WO2019109384A1 - 一种基于多尺度生境信息的苗期作物水肥检测和控制方法及装置 - Google Patents

一种基于多尺度生境信息的苗期作物水肥检测和控制方法及装置 Download PDF

Info

Publication number
WO2019109384A1
WO2019109384A1 PCT/CN2017/117190 CN2017117190W WO2019109384A1 WO 2019109384 A1 WO2019109384 A1 WO 2019109384A1 CN 2017117190 W CN2017117190 W CN 2017117190W WO 2019109384 A1 WO2019109384 A1 WO 2019109384A1
Authority
WO
WIPO (PCT)
Prior art keywords
fertilizer
polarization
water
hyperspectral
information
Prior art date
Application number
PCT/CN2017/117190
Other languages
English (en)
French (fr)
Inventor
张晓东
毛罕平
张雪威
邱白晶
张红涛
左志宇
高洪燕
倪纪恒
张怡雪
张放
Original Assignee
江苏大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 江苏大学 filed Critical 江苏大学
Priority to US16/646,369 priority Critical patent/US11406057B2/en
Publication of WO2019109384A1 publication Critical patent/WO2019109384A1/zh

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A01—AGRICULTURE; FORESTRY; ANIMAL HUSBANDRY; HUNTING; TRAPPING; FISHING
    • A01B—SOIL WORKING IN AGRICULTURE OR FORESTRY; PARTS, DETAILS, OR ACCESSORIES OF AGRICULTURAL MACHINES OR IMPLEMENTS, IN GENERAL
    • A01B79/00—Methods for working soil
    • A01B79/005—Precision agriculture
    • A—HUMAN NECESSITIES
    • A01—AGRICULTURE; FORESTRY; ANIMAL HUSBANDRY; HUNTING; TRAPPING; FISHING
    • A01C—PLANTING; SOWING; FERTILISING
    • A01C21/00—Methods of fertilising, sowing or planting
    • A01C21/007—Determining fertilization requirements
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
    • G01D21/00—Measuring or testing not otherwise provided for
    • G01D21/02—Measuring two or more variables by means not covered by a single other subclass

Definitions

  • the invention belongs to the technical field of facility agricultural biological information detection, and relates to a method and a device for detecting and controlling water and fertilizer of seedling crops based on multi-scale habitat information.
  • the water and fertilizer integrated equipment for small and medium-sized farmers is mostly a simple hand-held filling device, and the water and fertilizer integrated drip irrigation system has a low penetration rate.
  • the commonly used hand-held filling device not only manages extensively, but also causes serious waste of water and fertilizer resources and non-point source pollution, and wastes a lot of manpower, labor inefficiency, due to frequent application of fertilizer and water, facilities and vegetables are basically every three or four days. It is necessary to carry out the filling operation. It takes 1-2 days/person for each operation of the 10 mu greenhouse, resulting in high manpower and labor costs, and the economic benefits of facility planting are poor.
  • the present application has developed a multi-scale habitat information.
  • the method for detecting and controlling water and fertilizer in seedling crops can control the timing and irrigation amount of water and fertilizer irrigation according to the greenhouse environment and crop information.
  • the light and simplified water and fertilizer application device is developed to make it more cost-effective and practical.
  • the utilization rate of fertilizer and water in solar greenhouses and the reduction of labor intensity will help the development of high-efficiency, sustainable and intensive agriculture.
  • Intelligent monitoring technology for the growth process of greenhouse crops has become a key link in greenhouse production, and plant water stress status is an important basis for greenhouse intelligent water and fertilizer management.
  • Traditional experience detection and chemical analysis, crown temperature difference, image detection and other plant nutrition Moisture detection method, easy to be affected by environmental interference detection accuracy, easy to be misjudged, unable to meet the requirements can not meet the needs of modern facilities production, the present invention uses three-dimensional scanning imaging and micro-CT scanning imaging fusion of plant nutrition and moisture The internal and external macroscopic and microscopic morphological differences caused by stress are accurately phenotype.
  • the polarization-hyperspectral imaging technique can be used to detect the difference of nutrient water stress such as the apparent color texture of crops, combined with soil moisture content and ambient temperature and humidity illumination information.
  • the detection through the fusion and complementation of internal and external features of different scales, can achieve accurate detection of crop nutrient water stress, and based on this, decision-making control of water and fertilizer application is important to improve the intelligent technical level of greenhouse production of small and medium-sized farmers. Theoretical significance and application value.
  • the invention aims to provide a method and a device for detecting and controlling water and fertilizer of seedling crops based on multi-scale habitat information, so as to realize rapid and non-destructive accurate detection and water and fertilizer decision-making of water and fertilizer stress state of seedling crops, and provide scientific management for water and fertilizer of facilities. in accordance with.
  • the present invention achieves the above technical objects by the following technical means.
  • a method for detecting and controlling water and fertilizer in seedling crops based on multi-scale habitat information characterized in that it comprises the following steps:
  • step (3) normalizing the micro-CT characteristic variables, polarization-hyperspectral image feature variables and three-dimensional scanning imaging feature variables extracted in step (3), so that the characteristic value range is unified between 0-1;
  • step (3) Using the temperature, humidity, light and substrate moisture content information of the crop growth environment obtained in step (3), statistical analysis and calculation of temperature accumulation and illumination cumulative value since self-determination, combined with substrate moisture content and ambient temperature, humidity, illumination Information, calculate the transpiration of the plant, based on the collected crop micro-CT image, polarized hyperspectral image features, optimized characteristic variables of 3D laser scanning imaging, continuous tracking detection of crop nutrition and water stress, The fertilization irrigation model of the plant is obtained, and the model is input into the PLC control system;
  • step (3) micro-CT images, polarization hyperspectral image features, and three-dimensional laser scanning imaging characteristic parameters that characterize the growth state of the plant are obtained.
