A method for calibrating plant spatial pattern based on high-precision positioning

By carrying distance measurement sensors and multi-spectral cameras, a three-dimensional coordinate system and flight route are constructed, and the distance data is quickly obtained and the curved surface is fitted, which solves the problems of low efficiency and large error in plant space layout calibration in the existing technology, and realizes the High-precision and rapid calibration of plant spatial layout.

CN118015100BActive Publication Date: 2024-11-19CHONGQING INST OF GEOLOGY & MINERAL RESOURCES +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410216664.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-02-27
Publication Date
2024-11-19
Estimated Expiration
2044-02-27

AI Technical Summary

Technical Problem

The existing plant spatial pattern calibration methods are cumbersome and have low efficiency, especially in dense forest areas, which are difficult to quickly and accurately calibrate. In addition, the traditional method has a large error in the coordinate position of the plant in large-area samples.

Method used

Using a high-precision positioning method, the drone is equipped with a range-testing sensor and a multi-spectral camera to build a three-dimensional coordinate system, generate reference lines and flight lines, obtain distance data in real time through the drone and fit the surface, and define the virtual plants. Space location enables fast and accurate calibration of plant spatial layout.

Benefits of technology

It significantly reduces the difficulty and time of operation, improves calibration efficiency, and can accurately calibrate plants in large-area sample squares without on-site operation, with an error controlled within 0.2m. The traditional method requires at least one week and the error is as high as ten meters. above.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118015100B_ABST
    Figure CN118015100B_ABST
Patent Text Reader

Abstract

The present invention provides a method for calibrating plant spatial pattern based on high-precision positioning, the steps of which include: constructing a digital fence and a three-dimensional coordinate system according to the boundary of a sample area, setting a flight route of a drone, and matching the flight route with the points corresponding to the sample area in combination with a positioning system, controlling the drone to fly above the sample area according to the set path and obtaining the measured distance data in real time, importing the obtained distance data into a digital fence, and generating vertical lines corresponding to the obtained distance data, fitting the lower end points of all vertical lines to form a curved surface; defining an "umbrella-shaped" structure on the obtained curved surface as a unit as a virtual plant, and calibrating the coordinates at the vertex of the "umbrella-shaped" structure as the spatial position of the plant. The present invention not only greatly reduces the operational difficulty of plant spatial pattern calibration, but also significantly shortens the time required for plant spatial pattern calibration, improves implementation efficiency, and facilitates rapid and accurate calibration.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of ecological investigation and assessment, and specifically relates to a method for calibrating the spatial pattern of plants (trees) based on high-precision positioning. Background Art

[0002] The spatial pattern of plants mainly includes three types: uniform distribution, random distribution and clustered distribution. At present, in the field of ecological survey and assessment, the survey and assessment of plant spatial pattern mainly adopts the method of artificial sample plot survey, that is, a square or rectangular sample plot of a certain area is established in the field survey area with a sample plot rope, and one corner of the sample plot is defined as the origin. The sample plot frame connected to the origin is the X-axis and Y-axis, and the coordinate position of the plant in the sample plot on the X-axis and Y-axis relative to the origin is measured with a tape measure or steel tape measure. However, this method is cumbersome to operate, takes a lot of time, and the coordinate position of the plant obtained is very error-prone when the sample plot area is large.

[0003] Among the existing schemes, document CN110036863B discloses a method for measuring the spatial structure of topography and plant communities in karst rocky desertification areas, including the following steps: selecting a sample plot and dividing it into subsample plots; preparing instruments; measuring the sample plot, and extending the measurement of microtopography such as stone gullies, stone cracks, stone grooves and stone caves unique to the karst area, as well as characteristic topography such as ridges and valleys; measuring the spatial coordinates of trees with a breast diameter of ≥1 cm; calculating the distance between two trees; and determining the spatial distribution characteristics of plant communities. However, this method still requires operators to measure plant parameters one by one in the sample plot, and the implementation efficiency is still very low.

[0004] More importantly, since adjacent plants (trees) in many areas grow relatively densely and there are many cases of cross-growth of branches and leaves, even with the help of existing high-definition camera terminals and image recognition systems, the process of plant spatial pattern calibration still faces the technical difficulty of being unable to calibrate quickly and accurately. Summary of the invention

[0005] At least in order to solve the technical problems mentioned in the background technology, the present invention aims to provide a plant spatial pattern calibration method based on high-precision positioning.

[0006] The present invention adopts the following technical solution.

