Grid field governance technology based on UAV LiDAR images

Abstract

The present application relates to a method and device for generating a field governance plan based on LIDAR images. The method includes generating a vector map of the fields in the target area based on the panoramic RGB images of the fields in the target area and the actual division of the fields in the target area, and generating a digital surface model based on the panoramic LIDAR images; obtaining an elevation information image of the fields in the target area according to the vector map and the digital surface model; constructing a basic information table including the basic information of each field in the target area by using the elevation information image; associating the elevation information image with the basic information table to generate an elevation thematic map of the fields in the target area; and merging the fields in the elevation thematic map to generate a planning plan for the fields in the target area. The present application effectively improves the accuracy of elevation information, and when large-area fields are merged, it not only reduces the occupied area of the ridges, but also greatly improves the flatness of the fields in the merged planning plan.

Description

Technical Field

[0001] This application relates to the field of field improvement technology, and in particular to a method and device for generating a field improvement plan based on LIDAR images. Background Art

[0002] Farmland is the basic condition for growing food. By implementing measures such as expanding fields, reducing ridges and increasing fields, and shaping the edges of paddy fields for field improvement, the land utilization rate per unit area can be increased as a standard, thereby achieving an increase in rice production capacity.

[0003] Currently, there is little research on this work at home and abroad. In actual production, surveying instruments such as RTK and level are still widely used to conduct on-site surveys of the field elevation, and then the obtained data is processed by a computer. This method has certain limitations, mainly manifested in the following aspects:

[0004] (1) The traditional technology for obtaining field elevation is time-consuming and laborious, with low efficiency and low automation, and requires artificial-assisted decision-making. There are large errors in manual measurement.

[0005] (2) When dealing with large-area field improvement, the number of sampling points is still insufficient, and it is impossible to obtain the elevation information within pixels and dynamically obtain the elevation information of each point, resulting in limited identification of the flatness inside the field. Summary of the Invention

[0006] In order to solve the technical problems of large errors in elevation information in current field improvement and limited identification of the flatness inside the field due to insufficient sampling points in large-area field improvement, resulting in the inability to obtain elevation information within pixels and dynamically obtain the elevation information of each point, this application provides a method and device for generating a field improvement plan based on LIDAR images.

[0007] In the first aspect, this application provides a method for generating a field improvement plan based on LIDAR images, adopting the following technical solution:

[0008] A method for generating a field improvement plan based on LIDAR images includes:

[0009] Generating a vector map of the fields in the target area based on the panoramic RGB image of the fields in the target area and the actual division of the fields in the target area, where the vector map is marked with the vector boundaries of each field in the target area, and the vector boundary of each field has a unique code;

[0010] Generating a digital surface model based on the panoramic LIDAR image of the fields in the target area;

[0011] Adding the vector boundaries of each field plot and identifying corresponding unique codes for the vector boundaries on the obtained digital surface model according to the vector map to obtain the elevation information image of the field plots in the target area;

[0012] Using the elevation information image and the resolution to obtain the basic information of each field plot, and constructing a basic information table including the basic information of each field plot in the target area field plots; wherein, the basic information includes the maximum elevation, the minimum elevation, the average elevation, the field plot area and the corresponding unique code;

[0013] Associating the elevation information image and the basic information table through the unique code to generate the elevation thematic map of the field plots in the target area;

[0014] Invoking a preset merging rule and merging the field plots in the elevation thematic map according to the preset merging rule to generate a planning scheme for the field plots in the target area.

[0015] By adopting the above technical solution, in the process of obtaining the elevation information image, the elevation information of the panoramic LIDAR image can be continuously used, so that the elevation information of the field plots in the target area can be obtained with centimeter-level error from the elevation information image, effectively improving the accuracy of the elevation information; since each pixel point on the elevation information image corresponds to an elevation information, there are enough points with elevation information to support the merging of large-area field plots (leveling), which not only reduces the occupied area of the ridge, but also greatly improves the flatness of the field plots in the merged planning scheme.

[0016] Optionally, before generating the vector map of the target area field plots based on the panoramic RGB image of the target area field plots and the actual division situation of the target area field plots, it further includes:

[0017] Obtaining the LIDAR image and the RGB image of the target area field plots by using the LIDAR camera carried by the unmanned aerial vehicle, and respectively calculating the resolutions of the LIDAR image and the RGB image as the resolutions called when obtaining the basic information;

[0018] Solving the LIDAR image to obtain the panoramic LIDAR image of the target area field plots with position information and elevation information of pixel points, and performing stitching processing on the RGB image to obtain the panoramic RGB image.