  • the PLC control system accurately and quantitatively detects the model based on multi-feature fusion nitrogen and water stress. Determine the current water stress, nitrogen stress status and degree of the plant; based on the fertilization irrigation model, combined with the fertilization pump speed and fertilization flow rate, and the relationship between the fertilization pump speed and the fertilization flow rate, use PLC to control the inverter output frequency, In turn, the rotational speed of the fertilizing pump is controlled to achieve precise control of the fertilization ratio and irrigation amount.
  • the growth environment information includes temperature, humidity, light intensity of the greenhouse crop growth environment and water content information of the culture substrate, and the temperature transmitter, the humidity transmitter, the illuminance transmitter, and the moisture content transmitter are used. Obtain.
  • micro-scale micro-CT detection step is:
  • IPL software combined with target image analysis, obtain the characteristic parameters such as stomatal density, sponge thickness, palisade tissue, cilia density, vascular bundle profile structure and fault image gray scale of plant leaves and stems;
  • the perlite matrix was peeled off to generate a three-dimensional image of the root system, and the root volume, root width, root hair density and distribution parameters were derived by IPL language.
  • the sampling polarization angles of the polarizers are set to 0°, 45°, 90°, 135°, and 180°, respectively; the hyperspectral pre-filter is transparent.
  • the over-wavelength is 560 nm and 1450 nm, and the polarization-hyperspectral scanning imaging is performed in the horizontal direction and the vertical direction, respectively, to obtain polarization-hyperspectral characteristic images of the front view and the top view;
  • a black contour high-reflection target point with a diameter of 6 mm is pasted on the crop leaves and planting utensils to be scanned. Since the surface of the blade is curved, the shortest distance between the two target points is controlled at 20 mm when the target point is reflected;
  • the laser power is 65%
  • the shutter time is 7.2 ms
  • the resolution is 0.50 mm.
  • a water and fertilizer detection and control device for seedling crops based on multi-scale habitat information characterized in that it comprises a fertilization mechanism, an information acquisition system and a control system.
  • the fertilization mechanism is mainly composed of a main inlet pipe, a filter, a feed water pump, a water inlet electromagnetic valve, an output fertilizer pipe, a fertilization electromagnetic valve, a stirring motor, a mixed fertilizer irrigation, a fertilization pump, and a fattening pipe; wherein the main water inlet pipe Connected to the water source, the other end of the main inlet pipe is connected to the filter, and the outlet of the filter is connected with the feed water pump to provide the basic water source of the water and fertilizer supply system, and the connection between the outlet pipe of the feed water pump and the mixed fertilizer pipe is connected Inlet water solenoid valve to control the on and off of the water source pipeline; the fertilizer pump is connected to the upper part of the fertilizer tank through the pipeline, and there is a fertilization solenoid valve between the two to realize the on/off control of the fertilizer pipeline; the top of the fertiliser tank is installed There is a stirring motor, and the stirring motor has a stirring blade at the end of the output shaft, and is driven by
  • the information collection system includes an ambient temperature transmitter, an environmental humidity transmitter, an ambient light transmitter, a substrate moisture content transmitter, an EC sensor, two pH sensors, a liquid level sensor, a pressure transmitter, and crop information.
  • the ambient temperature transmitter, the ambient humidity transmitter, the ambient light transmitter, and the substrate moisture content transmitter are respectively connected to the analog module;
  • the EC sensor is connected in the fertilizer applicator pipeline for detecting the concentration of the nutrient solution, EC
  • the output end of the sensor is connected to the input end of the analog input module;
  • the two pH sensors are respectively installed at a distance of 10 cm from the liquid level of the mixed fertilizer tank and 20 cm from the bottom of the mixed fertilizer tank, and are respectively connected to the input end of the analog module.
  • the liquid level sensor is placed at the bottom of the fertilizer tank, and the signal output end of the liquid level sensor is also connected to the input end of the analog module.
  • the liquid level is judged based on the principle of different underwater pressure; the pressure transmitter is connected at the main pipe. In the middle of the road, to judge the pressure of the main pipeline, the output end thereof is connected with the input end of the analog module;
  • the crop information detecting system comprises a micro-CT scanning system, a polarization-hyperspectral imaging system and a three-dimensional laser scanning system;
  • the control system comprises a PLC controller, a touch screen, a frequency converter and an actuator, wherein the touch screen is connected with the PLC controller for human-computer interaction, input control mode and crop growth state information; wherein the input end of the frequency converter and the PLC The output end of the controller is connected, the output end of the frequency converter is connected with the fertilization pump, and the analog quantity module is connected with the PLC controller for realizing the multi-channel information acquisition control of the information acquisition system by the PLC controller;
  • the system combines the environmental information according to the water and fertilizer demand of the crop, and can realize the frequency control of the frequency converter through the PLC controller, thereby realizing the control of the speed of the fertilization pump to adjust the flow rate of the fertilization pipeline, and finally realize the master in the constant pressure main road. Precise control of the amount of irrigation and fertilization of the pipeline.
  • the sample holder is rotated, and is fixed to the bottom of the sample chamber by screws on the base, and the rotating center of the base of the rotating sample holder is mounted with a rotating shaft, and the end of the rotating shaft is fixed and fixed with a circle.