[0007] A method for calibrating plant spatial pattern based on high-precision positioning, comprising the following steps:

[0008] Step 1, determine the sample area according to the needs;

[0009] Step 2, construct a digital fence according to the boundary of the sample area;

[0010] Step 3, constructing a three-dimensional coordinate system in combination with the obtained digital fence, with two adjacent boundaries of the digital fence serving as the X-axis and Y-axis of the three-dimensional coordinate system, and the Z-axis of the three-dimensional coordinate system extending downward from the plane where the digital fence is located;

[0011] Step 4: Generate reference lines in the three-dimensional coordinate system, with the horizontal reference lines being parallel to the X-axis or the Y-axis, and the vertical reference lines being parallel to the Z-axis, and divide the range of the digital fence into several areas, with a reference line set every n meters;

[0012] Step 5, setting the flight route of the drone along the boundary of the digital fence and the horizontal reference line, and matching the flight route with the points corresponding to the sample area in combination with the positioning system; the drone is equipped with multiple straight and closely arranged ranging sensors, and the optical path of the ranging sensors is vertically downward;

[0013] Step 6, control the UAV to fly at a constant speed over the sample area according to the set path, and obtain the measured distance data in real time;

[0014] Step 7, importing the obtained distance data into the digital fence, and generating vertical lines corresponding to the obtained distance data;

[0015] Step 8, fitting the lower endpoints of all vertical lines to form a curved surface;

[0016] Step 9, define the "umbrella-shaped" structure on the obtained surface as a unit, as a virtual plant, all virtual plants together constitute the sample model, and calibrate the coordinates of the vertices of the "umbrella-shaped" structure as the spatial position of the plant.

[0017] In the present invention, during the process of measuring distance data, the flight path of the drone includes a plurality of straight line paths arranged in parallel.

[0018] As a preferred solution, the drone is equipped with 10-30 ranging sensors arranged in a straight line and closely together, and the optical path spacing between adjacent ranging sensors is no more than 15 cm.

[0019] According to one of the further preferred solutions, the drone is also equipped with a multispectral camera for identifying plant species, and the plant species information is simultaneously acquired during the process of acquiring the distance measurement data.

[0020] According to a further preferred solution, the drone is also equipped with a multispectral camera for identifying plant species; all vertices of the "umbrella-shaped" structures in step 9 are set as target points. After step 9 is implemented, the drone is controlled to fly again to the position directly above the plants in the sample area corresponding to each target point, so as to obtain the plant species corresponding to the "umbrella-shaped" structure and mark the obtained plant species information at the corresponding position.

[0021] Furthermore, step 10 is also included: selecting the data group with the largest value from the obtained distance data, and then taking the average value of the data group with the largest value as the height of the plane where the X-axis and the Y-axis are located from the ground, and then extending from the vertex of the "umbrella-shaped" structure to the ground to generate a virtual plant trunk.

[0022] Beneficial effects: The scheme of the present invention not only greatly reduces the operational difficulty of plant spatial pattern calibration, but also significantly shortens the time required for plant spatial pattern calibration, improves the implementation efficiency, and facilitates rapid and accurate calibration. More importantly, it can accurately calibrate plants in a large area (an area not less than 100m*100m) of sample plots without the need for operators to go to the site to calibrate one by one. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A schematic diagram of a three-dimensional coordinate system in an embodiment;

[0024] Figure 2 A schematic diagram of a UAV flight route set in the embodiment;

[0025] Figure 3 This is a schematic diagram of measuring distance data using a distance measuring sensor in an embodiment;

[0026] Figure 4 A schematic diagram of vertical lines corresponding to the distance data obtained by calibration in the embodiment;

[0027] Figure 5 It is a local schematic diagram of a curved surface formed by fitting the lower end points of vertical lines in the embodiment. DETAILED DESCRIPTION

[0028] The technical solution of the present invention is described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Example

[0029] A method for calibrating plant spatial pattern based on high-precision positioning, comprising the following steps:

[0030] Step 1, determine the sample area according to the needs;

[0031] Step 2: construct a digital fence based on the boundary of the sample area, such as Figure 1 and Figure 2 The rectangular frame shown by number 1 represents a digital fence;

[0032] Step 3, constructing a three-dimensional coordinate system in combination with the obtained digital fence, with two adjacent boundaries of the digital fence serving as the X-axis and Y-axis of the three-dimensional coordinate system, and the Z-axis of the three-dimensional coordinate system extending downward from the plane where the digital fence is located;

[0033] Step 4, generate reference line 2 in the three-dimensional coordinate system, the horizontal reference line is parallel to the X-axis or Y-axis, the vertical reference line is parallel to the Z-axis, and the range of the digital fence is divided into several areas, and a reference line is set every 0.1 meter;

