Improved drone monitoring method and system

Abstract

The present invention relates to a drone monitoring method for observing and monitoring an agricultural research station, the drone monitoring method comprising: (a) dividing the agricultural research station into a set of plots, each plot being defined by a geographic information system, and each plot being defined by coordinate information of a 2D polygonal boundary; (b) defining at least one 2D sampling point in each plot to acquire drone data, thereby acquiring a set of 2D sampling points; (c) superimposing the set of 2D sampling points onto a digital elevation model, thereby acquiring an altitude reference height for each of the 2D sampling points; (d) determining the altitude coordinates corresponding to each 2D sampling point based on the altitude reference height of the 2D sampling point; and (e) determining a 3D sampling point for each 2D sampling point by combining the altitude coordinates corresponding to the 2D sampling point with the 2D sampling point, wherein the altitude coordinates are determined by considering the height of the tree canopy at the 2D sampling point.

Description

Technical Field

[0001] The present invention relates to a drone monitoring method and system. Thereby, an object of the present invention is to provide a method for obtaining very detailed data of agricultural testing fields using an aerial drone.

Background Art

[0002] Drones equipped with cameras often capture images of a particular area with a high image overlap, resulting in a large amount of data for creating a map or 3D model of that area photogrammetrically. In the case of applications involving outdoor agriculture experiments (phenotyping), the geometric accuracy of the final results is extremely important for the end user to associate with specific field data.

[0003] Plant phenotyping in experimental fields is a study that quantifies and analyzes the physical and biochemical characteristics of plants / crops grown in multiple different environments. Before planting, researchers or breeders create an experimental design, which determines where in the field and which plants or plant varieties will be grown, and which specific treatment will be applied throughout the growing season. The experimental design consists of a collection of plots, i.e., a group of identical plants grown close to each other and receiving the same treatments. Phenotype data is typically collected or summarized at the plot level, taking into account the variation of different individual plants within a single plot.

[0004] Drone-based phenotyping involves using camera-equipped unmanned aerial systems (UAS), or drones, to capture images of a laboratory and calculate specific plant characteristics such as emergence count, coverage, height (health), biomass, health, and flowering time. The use of drones in phenotyping offers advantages over traditional survey methods, including the ability to measure plant traits more quickly, objectively, and without the need for human experts. One of the key techniques used in drone-based phenotyping is Structure from Motion (SfM), which involves creating a 3D model and orthomosaic of an area using highly overlapping images (typically 70-80%). With the use of precise GPS measurements, either on the drone itself or with reference targets placed in the field, the drone image products can be georeferenced with an accuracy of 2-3 cm. Such geographical references allow the experimental details to be overlaid on a digital plot layout, and based on this overlay, plot statistics for the drone products can be derived.

[0005] Direct and geographical reference is an alternative image processing technique that correlates real-world coordinates with image coordinates, and this alternative image processing technique is commonly applied to traditional aerial surveying. However, because of the low accuracy of all required viewing geometries of the drone camera, this alternative image processing technique is not widely applied to drone-based imagery, which is necessary to obtain accurate geometric results.

[0006] In particular, a major technical obstacle to using prior art methods and systems such as US 11,074,447 is the enormous amount of data, usually in the form of images, that needs to be acquired and processed. In fact, it is estimated that when acquiring a full view of an area, approximately 70% to 80% of the image data may be superfluous due to overlapping portions made to enable stitching of the images. This superfluous portion of the image data imposes serious limitations on data recording, data transfer, and / or data analysis associated with drone monitoring systems and methods. The present invention overcomes this problem and presents a solution that limits the data required for further use.

