WO2020013736A1

WO2020013736A1 - Remote investigative diagnosis of nitrogen supply to crops

Info

Publication number: WO2020013736A1
Application number: PCT/RU2019/050111
Authority: WO
Inventors: Виктор Иванович АБРАМОВ; Дмитрий Александрович АНДРЯКОВ; Сергей Геннадьевич КЛАДКО; Дмитрий Трофимович РУБИН; Дмитрий Михайлович МИХАЙЛОВ; Александр Владимирович ТРУФАНОВ
Original assignee: Общество с ограниченной ответственностью "АгроДронГрупп"
Priority date: 2018-07-11
Filing date: 2019-07-11
Publication date: 2020-01-16
Also published as: RU2693255C1

Abstract

The claimed solution relates to the field of agriculture. The technical effect is a more accurate determination of the condition of crops with the aid of photometric aerial photography carried out by an unmanned aerial vehicle. A reference region is formed next to a field to be diagnosed, said region having a uniform surface and being delineated by markers. Within the field, monitoring regions delineated by markers and exhibiting different degrees of crop density are formed. Aerial photography of a crop growth zone is carried out using an unmanned aerial vehicle equipped with a multi-channel camera. Multispectral images of the field are obtained which are referenced to the geographical coordinates of the areas photographed. Ground measurements are carried out in the monitoring regions of the field using a nitrogen testing device in order to determine the nitrogen content in the leaves of the crops. A spectral radiometer is used to carry out ground measurements of the spectral brightness values of point objects on the reference area and the spectral brightness values of a calibration panel, ground spectral brightness indices are determined, and an NDVI or GNDVI map is generated, reflecting the nitrogen content in the crops.

Description

REMOTE DIAGNOSTIC DIAGNOSTICS OF PLANT SUPPORT WITH NITROGEN

FIELD OF TECHNOLOGY

[0001] The claimed technical solution relates to the field of agriculture, in particular, to methods for reconnaissance diagnosis of plant conditions using unmanned aerial vehicles (UAVs) to obtain photometric data.

BACKGROUND

[0002] Diagnosis of the state of plants, for example, mineral, primarily nitrogen, plant nutrition is a priority area of agrochemical science and agricultural practice [1]. To determine the needs of crops for such basic nutrients as phosphorus and potassium, soil diagnostics was widely used, i.e. determination of the mobile forms of these elements in soils, on the basis of which the doses of the corresponding types and forms of fertilizers were calculated. Soil diagnostics was also used to identify the need of plants for nitrogen fertilizers, mainly in the pre-sowing period or at the beginning of active vegetation of crops. But unlike the content of mobile forms of phosphorus and potassium, which is characterized by relative stability even for several years, the content of nitrogen compounds available for plant nutrition in soils requires constant monitoring during each growing season due to the instability in time and the dynamics of this indicator, with one side, and the particular demands of plants for nitrogen almost throughout their entire growing season, on the other.

[0003] As a rule, nitrogen fertilizers are applied to the soil in early spring immediately before sowing spring crops or superficially at the beginning of the growing season of winter crops, focusing on data from an agrochemical survey of soils, including operational diagnostics. During critical periods of vegetation (tillering, branching, trumping –– stalking, earing – flowering, seed formation), vegetative fertilizing with nitrogen fertilizers is carried out to optimize the nitrogen nutrition of cereals and other crops according to the data of chemical methods of plant diagnostics — stem and leaf [2, 3]. [0004] In recent decades, physical, in particular photometric, methods for diagnosing nitrogen nutrition of crops, based on the relationship between the intensity of green color of plants and their availability with nitrogen, have become increasingly important, especially abroad, [4]. The photodetectors of diagnostic devices record either the concentration of chlorophyll in the indicator organs of plants, or the intensity of its fluorescence. As a result of thematic processing of data of contact or remote determination of these indicators, the demand of certain crops for nitrogen fertilizers in one or another period of their vegetation is calculated. In this case, the calculation of the so-called vegetation index (NDVI), which represents the ratio of the difference between the infrared and red spectra of electromagnetic reflection of sunlight or artificial light from plants to their sum, is most widely used.

