of the terrain in each sub-area. The results generated after processing are the divided sub-areas and the dominant slope direction of each sub-area.

[0048] In this embodiment, the spatial distribution patterns (e.g., density and direction) of the characteristic lines along the terrain's edge are first observed. Based on these patterns, the terrain area to be measured is then divided into multiple subareas, ensuring that the distribution and direction of the characteristic lines within each subarea are similar. Finally, the terrain inclination direction of each subarea is analyzed to determine the dominant slope direction. For example, in area A, the characteristic lines are dense in the northeast and trend northeast-southwest, while the southwest is sparse and trend east-west. Based on this, the area is divided into northeastern and southwestern subareas. The northeastern terrain primarily slopes southeast, with a dominant slope direction of southeast.

[0049] Step 123: Calculate the complexity value of each sub-region based on the density value of the terrain edge feature line and the corresponding dominant slope direction of each sub-region, and adjust the radar beam incident angle according to the complexity value.

[0050] Among them, the density value of the terrain edge feature line is the total length of the terrain edge feature lines per unit area; the complexity value is a value that quantifies the complexity of the sub-region terrain; the radar beam incident angle is the angle between the radar transmitted signal and the terrain surface. After processing, the radar beam incident angle is adjusted according to the complexity value.

[0051] In this application example, the density of the terrain edge feature lines in each sub-area is first calculated (total length divided by the sub-area area). A correction factor is then determined based on the dominant slope aspect (e.g., different coefficients result from different angles between the dominant slope aspect and a specific direction). The density value is then multiplied by the correction factor to obtain a complexity value. Finally, the radar beam incidence angle is adjusted based on this value; a larger value indicates a greater inclination. For example, if the feature lines in the central sub-area of ​​Area A have a total length of 100 meters and an area of ​​50 square meters, the density value is 100 ÷ 50 = 2 meters per square meter, the correction factor is 1.2, and the complexity value is 2 × 1.2 = 2.4, adjusting the inclination from 30 degrees to 45 degrees.

[0052] Step 124: Control the radar to scan each sub-area according to the adjusted radar beam incident angle, obtain the reflected signal of each sub-area, analyze the return time and strength of the reflected signal, and determine the position of the first reflection point.

[0053] Among them, radar is a device that transmits and receives detection signals; the reflected signal is the signal reflected by the radar signal after encountering the terrain; the return time is the time from the signal being transmitted to the signal being received; the strength is the intensity of the reflected signal; the position of the first reflection point is the spatial position of the terrain surface point where the reflected signal originally originated, and the position of the first reflection point is obtained after processing.

[0054] In the embodiment of the present application, the radar is first controlled to scan the sub-area at the adjusted incident angle and transmit a detection signal; then the reflected signal is received; then the signal return time (the longer the signal, the farther the distance) and the strength (reflecting the terrain's reflectivity) are analyzed; finally, the position of the first reflection point is calculated based on the return time and strength. For example, in the central sub-area of ​​area A, the radar transmits a signal at a 45-degree inclination. The return time of a certain reflected signal is 0.002 seconds, and the signal speed is 0.002 seconds. m / s, distance ×0.002÷2=300 meters, combined with the strength, the position is determined to be (200,300,150).

[0055] Step 125: Generate an initial point cloud data set based on the position of the first reflection point. In the initial point cloud data, for areas with complex terrain, increase the point cloud density in the areas with complex terrain by increasing the inclination of the radar beam incident angle to generate point cloud data with non-uniform resolution.

[0056] Among them, the initial point cloud data set is the set of all first reflection point positions; the complex terrain area is the area with large terrain undulations and complex structures (such as steep slopes and canyons); the radar beam incident angle inclination is the inclination angle between the radar beam and the terrain surface; the point cloud data with non-uniform resolution is the point cloud data with different point densities in different terrain areas. The result generated after processing is point cloud data with non-uniform resolution.

[0057] In this embodiment, all first reflection point locations are integrated to form an initial point cloud data set. Areas with complex terrain are then identified. These areas are then scanned again with the radar incident angle increased to obtain more reflection points. Finally, these newly added points are added to the initial set, increasing the point density in the complex areas and forming a non-uniform point cloud. For example, the initial point density in the center of Area A (the complex area) is 50 points per square meter. Increasing the incident angle increases the density by 30 points per square meter, bringing the total to 80 points per square meter. The simpler areas maintain a density of 30 points per square meter.

[0058] For example, in terrain processing for region A, we first analyze the terrain edge features (such as the junction of a canyon and a mountain), outline the region, and calculate the curvature change rate of each point on the outline (60 degrees / meter at the canyon edge, 0.2 degrees / meter on flat land). Points with large curvature change rates are connected to generate terrain edge feature lines. Next, we observe the distribution of feature lines (dense and north-south in the center, sparse and east-west in the east-west direction), and divide region A into central, eastern, and western sub-regions. The dominant slope direction in the central region is determined to be east-west, the eastern region is determined to be south-south, and the western region is determined to be north-south. Then calculate the density of characteristic lines in each sub-area: the total length of the central area is 240 meters, the area is 80 square meters, the density is 240÷80=3 meters / square meter, combined with the dominant slope correction coefficient of 1.1, the complexity value of 3×1.1=3.3, and the radar incident angle is adjusted to 50 degrees; the density in the east is 60÷40=1.5 meters / square meter, the correction coefficient is 1.0, the value is 1.5, and the incident angle is adjusted to 35 degrees; then control the radar to scan at the adjusted angle and receive the reflected signal. The return time of a certain signal is 0.003 seconds, and the distance is 3×10 8 ×0.003÷2=450 meters, determine the position of the first reflection point (150, 250, 180); finally, integrate the reflection points to form the initial point cloud. The central 80 square meters contain 4800 points (60 points / square meter), and the eastern 100 square meters contain 2500 points (25 points / square meter). Increase the incident angle of the central area to 60 degrees and scan again, adding 300 points, with a density of 60 + 300÷80 = 63.75 points / square meter. The original density of the eastern area is maintained to generate non-uniform point cloud data.

[0059] By executing steps 121 to 125, the embodiment of the present application generates terrain edge feature lines to accurately capture the terrain edge position and direction, providing a basis for regional division; the divided sub-regions and determined dominant slope directions make terrain analysis more targeted; the calculated complexity numerically quantifies the terrain complexity, and the radar incident angle adjusted accordingly makes the detection more in line with the terrain characteristics; the obtained first reflection point position provides basic data for point cloud generation; the non-uniform resolution point cloud data finally generated retains more details in complex terrain areas to reflect the real structure, and reduces data redundancy in simple areas to improve efficiency, taking into account both data accuracy and processing efficiency as a whole, and providing high-quality basic data for subsequent flight simulation terrain modeling and path planning.