[0019] Optionally, the calculation formula of the resolution GSD of the LIDAR image and the RGB image is as follows:

[0020]

[0021] Among them, L represents the maximum actual distance corresponding to the field of view angle of the LIDAR camera at a certain flight altitude, and P represents the number of pixels;

[0022]

[0023] Among them, H represents the flight altitude of the UAV, and θ represents the field of view angle of the LIDAR camera.

[0024] Optionally, before stitching the RGB images to obtain a panoramic RGB image, it further includes:

[0025] Filtering out invalid RGB images in the RGB images according to the valid RGB image determination condition to obtain high-quality RGB images;

[0026] Among them, the valid RGB image determination condition is the RGB image with the largest number of bytes in the images of the same scene;

[0027] The RGB images in the same scene are defined as N adjacent RGB images. The value of N depends on the overlap degree of adjacent RGB images, and the overlap degree between the RGB images after selecting one from N is greater than 0.

[0028] Optionally, before using the elevation information image and resolution to obtain the basic information of each plot in the target area plot and constructing a basic information table including the basic information of each plot in the target area plot, it further includes:

[0029] Using the threshold method to eliminate the outliers of the elevation information from the elevation information image of the target area plot to obtain an elevation information image that conforms to the actual situation of the plot.

[0030] Optionally, the preset merging rules include:

[0031] The area of each plot in the target area plot after merging does not exceed 40 mu;

[0032] Under the premise of keeping less or no construction on the north-south original ridge, if the elevation difference between adjacent plots in the target area plot is greater than 30 cm, they will not be merged;

[0033] The east-west ridge is always maintained on the horizontal line.

[0034] Optionally, when generating a digital surface model based on the panoramic LIDAR image of the target area plot, call the LAS Dataset to Raster tool to convert the LAS dataset of the panoramic LIDAR image into a raster dataset, and the internal sampling value setting of the LASDataset to Raster tool is set according to the planting interval of ground plants.

[0035] Second aspect, the present application provides a device for generating a field management solution based on LIDAR images, adopting the following technical solution:

[0036] A device for generating a field management solution based on LIDAR images, comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein when the processor executes the program, it implements the method for generating a field management solution based on LIDAR images as described in the first aspect.

[0037] By adopting the above technical solution, during the process of obtaining the elevation information image, the elevation information of the panoramic LIDAR image can be continuously used, so that the elevation information of the fields in the target area can be obtained from the elevation information image with centimeter-level error, effectively improving the accuracy of the elevation information; since each pixel point on the elevation information image corresponds to an elevation information, there are a sufficient number of points with elevation information to support the merging of large-area fields during the merging (leveling) of fields, not only reducing the occupied area of the field ridges, but also greatly improving the flatness of the fields in the merged planning solution.

[0038] In summary, the present application includes at least one of the following beneficial technical effects:

[0039] 1. During the process of obtaining the elevation information image, the elevation information of the panoramic LIDAR image can be continuously used, so that the elevation information of the fields in the target area can be obtained from the elevation information image with centimeter-level error, effectively improving the accuracy of the elevation information;

[0040] 2. Since each pixel point on the elevation information image corresponds to an elevation information, there are a sufficient number of points with elevation information to support the merging of large-area fields during the merging of fields, not only reducing the occupied area of the field ridges, but also greatly improving the flatness of the fields in the merged planning solution. On the premise of facilitating mechanical standard operations, it reduces the occupied area of the pond ridges and improves the land utilization rate per unit area;

[0041] 3. When generating the digital surface model, the setting of the sampling value parameter takes into account the spacing of the planted plants in the fields of the target area, so that the generated digital surface model can better reflect the influence of the height of the planted plants on the ground, improving the calculation efficiency and accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 is a flowchart of a method for generating a field management solution based on LIDAR images provided by an embodiment of the present application;

[0043] Figure 2 is an elevation thematic map of the fields in the target area in an embodiment of the present application;

[0044] Figure 3It is the merged elevation thematic map in the embodiments of the present application.