[0007] US 11,074,447 discloses a land analysis system that uses drone-captured images to detect plant health and / or soil moisture levels at a site. For example, the system instructs a drone to fly along a flight path, capture images of the land below, and measure altitude data. The system processes the images, for example, using artificial intelligence, to identify locations where plant material may be present. The system then further processes the images to identify plant health at the identified locations. The system further uses the altitude data to determine the strata of plants at the identified locations. Optionally, the system can further process the images to determine soil moisture levels at the identified locations. The method disclosed in US 11,074,447 requires a high degree of data redundancy to enable the stitching of acquired images to cover the entire experimental field.

[0008] Many phenotyping algorithms require extremely high-resolution images (such as mm or sub-mm resolution) to calculate the characteristics of specific plants or soils. Data acquisition and processing (SfM) of such high-resolution drone data is extremely costly and operationally complex. These cost and operational complexity issues may be considered the biggest limiting factor in further applying high-resolution drone data and derived trait algorithms to large-scale plant phenotyping.

[0009] The present invention aims to provide a methodology for drone-based image acquisition and processing, thereby enabling users to acquire images much faster than current state-of-the-art technologies, with a limited and acceptable compromise regarding the geometrical accuracy of the results, while significantly reducing the amount of data required. This will allow users to deploy solutions for agricultural experimental fields on a large scale in agricultural contexts, while drastically reducing overall costs. [Overview of the Initiative]

[0010] The present invention relates to a drone monitoring method for observing and monitoring agricultural testing fields as described in claim 1, and a drone monitoring system as described in claim 10. Further preferred embodiments are disclosed in the dependent claims herein and in the following claims.

[0011] The present invention enables the acquisition of plant data in agricultural testing fields without the need for stitching, and therefore without overlap. In fact, in one preferred embodiment of the present invention, the data acquired at the 3D sampling points includes up to 20%, preferably up to 10%, more preferably up to 5%, and even more preferably up to 2%, overlap, and most preferably, the data acquired at the 3D sampling points includes no overlap whatsoever. Furthermore, data is not required for locations in between the plots. Alternatively or additionally, data may be acquired two or more times for up to 20% of the area of ​​a plot, preferably up to 10%, more preferably up to 5%, and even more preferably up to 2%, and most preferably, no data is obtained twice or more times for any area of ​​a plot. The data preferably includes image data. Alternatively or additionally, the data includes LiDAR data. LiDAR data, or Light detection and ranging (lidar) data, is generally known to those skilled in the art. LiDAR is a technique used to create high-resolution models of the ground using sensors that detect reflections of pulsed laser beams. The reflections are recorded as millions of individual points called a "point cloud," which represents the three-dimensional positions of objects on the surface, including buildings, vegetation, and the ground.At the same time, the present invention makes it possible to acquire data at very low height, and therefore the data becomes very detailed, and the data has a resolution of, for example, 10 cm or better, preferably 3 cm or better, such as 2.5 cm, 2.0 cm, 1.5 cm, 1.0 cm, or any value in between, preferably 10 mm or better, such as 10 mm, 9 mm, 8 mm, 7 mm, 6 mm, 5 mm, or any value in between, and even more preferably 5 mm or better, such as 4 mm, 3 mm, 2 mm, or any value in between, or any value in between.

[0012] The present invention is particularly useful in experimental fields where there is a need to obtain more detailed and localized information than to obtain a complete overview of the field, including parts of the field that do not belong to the experiment. This is especially useful in agricultural testing fields, where monitoring the characteristics of individual plant specimens is important, and in some cases, complete statistics of the entire plot are less important.

[0013] The present invention enables the acquisition of drone data, particularly drone images and / or drone LiDAR data, by using prior knowledge of the exact location of individual plots in the testing fields to define locations, i.e., the 3D sampling points. This can be done at a relatively close distance to plants, increasing the spatial detail of the images while eliminating the effect of the low accuracy of the sensor when investigating shape. The lower the sensor's accuracy for a given object, the smaller the absolute impact of errors in the viewing geometry becomes, until the error is within an acceptable range. To accurately predetermine the drone's altitude, an accurate digital elevation model and an indicator of the plant height at the moment of drone flight are used.For example, in contrast to a method in which the altitude of the 3D sampling point, i.e., the 3D point from which data is acquired, is not determined prior to the drone reaching the 2D or 3D sampling point, it should be noted that in the present invention, for example, the drone predefines the altitude of the 3D sampling point on the basis of prior knowledge of the plot location and the 2D sampling point. In one embodiment, the drone is configured to acquire and / or measure the altitude above the experimental field and steer the drone toward the predefined altitude based on the acquired and / or measured altitude, i.e., the drone includes a steering algorithm to ensure that it positions itself at the predetermined altitude.