[0034] Step 5: Set the drone flight path along the digital fence boundary and lateral reference line ( Figure 2 The flight path of the UAV includes several parallel straight paths, and the flight path is matched with the corresponding points / positions of the sample area in combination with the positioning system; the UAV is equipped with multiple straight and closely arranged distance measuring sensors, and the optical path of the distance measuring sensors is vertically downward; when the UAV completes the entire sample area according to the flight path, the measurement area of ​​the distance measuring sensor covers the crown width of all plants (trees) in the sample area, except for the small trees that are completely blocked;

[0035] Step 6, control the drone to fly at a constant speed above the sample area according to the set path, and obtain the measured distance data in real time (the distance data refers to the distance between the light emitting end of the ranging sensor and the object to be measured; when the light emitted by the ranging sensor is blocked by leaves, the leaves are the object to be measured; when the light emitted by the ranging sensor is blocked by the ground, the ground is the object to be measured); combined with Figure 3 As shown, the drone is equipped with 12 straight and closely arranged distance measuring sensors, and the optical path spacing of adjacent distance measuring sensors is set to 10cm. When the drone flies at a constant speed above the sample area according to the set path, the width area that can be covered by the distance measuring sensors on the drone at the same time is as shown by the numbers 10 / 11 / 12 / 13. Taking two plants side by side (tree one and tree two) as an example, if the drone needs to fly back and forth four times to make the measurement range completely cover the two plants, then the width area measured by the drone during the first flight corresponds to number 10, the width area measured by the drone during the second flight corresponds to number 11, the width area measured by the drone during the third flight corresponds to number 12, and the width area measured by the drone during the fourth flight corresponds to number 13. Accordingly, the measured distance data are H1, H2...H48 in sequence;

[0036] Step 7, import the obtained distance data into the digital fence, and generate vertical lines corresponding to the obtained distance data, that is, draw corresponding vertical lines according to the distance value (also representing the optical path between the distance measuring sensor and the object to be measured), and the model section is obtained as follows Figure 4 As shown;

[0037] Step 8: Fit the lower endpoints of all vertical lines to form a surface, such as Figure 5 As shown;

[0038] Step 9, define the "umbrella-like" structure on the obtained surface as a unit. This "umbrella-like" structure is consistent with the crown outline of the plant (tree) and is used as a virtual plant (such as Figure 5 The "umbrella-shaped" structure in the middle left part represents virtual plant 1, which corresponds to tree 1 in the sample area. The "umbrella-shaped" structure in the right part represents virtual plant 2, which corresponds to tree 3 in the sample area. The sample model is composed of all virtual plants, and the coordinates of the vertices of the "umbrella-shaped" structure are calibrated as the spatial position of the plant.

[0039] Step 91: The drone is also equipped with a multispectral camera for identifying plant species (such as a Phantom4-M multispectral camera); all vertices of the "umbrella-shaped" structures in step 9 are set as target points. After step 9 is implemented, the drone is controlled to fly again to the position directly above the plants in the sample area corresponding to each target point (such as flying directly above tree one and tree two, taking an image of the tree and identifying its type), so as to obtain the plant species corresponding to the "umbrella-shaped" structure, and mark the obtained plant species information at the corresponding position; for example, tree one (X2, Y2, Z2) is marked as holly tree (X2, Y2, Z2), and tree two (X3, Y3, Z3) is marked as soapberry tree (X3, Y3, Z3);

[0040] Step 10: Filter out the data group with the largest value from the obtained distance data ( Figure 5 Among them, the largest values ​​screened out are H1, H27, H47, and H48. These four data are taken as a group, and their average value is calculated = (H1+H27+H47+H48) / 4). Then, the average value of the data group with the largest value is used as the height of the plane where the X-axis and Y-axis are located from the ground, and then the virtual plant trunk 8 is extended from the top of the "umbrella-shaped" structure to the ground.

[0041] The above content of this embodiment mainly schematically embodies the method of calibrating the spatial pattern of two plants. By referring to this method, other plants in the sample area can be calibrated one by one, which is mainly suitable for areas with relatively flat terrain.

[0042] In most cases, the crown width 3 of plants (trees) is circular or nearly circular in the vertical direction and has an "umbrella-like" structure in the vertical direction. Figure 4 A1, the correspondence between the two crowns of trees Figure 4A2) is higher than other parts of the crown, and the crown center 4 of the plant (tree) is close to the center of the circle in the downward direction. One of the key points of the present invention is to cleverly utilize this inherent characteristic of plants (trees). After combining drone ranging (height) and virtual shaping, the spatial position of the plant can be quickly and accurately determined, which can not only quickly and accurately calibrate the geometric center of the plant's crown, but also quickly determine the crown outline of the plant.