[0005] Plant biomass detectors of electromagnetic radiation are photometric devices of various designs used as portable (handheld) devices (European “YARA”, “CropCircle”, American SSM-200, SSM-1000, “GreenSeeker”, domestic models - single-beam and two-beam “Spectrolux”), as well as in the form of mobile N-sensors installed on fertilizer application units (“YARA”, “ALS”), and multi-zone photometers installed on aircraft or space platforms. Of the types of satellite imagery, photography and TV-shooting with fixed wavelengths of 0.3 - 1, 1 μm, spectrometric indication - 0.3 - 3.0 microns, infrared indication - 3 - 300 microns, microwave indication - 0.3 are distinguished. - 10 cm, radar indication - 10 - 70 cm. For example, the Russian satellite “Resurs-02D” is equipped with a multi-zone scanning device “ADAPTON” with a spectral range from 0.5 to 2.4 microns with a resolution on the terrain of 30 m, video spectrometer equipment "BC" with a spectral range from 0.4 to 1.0 μm and a resolution of 30 m, scanning device ultra-high resolution VZOR with a spectral range of 0.5 - 0.9 μm and a resolution of 2 m in panchrome and 4 m in a spectrum with a total number of spectral channels of 266. Spectrometric or radar information obtained in one way or another is used by different sectors of the people farms. Studies have shown that ground-based and aerospace surveys can be successfully used to diagnose nitrogen nutrition of plants [5].

[0006] However, despite significant advantages compared to traditional chemical diagnostics, both ground-based and space-based indications of nitrogen supply to crops have certain limitations: ground-based - in terms of coverage, aerospace - in time parameters. At the same time it was It is shown that for photometric inspection of crops, the use of low-flying aircraft, in particular helicopters, is most suitable, although its use has a significant limitation - the cost of diagnostic work. In this regard, the use of unmanned aerial vehicles - UAVs equipped with appropriate photometric equipment - seems to be the most promising for the operational diagnostics of nitrogen nutrition of plants.

[0007] Particular applications of UAVs for diagnosing, in particular, plant nutrition, are known, for example, from US20150254800A1 (FI2 Solutions LLC, 09/10/2015). This decision reveals the general principles of using the NDVI index obtained by UAV aerial surveys, with its subsequent correlation for predicting the nitrogen saturation index of plants. This approach has a rather large error in accuracy, since it does not provide for the correction of the resulting orthomosaic, which is used to construct the NVDI index map, and also does not provide for the calibration of UAV equipment based on ground-based photometric diagnostics.

SUMMARY OF THE INVENTION

[0008] The technical problem to be solved is to provide a new, more accurate and universal method for reconnaissance diagnostics of plant conditions using UAVs, as well as expanding the ability to analyze various indicators of plant conditions.

[0009] The technical result is to increase the accuracy of determining the state of plants using photometric information obtained during aerial surveys of UAVs by calibrating the photometric equipment of UAVs based on ground photometric diagnostics and correction of aerial survey data by reducing them to spectral brightness coefficients (CSW) measured by a ground-based spectroradiometer at reference sites.

[0010] The claimed result is achieved due to the method of reconnaissance diagnostics of the state of plants in the fields, containing stages in which:

- form a reference site next to the diagnosed field, and the site contains a uniform surface and is marked with markers;

- in the field of the diagnosed field, control areas are formed marked with markers, and the control areas have a different degree of plant density; - carry out aerial photography of the zone of plant growth using UAVs equipped with a multi-channel camera, and the coverage area of UAV surveys is 60%;

- using UAVs get multispectral images of the field with reference to the geographic coordinates of the survey areas;

- perform ground-based measurement of plants using the N-tester at the mentioned control field fields to obtain data on the nitrogen content in the leaves of plants;

- perform, using a spectroradiometer, ground-based measurement of the spectral brightness values of point objects on the reference site and the spectral brightness values of the calibration panel;

- determine the ground-based spectral brightness coefficients (CSW) by dividing the results of measuring the areas on the reference site with a spectroradiometer by the measurement result of the calibration panel, and bringing the result to a range of 0-1 (0-100%).