  • the sample holder is rotated, and the rotating shaft drives the sample holder to rotate 360°, and the X-ray emitter fixed in the middle of the firing chamber realizes the tomographic scanning process of the sample by the tilting motion.
  • the polarization-hyperspectral image detection system includes a control system, a two-coordinate sample stage, an image acquisition system, and a light source system;
  • the image acquisition system comprises two polarization-hyperspectral imaging systems, an image collector, a vertical arm and a cantilever;
  • the vertical arm is composed of a first base, a pole with a lead screw and a first slider, the first base It is fixed to the left side of the bottom of the light box by screws.
  • the upper part of the first base is connected with the vertical rod through a hinge.
  • the vertical rod can swing left and right with the hinge as the center to complete the spatial position adjustment of the imaging device;
  • the first pole is mounted on the pole a slider;
  • the first polarization-hyperspectral imaging system is mounted on the first slider, and the first slider can be driven up and down along the vertical rod by the lead screw to drive the first polarization-hyperspectral imaging system to find the optimal detection position.
  • the cantilever is composed of a second base, a crossbar with a lead screw and a second slider.
  • the second base is fixed to the upper part of the right side plate of the light box by screws, and the second base is connected with the cross bar by a hinge, and the cross bar can
  • the upper and lower swings are centered on the hinge to complete the spatial orientation adjustment of the imaging device;
  • the second slider is mounted on the crossbar, the second polarization-hyperspectral imaging system is mounted on the second slider, and the second slider can be wired
  • the bar drive moves left and right along the horizontal bar in the horizontal direction, and drives the second polarization-hyperspectral imaging system to find the optimal detection position, thereby realizing the acquisition of the polarized hyperspectral image information in the overhead direction;
  • the light source system is composed of a visible light-near-infrared light source and a pan/tilt head, and a cloud platform is installed at a bottom end and a top end of the pole, a right end and a left end of the pole, respectively, and each visible light source is equipped with a visible light-near-infrared light source.
  • the visible-near-infrared light source can be set at the elevation angle through the gimbal to achieve clear and uniform imaging of the plant;
  • the two-coordinate sample table is fixed at a geometric center position of a bottom plane of the light box, and a sample bracket is mounted on a top end of the vertical screw for placing the sample to be tested, and the sample holder can be driven by the movement of the horizontal lead screw and the vertical lead screw.
  • the frame realizes the uniform displacement in the horizontal and vertical directions, and can cooperate with the image acquisition control system to realize the scanning imaging of the first polarization-hyperspectral imaging system and the second polarization-hyperspectral imaging system;
  • the polarization-hyperspectral imaging system is composed of a front polarizing plate, a polarizing plate driving device, a pre-filter, a filter switching device, a spectrograph and an imaging system from front to back, respectively, and the polarizing plate is in the whole system.
  • the polarization drive is driven by 360° rotation to set the arbitrary polarization angle.
  • the spectrograph and imaging system can realize the setting of the polarization angle and the acquisition of the step polarization information;
  • the polarizer is 560nm and
  • the 1450nm narrow-band filter and the filter adopt the method of rotary switching, and cooperate with the spectrograph and imaging system to realize the acquisition of the hyperspectral nutrient and water stress characteristic images of the crop sample in the front view and the top view;
  • the control system includes a control computer, a light source controller, an image collector, and a motion controller;
  • the light source controller is connected to the visible light-near-infrared light source to realize light source control of different light intensity and light quality;
  • the image acquisition device connects two polarization-hyperspectral imaging systems and a control computer, and the control computer issues instructions to realize imaging information collection and analysis of the front-view and top-view polarization-hyperspectral imaging systems;
  • the motion controller is connected to the two-coordinate sample table, the vertical arm, the cantilever and the pan/tilt; at the same time, the motion controller is connected to the control computer, and the control computer issues a command to realize the lifting and horizontal displacement control of the two-coordinate sample table, the opposite arm and the cantilever
  • the slider drive control, as well as the tilt and tilt control of the pan/tilt is connected to the control computer, and the control computer issues a command to realize the lifting and horizontal displacement control of the two-coordinate sample table, the opposite arm and the cantilever.
  • the invention aims at precise fertilization irrigation for the current dynamic ratio of nutrient solution of greenhouse crop seedlings, and the technical scheme relates to: 1 precise dispensing technology of nutrient solution: by controlling the inflow of mother liquor, the precise concentration of nutrient solution is realized. The relationship model between nutrient solution concentration and variable frequency control parameters was established under different nutrient solution ratios, and the regulation methods of nutrient solution concentration and application amount were given. 2Irrigation volume and fertilization amount decision: Through the research on the growth state of crops under different greenhouse conditions such as different light intensity, light accumulation, temperature and humidity, combined with the influence of water and fertilizer on crop growth information; formulate irrigation amount and fertilization amount under greenhouse conditions Decision making optimization program.
  • the present invention has the following advantageous effects.
  • the invention combines crop information detection technology, greenhouse environment information detection technology, liquid fertilizer concentration precision blending technology and irrigation fertilization amount control decision-making control technology, and proposes to be able to according to main environmental information such as greenhouse temperature, humidity, illumination and substrate moisture content. Combined with multi-information crop nutrient information, precise control of irrigation timing, water and fertilizer irrigation and matching crop fertilization and irrigation control strategies.
  • a method for detecting and controlling water and fertilizer in seedling crops based on multi-scale habitat information was proposed to solve the problem of dynamic control of water and fertilizer according to crop demand, and overcome the limitations of fertilization control based on environmental factor information at present. Fertilization, thus greatly reducing the amount of fertilizer, reducing labor costs and improving economic efficiency.