[0060] In a possible embodiment, step 123 of calculating the complexity value of each sub-region based on the density value of the terrain edge feature line and the corresponding dominant slope aspect of each sub-region includes: Step a1: Count the total length of all terrain edge feature lines in each sub-region, divide the total length by the area of ​​the corresponding sub-region, and obtain the density value of the terrain edge feature lines in the unit sub-region.

[0061] Among them, the total length of the terrain edge feature lines refers to the sum of the lengths of all terrain edge feature lines in each sub-region; the sub-region area is the area occupied by each sub-region; the density value of the terrain edge feature lines in a unit sub-region is the value obtained by dividing the total length by the sub-region area, which is used to reflect the density of the terrain edge feature lines in the sub-region. The result obtained after processing is this density value.

[0062] In an embodiment of the present application, the lengths of all terrain edge feature lines in each sub-area are first counted, and these lengths are added together to obtain the total length. Then, the area of ​​the sub-area is measured, and finally, the total length is divided by the area of ​​the corresponding sub-area to obtain the density value of the terrain edge feature lines in the unit sub-area. For example, in sub-area B, the lengths of all terrain edge feature lines are 50 meters, 80 meters, and 120 meters, respectively. The total length is 50+80+120=250 meters. The area of ​​the sub-area is 200 square meters, and the density value is 250÷200=1.25 meters / square meter.

[0063] Step a2: Analyze the angle range between the corresponding dominant slope direction of each sub-area and the north direction, and divide the angle range into multiple angle intervals.

[0064] Among them, the dominant slope direction is the main inclination direction of the terrain in the sub-area; the due north direction is the geographical due north; the angle range is the angle range formed between the guiding slope direction and the due north direction; the angle interval is the multiple continuous angle segments divided by the angle range of 0-360 degrees. After processing, the divided angle interval is obtained.

[0065] In the embodiment of the present application, the dominant slope direction of each sub-area is first determined, and then the angle between the dominant slope direction and the north direction is measured to obtain the angle range. Finally, the angle range of 0-360 degrees is divided into multiple angle intervals according to fixed angle intervals (such as every 45 degrees). For example, the angle between the dominant slope direction of sub-area C and the north direction is 70 degrees, and 0-360 degrees is divided into intervals such as 0-45 degrees, 46-90 degrees, and 91-135 degrees. 70 degrees belongs to the angle interval of 46-90 degrees.

[0066] Step a3: Determine the corresponding aspect coefficient according to the angle range of the dominant aspect of each sub-area.

[0067] The angle interval of the dominant aspect refers to the angle segment corresponding to the angle between the dominant aspect of each sub-area and the north direction. The aspect coefficient is a value set based on the angle interval to correct for the complexity of the terrain. Different intervals correspond to different coefficients. After processing, the aspect coefficient of each sub-area is obtained.

[0068] In the embodiment of the present application, it is first determined to which angle range the dominant slope aspect of each sub-area belongs, and then the slope aspect coefficient corresponding to the sub-area is determined according to a preset correspondence (such as 1.0 for the 0-45 degree range and 1.2 for the 46-90 degree range). For example, the dominant slope aspect angle of sub-area C belongs to the 46-90 degree range, and the slope aspect coefficient corresponding to this range is 1.2. Therefore, the slope aspect coefficient of sub-area C is 1.2.

[0069] Step a4: Calculate the average curvature of the terrain edge feature lines in each sub-area, and divide the average curvature into multiple levels. Each level of the average curvature corresponds to a different complexity correction factor.

[0070] Among them, the average curvature of the terrain edge characteristic line refers to the average curvature of all terrain edge characteristic lines in the sub-area. The curvature can be expressed by the ratio of the actual length of the line to the straight-line distance; the level is the multiple continuous intervals into which the average curvature is divided; the complexity correction factor is a value set according to the level for correcting the complexity of the terrain. Different levels correspond to different factors. After processing, the complexity correction factor of each sub-area is obtained.

[0071] In an embodiment of the present application, the curvature of each terrain edge characteristic line in each sub-area is first calculated (actual length ÷ straight-line distance), and then these curvatures are added and divided by the number of characteristic lines to obtain the average curvature. The average curvature is then divided into multiple levels (such as 1.0-1.5 for level 1, 1.6-2.0 for level 2), and finally the complexity correction factor is determined according to the preset value corresponding to the level (such as level 1 corresponds to 1.0, level 2 corresponds to 1.3). For example, the curvatures of the two characteristic lines in sub-area D are 1.4 and 1.6 respectively, and the average curvature is (1.4+1.6)÷2=1.5, which belongs to the 1.0-1.5 level, and the corresponding complexity correction factor is 1.0.

[0072] Step a5: The product of the density value of the terrain edge feature line in the unit sub-region, the corresponding aspect coefficient, and the corresponding complexity correction factor is used as the complexity value of each sub-region.

[0073] Among them, the density value of the terrain edge feature lines in the unit sub-area is a value reflecting the density of the edge lines; the aspect coefficient is a value correcting the influence of the aspect; the complexity correction factor is a value correcting the influence of line bending; the complexity value is the product of these three values, which is used to comprehensively reflect the complexity of the sub-area. The result obtained after processing is the complexity value of each sub-area.

[0074] In this embodiment of the present application, the density value obtained in step a1, the aspect coefficient obtained in step a3, and the complexity correction factor obtained in step a4 are first obtained, and then these three values ​​are multiplied together. The result is the complexity value of the corresponding sub-area. For example, the density value of sub-area D is 1.8 meters per square meter, the aspect coefficient is 1.1, and the complexity correction factor is 1.0, then the complexity value is 1.8×1.1×1.0=1.98.

[0075] For example, when processing sub-area E in area A, in step a1, the lengths of the three terrain edge feature lines in the sub-area are calculated to be 60 meters, 90 meters, and 150 meters, respectively. The total length is 60 + 90 + 150 = 300 meters. The area of ​​the sub-area is 200 square meters, and the density value is 300 ÷ 200 = 1.5 meters per square meter. In step a2, the angle between the dominant slope and due north is determined to be 60 degrees, and the angle range is divided into intervals of 0-45 degrees and 46-90 degrees, with 60 degrees belonging to the 46-90 degree interval. In step a3, this interval corresponds to an aspect coefficient of 1.2. In step a4, the curvature of the three feature lines is calculated to be 1.6, 1.8, and 2.0, respectively, with an average curvature of (1.6 + 1.8 + 2.0) ÷ 3 = 1.8. The classification level of 1.6-2.0 corresponds to level 2, and the correction factor is 1.3. In step a5, the complexity value is 1.5 × 1.2 × 1.3 = 2.34.