[0045] Figure 4 It is the structural block diagram of a generating device for a field governance scheme based on LIDAR images provided by the embodiments of the present application. Specific embodiments

[0046] In order to make the objectives, technical solutions and advantages of the present application more clear and understandable, the following further elaborates on the present application in conjunction with the attached Figures 1-4 drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0047] The embodiments of the present application disclose a method for generating a field governance scheme based on LIDAR images, as Figure 1 shown, including the following steps:

[0048] Step S100: Obtain the LIDAR image and RGB image of the fields in the target area through the LIDAR camera carried by the drone, and calculate the resolutions of the LIDAR image and RGB image respectively for calculating the area of each field; in the LIDAR image, the value of the pixel point is the elevation value of the corresponding physical position (the position determined by longitude and latitude) relative to the sea level.

[0049] The formulas for calculating the resolutions GSD of the LIDAR image and RGB image are as follows:

[0050]

[0051] where L represents the maximum actual distance corresponding within the field of view angle of the LIDAR camera at a certain flight altitude, and P represents the number of pixels;

[0052]

[0053] where H represents the flight altitude of the drone, and θ represents the field of view angle of the LIDAR camera.

[0054] Step S200: Call the Zhitu software to calculate the acquired LIDAR images and stitch the RGB images, respectively obtaining a panoramic LIDAR image and a panoramic RGB image of the target area's fields with position information (latitude and longitude information) and elevation information for each pixel point; in this embodiment, the called Zhitu software is, for example, the DJI Zhitu software. Set the point cloud effective distance in the software (point clouds beyond the effective distance from the LIDAR camera on the drone will be filtered in post-processing), and the other parameters can be directly defaulted. The point cloud effective distance is modified according to the actual situation. For example, if the drone flight altitude is 150m, the point cloud effective distance setting range can be in [150m, 200m], where 200m is the allowable error value obtained based on experience.

[0055] In this embodiment, to improve the quality of the acquired panoramic RGB images, before stitching, it is necessary to first filter out the invalid RGB images (low-quality RGB images) in the RGB images according to the determination conditions for valid RGB images (high-quality RGB images) to obtain high-quality RGB images; in this embodiment, the determination condition for valid RGB images is the RGB image with the largest number of bytes in the same scene; among them, the RGB images in the same scene are defined as N adjacent RGB images, and the value of N depends on the overlap degree of adjacent RGB images. After selecting N = 1, the overlap degree between the RGB images is greater than 0, and the aerial photography requirement is that the overlap degree shall not be less than 70%, so generally N is taken as 3 or 4.

[0056] Step S300: Generate a vector map of the target area's fields based on the stitched panoramic RGB image and the actual division of the target area's fields (the original vector boundary data of each field), and generate a digital surface model based on the panoramic LIDAR image; among them, the generated vector map is marked with the vector boundaries of each field in the target area's fields, and each vector boundary of the field has a unique code.

[0057] Step S400: Add the vector boundaries of each field in the target area's fields and the corresponding unique codes for the vector boundaries to the obtained digital surface model according to the vector map to obtain an elevation information image of the target area's fields; that is, in the elevation information image, any field has the elevation information of the field, the vector boundary of the field, and the corresponding unique code.

[0058] Step S500, using the obtained elevation information image, the resolution of the LIDAR image and the resolution of the RGB image to obtain the basic information of each grid field in the target area, and constructing a basic information table including the basic information of each grid field in the target area; wherein the basic information of the grid field includes the maximum elevation, the minimum elevation and the average elevation of the grid field, the grid field area and the unique code corresponding to the vector boundary of the grid field; in this way, the basic information table and the vector boundary data are associated, so as to facilitate the visualization of the statistical data on the elevation information image.

[0059] Since the flight process is easily affected by factors such as flying birds, resulting in abnormalities in the collected elevation values, in order to improve the accuracy of the basic information obtained, before the program executes step S500, the elevation information image of the grid fields in the target area is obtained, and the abnormal values ​​of the elevation information are removed from the elevation information image of the grid fields in the target area using the threshold method to obtain an elevation information image that is more in line with the actual elevation of the grid fields.