[0043] The solution of the present invention not only greatly reduces the operational difficulty of plant spatial pattern calibration, but also significantly shortens the time required for plant spatial pattern calibration, improves implementation efficiency, and facilitates rapid and accurate calibration (after the flight path of the drone is set, the spatial pattern calibration of thousands of trees with a sample area of ​​500m*500m can be completed in only half a day, and the (coordinate) error can be controlled within 0.2m, while the traditional manual sample survey takes at least a week and the error is as high as more than ten meters). More importantly, it is possible to accurately calibrate plants in a large area (an area of ​​not less than 100m*100m) of sample plots without the need for operators to go to the site to calibrate one by one. Example

[0044] A method for calibrating plant spatial pattern based on high-precision positioning, the main steps of which refer to Example 1, and the main difference between the method and Example 1 is that: in the process of obtaining the measured distance data, the plant species information is obtained synchronously (in Example 1, the plant species information is obtained after the spatial coordinate position of the plant is calibrated). Specifically:

[0045] Step 1, determine the sample area according to the needs;

[0046] Step 2, construct a digital fence according to the boundary of the sample area;

[0047] Step 3, constructing a three-dimensional coordinate system in combination with the obtained digital fence, with two adjacent boundaries of the digital fence serving as the X-axis and Y-axis of the three-dimensional coordinate system, and the Z-axis of the three-dimensional coordinate system extending downward from the plane where the digital fence is located;

[0048] Step 4: Generate reference lines in the three-dimensional coordinate system. The horizontal reference lines are parallel to the X-axis or Y-axis, and the vertical reference lines are parallel to the Z-axis. The digital fence is divided into several areas, and a reference line is set every 0.2 meters.

[0049] Step 5, setting the UAV flight path along the boundary of the digital fence and the horizontal reference line. The UAV flight path includes several parallel straight paths, and the flight path is matched with the points corresponding to the sample area in combination with the positioning system; the UAV is equipped with multiple straight and closely arranged ranging sensors, and the optical path of the ranging sensors is vertically downward;

[0050] Step 6, control the drone to fly at a constant speed above the sample area according to the set path, and obtain the measured distance data in real time (the distance data refers to the distance between the light emitting end of the ranging sensor and the object to be measured; when the light emitted by the ranging sensor is blocked by leaves, the leaves are the object to be measured, and when the ranging sensor is located above the tree, the light it emits can basically be blocked by the leaves on the top layer or the leaves close to the top layer; when the light emitted by the ranging sensor is blocked by the ground, the ground is the object to be measured); combined with Figure 3 As shown, the drone is equipped with 12 straight and closely arranged distance measuring sensors, and the optical path spacing of adjacent distance measuring sensors is set to 20cm. When the drone flies at a constant speed above the sample area according to the set path, the width area that can be covered by the distance measuring sensors on the drone at the same time is as shown by the numbers 10 / 11 / 12 / 13. Taking two plants side by side (tree one and tree two) as an example, if the drone needs to fly back and forth four times to make the measurement range completely cover the two plants, then the width area measured by the drone during the first flight corresponds to number 10, the width area measured by the drone during the second flight corresponds to number 11, the width area measured by the drone during the third flight corresponds to number 12, and the width area measured by the drone during the fourth flight corresponds to number 13. Accordingly, the measured distance data are H1, H2...H48 in sequence;

[0051] In this step, the drone is also equipped with a multispectral camera for identifying plant species, and the plant species information is obtained synchronously during the process of obtaining the measured distance data;

[0052] Step 7, import the obtained distance data into the digital fence, and calibrate the vertical lines corresponding to the obtained distance data, that is, draw the corresponding vertical lines according to the distance value (also representing the optical path between the ranging sensor and the object to be measured), and then the model section is obtained as follows Figure 4 As shown;

[0053] Step 8: Fit the lower endpoints of all vertical lines to form a surface, such as Figure 5 As shown;

[0054] Step 9, define the “umbrella-like” structure on the obtained surface as a unit, as a virtual plant (such as Figure 5 The "umbrella-shaped" structure in the middle left part represents virtual plant 1, which corresponds to tree 1 in the sample area. The "umbrella-shaped" structure in the right part represents virtual plant 2, which corresponds to tree 3 in the sample area. The sample model is composed of all virtual plants, and the coordinates of the vertices of the "umbrella-shaped" structure are calibrated as the spatial position of the plant.