[0076] By executing steps a1 to a5, the embodiment of the present application calculates density values, divides aspect intervals, determines aspect coefficients, and calculates average curvature and complexity correction factors in a step-by-step manner. Finally, the three are multiplied together to obtain a complexity value. This comprehensively considers the density of terrain edges, aspect characteristics, and line curvature characteristics, making the quantified complexity result more consistent with actual terrain characteristics and accurately reflecting the terrain complexity of the sub-region. This provides a scientific and reliable basis for subsequent adjustment of relevant parameters based on complexity, thereby improving the accuracy and pertinence of terrain analysis.

[0077] In one possible example, step 125 of generating an initial point cloud data set based on the position of the first reflection point, and increasing the point cloud density of the complex terrain area in the initial point cloud data by increasing the inclination of the radar beam incident angle to generate point cloud data with non-uniform resolution, includes: b1. Arrange the position information of all first reflection points according to three-dimensional space coordinates to generate an initial point cloud data set.

[0078] Among them, the position information of the first reflection point is the three-dimensional spatial coordinates of the point corresponding to the signal reflected from the terrain surface for the first time detected by the radar. The three-dimensional spatial coordinates are the numerical values ​​that describe the position of the point in the three-dimensional space. The initial point cloud data set is a data set formed by arranging the position information of all first reflection points according to the three-dimensional spatial coordinates. It is used to preliminarily reflect the three-dimensional structure of the terrain. The result obtained after processing is the initial point cloud data set.

[0079] In an embodiment of the present application, the position information of all the first reflection points is first collected. This information includes the specific values ​​of each point in the x, y, and z directions. Then, these points are arranged in order of size of the three-dimensional space coordinates (first by the x coordinate, then by the y coordinate, and finally by the z coordinate). Finally, the arranged points are integrated to form an initial point cloud data set. For example, in area A, the coordinates of the three first reflection points are obtained as (10, 20, 5), (15, 25, 8), and (8, 18, 6). After arranging them from small to large according to the x coordinate, an initial point cloud data set containing these three points is generated.

[0080] b2. Allocate each initial point cloud data in the initial point cloud data set to a corresponding sub-region.

[0081] Among them, the initial point cloud data set is a data set consisting of the three-dimensional coordinates of the first reflection point, the initial point cloud data is the three-dimensional coordinates of a single point in the set, the sub-region is a pre-divided small area with similar terrain features, and the corresponding sub-region refers to the small area where the initial point cloud data is located (that is, the coordinates of the point fall within the range of the sub-region). The result obtained after processing is the initial point cloud data assigned to each sub-region.

[0082] In the embodiment of the present application, the range of each sub-area (i.e., the starting and ending values ​​of the x and y coordinates) is first clarified, and then the x and y coordinates of each point in the initial point cloud data set are checked one by one to determine which sub-area range the point falls within, and finally the point is assigned to the corresponding sub-area. For example, the x range of sub-area B in area A is 0-200, and the y range is 0-200. The coordinates of a point are (150, 180, 10), and its x and y coordinates fall within sub-area B, so it is assigned to sub-area B.

[0083] b3. Extracting a sub-region as a complex terrain region from all sub-regions allocated with initial point cloud data, controlling the target synthetic aperture radar to scan the complex terrain region multiple times according to the target radar beam incident angle inclination to obtain position information of the second reflection point, wherein the target radar beam incident angle inclination is greater than the adjusted radar beam incident angle.

[0084] Among them, the initial point cloud data is the coordinates of the first reflection point assigned to the sub-area, the sub-area of ​​the complex terrain area is a sub-area with large terrain undulations and complex structures, the target synthetic aperture radar is a device used to transmit and receive detection signals, the target radar beam incident angle inclination is an inclination angle larger than the previously adjusted angle (used to increase the point cloud density), the position information of the second reflection point is the three-dimensional coordinates of the reflection point newly obtained after multiple scans, and the result obtained after processing is the position information of the second reflection point.

[0085] In an embodiment of the present application, first, from the sub-areas assigned with initial point cloud data, sub-areas of complex terrain areas are identified according to the complexity values, and then the target synthetic aperture radar is controlled to use a target radar beam incident angle inclination larger than before to scan these areas multiple times, and a new reflection signal is obtained each time the scan is performed. Finally, the position information of the second reflection point is determined based on these signals. For example, sub-area B of area A is an area with complex terrain. The previous inclination was 50 degrees, and the target inclination was set to 65 degrees. After scanning it twice, the coordinates of 25 second reflection points are obtained.

[0086] b4. Add the position information of the second reflection point to the initial point cloud data of the corresponding complex terrain area to obtain target point cloud data of the complex terrain area, so as to increase the point cloud density of the complex terrain area.

[0087] Among them, the position information of the second reflection point is the newly acquired three-dimensional coordinates of the reflection point, the initial point cloud data of the complex terrain area is the original coordinates of the first reflection point in the area, the target point cloud data is the coordinate set of all reflection points in the area after adding the second reflection point, and the point cloud density is the number of reflection points per unit area. The result obtained after processing is the target point cloud data of the complex terrain area.

[0088] In an embodiment of the present application, the position information of the second reflection point obtained in step b3 is first collected, and then these coordinates are added to the initial point cloud data of the corresponding complex terrain area, merged with the original first reflection point coordinates, and finally the target point cloud data of the area is obtained. Due to the addition of new reflection points, the point cloud density (total number of points ÷ area of ​​the area) is improved. For example, sub-area B of area A initially has 10 points (area 20 square meters, density 0.5 / square meter). After adding 25 second reflection points, the total number of points is 35, and the density becomes 35 ÷ 20 = 1.75 / square meter.

[0089] b5. By comparing the point cloud density corresponding to each sub-area, verify that the point cloud density of the complex terrain area is higher than that of the non-complex terrain area in the terrain area to be measured. After verification, generate point cloud data with non-uniform resolution based on the target point cloud data of the complex terrain area and the initial point cloud data of the non-complex terrain area.

[0090] Among them, the point cloud density is the number of reflection points per unit area. The point cloud density of the complex terrain area is the density of the target point cloud data in the area. The non-complex terrain area is a sub-area with flat terrain, and its initial point cloud data is the coordinates without adding the second reflection point. The point cloud data with non-uniform resolution is a data set in which the point cloud density in the complex terrain area is higher than that in the non-complex terrain area. The result obtained after processing is point cloud data with non-uniform resolution.