[0060] When using the threshold method to remove abnormal values ​​of elevation information from the elevation information image of the grid field in the target area, the method of setting the threshold during threshold segmentation includes but is not limited to the empirical threshold method, the automatic threshold method, the color space transformation method, etc. For example, the empirical threshold method is used to count the elevations of the target grid field and other objects (such as roads, canals, trees, etc.) in the elevation information image based on the prior knowledge of previous processing of elevation information images; the values ​​that exceed the preset value of the average elevation value of the target grid field (such as -1 meter or other values, determined by the terrain) to the preset value of the average elevation value of the target grid field (such as +1 meter or other values, determined by the terrain) are regarded as abnormal points and removed, and the threshold interval is selected through the judgment of pixels and the analysis of images, and then compared through experiments, and finally a better threshold is selected for threshold setting.

[0061] Step S600: associate the elevation information image of the target area grid field with the basic information table through a unique code to produce an elevation thematic map of the target area grid field, such as Figure 2 shown.

[0062] Step S700: call the preset merging rule from the system, and merge the grid fields in the elevation thematic map according to the preset merging rule to generate a planning scheme for the grid fields in the target area, that is, the merged elevation thematic map, such as Figure 3 As shown; afterwards, bulldozers and other machinery can be used to manage the grid fields according to the planning scheme of the grid fields in the target area, reduce the ridges in the grid fields in the target area, and increase the arable area of ​​the grid fields.

[0063] In the embodiment of the present application, a preset merging rule is formulated according to the actual production needs and the code generated by the rule is stored in the system, so that the program of the method for generating a grid field management solution based on LIDAR images can call the preset merging rule for automatic merging; wherein, the preset merging rule formulated according to the actual production needs includes but is not limited to the following:

[0064] (1) To facilitate management by farmers, the area of ​​each grid field in the target area after transformation shall not exceed 40 mu;

[0065] (2) In order to save reconstruction costs, the original north-south ridges were changed under the premise of keeping construction as little as possible, and adjacent fields in the target area with elevation differences greater than 30 cm were not merged;

[0066] (3) In order to facilitate mechanized operations and ensure consistency in grid layout, the east-west ridges are kept on a horizontal line as much as possible.

[0067] In the embodiment of the present application, the implementation and parameter setting of steps S300 to S500 are achieved by calling ARCGIS10.7; when generating the digital surface model, the LAS Dataset to Raster tool is used to convert the established LAS dataset (panoramic LIDAR image) into a raster dataset. The parameter settings in the LAS Dataset to Raster tool are set according to the actual situation of the plants on the ground. For example, the planting interval of the plants is about 30 cm, so the sampling value (parameter in the tool) is set to 0.3 m, which can better reflect the influence of the height of the plants planted on the ground and improve the calculation efficiency and accuracy.

[0068] The actual application results show that the grid field management method based on UAV LIDAR images proposed in this application can not only adapt to large-scale grid field transformation and quickly and easily obtain grid field elevation information; but also can obtain the elevation information of the grid fields in the target area with a centimeter-level error, reflecting the actual altitude information of each grid field in the target area. Based on such high-precision elevation information, large-scale leveling can be achieved for the regional scale, and fine leveling can be achieved for the farmland scale. The grid fields are planned according to the actual situation of the grid fields in the target area (such as parameter setting needs to consider the plants planted in the grid fields) and actual production needs. Under the premise of facilitating standard mechanical operations, the area occupied by the pond stems is reduced, the land utilization rate per unit area is improved, the soaking time and water consumption are reduced, and the goals of saving resources, saving costs and increasing efficiency, and increasing production and income are achieved.

[0069] In addition, the embodiment of the present application also discloses a device for generating a grid field management solution based on LIDAR images, which is deployed in a server-side and a user-side system; specifically, the device includes: one or more processors and a memory, such as Figure 4As shown, a processor 200 and a memory 100 are taken as examples. The processor 200 and the memory 100 can be connected through a bus or other means. For example, taking the connection through a bus as an example.

[0070] The memory 100, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs, such as a method for generating a grid field governance solution based on LIDAR images in the embodiments of the present application. The processor 200 realizes the method for generating a grid field governance solution based on LIDAR images in the above embodiments of the present application by running the non-transitory software programs and instructions stored in the memory 100.

[0071] The memory 100 may include a program storage area and a data storage area. Among them, the program storage area can store an operating device and application programs required for at least one function; the data storage area can store data required for executing the method for generating a grid field governance solution based on LIDAR images in the above embodiments. In addition, the memory 100 may include a high-speed random access memory and may also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some embodiments, the memory may optionally include a memory remotely provided relative to the processor, and these remote memories can be connected to the terminal through a network. Examples of the above network include but are not limited to the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0072] The non-transitory software programs and instructions required to implement the method for generating a grid field governance solution based on LIDAR images in the above embodiments are stored in the memory. When executed by one or more processors, the method for generating a grid field governance solution based on LIDAR images in the above embodiments is executed. For example, the method steps S100 to step S700 described above are executed. Figure 1 in the above.