  • the fertilizing device of the invention uses the frequency converter to control the speed of the fertilizing pump to adjust the flow rate of the liquid fertilizer pipeline, and sends the liquid fertilizer to the constant pressure main pipeline, realizing the real-time dynamic and precise control of the liquid fertilizer concentration ratio and the irrigation amount.
  • the frequency conversion technology based on crop habitat information feedback is used to realize the automatic proportioning and variable filling operation of greenhouse water and fertilizer.
  • the constant pressure variable frequency water and fertilizer ratio control technology based on crop habitat information feedback has not been seen in greenhouse vegetable production.
  • the present invention uses micro-CT to scan microscopic morphological characteristics of characteristic variables such as stomatal, corpus cavernosum, palisade tissue, cilia, vascular bundle, root volume, root and root hair density under crop nutrient and water stress; using double-position polarization-hyperspectral Imaging system, obtaining macroscopic morphological features such as crown width, plant height and leaf dip angle of water stress; and leaf vein distribution, average gray scale, leaf edge shadow area and 0° of nutrient and moisture sensitive wavelengths of 560 nm and 1450 nm hyperspectral images, Polarization state of 560nm, 1450nm characteristic image, stock vector, Mueller matrix variable, plant volume, leaf area, stem diameter, plant height, etc.
  • characteristic variables such as stomatal, corpus cavernosum, palisade tissue, cilia, vascular bundle, root volume, root and root hair density under crop nutrient and water stress
  • double-position polarization-hyperspectral Imaging system obtaining macroscopic
  • FIG. 1 is a schematic view showing the structure of a water and fertilizer detecting and controlling device for a seedling crop according to the present invention.
  • FIG. 2 is a schematic view showing the structure of a micro-CT apparatus used in the present invention.
  • FIG. 3 is a schematic view showing the structure of a polarization-hyperspectral imaging system employed in the present invention.
  • FIG. 4 is a schematic structural view of a three-dimensional laser scanning system used in the present invention.
  • FIG. 5 is a flow chart of a method for detecting and controlling water and fertilizer of seedling crops based on multi-scale habitat information according to the present invention.
  • Figure 6 Three-dimensional spatial grid model of lettuce.
  • the fertilization mechanism is mainly composed of main inlet pipe 1, filter 2, inlet pump 3, inlet solenoid valve 4, fertilizer pipe 5, fertilization solenoid valve 6, stirring motor 22, fertilizer irrigation 23, fertilizer pump 24, and
  • the fertilizer pipe 26 is composed; wherein the main water inlet pipe 1 is connected to the water source, and the other end of the main water inlet pipe 1 is connected to the filter 2, and the water outlet of the filter 2 is connected with the feed water pump 3 to provide a basic water source for the water and fertilizer supply system,
  • a water inlet solenoid valve 4 is connected between the outlet line of the inlet water pump 3 and the fertilizer line to control the on and off of the water source line;
  • the fertilizer pump 24 is connected to the upper part of the fertilizer tank 23 through the pipeline, and there is a
  • the fertilizing electromagnetic valve 6 is used for realizing the on-off control
  • the information collection system includes an ambient temperature transmitter 12, an ambient humidity transmitter 13, an ambient light transmitter 14, a substrate moisture content transmitter 15, an EC sensor 7, and two pH sensors 19-1 and 19-2.
  • the ambient temperature transmitter 12, the ambient humidity transmitter 13, and the ambient light transmitter 14 are installed inside the greenhouse, and the substrate moisture content transmitter 15 is installed in the substrate of the crop perlite potted; the ambient temperature transmitter 12
  • the ambient humidity transmitter 13, the ambient light transmitter 14, and the substrate moisture level transmitter 15 are respectively coupled to the analog module 16.
  • the EC sensor 7 is connected in the fertilizer applicator line for detecting the concentration of the nutrient solution.
  • the output of the EC sensor 7 is connected to the input of the analog input module 16, and the two pH sensors 19-1 and 19-2 are respectively installed. 10cm from the liquid level of the fertilizer tank 23 and 20cm from the bottom of the fertilizer tank, the two pH sensors 19-1 and 19-2 arranged in the upper and lower sides of the fertilizer tank 23 are respectively connected to the input end of the analog module 16;
  • the difference between the upper and lower pH sensors 19-1 and 19-2 is compared, the uniformity of the fertilizer agitation is judged, the rotational speed required for the agitating motor is determined, and the agitating motor and the agitating device are self-started and stopped.
  • the liquid level sensor 20 is placed at the bottom of the fertilizer tank 23, and the liquid level is judged based on the principle of different underwater pressure.
  • the 4-20 mA signal output end is connected to the input end of the analog quantity module 16, and the pressure transmitter 25 is connected. In the middle of the main line to determine the main line pressure, the output of the pressure transmitter 25 is connected to the input of the analog module 16.
  • the crop information detection system includes a micro-CT scanning system, a polarization-hyperspectral imaging system, and a three-dimensional laser scanning system for detecting crop information indicating micro-scale, leaf-scale, and canopy scales of plants under nitrogen and water-fertilization stress levels.
  • 27 is a rotating sample holder, which is fixed to the bottom of the detecting sample chamber by screws on the base, and the rotating center of the rotating sample holder base is mounted with a rotating shaft and a rotating shaft end.
  • the circular sample holder is fixedly mounted, and the rotating shaft drives the sample holder to rotate 360° when detecting, and the X-ray emitter 29 fixed in the middle of the firing chamber realizes the tomographic scanning process of the sample by the pitching motion.