[0091] In an embodiment of the present application, the point cloud density of each sub-area is first calculated separately (target point cloud data is used for complex terrain areas, and initial point cloud data is used for non-complex areas), and then these densities are compared to verify whether the density of the complex terrain area is higher than that of the non-complex area. After the verification is passed, the target point cloud data of the complex terrain area and the initial point cloud data of the non-complex area are integrated to generate point cloud data with non-uniform resolution. For example, the density of sub-area B (complex) in area A is 1.75 points / square meter, and the density of sub-area C (non-complex) is 0.25 points / square meter. After comparison, the density of B is higher, and a non-uniform data set is formed after integration.

[0092] For example, in the terrain processing of area A, step b1 first collects the coordinates of the first reflection points (10, 20, 5), (15, 25, 8), etc., and arranges them in three-dimensional coordinates to form an initial point cloud data set; step b2 distributes these points to sub-areas, with 10 points in sub-area B (area 20 square meters) and 5 points in sub-area C (area 20 square meters); step b3 identifies sub-area B as an area with complex terrain, scans it twice with an inclination of 65 degrees (greater than the previous 50 degrees), and obtains 25 second reflection points; step b4 adds these 25 points to sub-area B, for a total of 35 points with a density of 35 ÷ 20 = 1.75 points per square meter; step b5 calculates the density of sub-area C as 5 ÷ 20 = 0.25 points per square meter. After comparison, the density of B is higher (verified), and the data of the two areas are integrated to generate point cloud data with non-uniform resolution.

[0093] By executing steps b1 to b5, the embodiment of the present application forms an initial point cloud by arranging the coordinates of the reflection points and assigning them to sub-areas, providing a basis for targeted processing; multiple scans with greater inclinations are used to obtain more reflection points in areas with complex terrain, directly increasing their point cloud density; and verification ensures that the density of complex areas is higher than that of simple areas. The non-uniform resolution point cloud data ultimately generated can not only meticulously reflect the details of complex terrain, but also avoid data redundancy in simple areas, taking into account both data accuracy and processing efficiency, and providing high-quality data support for subsequent three-dimensional terrain modeling.

[0094] In one possible embodiment, the high-resolution terrain features include texture information and geometric information. S13: A dual-branch processing model is constructed, wherein the first branch processes the target multispectral image and the second branch processes the point cloud data. The high-resolution terrain features output by the two branches are fused through a weighted mechanism to obtain fused features, including: Step 131: Decompose the target multispectral image into single-band images corresponding to multiple bands based on the first branch, determine the spatial distribution pattern of pixel values ​​in each single-band image, and extract texture information corresponding to the terrain features based on the spatial distribution pattern.

[0095] Among them, the first branch is the part that processes the target multispectral image. The target multispectral image is a terrain image containing multiple wavelengths of light. The single-band image is the image corresponding to a single wavelength after the multispectral image is decomposed. The spatial distribution pattern of pixel values ​​is the arrangement and change law of pixel values ​​in the single-band image. The texture information is the surface features related to the terrain features extracted based on this pattern. The result obtained after processing is the texture information corresponding to the terrain features.

[0096] In an embodiment of the present application, the target multispectral image is first decomposed into multiple single-band images according to different light wavelengths based on the first branch, and then the high and low distribution and changes of pixel values ​​in each single-band image are observed to determine its spatial distribution pattern. Finally, according to the aggregation or dispersion characteristics of pixels in the pattern, texture information corresponding to the terrain features is extracted. For example, the multispectral image of area A is decomposed into single-band images of red light and blue light. The pixel values ​​of flat land in the red light band are evenly distributed, while the pixel values ​​of steep slopes vary greatly. Based on this, texture information reflecting the roughness of different terrain surfaces is extracted.

[0097] Step 132: Based on the second branch, the point cloud data is organized into multiple data groups according to the spatial position relationship, the distribution density and the corresponding relative position relationship of the point cloud data of each data group are analyzed, and the geometric information corresponding to the terrain features is extracted.

[0098] Among them, the second branch is the part that processes point cloud data. Point cloud data is terrain data composed of a large number of three-dimensional point coordinates. The data group is a point cloud subset organized by spatial position (such as close distance). The distribution density is the number of points per unit volume in the data group. The relative position relationship is the distance and orientation relationship of the points in the group. The geometric information is the features related to the three-dimensional structure of the terrain extracted based on density and position. The result obtained after processing is the geometric information corresponding to the terrain features.

[0099] In an embodiment of the present application, the point cloud data is first grouped according to the spatial position relationship based on the second branch (such as grouping points with close distances together) to form multiple data groups. Then, the point distribution density of each data group (the total number of points divided by the spatial volume within the group) and the relative positions between the points (such as height difference, distance) are analyzed. Finally, based on this information, geometric information corresponding to the terrain features is extracted. For example, in the point cloud data of area B, the points in a certain data group are dense and the height difference is as high as 30 meters. The geometric information reflecting the steep terrain of the area is extracted.

[0100] Step 133: Determine the credibility weight of the texture information based on the shooting parameters of the target multispectral image, and determine the credibility weight of the geometric information based on the acquisition parameters of the point cloud data.

[0101] Among them, the shooting parameters are the conditions when shooting multispectral images (such as lighting and device resolution), the acquisition parameters are the conditions when obtaining point cloud data (such as radar accuracy and number of scans), the credibility weight is a numerical value that measures the reliability of information, the credibility weight of texture information is the weight determined according to the shooting parameters, and the credibility weight of geometric information is the weight determined according to the acquisition parameters. The result obtained after processing is the credibility weight of texture information and geometric information respectively.

[0102] In an embodiment of the present application, the credibility weight of the texture information is first determined based on the shooting parameters of the target multispectral image. If the lighting is sufficient and the device resolution is high, the texture information is more reliable and the weight is high, otherwise it is low. Then, the credibility weight of the geometric information is determined based on the acquisition parameters of the point cloud data. If the radar accuracy is high and the number of scans is large, the geometric information is more reliable and the weight is high, otherwise it is low. For example, when the multispectral image of area A is shot, the lighting is good, the weight of the texture information is set to 0.5. When the radar accuracy is high during point cloud data acquisition, the weight of the geometric information is set to 0.5.

[0103] Step 134: weightedly combine the texture information and the geometric information according to the corresponding credibility weights to obtain a fusion feature.

[0104] Among them, texture information is the terrain surface features extracted from multispectral images, geometric information is the three-dimensional structural features extracted from point cloud data, credibility weight is a numerical value to measure the reliability of the two, weighted merging is the process of combining the two types of information according to the weight, fusion features are the comprehensive terrain features obtained after merging, and the result obtained after processing is the fusion feature.