- transmit aerial survey data and ground-based measurements to a computing device, and process them, during which:

- form, according to aerial survey data, a multispectral orthophotomap of the diagnosed field;

- calculate the atmospheric corrected QWP for each pixel of the orthomosaic, by regressing the initial brightness values of the pixels obtained directly during aerial photography, to the QWS obtained from ground-based measurements of the reference sites included in the orthomosaic;

- carry out the construction of a map of the vegetation index NDVI or GNDVI, displaying the nitrogen content in the plants of the diagnosed field; and

- perform calibration of the obtained maps using the data of ground measurement of reference sites.

[OOP] In one particular example of the implementation of this solution, at the stage of constructing the vegetation index maps, at least one of the following indicators is additionally determined from the group: the amount of aboveground plant phytomass, seedling density.

DESCRIPTION OF DRAWINGS

[0012] FIG. 1 illustrates a general scheme for implementing the claimed solution.

[0013] FIG. 2 illustrates the process of transmitting data to a main computing device.

[0014] FIG. 3 illustrates a general view of UAV elements. [0015] FIG. 4 illustrates a general view of a computing device.

[0016] FIG. 5 illustrates a flowchart of a plant mapping state mapping algorithm.

DETAILED DESCRIPTION OF THE INVENTION

[0017] As shown in FIG. 1, the main elements that ensure the implementation of the claimed method for diagnosing the state of plants on the field (10) are UAVs (20), ground-based spectrometer (30) and N-tester (30).

[0018] As a ground-based spectrometer (30), for example, OceanOptics USB 2000+, Maya, and analogs covering the entire wavelength range of a multispectral camera mounted on a UAV and having the necessary sensitivity index can be used. As the N-tester (40), for example, Spad 502 Plus, Yara or their analogs can be used.

[0019] UAV (20) can be a helicopter or aircraft type aircraft, for example, UAV (20) developed by AgroDronGrupp LLC, such as: Phantom 4 Pro or Orlan (htip: /7agrodronegroup.ni/) ·

[0020] At the first stage of the implementation of the claimed solution, ground work is performed before receiving aerial survey data using UAVs (20).

[0021] Next to the diagnosed field (10), one or more test sites (ON, 111) are prepared, each of which is marked with white markers, which ensures their accurate identification in the pictures. The platform (PO, 11) is prepared with a size of approximately 7x5 meters and should contain a uniform surface of medium brightness. As a test site (PO 111), for example, a dirt or asphalt road, a plot of a plowed field without vegetation, can serve. When performing aerial photography, the test site (software, 1 11) should fall into the shooting area.

[0022] Calibration of photometric instruments (cameras, spectrometers), i.e. the creation of scales that should show the level of nitrogen supply to plants of a particular crop is the most labor-intensive, important and responsible operation in the entire technological chain of practical diagnostics of nitrogen nutrition of plants.

[0023] Field experiment designs should consist of variants representing doses of ammonium nitrate or other nitrogen fertilizers in the active substance, i.e. nitrogen (N) from zero (N0) to presumably optimal or above optimum, for example, N 120-150 kg / ha. Since large instruments are not required for calibration sowing area, then field experiments are laid with in-field control sites for each option with an area of several square meters, for example, 40 m2 (4 mx 10 m). If any other tasks are set before the field experiment, for example, ecological or economic, then each variant of the fertilizer of such a scientific experiment is laid down in triplicate - quadruple repetition. If the task is limited only by the calibration of the photometric device, then the so-called production (scientific-production) field experience is bookmarked in one or two repetitions to simplify technological operations.

[0024] In different places of the field (10), in particular, with low, high and medium density of plants, the in-field control sites (101) - (106) are marked with markers that are clearly distinguishable in aerial photography. The intra-field pads (101) to (106) are preferably made in an amount of 7-10 pcs. 5x5 meters in size.

[0025] The next step is to prepare the UAV (20) for aerial photography. The UAV sensors (20) are tuned for shooting, during which the UAV camera sensors (20) are calibrated and configured using the certified calibration panel. Next, the flight mission with the coordinates of the field (10) is loaded into the memory of the UAV (20). The flight altitude is set, which depends on the aircraft used to perform aerial photography. For example, in the case of using a UAV in the form of a copter, its flight altitude is set no higher than 100 m, and the shooting overlap is 60%. When using an aircraft-type UAV, its altitude may exceed the mentioned value. After the UAV is configured (20), it is launched to perform aerial photography in an automated mode.