  • the polarization-hyperspectral image detection system is shown in FIG. 3, and includes a control system, a dual coordinate sample stage 32, an image acquisition system, and a light source system; wherein the image acquisition system includes two polarization-hyperspectral imaging systems 35 and an image collector.
  • the vertical arm 33 is composed of a first base 33-1, a lead rod 33-2 with a lead screw and a first slider 33-3, and the first base 33-1 is fixed by screws
  • the upper portion of the first base 33-1 is connected to the upright 33-2 by a hinge, and the upright 33-2 can swing left and right with the hinge as the center to complete the spatial orientation adjustment of the imaging device
  • a first slider 33-3 is mounted on the upright 33-2; the first polarization-hyperspectral imaging system 35-1 is mounted on the first slider 33-3, and the first slider 33-3 can be driven by the lead screw Moving up and down along the vertical rod 33-2, the first polarization-hyperspectral imaging system 35-1 is driven to find the optimal detection position, and the polarization hyperspectral image information in the main viewing direction is acquired.
  • the cantilever 34 is composed of a second base 34-1, a crossbar 34-2 with a lead screw, and a second slider 34-3.
  • the second base 34-1 is fixed to the upper part of the right side plate of the optical box 42 by screws.
  • the second base 34-1 is connected to the crossbar 34-2 through a hinge, and the crossbar 34-2 can swing up and down with the hinge as the center to complete the spatial orientation adjustment of the image forming apparatus;
  • the crossbar 34-2 is mounted with the second
  • the slider 34-3, the second polarization-hyperspectral imaging system 35-2 is mounted on the second slider 34-3, and the second slider 34-3 can be driven by the lead screw along the horizontal bar 34-2 in the horizontal direction Moving, the second polarization-hyperspectral imaging system 35-2 is driven to find the optimal detection position, and the polarization hyperspectral image information in the overhead direction is acquired.
  • the light source system is composed of a visible light-near-infrared light source 37 and a pan/tilt head 36.
  • a pan/tilt 36 is installed at the bottom end and the top end of the upright bar 33-2, and the right end and the left end of the upright bar 34-2, respectively.
  • the visible-near-infrared light source 37 is separately mounted, and the visible-near-infrared light source 37 can be set by the pan and tilt angle 36 to achieve clear and uniform imaging of the plant.
  • the two-coordinate sample table 32 is fixed at a geometric center position of a bottom plane of the light box 42, wherein 32-1 is a horizontal lead screw, a 32-2 position vertical lead screw, and a sample bracket is mounted on a top end of the vertical lead screw 32-2.
  • 32-1 is a horizontal lead screw
  • 32-2 position vertical lead screw
  • a sample bracket is mounted on a top end of the vertical lead screw 32-2.
  • the movement of the horizontal lead screw 32-1 and the vertical lead screw 32-2 can drive the sample holder to achieve uniform displacement in the horizontal and vertical directions, and can realize the push-type polarization with the image acquisition control system.
  • the polarization-hyperspectral imaging system 35 is composed of a front polarizing plate, a polarizing plate driving device, a pre-filter, a filter switching device, a spectrograph and an imaging system from front to back, respectively, and the polarizing plate is entirely At the forefront of the system, 360° rotation is driven by the polarization driving device to set the arbitrary polarization angle.
  • the spectrograph and imaging system can realize the setting of the polarization angle and the acquisition of the step polarization information.
  • the polarizer is 560 nm.
  • the filter adopts the rotary switch mode, and the acquisition of the hyperspectral nutrient and water stress characteristic images of the crop sample by the spectrograph and the imaging system;
  • the control system includes a control computer 41.
  • the spectral imaging system 35 and the control computer 41 are commanded by the control computer 41 to enable imaging information acquisition and analysis of the front-view and top-view polarization-hyperspectral imaging systems 35.
  • the motion controller 38 is connected to the two-coordinate sample table 32, the vertical arm 33, the cantilever 34 and the pan/tilt 36; at the same time, the motion controller 38 is connected to the control computer 41, and the control computer 41 issues an instruction to realize the lifting of the two-coordinate sample table 32. And horizontal displacement control, opposite arm 33, slider drive control of the boom 34, and pitch angle control of the pan/tilt head 36.
  • the three-dimensional scanning imaging system is composed of a PC 43, a FireWire adapter 44, a FireWire cable 45, a handheld three-dimensional scanning head 46, and a power module 47, wherein the handheld three-dimensional scanning head 46 is connected to the FireWire adapter 44 via a FireWire cable 45, and The three-dimensional scanning control and information collection of the handheld three-dimensional scanning head 46 are realized by the PC software through the FireWire adapter 44.
  • the power module 47 is connected to the FireWire adapter 44 to supply power to the PC; the power module 47 and the handheld three-dimensional scanning
  • the head 46 is connected to provide power to the handheld three-dimensional scanning head 46.
  • the control system includes a PLC controller 18, a touch screen 21, a frequency converter 9 and an actuator, wherein the touch screen 21 is connected to the 422 communication port of the PLC controller 18 for human-computer interaction, input control mode and crop growth status information.
  • the input end of the frequency converter 9 is connected with the output end of the PLC controller 18, the output end of the frequency converter 9 is connected with the fertilizing pump 24, and the analog quantity module 16 is connected with the PLC controller 18 for implementing the PLC controller 18 Multi-channel information collection for information collection systems.