[0014] The present invention provides an improved system and method for high-resolution images without redundancy due to overlap. The invention aims to provide high-resolution images from pre-determined viewpoints, each viewpoint linked to a plot. This also allows for the identification of plots from data acquired at 3D sampling points, preferably from metadata that may be included in such data.

[0015] Furthermore, an additional advantage of the present invention compared to conventional methods is that it enables the handling of crops or plants that move due to wind or other weather conditions. This means that because the sampling points are predefined, the location can be specified upfront, and the user can be confident that they are photographing the same crop in multiple different imaging sessions, even if the crop has moved due to weather conditions. This is difficult to achieve with conventional methods that attempt to acquire 3D models by stereoscopic imaging. [Brief explanation of the drawing]

[0016] [Figure 1] The diagram shows an overview of the experimental site. [Figure 2] This diagram illustrates the layout of the experimental site, where the boundaries of individual sections overlap. [Figure 3] This diagram illustrates the 2D sampling points of drone data based on the 2D position of individual sections within the experimental area. [Figure 4] This diagram illustrates drone data points based on structure derived from motion during high-resolution and low-altitude flight. [Figure 5] This diagram illustrates drone data points based on structure derived from motion during low-resolution and high-altitude drone flight. [Modes for carrying out the invention]

[0017] The present invention relates to a drone monitoring method for observing and monitoring agricultural research stations.

[0018] The method is, (a) A step of dividing the test site into a set of sections, each of which is defined by a geographic information system, and which is defined by coordinate information of a 2D polygonal boundary.

[0019] In the context of this application, a test site, also referred to as a laboratory, typically comprises multiple sections, in which a particular type of plant, such as a plant species or a plant variation, is grown according to predefined experimental growing conditions. The plant types and / or experimental growing conditions may differ for each of the different sections. The idea of ​​dividing the test site into multiple sections, which may usually be located very close to each other, is to allow for the experimental testing of differences in characteristics for multiple different types of plants and / or multiple different growing conditions while minimizing variations in other variables, such as weather conditions. To enable clear distinction between the multiple sections, the boundaries of these sections must be clearly defined. Basically, this is done using 2D polygonal boundaries. It should be noted that certain other types of shapes, such as elliptical or circular boundaries, may be easily approximated using 2D polygonal boundaries. However, polygonal boundaries are preferred in order to easily connect the plots to their geographical locations. Figure 1 shows an overview of experimental site (1), and Figure 2 shows an overview of experimental site (1) with 2D polygonal (rectangular) boundaries (3A, 3B, 3C, 3D, 3E, 3F) overlaid for the individual plots (2A, 2B, 2C, 2D, 2E, 2F).

[0020] The method is, (b) further comprising the step of defining at least one 2D sampling point within each section to acquire drone data, thereby acquiring a set of 2D sampling points.

[0021] These 2D sampling points (4A, 4B, 4C, 4D, 4E, 4F) are illustrated in Figure 3 for each section having its own boundary line (3A-3F). In a preferred embodiment, multiple sections each include a boundary line of the same shape, and preferably, the drone includes a data acquisition sensor having a shape corresponding to the shape of the boundary line. This makes it possible to acquire optimal data for each section using a minimum number of sampling points.