[0073] The above are all preferred embodiments of the present application. The protection scope of the present application is not limited thereby. Any feature disclosed in this specification (including the abstract and drawings), unless specifically described, can be replaced by other equivalent or similar-purpose alternative features. That is, unless specifically described, each feature is only an example of a series of equivalent or similar features.

Claims

1. A method for generating a field governance solution based on LIDAR images, characterized in that, it includes: Generating a vector map of the fields in the target area based on the panoramic RGB image of the fields in the target area and the actual division of the fields in the target area. Among them, the vector map is marked with the vector boundaries of each field in the target area, and the vector boundary of each field has a unique code; Generating a digital surface model based on the panoramic LIDAR image of the fields in the target area; Adding the vector boundaries of each field and the corresponding unique code for the vector boundary identification on the obtained digital surface model according to the vector map to obtain the elevation information image of the fields in the target area; Using the elevation information image and resolution to obtain the basic information of each field, and constructing a basic information table including the basic information of each field in the target area; among them, the basic information includes the maximum elevation, minimum elevation and average elevation, field area and the corresponding unique code; Associating the elevation information image and the basic information table through the unique code to generate an elevation thematic map of the fields in the target area; Calling a preset merging rule and merging the fields in the elevation thematic map according to the preset merging rule to generate a planning solution for the fields in the target area.

2. The generating method according to claim 1, characterized in that, Before generating the vector map of the fields in the target area based on the panoramic RGB image of the fields in the target area and the actual division of the fields in the target area, it further includes: Obtaining the LIDAR image and RGB image of the fields in the target area through a LIDAR camera carried by a drone, and respectively calculating the resolutions of the LIDAR image and RGB image, which are used as the resolutions called when obtaining basic information; Calculating the panoramic LIDAR image of the fields in the target area with position information and elevation information of pixel points from the LIDAR image, and stitching the RGB image to obtain a panoramic RGB image.

3. The generating method according to claim 2, characterized in that, The calculation formula of the resolution GSD of the LIDAR image and the RGB image is as follows: Among them, L represents the maximum actual distance corresponding to the field of view angle range of the LIDAR camera at a certain flight altitude, and P represents the number of pixels; Among them, H represents the flight altitude of the drone, and θ represents the field of view angle of the LIDAR camera.

4. The generating method according to claim 2, characterized in that, Before stitching the RGB image to obtain a panoramic RGB image, it further includes: Filtering the invalid RGB images in the RGB images according to the valid RGB image determination condition to obtain high-quality RGB images; Among them, the valid RGB image determination condition is the RGB image with the largest number of bytes in the images under the same scene; The RGB images under the same scene are defined as N adjacent RGB images, and the value of N depends on the overlap degree of adjacent RGB images, and the overlap degree between the RGB images after selecting N as 1 is greater than 0.

5. The generating method according to claim 1, characterized in that, Before obtaining the basic information of each grid field by using the elevation information image and the resolution and constructing a basic information table including the basic information of each grid field in the target area, the method further includes: The threshold method is used to remove abnormal values ​​of the elevation information from the elevation information image of the grid field in the target area, so as to obtain an elevation information image whose elevation information conforms to the actual situation of the grid field.

6. The generation method according to claim 1, It is characterized in that The preset merging rules include: The area of ​​each grid field in the merged target area shall not exceed 40 mu; Under the premise of keeping construction to a minimum or not at all, the existing north-south ridges will not be merged if the elevation difference between adjacent fields in the target area is greater than 30 cm; The east-west ridges always remain on a horizontal line.

7. The generation method according to claim 1, It is characterized in that When generating a digital surface model based on the panoramic LIDAR image of the target area grid field, the LASDataset to Raster tool is called to convert the LAS dataset of the panoramic LIDAR image into a raster dataset, and the sampling value setting in the LASDataset to Raster tool is set according to the planting interval of the plants on the ground.

8. A device for generating a grid field management solution based on LIDAR images, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, a method for generating a grid field management solution based on LIDAR images as described in any one of claims 1 to 7 is implemented.

Images