[0055] Since the plant species information is obtained in step 6, there will be multiple plant species information corresponding to the same plant. For example, when the drone measured the width area corresponding to number 10 during the first flight, tree one was identified as a holly tree. When the drone measured the width area corresponding to number 11 during the second flight, tree one was identified as a holly tree. When the drone measured the width area corresponding to number 12 during the third flight, tree two was identified as a soapberry tree. When the drone measured the width area corresponding to number 13 during the fourth flight, tree two was also identified as a soapberry tree. In order to improve the accuracy of the calibration result, the plant species information corresponding to the vertex of the "umbrella-shaped" structure (the crown of the plant) is used as the standard, that is, when the drone measured the width area corresponding to number 11 during the second flight, tree one was identified as a holly tree. Finally, tree one (X2, Y2, Z2) is calibrated as a holly tree (X2, Y2, Z2), and tree two (X3, Y3, Z3) is calibrated as a soapberry tree (X3, Y3, Z3).

[0056] Step 10: Filter out the data group with the largest value from the obtained distance data ( Figure 5 Among them, the largest values ​​screened out are H1, H27, H47, and H48. These four data are taken as a group, and their average value is calculated = (H1+H27+H47+H48) / 4). Then, the average value of the data group with the largest value is used as the height of the plane where the X-axis and Y-axis are located from the ground, and then the virtual plant trunk 8 is extended from the top of the "umbrella-shaped" structure to the ground.

Claims

1. A method for calibrating plant spatial pattern based on high-precision positioning, characterized in that the steps include: Step 1, determine the sample area according to the needs; Step 2, construct a digital fence according to the boundary of the sample area; Step 3, constructing a three-dimensional coordinate system in combination with the obtained digital fence, with two adjacent boundaries of the digital fence serving as the X-axis and Y-axis of the three-dimensional coordinate system, and the Z-axis of the three-dimensional coordinate system extending downward from the plane where the digital fence is located; Step 4: Generate reference lines in the three-dimensional coordinate system, with the horizontal reference lines being parallel to the X-axis or the Y-axis, and the vertical reference lines being parallel to the Z-axis, and divide the range of the digital fence into several areas, with a reference line set every n meters; Step 5, setting the flight route of the drone along the boundary of the digital fence and the horizontal reference line, and matching the flight route with the points corresponding to the sample area in combination with the positioning system; wherein the drone is equipped with a plurality of ranging sensors arranged in a straight line and closely, and the optical path of the ranging sensors is vertically downward; Step 6, control the drone to fly at a constant speed above the sample area according to the set path, and obtain the measured distance data in real time, where the distance data refers to the distance between the light emitting end of the distance measuring sensor and the object to be measured; Step 7, importing the obtained distance data into the digital fence, and generating vertical lines corresponding to the obtained distance data; Step 8, fitting the lower endpoints of all vertical lines to form a curved surface; Step 9, define the "umbrella-shaped" structure on the obtained surface as a unit, as a virtual plant, all virtual plants together constitute the sample model, and calibrate the coordinates of the vertices of the "umbrella-shaped" structure as the spatial position of the plant.

2. The method for calibrating plant spatial pattern according to claim 1, characterized in that: In the process of measuring distance data, the UAV flight path includes several parallel straight line paths.

3. The method for calibrating plant spatial pattern according to claim 2, characterized in that: The drone is equipped with 10-30 ranging sensors arranged in a straight line and closely together, and the optical path spacing between adjacent ranging sensors is no more than 15cm.

4. The method for calibrating plant spatial pattern according to claim 3, characterized in that: The drone is also equipped with a multispectral camera for identifying plant species, which can simultaneously obtain plant species information while obtaining distance measurement data.

5. The method for calibrating plant spatial pattern according to claim 3, characterized in that: The drone is also equipped with a multispectral camera for identifying plant species. The vertices of all the "umbrella-shaped" structures in step 9 are set as target points. After step 9 is implemented, the drone is controlled to fly again to the position directly above the plants in the sample area corresponding to each target point to obtain the plant species corresponding to the "umbrella-shaped" structure, and the obtained plant species information is marked at the corresponding position.

6. The method for calibrating plant spatial pattern according to any one of claims 1 to 5, characterized in that: The method also includes step 10: selecting the data group with the largest value from the obtained distance data, and then taking the average value of the data group with the largest value as the height of the plane where the X-axis and the Y-axis are located from the ground, and then extending from the vertex of the "umbrella-shaped" structure to the ground to generate a virtual plant trunk.