[0022] In a preferred embodiment, the 2D sampling points essentially include the plot centers of each section, that is, for each section, the 2D polygonal boundary may preferably define a center that is a geometric center, and the geometric center may be selected as the 2D sampling point. Preferably, a single 2D sampling point is used for each section. However, in some of the plurality of embodiments, the at least one 2D sampling point in each section includes two or more 2D sampling points. Although the present invention focuses on obtaining relevant data using a minimum number of samples, in case of large and / or elongated plots, for example, or when it is necessary to obtain data from different locations and / or under different angles for the same plot, more than one sampling point may be used when volumetric data is required.

[0023] The method further comprises (c) overlaying the set of 2D sampling points with a digital elevation model, thereby obtaining an altitude reference height for each of those 2D sampling points.

[0024] A digital elevation model (DEM) may be obtained, for example, from data obtained by a drone inspection round. The drone may use conventional techniques such as structure extraction technology from motion to obtain data for the entire field. Such drone inspections may be performed only once or a few times at a low resolution, while it should be noted that the present invention may be easily repeated daily or hourly at a high resolution. Alternatively, the DEM may be obtained from a database such as a geographic information system database. For example, data from a Google Earth database may be used. Google Earth is a computer program that renders a 3D representation of the Earth mainly based on satellite images. The program maps the Earth by superimposing satellite images, aerial photography, and GIS data on a 3D globe, enabling users to view the landscape from various angles. Additionally or alternatively, local governments or companies may use high-precision (2 - 3 [cm]) digital terrain model (DTM) data obtained from an aerial survey campaign using a photogrammetric camera or LIDAR.

[0025] Figure 4 illustrates drone data points (5A, 5B, 5C, 5D, 5E, 5F) based on structure from motion in the case of high resolution and low altitude flight, where the overlap rate can be very large. Figure 5 illustrates drone data points (6A, 6B, 6C, 6D, 6E, 6F) based on structure from motion in the case of low resolution and high altitude drone flight, where the resolution may be too low to derive specific characteristics from the image data, for example. This makes it possible to understand that a large amount of data is obtained that is not useful for acquiring plant characteristics but is necessary to create a stitched view in the test field. Furthermore, it should be noted that 2D sampling points in structure from motion techniques are usually not predefined. Although such techniques lead to high data load and computationally intensive work, such techniques may be used to acquire DEMs and altitude reference heights.

[0026] The method is, (d) For each 2D sampling point, the steps of determining an altitude coordinate corresponding to the 2D sampling point based on the altitude reference height of the 2D sampling point, (e) For each 2D sampling point, the step of determining a 3D sampling point by combining the altitude coordinates corresponding to the 2D sampling point with the 2D sampling point, further includes: The altitude coordinates are determined by considering the canopy height at the 2D sampling point.

[0027] It is important to note that the altitude coordinates of each 2D sampling point are obtained before the actual acquisition of data by the drone. The altitude coordinates depend on the altitude reference height and the height of the plant canopy. In some of the multiple embodiments, for example, when the DEM is acquired by a recent drone inspection round, the altitude reference height already includes the height of the canopy. On the other hand, in other embodiments, the height of the canopy is provided separately from this method, for example, by the user or by calculation by a plant growth model. In one embodiment, the height of the canopy at a 2D sampling point is provided using a digital surface model. Alternatively or additionally, the height of the canopy at a 2D sampling point is provided by user input.

[0028] In several embodiments, the DEM is one of a digital surface model, a digital terrain model, or a digital canopy height (DCH) model. Examples of such models can be found, for example, in "Forest Assessment Using High Resolution SAR Data in X-Band", by Perko et al., Remote Sensing 3(4):792-815, April 2011, DOI:10.3390 / rs3040792. It should be noted that the DCH may typically be obtained from the difference between the DSM and DTM. The DSM may be calculated directly from drone flight data acquired near the drone sampling flight, or it may be preferable to calculate it by combining the DTM obtained from previous data with a plant-height estimation or DCH model. The altitude from the canopy height is determined by the required resolution and the drone image sensor used. For example, if the tree canopy is very low due to the low height of the plants, the DCH can be discarded.