  • PLC controller 18 has built-in multi-featured fusion nitrogen and water stress accurate quantitative detection model, fertilization irrigation model model fertilization pump rotation speed and fertilization flow rate, and the relationship between fertilization pump speed and fertilization flow rate, PLC controller 18 according to more Characteristic fusion nitrogen and water stress precise quantitative detection model to determine the current water stress, nitrogen stress status and extent of the plant; and then based on the fertilization irrigation model, combined with fertilization pump speed and fertilization flow, and fertilization pump speed and fertilization flow
  • the relationship model uses PLC to control the output frequency of the inverter, and then controls the rotation speed of the fertilizing pump to achieve precise control of the fertilization ratio and irrigation amount.
  • the method for detecting and controlling water and fertilizer of seedling crops based on multi-scale habitat information mainly comprises the following steps:
  • Step 1 Adopt a standard nutrient solution formula, use perlite as a culture medium, and use a soilless culture to colonize greenhouse vegetable crops, and adopt greenhouse standardization management to manage the nutrient elements and water supply in the seedling stage.
  • Step 2 Select plants after one week of colonization for water stress and nitrogen stress samples
  • the water supply is divided into 5 different levels according to 100%, 80%, 60%, 40% and 20% of the standard supply under the condition of constant nutrient elements, each level 20 a sample of 100 samples for the cultivation of water stress samples;
  • the nitrogen supply is divided into 4 different levels according to 200%, 100%, 50% and 25% of the standard supply, 20 samples per level. A total of 80 samples were used for nitrogen stress samples.
  • Step 3 After 3 days of water stress at the seedling stage, the micro-scale micro-CT, leaf-scale polarization-hyperspectral imaging scan and canopy-scale plants were carried out at 7 days, 15 days and 21 days after stress. Multi-information acquisition of canopy three-dimensional scanning imaging, while obtaining temperature, humidity, light intensity and water content information of the cultivation substrate in the greenhouse crop growth environment. The above information is obtained by following the steps below:
  • Temperature, humidity, illumination, and substrate moisture content information are obtained using a temperature transmitter, a humidity transmitter, an illuminance transmitter, and a moisture content transmitter.
  • IPL software combined with target image analysis, obtain the characteristic parameters such as stomatal density, sponge thickness, palisade tissue, cilia density, vascular bundle profile structure and fault image gray scale of plant leaves and stems;
  • the perlite matrix was peeled off to generate a three-dimensional image of the root system, and the root volume, root width, root hair density and distribution parameters were derived by IPL language.
  • the sample After acquiring the scanned image of the micro-CT and performing feature extraction, the sample is sequentially taken out for scanning imaging of the polarization-hyperspectral image, and the following steps are performed:
  • the sampling polarization angles of the polarizing plates are set to: 0°, 45°, 90°, 135°, 180°;
  • the high-spectral pre-filter has a wavelength of 560 nm and 1450 nm, and is subjected to push-type polarization-hyperspectral scanning imaging in the horizontal direction and the vertical direction, respectively, to obtain polarization-hyperspectral characteristic images of the front view and the top view;
  • the sample is sequentially taken out for scanning imaging of the polarization-hyperspectral image: before the 3D laser scanning data is collected, the laser power, shutter time and acquisition software of the scanner sensor are determined in advance. Resolution to ensure the clarity of the 3D model.
  • the laser power was set to 65%
  • the shutter time was 7.2 ms
  • the resolution was 0.50 mm
  • the three-dimensional shape of the plants collected under the parameters was set.
  • a black contour high-reflection target point with a diameter of 6 mm is pasted on the crop leaves and planting utensils to be scanned. Since the surface of the blade is curved, the shortest distance between the two target points is controlled at 20 mm when the target point is reflected;
  • Step 4 Perform routine physical and chemical testing: weigh the dry and wet weight of the sample to determine the true moisture content of the plant; use the AutoAnalyzer 3 continuous flow analyzer produced by SEAL, UK, to determine the total nitrogen content of the sample by Kjeldahl nitrogen nutrition test; Scanning electron microscopy and ultra-depth-depth 3D microscopic imaging technology to obtain the stomatal and cilia density of the plant, as well as the thickness of the sponge and palisade tissue, the distribution density of the vascular bundle and the diameter of the tube;
  • Step 5 normalizing the micro-CT characteristic variable, the polarization-hyperspectral image feature variable and the three-dimensional scanning imaging feature variable extracted in step 3, so that the characteristic value range is unified between 0-1;
  • Step 7 Using SVR support vector machine regression method to perform feature layer fusion, and establish characteristic variables such as stomatal, sponge, palisade tissue, cilia, vascular bundle, root volume, root and root hair density, and polarization based on micro-CT system.
  • Step 8 Using the temperature, humidity, illumination and substrate moisture content information of the crop growth environment obtained in step 3, statistically analyzing and calculating the temperature accumulation and the cumulative value of the light since the colonization, combined with the substrate moisture content and the ambient temperature, humidity, and illumination information, Based on the calculated transpiration of plants, based on the micro-CT images, polarized hyperspectral image features and 3D laser scanning imaging characteristics of crops, the crops were continuously and continuously monitored for crop nutrient and water stress, and the fertilization amount of plants was obtained. Model and input the model into the PLC control system;
  • Step 9 Under the condition of constant pressure of the main road, establish a relationship model between fertilization flow rate and fertilization amount, fertilization pump speed and fertilization flow rate, and input the model between fertilization flow rate and fertilization amount, fertilization pump speed and fertilization flow rate.
  • PLC control system
  • Step 10 According to the step (3), obtain a micro-CT image, a polarization hyperspectral image feature, a three-dimensional laser scanning imaging characteristic parameter for characterizing the growth state of the plant, and an accurate quantitative detection model of the PLC control system according to the multi-feature fusion nitrogen and water stress. Determine the current water stress, nitrogen stress status and degree of the plant; based on the fertilization irrigation model, combined with the fertilization pump speed and fertilization flow rate, and the relationship between the fertilization pump speed and the fertilization flow rate, use PLC to control the inverter output frequency, In turn, the rotational speed of the fertilizing pump is controlled to achieve precise control of the fertilization ratio and irrigation amount.