[0105] In an embodiment of the present application, texture information and geometric information are first obtained, then the texture information is multiplied by its credibility weight, and the geometric information is multiplied by its credibility weight. Finally, the two results are added together to obtain a fusion feature. For example, the quantization value of the texture information is 0.8 (weight 0.5), and the quantization value of the geometric information is 0.7 (weight 0.5). The fusion feature is 0.8×0.5+0.7×0.5=0.75, which comprehensively reflects the surface and structural characteristics of the terrain.

[0106] For example, when processing terrain data in area A, step 131 decomposes the multispectral image into single-band images of red and green light. In the red band, the pixels on flat ground are evenly distributed, while the pixels on steep slopes vary greatly, so texture information (quantization value 0.8) is extracted. Step 132 groups the point cloud data by location. A certain group of points is dense and has a large height difference, so geometric information (quantization value 0.7) is extracted. Step 133 determines that the credibility weights of texture and geometric information are both 0.5 based on the shooting parameters (good lighting) and acquisition parameters (high radar accuracy). Step 134 calculates the fusion feature as 0.8×0.5+0.7×0.5=0.75, obtaining a fusion feature that integrates the two types of information.

[0107] By executing steps 131 to 134, the embodiment of the present application extracts texture information of the terrain surface through the first branch, extracts geometric information of the terrain structure through the second branch, assigns reasonable credibility weights in combination with the shooting and acquisition parameters, and finally obtains fusion features through weighted merging, which not only retains the detailed features of the terrain surface but also integrates the three-dimensional structural information, making the description of the terrain features more comprehensive; at the same time, the introduction of credibility weights allows highly reliable information to account for a larger proportion in the fusion, improves the accuracy of the fusion features, and provides high-quality basic data for the subsequent generation of accurate three-dimensional terrain models.

[0108] In one possible embodiment, step 131, based on the first branch, decomposing the target multispectral image into single-band images corresponding to the multiple bands, determining a spatial distribution pattern of pixel values ​​in each single-band image, and extracting texture information corresponding to the terrain features based on the spatial distribution pattern, includes: c1. Split the target multispectral image into multiple independent single-band images according to the preset light wavelength range.

[0109] Among them, the target multispectral image is a terrain image containing information on multiple light wavelengths, the preset light wavelength range is a pre-set interval of different light wavelengths, and the single-band image is an image corresponding to a single interval after the multispectral image is split according to the preset wavelength range. The result generated after processing is multiple independent single-band images.

[0110] In an embodiment of the present application, the target multispectral image is first split into multiple independent single-band images according to a preset light wavelength range, and each single-band image corresponds to a wavelength range. For example, the multispectral image of area A is split according to the preset wavelength ranges of red light (600-700 nanometers) and green light (500-600 nanometers) to obtain a red light single-band image and a green light single-band image.

[0111] c2. Based on the values ​​of adjacent pixels in each single-band image, determine the locations where the difference in the values ​​of adjacent pixels in each single-band image is greater than a preset difference threshold. Divide the corresponding single-band image based on the locations where the difference in the values ​​is greater than the preset difference threshold, and record the distribution characteristics of each area in the single-band image.

[0112] Among them, the numerical difference of adjacent pixels is the difference between the numerical values ​​of two adjacent pixels in a single-band image. The preset difference threshold is a pre-set value for judging whether the pixel change is obvious. The position where the numerical difference is greater than the preset difference threshold is the boundary position where the pixel change is obvious. The distribution characteristic is the arrangement pattern of pixel values ​​in each area after division. The result generated after processing is the divided area and the distribution characteristics of each area.

[0113] In an embodiment of the present application, the numerical difference between adjacent pixels in each single-band image is first calculated, the difference is compared with a preset difference threshold, and the position where the difference is greater than the threshold is determined. The single-band image is divided into multiple regions based on these positions, and the distribution characteristics of each region are recorded. For example, in the red light single-band image of region A, the value of pixel A is 80, the value of the adjacent pixel B is 50, the difference is 30, and the preset difference threshold is 20. 30 is greater than 20, so this position is determined to be the boundary, and three regions are divided accordingly. The pixel values ​​in region 1 are concentrated between 70 and 80, and the values ​​in region 2 are dispersed.

[0114] c3. Determine the spatial distribution pattern of pixel values ​​in each single-band image based on the distribution characteristics of each area.

[0115] Among them, the distribution characteristics of the region are the arrangement rules of pixel values ​​in the region (such as concentrated or dispersed), and the spatial distribution pattern of pixel values ​​is the overall pixel arrangement pattern of the region summarized based on the distribution characteristics (such as uniform or strip-shaped). The result generated after processing is the spatial distribution pattern of pixel values ​​in each single-band image.

[0116] In an embodiment of the present application, the distribution characteristics of each area in each single-band image are first sorted out, and the overall arrangement rules of each area are summarized based on these characteristics. The areas with similar rules are merged to determine the spatial distribution pattern of the pixel values ​​of the entire single-band image. For example, in the green light single-band image of area A, the pixel values ​​in area 1 are uniform and vary little, while the values ​​in area 2 vary greatly and are scattered, which is summarized as a spatial distribution pattern of "uniform distribution + scattered distribution".

[0117] c4. Based on the shape, terrain undulation and terrain morphology of each area in the spatial distribution pattern, texture information corresponding to the terrain features is extracted from the single-band image.

[0118] Among them, the spatial distribution pattern is the overall arrangement law of pixel values, the shape of the area is the appearance of the divided area (such as block or strip), the terrain undulation is the height change of the terrain, the terrain morphology is the overall appearance of the terrain (such as flat or steep), and the texture information is the surface texture corresponding to the terrain characteristics extracted from the distribution pattern. The result generated after processing is the texture information corresponding to the terrain characteristics.

[0119] In an embodiment of the present application, the shape, terrain undulation and terrain morphology of each area in the spatial distribution pattern are first analyzed, and texture information corresponding to the terrain features is extracted from the single-band image based on this information. For example, in the red light single-band image of area A, area 1 is blocky, with small undulations and a flat shape, and a "smooth" texture is extracted; area 2 is long strip-shaped, with large undulations and a steep shape, and a "rough" texture is extracted.

[0120] For example, when processing the target multispectral image of area A, step d1 splits the image into two single-band images according to the red light (600-700 nanometers) and green light (500-600 nanometers) ranges; step d2 calculates the adjacent pixel values ​​of 80 and 50 in the red light single-band image, with a difference of 30, and a preset threshold of 20, 30>20, to determine the boundary position and divide area 1 (pixel value 70-80, concentrated distribution) and area 2 (pixel value 30-50, dispersed distribution) into "concentrated distribution + dispersed distribution" based on the distribution characteristics; step d4 analyzes area 1, which is blocky, with small undulations and flat shape, and extracts "smooth" texture; area 2 is long, with large undulations and steep shape, and extracts "rough" texture.