[0029] In one embodiment, the altitude coordinate is determined as a minimal height above the canopy height, which is the minimum safety height. To obtain the most detailed and accurate data, the altitude coordinate is preferably minimal. However, several safety measures are necessary to ensure that the drone itself does not affect the data. For example, if the drone is an aerial drone, air displacement from the drone's rotors should not affect data acquisition. Therefore, a minimum altitude of approximately 1 m above the canopy may be required.

[0030] In one embodiment, the altitude coordinate is a value between a minimum height above the canopy height and a maximal height above the canopy height, where the minimum height is the safety minimum height, and the maximal height is determined by the drone's resolution, the required image detail, the required geometric accuracy, and / or the image position and field of view accuracy. The required image detail is essentially determined by the characteristics of the plant being surveyed. The required geometric accuracy is essentially determined by the size of the plots and the distances between multiple plots. The field of view and position accuracy essentially depend on the characteristics of the drone and / or the characteristics of the sensors mounted on the drone. Better field of view and position accuracy allows for flying higher than the canopy. In several embodiments of the present invention, the maximum height from the height of the tree canopy is, for example, 20[m], 19[m], 18[m], 17[m], 16[m], 15[m], etc., up to 20m or less, or any value in between; more preferably, for example, 15[m], 14[m], 13[m], 12[m], 11[m], 10[m], etc., up to 15m or less, or any value in between; even more preferably, up to 10[m], 9[m], 8[m], 7[m], 6[m], 5[m], or any value in between; and even more preferably, for example, 5[m], 4[m], 3[m], 2[m], 1[m], etc., up to 5m or less, or any value in between or less. In several embodiments of the present invention, the minimum height from the tree canopy is at least 1 cm, such as 1 cm, 2 cm, 3 cm, 4 cm, 5 cm, or any value in between, and more preferably at least 5 cm, such as 5 cm, 10 cm, 15 cm, 20 cm, 25 cm, 30 cm, 35 cm, 40 cm, 45 cm, 50 cm, or any value in between or greater.

[0031] In one preferred embodiment, a drone monitoring method is (f) Preferably, the step of sending the drone to the 3D sampling point sequentially, and preferably in a single data acquisition round, such as in a single flight if the drone is an aerial drone, (g) The further step of acquiring data at each of the 3D sampling points by a drone.

[0032] The data may preferably be acquired by the drone while it is moving or while it is stationary for a period of time.

[0033] The present invention also relates to a drone monitoring system comprising a processing subsystem and at least one drone, wherein the processing subsystem is configured to perform the method described in any one of the preceding claims, and the drone is configured to move to 3D sampling points, preferably sequentially and preferably in a single flight, and to acquire data at each of the 3D sampling points, thereby preferably being an aerial drone.

[0034] The present invention also relates to a computer-readable file containing a set of 3D sampling points acquired using a method and / or a system according to the present invention. The computer-readable file is preferably a Keyhole Markup Language (KML) file. KML is a file format used to display geographical data in Earth browsers such as Google Earth. KML is based on the XML standard and uses a tag-based structure with nested elements and attributes. All tags are case-sensitive and must be written as described in the KML reference, which indicates which tags are optional. Within a given element, the tags must be written in the order shown in the reference. Preferably, this includes an ordering that determines the drone's trajectory for the set of 3D sampling points.

[0035] In several embodiments, data acquisition includes the acquisition of image data, which optionally includes a set of images acquired at a number of different angles. Alternatively or additionally, data acquisition includes the acquisition of LIDAR data.

[0036] The drone monitoring method is preferably, (h) Preferably, (h1) A step of georeferencing the data based on the location of the sensor, and / or (h2) A step of geographically referencing the data based on sensor exterior orientation parameters, sensor interior parameters, and the digital elevation model, and / or (h3) A step of executing a trait algorithm to extract plant traits from the data. This includes further steps in analyzing the data acquired by the drone.