Landscapes

  • Life Sciences & Earth Sciences (AREA)
  • Soil Sciences (AREA)
  • Environmental Sciences (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Mechanical Engineering (AREA)
  • Investigating Or Analysing Materials By Optical Means (AREA)

Abstract

一种基于多尺度生境信息的苗期作物水肥检测和控制方法及装置，基于作物冠层尺度的三维激光扫描信息、叶片尺度的偏振-高光谱成像信息、微尺度的micro-CT扫描信息，对作物水肥胁迫的多尺度特征进行融合分析，结合作物生长温室内的温度、湿度、光照和基质含水率的实时反馈，通过多信息融合建模，实现对作物水肥胁迫和需水需肥信息的综合判断和反馈，并给出了施肥灌溉量的决策信息。基于水肥决策信息，水肥控制系统基于变频调速技术和管道恒压控制技术，通过动态控制灌溉泵和施肥泵的转速，控制管路压力、施肥流量，结合EC值的动态反馈，实现液肥配比和灌溉量精确控制。

Description

一种基于多尺度生境信息的苗期作物水肥检测和控制方法及装置 技术领域
本发明属于设施农业生物信息探测技术领域，涉及一种基于多尺度生境信息的苗期作物水肥检测和控制方法及装置。
背景技术
我国设施面积已达410万公顷，但适用于日光温室、钢架大棚的水肥一体化技术装备研究较少，大型现代化温室采用的进口水肥灌施系统能够实现多路肥料的精确混合施用和科学管理，但是通常体积较大，操作和管理复杂，价格也较高，不适合普通农户。而国产小型施肥系统，水肥管理粗放，尤其是缺少作物营养、水分需求信息，因此不能做到按需施肥灌溉、不能达到节水节肥的目的，且大多没有混肥搅拌和沉积排出装置，易发生沉积和堵塞，故障率较高。
中小农户的水肥一体化装备大多为简易型手持式的灌施装置，水肥一体化滴灌系统普及率较低。而普遍使用的手持式灌施装置不仅管理粗放，造成水肥资源的严重浪费和面源污染，且浪费大量的人力，劳动效率低下，由于设施肥水施用频繁，设施果蔬基本上每隔三四天就需要进行灌施作业，10亩大棚每次作业需要1-2天/人，导致人力和用工成本较高，设施种植的经济效益较差。
针对目前中小农户日光温室生产水肥灌施自动化程度低，水肥不能精确按需调控，以及大水大肥造成的农产品品质下降和面源污染等问题，本发明申请开发了一种基于多尺度生境信息的苗期作物水肥检测和控制方法，能依据温室环境和作物信息控制水肥灌溉时机、灌溉量；针对日光温室种植农艺要求，研发轻简化水肥灌施装置，使之更加低成本和实用化，提高日光温室肥水的利用率及降低人工劳动强度，有助于精确农业实现高效、持续、集约的发展。
温室种植作物的生长过程的智能监测技术已成为温室生产的关键环节，而植株水分胁迫状态是温室智能化水肥管理的重要依据，传统经验检测法和化学分析、冠气温差、图像检测等植株营养水分检测方法，易受环境干扰检测精度较低等问题，容易出现误判，无法满足要求无法满足现代化设施生产的需求，本发明利用三维扫描成像和micro-CT扫描成像融合对植株的营养和水分胁迫导致的内外宏观和微观形态差异进行精确表型，利用偏振-高光谱成像技术则能够对作物表观颜色纹理偏振态等营养水分胁迫的差异进行 探测，结合土壤含水率和环境温湿度光照信息的探测，通过不同尺度的内外特征的融合互补能够实现对作物营养水分胁迫的精确探测，并以此为依据进行水肥灌施的决策控制，对提高中小农户温室生产的智能化技术水平具有重要的理论意义和应用价值。
发明内容
本发明的目的在于提供一种基于多尺度生境信息的苗期作物水肥检测和控制方法及装置，以实现对苗期作物水肥胁迫状态的快速无损精确探测和水肥决策，为设施水肥的科学管理提供依据。
本发明是通过以下技术手段实现上述技术目的的。
一种基于多尺度生境信息的苗期作物水肥检测和控制方法，其特征在于，包括如下步骤：
(1)采取标准营养液配方，利用珍珠岩作为培养基质，采用无土栽培定植温室蔬菜作物，采用温室标准化管理方式进行管理，保证苗期作物营养元素和水分的正常供给；
(2)选取定植一周后的植株分别进行水分胁迫、氮素胁迫样本培育；
(3)苗期样本水分胁迫、营养胁迫分别在胁迫后通过微尺度的micro-CT、叶片尺度的偏振-高光谱成像扫描、冠层尺度的植株冠层三维扫描成像进行植株生长状态的连续检测与信息采集，同时获取植株生长环境信息；
(4)进行常规理化检测：称量样本的干湿重，确定植株的含水率真值；用凯氏定氮法测定样品全氮含量；采用扫描电镜和超景深3D显微成像技术，获取植株的气孔和纤毛密度，以及海绵体和栅栏组织厚度，纤维管束分布密度和管径实测值；
(5)对步骤(3)提取的micro-CT的特征变量、偏振-高光谱图像特征变量和三维扫描成像特征变量进行归一化处理，使其特征值范围统一在0-1之间；