[0121] By executing steps c1 to c4, the embodiment of the present application splits the multispectral image into single-band images, providing a basis for targeted analysis of terrain features at different wavelengths; calculates pixel differences and divides regions to accurately locate boundaries where terrain changes are obvious; summarizes spatial distribution patterns and extracts the overall regularity of pixel arrangement; and finally extracts texture information that accurately reflects the surface features of the terrain (such as smoothness and roughness), providing realistic surface texture data for subsequent terrain modeling, thereby improving the accuracy of terrain feature description.

[0122] In a possible embodiment, S14, dynamically associating the vegetation index distribution of the terrain area to be measured with the elevation information in the fusion feature to generate a spatiotemporal dynamic mapping result of the vegetation coverage area terrain and crop type, includes: Step 141 : Select an image of a specific band from the target multispectral image, and obtain the vegetation index distribution of the terrain area to be measured by calculating the proportional relationship of the pixel values ​​corresponding to vegetation in the single-band image of the specific band.

[0123] Among them, the target multispectral image is a terrain image containing information on multiple light wavelengths; the image of a specific band is an image corresponding to a light wavelength that has a significant vegetation response, selected from the target multispectral image; the pixel value corresponding to the vegetation is the pixel value of the vegetation area in the image of the specific band; the proportional relationship is the ratio of the vegetation pixel values ​​in images of different specific bands; the vegetation index distribution is a numerical distribution calculated through the proportional relationship, reflecting the vegetation growth conditions in the terrain area to be measured; the result generated after processing is the vegetation index distribution of the terrain area to be measured.

[0124] In an embodiment of the present application, first, specific band images (such as red light and near-infrared bands) that have obvious vegetation responses are selected from the target multispectral image, and then the proportional relationship of the pixel values ​​corresponding to the vegetation in these specific band images is calculated. Specifically, the vegetation pixel value of the red light band is subtracted from the vegetation pixel value of the near-infrared band, and then divided by the sum of the vegetation pixel values ​​of these two bands. Finally, the vegetation index distribution of the terrain area to be measured is obtained according to the calculation results. For example, in area A, red light and near-infrared band images are selected. At a certain position, the red light pixel value is 20, and the near-infrared pixel value is 80. The proportional relationship is (80-20) ÷ (80+20) = 60 ÷ 100 = 0.6. The vegetation index at this position is 0.6, thereby obtaining the vegetation index distribution of the entire area.

[0125] Step 142: extract elevation information from the fused features, and determine the range of elevation values ​​in the terrain area to be measured based on the elevation information.

[0126] Among them, the fusion feature is the terrain information that integrates the features of multispectral images and point cloud data; the elevation information is the height value of each location in the terrain; the range of elevation value variation is the height difference between the highest and lowest points in the terrain area to be measured; the result generated after processing is the range of elevation value variation in the terrain area to be measured.

[0127] In an embodiment of the present application, the elevation information (i.e., height value) of each location is first extracted from the fused features, and then the maximum and minimum values ​​in these elevation information are found. Finally, the range of variation of the elevation value is obtained by subtracting the minimum value from the maximum value. For example, in area A, the maximum value of the elevation information extracted from the fused features is 400 meters, the minimum value is 150 meters, and the range of variation of the elevation value is 400-150=250 meters.

[0128] Step 143: Correlate the vegetation index distribution with the elevation information according to the spatial position, and obtain the correlation degree between the vegetation index distribution value and the elevation value at each spatial position.

[0129] Among them, the vegetation index distribution is the numerical distribution reflecting the vegetation growth situation; the elevation information is the height value of each terrain location; the spatial position is the specific location of each point in the terrain; the degree of association is the closeness of the connection between the vegetation index value and the elevation value at the same spatial location; the result generated after processing is the degree of association between the vegetation index and elevation at each spatial location.

[0130] In an embodiment of the present application, the vegetation index value and the corresponding elevation value of each spatial position are first determined, and then the degree of correlation between the two values ​​is calculated (the degree of correlation is high when the value change trends are consistent, and the degree of correlation is low when the trends are opposite). Finally, the degree of correlation for each spatial position is recorded. For example, in area A, the vegetation index at position B is 0.5 and the elevation is 200 meters, and the vegetation index at position C is 0.75 and the elevation is 350 meters. Both values ​​increase with increasing position, and the degree of correlation is 0.8; the vegetation index at position D is 0.3 and the elevation is 250 meters. The value changes are inconsistent, and the degree of correlation is 0.4.

[0131] Step 144: Adjust the degree of correlation according to the range of change of the elevation value, and generate a spatiotemporal dynamic mapping result based on the adjusted degree of correlation and the growth law of the crop type under the elevation information and vegetation index distribution.

[0132] The range of elevation values ​​represents the overall undulation of the terrain; the degree of correlation represents the closeness of the connection between the vegetation index and the elevation; the crop type represents the plant species planted on the terrain; the growth pattern represents the growth characteristics of different crops at different altitudes and vegetation indices; the spatiotemporal dynamic mapping result reflects the corresponding relationship between the terrain and crop types in the vegetation coverage area over time; and the result generated after processing is the spatiotemporal dynamic mapping result.

[0133] In an embodiment of the present application, the degree of association is first adjusted according to the range of change of the elevation value (the degree of association is increased in areas with large elevation changes, and the degree of association is reduced in areas with small changes), and then combined with the growth patterns of different crops (for example, a certain crop grows well at high altitudes and high vegetation indices), and finally the association results at different time points are recorded to generate a correspondence between the terrain of the vegetation coverage area and the crop type over time, that is, a spatiotemporal dynamic mapping result. For example, in area A, the degree of association of the mountainous area with large elevation changes is adjusted to 0.9. In spring, a certain location (elevation 300 meters, index 0.8) corresponds to crop B, and in summer, the index of the location becomes 0.9, which still corresponds to crop B.

[0134] For example, in processing area A, step 141 selects red and near-infrared band images. At location B, the red pixel value is 30, the near-infrared pixel value is 90, and the vegetation index is calculated as (90-30) ÷ (90+30) = 60 ÷ 120 = 0.5. At location C, the red pixel value is 10, the near-infrared pixel value is 70, and the vegetation index is (70-10) ÷ (70+10) = 60 ÷ 80 = 0.75, resulting in a vegetation index distribution. Step 142 extracts elevation information from the fused features, with a maximum value of 400 meters and a minimum value of 150 meters, for an elevation range of 400-150 = 250 meters. In step 143, the values ​​at locations B (0.5, 200 meters) and C (0.75, 350 meters) both increase with increasing location, with a correlation of 0.8. Location D (0.6, 220 meters) has a higher elevation and index than location B, with a correlation of 0.7. In step 144, the elevation of the mountain in area A changes by 150 meters, and the correlation degree is adjusted to 0.9; the elevation of the flat land changes by 50 meters, and the correlation degree is adjusted to 0.6. Considering that crop B grows best at high altitudes and high indexes, position B corresponds to crop B in spring, and the index of this position changes to 0.6 in summer, still corresponding to crop B, to generate a spatiotemporal dynamic mapping result.