[0026] The speed of movement of the UAV (20) depends on the flight altitude and speed of its operation (taking pictures of predetermined sections of the field), which is determined by the autonomous route of the UAV (20), in particular, the trajectory of its operation, as well as longitudinal overlap - frame rate , and transverse overlap - the distance between the zones of passage.

[0027] During the aerial survey, ground work is also carried out, which consists in measuring the spectral brightness coefficients of the surface of the test site (PO 111) with a uniform surface of medium brightness (7-10 points) with a ground spectrometer (30). The obtained indicators are then used to make atmospheric correction in the spectral aerial survey results.

[0028] Measurements at the test site (ON, 111) are alternated with frequent measurements of a certified calibration panel (at least one measurement of the panel spectrum before and after 10 measurements of the object in stable lighting or measurement of the panel spectrum up to and after each measurement of the object with changing lighting, for example, variable cloud cover).

[0029] At the in-field control sites (101) to (106), an N-tester (40) measures the nitrogen content in the leaves of the plants. The number of measurements can be different and depends on the type of plant, for example, for winter wheat in flag leaves at least 30 measurements are performed.

[0030] First of all, in special field experiments, the nature of the dependence of the state of crops on increasing doses of nitrogen fertilizers was revealed, since this indicator should be the basis for logical and statistical estimates of the photometric diagnostics of plant nitrogen nutrition. As a result of research, the dependence of photometric indicators on increasing doses of nitrogen fertilizers applied for crops, the relationship with other diagnostic indicators studied in these field experiments, the relationship with yield and crop quality was established.

[0031] It is characteristic that the readings of N-testers during the growing season of crops are somewhat reduced, which is associated with a gradual decrease in the content of chlorophyll, i.e. the transition of plants from a neoplasm of organic substances to their transportation from vegetative organs to generative ones with an appropriate biochemical transformation. This is evidenced, in particular, by the photometry results of white mustard (Fig. 4). During the transition from the phase of the beginning of flowering to the phase of seed formation, the readings of the Yar N tester noticeably decreased, and the dependence on the doses of nitrogen fertilizers applied under the crop even increased, due to the high dependence of the formation of mustard biomass on the supply of plants with nitrogen. In the phase of the beginning of flowering, the correlation coefficient of the readings of the photometer for the diagnosis of mustard leaves with doses of nitrogen was 0.78, in the phase of the end of flowering - 0.91, and in the phase of seed formation reached 0.94. It follows that the readings of photometers reflect the real supply of plants with nitrogen nutrition, which in turn affects the yield of crops. In other words, the high statistical and biologically proven reliability of the dependence of the N-tester readings on the doses of nitrogen fertilizers serves as the scientific basis for the diagnosis of nitrogen nutrition of plants, which makes it possible to abandon the complex and labor-intensive manual procedures of plant diagnostics unsafe for health, and to a certain degree to robotize diagnostic processes.

[0032] Also, at each in-field site (101) to (106), (using a video camera, smartphone, etc.): a general photo of the field, which characterizes the uniformity of seedlings, with reference to GNSS coordinates; a photograph of plants that fit inside the agronomist’s accounting framework (agronomist line) laid on the plants characterizes the density of the seedlings; A photograph of the most typical 5 plants on a white background characterizes the degree of plant development.

[0033] As shown in FIG. 2, after completing the UAV flight mission (20), the aerial survey data is transmitted together with the GNSS data for the coordinates of the images and the parameters of the external orientation (coordinates of the image centers, roll, pitch, height, etc.) to a computer computing device (50). The computing device (50) also transmits data from ground-based measurements carried out using a spectroradiometer (30) and an N-tester (40).

[0034] FIG. Figure 3 shows a general example of a UAV circuit (20). The UAV (20) contains a computing unit (201), which provides processing of the received data from the camera (205), and can be a processor or microcontroller.

[0035] The UAV memory (202) (20) may represent one or more means providing storage of the necessary program logic for performing the flight mission, as well as storage of images captured by the camera (205). The memory (202) is preferably a combination of RAM and ROM, and can be various known means of a volatile and / or non-volatile type, for example, integrated or removable flash memory (memory card), EEPROM, HDD, SSD, DRAM, SRAMH T .P.