[0037] Georeferencing the data is, for example, based on the center of the image when the data includes image data, so that the center of the image is used to geographically reference the image. However, in several embodiments, georeferencing the data may involve geographical referencing of any data point within the data; for example, when the data includes image data, each pixel in the image may be geographically referenced.

[0038] Therefore, the sensor external orientation parameters may typically include, for example, a geographical location obtained using GPS or other positioning systems or tools. The sensor external orientation parameters may also typically include external orientation indicating the angle at which the data is acquired. These angles may typically represent the roll, pitch, and / or yaw of the drone or the data acquisition sensor mounted on the drone.

[0039] Internal orientation parameters include, for example, CCD characteristics, pixel density, focal length, lens distortion, ppx (pixel density in the x-direction), ppy (pixel density in the y-direction perpendicular to the x-direction), and lens properties, when a wide-angle lens or zoom lens is used to acquire data.

[0040] The DEM may be used, for example, as disclosed in "High Throughput Field Phenotyping for Plant Height Using UAV-Based RGB Imagery in Wheat Breeding Lines: Feasibility and Validation", by Volpato et al., Front. Plant Sci., 16 February 2021, Sec. Plant Breeding, Volume 12 - 2021, https: / / doi.org / 10.3389 / fpls.2021.591587.

[0041] Preferably, the geographical reference of the data is, for example, - "Direct georeferencing on small unmanned aerial platforms for improved reliability and accuracy of mapping without the need for ground control points", by Mian et al., The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XL-1 / W4, 2015, International Conference on Unmanned Aerial Vehicles in Geomatics, 30 Aug-02 Sep 2015, Toronto, Canada, and / or - "Static error budget analysis for a land-based dual-camera mobile mapping system", by Jiann-Yeou Rau et al., Journal of the Chinese Institute of Engineers, Vol. 34, No. 7, October 2011, 849-862 As explained in [the document], this includes direct georeferencing of the data.

[0042] In one preferred method, the data acquired at the 3D sampling point includes up to 20%, preferably up to 10%, more preferably up to 5%, and even more preferably up to 2% overlap, and most preferably, the data acquired at the 3D sampling point includes no overlap.

[0043] In one preferred embodiment, the drone monitoring method includes a step of obtaining data at different times, repeating steps (d), (e), and other steps at different times, thereby enabling observation and monitoring of the agricultural research station over a long period of time, and optionally, if the digital elevation model changes between those different times, step (c) is also repeated at different times. It should be noted that this uses the same set of 2D sampling points, i.e., the set of 2D sampling points defined in steps (a) and (b), enabling, for example, geographical reference across repeated data acquisition rounds. Preferably, the data is obtained in the same manner across repeated data acquisition rounds.

[0044] It should be noted that in the present invention, the drone may be any type of drone capable of being piloted to a 3D sampling point including a predefined altitude. However, preferably, the drone is an aerial drone, i.e., a drone capable of flight.

[0045] Preferably, the drone includes an altitude-steerable sensor or camera, i.e., one whose altitude can be controlled. Preferably, the drone includes a locator that can maintain tracking of its position, and this locator may include a GPS chip.

[0046] Preferably, the drone includes a locator with a value within 10 cm, such as 10 cm, 9 cm, 8 cm, 7 cm, 6 cm, 5 cm, 4 cm, 3 cm, 2 cm, or any value between these, or a value with higher precision than these. Preferably, the drone is a real-time kinematic (RTK) drone. Real-time kinematic positioning (RTK) is a surveying application that corrects the errors common in current GNSS systems. This allows the drone to determine its position with an accuracy of 2-3 cm.