[0135] By executing steps 141 to 144, the embodiment of the present application reflects the vegetation growth situation by calculating the vegetation index distribution, determines the elevation change range to reflect the terrain undulation, establishes the correlation degree between vegetation and elevation and dynamically adjusts it, and finally generates a spatiotemporal dynamic mapping result in combination with the crop growth law, so that the terrain model not only contains terrain height information, but also incorporates the dynamic change characteristics of vegetation and crops, providing a more comprehensive and practical reference basis for flight simulation terrain modeling and path planning.

[0136] Figure 2 A schematic diagram of a multi-spectral fusion flight simulation terrain generation system provided in an embodiment of the present application is shown in FIG. Figure 2 As shown, the system includes: The acquisition module 21 is used to collect original multispectral images and original synthetic aperture radar data of the terrain area to be measured, identify cloud coverage based on the pixel value distribution of the original multispectral images, and screen target multispectral images and target synthetic aperture radar data of the corresponding time window with cloud coverage below a set threshold.

[0137] The extraction module 22 is used to extract terrain edge features from the target multispectral image and dynamically adjust the radar beam incident angle based on the terrain edge features to generate point cloud data with non-uniform resolution. During the generation process, the point cloud density in complex terrain areas, such as steep slopes and canyon areas, is increased by increasing the inclination of the radar beam incident angle.

[0138] Construction module 23 is used to construct a dual-branch processing model, in which the first branch processes the target multispectral image and the second branch processes the point cloud data. The high-resolution terrain features output by the two branches are fused through a weighted mechanism to obtain fused features.

[0139] The association module 24 is used to dynamically associate the vegetation index distribution of the terrain area to be measured with the elevation information in the fusion feature to generate a spatiotemporal dynamic mapping result of the terrain and crop types in the vegetation coverage area.

[0140] The generation module 25 is used to generate a three-dimensional flight simulation terrain model with semantic labels based on the spatiotemporal dynamic mapping results, which is used for planning the flight simulation path.

[0141] Figure 2 The multi-spectral fusion flight simulation terrain generation system can perform Figure 1 The implementation principles and technical effects of the multispectral fusion flight simulation terrain generation method described in the illustrated embodiment will not be elaborated upon. The specific manner in which the various modules and units of the multispectral fusion flight simulation terrain generation system in the aforementioned embodiment perform their operations has been described in detail in the relevant embodiments of the method and will not be further elaborated here.

[0142] In one possible design, Figure 2 The multi-spectral fusion flight simulation terrain generation system of the embodiment shown can be implemented as a computing device, such as Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32 .

[0143] The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32. The processing component 32 is used to execute the specific process of the multi-spectral fusion flight simulation terrain generation method.

[0144] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented as one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above method.

[0145] The storage component 31 is configured to store various types of data to support operations on the terminal. The storage component can be implemented by any type of volatile or non-volatile memory device, or a combination thereof, such as random access memory (RAM), static random access memory (SRAM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0146] Of course, a computing device may also include other components, such as input / output interfaces, display components, and communication components. The input / output interfaces provide an interface between the processing component and peripheral interface modules, which may be output devices, input devices, etc. The communication components are configured to facilitate wired or wireless communication between the computing device and other devices.

[0147] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. In this case, the computing device can refer to a cloud server, and the above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from the cloud computing platform.

[0148] The present application also provides a computer storage medium storing a computer program, wherein the computer program can achieve the above-mentioned Figure 1 The flight simulation terrain generation method based on multi-spectral fusion in the illustrated embodiment.

[0149] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0150] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of this embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0151] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0152] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A multi-spectral fusion flight simulation terrain generation method, characterized in that: include: Collecting original multispectral images and original synthetic aperture radar data of the terrain area to be measured, identifying cloud cover based on the pixel value distribution of the original multispectral images, and screening target multispectral images and target synthetic aperture radar data of the corresponding time window with cloud cover below a set threshold; Extracting terrain edge features from the target multispectral image, and dynamically adjusting the radar beam incident angle based on the terrain edge features to generate point cloud data with non-uniform resolution. During the generation process, increasing the inclination of the radar beam incident angle increases the point cloud density in areas with complex terrain, such as steep slopes and canyon areas. A dual-branch processing model is constructed, where the first branch processes the target multispectral image and the second branch processes the point cloud data, and the high-resolution terrain features output by the two branches are fused through a weighted mechanism to obtain a fused feature; Dynamically associating the vegetation index distribution of the terrain area to be measured with the elevation information in the fusion feature to generate a spatiotemporal dynamic mapping result of the terrain and crop types in the vegetation coverage area; Based on the spatiotemporal dynamic mapping results, a three-dimensional flight simulation terrain model with semantic labels is generated for flight simulation path planning.

2. The method for generating terrain for flight simulation based on multi-spectral fusion according to claim 1, characterized in that: The dynamically adjusting the radar beam incident angle based on the terrain edge features to generate point cloud data with non-uniform resolution, and increasing the inclination of the radar beam incident angle during the generation process to improve the point cloud density in the complex terrain area, includes: By analyzing the terrain edge features, the contour of the terrain area to be measured is generated, the curvature change rate of the contour is calculated, and the terrain edge characteristic line is generated based on the curvature change rate; Based on the spatial distribution pattern of the terrain edge characteristic lines, the terrain area to be measured is divided into a plurality of sub-areas with similar terrain characteristics, and the corresponding dominant slope direction of each sub-area is determined; Calculating a complexity value of each sub-area based on a density value of the terrain edge characteristic line and a corresponding dominant slope aspect of each sub-area, and adjusting a radar beam incident angle according to the complexity value; Controlling the radar to scan each of the sub-areas according to the adjusted radar beam incident angle, obtaining a reflection signal from each of the sub-areas, and analyzing the return time and strength of the reflection signal to determine the position of the first reflection point; An initial point cloud data set is generated based on the position of the first reflection point. In the initial point cloud data, the point cloud density of the complex terrain area is increased by increasing the inclination of the radar beam incident angle to generate point cloud data with non-uniform resolution.