[0036] The navigation system (203) is a receiver of GNSS signals, in particular GPS, GLONASS, Galileo, Compassnnn combinations thereof. The navigation system (203) also provides coordinates for photographs.

[0037] As the means of transmitting and receiving information (204), solutions can be applied to provide a data channel of a wired and / or wireless type, for example, a GSM / GPRS / LTE / 5G modem, Wi-Yesodul, Bluctoothn / or ВЕЕmodule, module satellite communications, etc.

[0038] The UAV camera (205) is a multispectral (5 channels) or hyperspectral camera (up to 40 channels) providing images in the visible (RGB) and near infrared (NIR) ranges.

[0039] The UAV (20) may also contain a battery charge monitoring sensor (206) that monitors the battery charge (207) and provides optimization of the UAV flight time (20) and its return for charging / replacing the battery (207). [0040] The UAV (20) moves using the propeller group (208).

UAV elements (20) are interconnected via a data bus (210), which ensures the transmission of the necessary control signals and information flows.

[0041] FIG. 4 shows an example implementation of a computing device (50). The device (50) can be selected from a number of known solutions and can be, without limitation, a personal computer, laptop, tablet, smartphone, server, mainframe, etc.

[0042] In general, device (50) comprises components such as: one or more processors (501), at least one memory (502), data storage means (503), input / output interfaces (504), means B / In (505), means of network interaction (506).

[0043] The processor (501) of the device performs the basic computing operations necessary for the operation of the device (50) or the functionality of one or more of its components. The processor (501) executes the necessary machine-readable instructions contained in the random access memory (502), providing the execution of the required logical data processing operations.

[0044] The memory (502), as a rule, is made in the form of RAM and contains the necessary program logic that provides the required functionality.

[0045] The storage medium (503) may be in the form of HDD, SSD disks, an array raid, network storage, flash memory, optical information storage devices (CD, DVD, MD, Blue-Ray disks) and the like.

[0046] Interfaces (504) are standard means for connecting and operating the device (50), for example, USB, RS232, RJ45, LPT, COM, HDMI, PS / 2, Lightning, FireWire, etc. The choice of interfaces (504) depends on the specific design of the device (50).

[0047] The following can be used as I / O data means (505): keyboard, joystick, display (touch screen), projector, touchpad, mouse, trackball, light pen, speakers, microphone, etc.

[0048] Network communication tools (506) are selected from devices capable of receiving and transmitting network data, for example, an Ethernet card, a WLAN / Wi-Fi module, a Bluetooth module, a BLE module, an NFC module, IrDa, an RFID module, GSM / GPRS / LTE / 5G modem, etc. Using means (505), the organization of data exchange via a wired or wireless data channel is provided, for example, WAN, PAN, LAN (LAN), Intranet, Internet, WLAN, WMAN, UMTS, GSM, 4G / 5G. [0049] The components of the device (50) are interfaced via a common data bus (510).

[0050] In FIG. 5 is a flowchart of plant diagnostics based on data obtained during ground scanning and aerial surveys, which are transmitted (601) to a computing device (50). Next, at step (602), the primary processing of the mentioned data is performed using a computing device (50), during which, using photogrammetric algorithms, a brightness-balanced digital spectral orthophotomap of the diagnosed field (10) is created.

[0051] In the next step (603), the results of ground-based and airborne spectrometric measurements are given in the spectral brightness coefficients (QPS), and atmospheric correction is performed. Such correction can be performed in camera software (205) to ensure correction of newly acquired images from UAVs (20), or in specialized GIS software using regression of the average values of the brightness of the reference area (software, 111) in each spectral zone (visible and infrared ranges) to the average values of the QWS of the reference sites for ground measurements.

[0052] Next, at step (605), one or more plant state maps are constructed based on spectrometric and photographic measurements of the diagnosed field (10).

[0053] The construction of a map of the nitrogen content in plants is carried out on the basis of calculations according to the formulas:

R800-R670 ^ _{LTTT TT} R750-R550.

NDVI = - R800 + R670 or GNDV1 = - R750 + R550) ',

where RXXX is the spectral brightness value at the wavelength.