Claims

1. A drone monitoring method for observing and monitoring agricultural research stations, wherein the drone monitoring method is (a) A step of dividing the agricultural experimental station into a set of plots, each of which is defined by a geographic information system, and each plot is defined by coordinate information of a 2D polygonal boundary, (b) The steps of defining at least one 2D sampling point in each section to acquire drone data, thereby acquiring a set of 2D sampling points, (c) The steps of superimposing the set of 2D sampling points onto a digital elevation model, thereby obtaining an elevation reference height for each of the 2D sampling points, (d) For each 2D sampling point, the steps of determining the altitude coordinates corresponding to the 2D sampling point based on the altitude reference height of the 2D sampling point, (e) For each 2D sampling point, the step of determining a 3D sampling point by combining the altitude coordinates corresponding to the 2D sampling point with the 2D sampling point, The altitude coordinates are determined by considering the height of the tree canopy at the 2D sampling point. Drone monitoring methods.

2. The drone monitoring method according to claim 1, wherein the height of the tree canopy at the 2D sampling point is provided using a digital surface model.

3. The drone monitoring method according to claim 1, wherein the height of the tree canopy at the 2D sampling point is provided by user input.

4. The drone monitoring method according to claim 1, wherein the altitude coordinate is determined as the minimum height from the height of the tree canopy, and the minimum height is the minimum safety height.

5. The drone monitoring method according to claim 1, wherein the altitude coordinate is a value between the minimum height from the height of the tree canopy and the maximum height from the height of the tree canopy, the minimum height being a safety minimum height, and the maximum height being determined by the drone's resolution, the required image detail, the required geometric accuracy, and / or the image position and field of view accuracy.

6. The drone monitoring method is: (f) Preferably, the step of sending the drone to the 3D sampling point in sequence and preferably in one data acquisition round, (g) The step of acquiring data at each of the 3D sampling points by the drone, The drone monitoring method according to claim 1, further comprising:

7. The drone monitoring method is: (h) Preferably, (h1) A step of geographically referencing the data based on the location of the sensor, and / or (h2) A step of geographically referencing the data based on external sensor orientation parameters, internal sensor parameters, and the digital elevation model, and / or (h3) A step of executing an algorithm for extracting plant characteristics from the aforementioned data, The drone monitoring method according to claim 6, further comprising the step of analyzing the data acquired by the drone.

8. The drone monitoring method according to claim 6, wherein the data acquired at the 3D sampling points includes a maximum of 20%, preferably a maximum of 10%, more preferably a maximum of 5%, and even more preferably a maximum of 2% overlap, and most preferably the data acquired at the 3D sampling points does not include overlap.

9. The drone monitoring method according to claim 6, wherein data is acquired two or more times for a maximum of 20% of the area of ​​the section, preferably for a maximum of 10% of the area of ​​the section, more preferably for a maximum of 5% of the area of ​​the section, even more preferably for a maximum of 2% of the area of ​​the section, and most preferably, data is not acquired two or more times for any area of ​​the section.

10. The drone monitoring method according to claim 1, comprising the step of repeating steps (d), (e), and any further step at different times to acquire data at those different times, thereby enabling observation and monitoring of the agricultural research station over a long period of time, wherein step (c) is also repeated at different times if the digital elevation model is changed between those different times, as is optional.

11. A drone monitoring system comprising a processing subsystem and at least one drone, wherein the processing subsystem is configured to perform the drone monitoring method according to any one of claims 1 to 10, and the drone is configured to move to the 3D sampling points, preferably in sequence and preferably in a single flight, and to acquire data at each of the 3D sampling points.

12. The drone monitoring system according to claim 11, wherein the drone is an aerial drone.

13. A computer-readable file comprising a set of 3D sampling points acquired using the drone monitoring method described in any one of claims 1 to 10 and / or the drone monitoring system described in claim 11.

14. A computer-readable file according to claim 13, which is a KML file.

15. The set of 3D sampling points includes an ordering that determines the path of the drone, as described in claim 13, in a computer-readable file.

Images