3. The method for generating terrain for flight simulation based on multi-spectral fusion according to claim 2, characterized in that: Calculating the complexity value of each sub-region based on the density value of the terrain edge feature line and the corresponding dominant slope direction of each sub-region includes: Counting the total length of all the terrain edge feature lines in each sub-region, and dividing the total length by the area of ​​the corresponding sub-region to obtain a density value of the terrain edge feature lines in the unit sub-region; Analyze the angle range between the corresponding dominant slope direction of each sub-area and the true north direction, and divide the angle range into multiple angle intervals; Determine the corresponding aspect coefficient according to the angle interval of the dominant aspect of each sub-area; Calculating the average curvature of the terrain edge characteristic lines in each sub-area, and dividing the average curvature into multiple levels, each level of the average curvature corresponding to a different complexity correction factor; The product of the density value of the terrain edge characteristic line in the unit sub-region, the corresponding slope aspect coefficient, and the corresponding complexity correction factor is used as the complexity value of each sub-region.

4. The method for generating terrain for flight simulation based on multi-spectral fusion according to claim 2, characterized in that: Generating an initial point cloud data set based on the position of the first reflection point, and increasing the point cloud density of a complex terrain area in the initial point cloud data by increasing the inclination of the radar beam incident angle to generate point cloud data with non-uniform resolution, includes: Arranging the position information of all the first reflection points according to three-dimensional space coordinates to generate an initial point cloud data set; Allocating each initial point cloud data in the initial point cloud data set to a corresponding sub-region; extracting a subregion as a complex terrain region from all subregions allocated with initial point cloud data, controlling the target synthetic aperture radar to scan the complex terrain region multiple times according to a target radar beam incident angle inclination to obtain position information of a second reflection point, wherein the target radar beam incident angle inclination is greater than the adjusted radar beam incident angle; Adding the position information of the second reflection point to the initial point cloud data of the corresponding complex terrain area to obtain target point cloud data of the complex terrain area, so as to increase the point cloud density of the complex terrain area; By comparing the point cloud densities corresponding to each of the sub-areas, it is verified that the point cloud density of the complex terrain area is higher than the point cloud density of the non-complex terrain area in the terrain area to be measured. After the verification is passed, point cloud data with non-uniform resolution is generated based on the target point cloud data of the complex terrain area and the initial point cloud data of the non-complex terrain area.

5. The method for generating terrain for flight simulation based on multi-spectral fusion according to claim 1, characterized in that: The high-resolution terrain features include texture information and geometric information; The dual-branch processing model is constructed, where the first branch processes the target multispectral image and the second branch processes the point cloud data. The high-resolution terrain features output by the two branches are fused through a weighted mechanism to obtain fused features, including: Decomposing the target multispectral image into single-band images corresponding to the plurality of bands based on the first branch, determining a spatial distribution pattern of pixel values ​​in each of the single-band images, and extracting texture information corresponding to the terrain features based on the spatial distribution pattern; Based on the second branch, the point cloud data is organized into multiple data groups according to spatial position relationships, and the distribution density and relative position relationship of the point cloud data of each data group are analyzed to extract geometric information corresponding to the terrain features; Determining the credibility weight of the texture information based on the shooting parameters of the target multispectral image, and determining the credibility weight of the geometric information based on the acquisition parameters of the point cloud data; The texture information and the geometric information are weighted and merged according to corresponding credibility weights to obtain a fusion feature.

6. The method for generating terrain for flight simulation based on multi-spectral fusion according to claim 5, characterized in that: The method of decomposing the target multispectral image into single-band images corresponding to a plurality of bands based on the first branch, determining a spatial distribution pattern of pixel values ​​in each of the single-band images, and extracting texture information corresponding to terrain features based on the spatial distribution pattern includes: Splitting the target multispectral image into multiple independent single-band images according to a preset light wavelength range; Based on the values ​​of adjacent pixels in each of the single-band images, determining positions where a difference in the values ​​of adjacent pixels in each of the single-band images is greater than a preset difference threshold, dividing the corresponding single-band image according to the positions where the difference in the values ​​is greater than the preset difference threshold, and recording the distribution characteristics of each area in the single-band image; Determining a spatial distribution pattern of pixel values ​​in each of the single-band images according to the distribution characteristics of each region; According to the shape, terrain relief and terrain morphology of each area in the spatial distribution pattern, texture information corresponding to the terrain features is extracted from the single-band image.

7. The method for generating terrain for flight simulation based on multi-spectral fusion according to claim 1, characterized in that: The method of dynamically associating the vegetation index distribution of the terrain area to be measured with the elevation information in the fusion feature to generate a spatiotemporal dynamic mapping result of the terrain and crop types in the vegetation coverage area includes: Selecting an image of a specific band from the target multispectral image, and obtaining a vegetation index distribution of the terrain area to be measured by calculating a proportional relationship of pixel values ​​corresponding to vegetation in a single-band image of the specific band; Extracting elevation information from the fused features, and determining a variation range of elevation values ​​in the terrain area to be measured based on the elevation information; Associating the vegetation index distribution with the elevation information according to spatial positions to obtain a correlation degree between the vegetation index distribution value and the elevation value at each spatial position; According to the variation range of the elevation values, the degree of association is adjusted, and based on the adjusted degree of association, combined with the growth pattern of the crop type under the elevation information and the vegetation index distribution, a spatiotemporal dynamic mapping result is generated.

8. A multi-spectral fusion flight simulation terrain generation system, characterized by: include: An acquisition module is configured to acquire original multispectral images and original synthetic aperture radar data of the terrain area to be measured, identify cloud coverage based on the pixel value distribution of the original multispectral images, and screen target multispectral images and target synthetic aperture radar data of the corresponding time window whose cloud coverage is lower than a set threshold; an extraction module for extracting terrain edge features from the target multispectral image, dynamically adjusting the radar beam incident angle based on the terrain edge features to generate point cloud data with non-uniform resolution, and increasing the point cloud density in areas with complex terrain, such as steep slopes and canyons, by increasing the inclination of the radar beam incident angle during the generation process; A construction module is used to construct a dual-branch processing model, wherein the first branch processes the target multispectral image and the second branch processes the point cloud data, and the high-resolution terrain features output by the two branches are fused through a weighted mechanism to obtain a fused feature; An association module is used to dynamically associate the vegetation index distribution of the terrain area to be measured with the elevation information in the fusion feature to generate a spatiotemporal dynamic mapping result of the terrain and crop types in the vegetation coverage area; A generation module is used to generate a three-dimensional flight simulation terrain model with semantic labels based on the spatiotemporal dynamic mapping result, which is used for planning the flight simulation path.

9. A computing device, characterized in that It includes a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a multi-spectral fusion flight simulation terrain generation method as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that A computer program is stored, and when the computer program is executed by a computer, the method for generating multi-spectral fusion flight simulation terrain according to any one of claims 1 to 7 is implemented.

Images