[0054] A map of the above-ground plant phytomass and seedling density can be calculated based on a regression of the NDVI or GNDVI values to the values of these parameters determined at the in-field control sites.

[0055] Index maps are created in the GIS of the computing device (50) and calibrated at step (606) using regression on test sites (software, 111) marked with markers that are clearly visible in the pictures.

[0056] Ground measurements using the N-tester allow you to compare the NDVI and GNDVI indices with an agronomic indicator, the nitrogen content index. Several dozen measurements allow linear regression and instead of NDVI or GNDVI provide a map of the nitrogen content in plants. [0057] An embodiment of the present method is presented in more detail below. The studies were carried out on the basis of the Central Field Experimental Station of the All-Russian Research Institute of Agricultural Chemistry. D.N. Pryanishnikova (Moscow region) by setting up a field experiment with increasing doses of nitrogen fertilizers applied in the spring of 2017 under winter wheat of the Moscow variety 3. The experiment was repeated 3-fold, the size of the on-field control sites was 4 x 15 m. The experimental design includes 5 options: 1) . Control - N0, 2) N30, 3) N60, 4) N90, 5) N120. In the experiment, the traditional agricultural technique of cultivating winter wheat was used, which consists in combating harmful organisms by using chemical plant protection products.

[0058] In the course of the study, NDVI values were calculated, calculated using data obtained from UAVs, with other indicators for winter wheat (tubing phase, 2017), which are presented in Table 1.

Table 1

[0059] To determine the accuracy of the remote diagnostics of nitrogen supply using UAVs, it is necessary to compare it with the already proven methods of photometric diagnostics. These include, in particular, contact photometry using the N-tester "Yara". Studies have shown that with a coefficient of correlation of NDVI readings with UAVs with the readings of the Yara photometer at 0.97 (Table 1), the determination coefficient, that is, the direct relationship between these indicators is 0.94, or 94%. Thus, the difference between the data on the supply of winter wheat with nitrogen nutrition obtained by the ground method from the Yara photometer and remote using a photometer mounted on UAVs is only 6%, some of which (about 2%) the inaccuracy of the Yara photometer itself and only 4% for the error of remote photometry.

[0060] Carrying out atmospheric correction of images allows one to compare aerial photographs not only within the maps for one date, but also between dates, which makes monitoring the condition of crops much more efficient and allows obtaining reliable maps of the dynamics of quantitative indicators.

List of references:

1. Pryanishnikov D.N. Selected Works. T. 1. - M .: Kolos, 1965 .-- 767 p.

2. The methodology of field experiments to optimize the nitrogen nutrition of cereals, sugar beets and potatoes based on operational soil and plant diagnostics. Team of authors. - M. All-Russian Research Institute of Agrochemistry and Soil Science D.N. Pryanishnikova, 1985 .-- 92 p.

3. Tserling VV Diagnostics of nutrition of crops: a Handbook. - M .: Agropromizdat, 1990 .-- 235 p.

4. Osipov Yu.F., Ivanitsky Y.V., Shirinyan M.Kh., Afanasyev R.A., Galitsky V.V. Use of the N-tester Yara device for the diagnosis of nitrogen nutrition of winter wheat. Fertility, 2011. N ° 1.

5. Afanasyev R.A. Agrochemical support of precision farming // Problems of Agrochemistry, 2008. N ° 3. P. 46-53.

Claims

CLAIM

1. A method for reconnaissance diagnostics of the state of plants in the fields, comprising the steps of: forming a reference site next to the field being diagnosed, the site containing a uniform surface and marked with markers;

- in the field of the diagnosed field, control areas are formed marked with markers, and the control areas have a different degree of plant density;

- carry out aerial photography of the zone of plant growth using UAVs equipped with a multi-channel camera, and the coverage area of UAV surveys is 60%;

- determine the ground-based spectral brightness coefficients (CSW) by dividing the results of measuring the sections on the reference site with a spectroradiometer by the measurement result of the calibration panel, and bringing the result to a range of 0-1 (0-100%);

2. The method according to claim 1, characterized in that at the stage of constructing the maps of the vegetation index, at least one of the following indicators is additionally selected from the group: the amount of aboveground plant phytomass, seedling density.