(6)对步骤(5)提取的归一化特征参数，利用主成分分析，结合分段逐步回归法进行特征降维和优化，以相关性和独立性为原则，在显著性水平α＝0.005时，当变量进入时模型的F>4.35则保留，变量回判时模型的F<2.95则剔除，同时保证R 2>0.9，以获取相关性最大、多重共线性最小、相对检测误差最小为优化原则，进行特征寻优，获取最优化的特征变量作为植株水分胁迫、氮素胁迫诊断的特征变量；
(7)利用SVR支持向量机回归法进行特征层融合，建立基于micro-CT系统获取的气孔、海绵体、栅栏组织、纤毛、维管束、根系体积、主根和根毛密度特征变量，以及基于偏振-高光谱图像系统获取的冠幅、株高、叶倾角、560nm和1450nm高光谱水分敏感波长下的叶脉分布、平均灰度、叶缘阴影面积；0°、45°、90°、135°、180°特征 偏振角下，560nm和1450nm特征图像的偏振态、stock向量、穆勒矩阵变量，以及三维激光扫描的植株体积、叶面积、茎粗特征变量多特征融合氮素和水分胁迫精确定量检测模型，并将多特征融合氮素和水分胁迫精确定量检测模型输入PLC控制系统；
(8)利用步骤(3)获取的作物生长环境的温度、湿度、光照和基质含水率信息，统计分析计算自定植以来的温度累计、光照累计值，结合基质含水率和环境温度、湿度、光照信息，计算植株的蒸腾量，在此基础上，基于采集的作物micro-CT图像、偏振高光谱图像特征、三维激光扫描成像的最优化的特征变量，进行作物营养和水分胁迫的跟踪连续检测，得出植株的施肥灌溉量模型，并将该模型输入PLC控制系统；
(9)在主管路恒压的条件下，建立施肥流量与施肥量、施肥泵转速与施肥流量之间的关系模型，并将施肥流量与施肥量、施肥泵转速与施肥流量之间的模型输入PLC控制系统；
(10)根据步骤(3)来获取表征植株的生长状态的micro-CT图像、偏振高光谱图像特征、三维激光扫描成像特征参数，PLC控制系统根据多特征融合氮素和水分胁迫精确定量检测模型判定植株当前的水分胁迫、氮素胁迫状态和程度；再基于施肥灌溉量模型，结合施肥泵转速与施肥流量、以及施肥泵转速与施肥流量之间的关系模型，利用PLC控制变频器输出频率，进而控制施肥泵的转速，实现施肥配比和灌溉量的精确节控制。
进一步地，所述生长环境信息包括温室作物生长环境的温度、湿度、光照强度和栽培基质的含水率信息，采用温度变送器、湿度变送器、光照度变送器、含水率变送器来获取。
进一步地，所述微尺度的micro-CT检测步骤为：
①将植株样本，放入Micro-CT扫描系统的样本仓的旋转样本托架上，启动Micro-CT扫描系统，进行扫描，获得样本的CT剖面；
②利用IPL软件，进行样本断层图片的边界和轮廓选取；
③选取不同断层切面进行图像分析，根据断层图片中目标灰阶的不同，调节高、低阈值，选定目标阈值范围，并二值化样本目标的断层图像；
④利用IPL软件结合目标图像分析，获取植株叶片和茎秆的气孔密度、海绵体厚度、栅栏组织、纤毛密度、维管束的剖面结构和断层图像灰度等特征参数；
⑤基于选择的边界和阈值，剥离珍珠岩基质，生成根系三维图像，执行IPL语言导出根系体积和主根、根毛密度及分布参数。
进一步地，叶片尺度的偏振-高光谱扫描成像检测步骤：
①将样本置于偏振-高光谱成像系统的双坐标样本台上，设置可见光-近红外光源系统37的波长范围为300-2200nm，设置光强范围为6500lux，调整成像系统的几何中心与位移台水平和垂直轴线XZ轴几何中心一致，进行偏振-高光谱图像的扫描成像；
②利用具有前置偏振滤光片组的两台高光谱成像系统，设置偏振片的采样偏振角分别为0°、45°、90°、135°、180°；高光谱前置滤光片透过波长为560nm、1450nm，分别在水平方向和垂直方向上，进行推帚式的偏振-高光谱扫描成像，获取主视和俯视方向的偏振-高光谱特征图像；
③通过坐标匹配和主视/俯视特征图像融合，提取营养和水分胁迫样本的主视和俯视视场下的高光谱特征图像，并提取植株冠幅、株高、叶倾角图像；
④基于前置560nm和1450nm滤光片，提取特征波长下的冠层高光谱特征图像，提取叶面的560nm和1450nm高光谱营养水分敏感波长下的叶脉分布，平均灰度、叶缘阴影面积等特征参数；
⑤基于获取的0°、45°、90°、135°、180°特征偏振角的560nm、1450nm的偏振高光谱图像，提取氮素和水分胁迫植株样本的偏振态、斯托克向量、穆勒矩阵变量。
进一步地，冠层尺度的三维激光扫描成像检测步骤：
①首先在所要扫描的作物叶片和栽植器皿上方粘贴直径为6mm的黑色轮廓高反射目标点，由于叶片表面弯曲，在贴反射目标点时，两目标点之间的最短距离控制在20mm；
②运行扫描系统，用三维激光扫描仪测量校准板，以纠正传感器参数，确保数据采集精度；
③最后通过手持扫描的方式，依次获取所有作物样本的三维数据。
进一步地，在三维激光扫描成像检测时，激光功率为65％，快门时间为7.2ms，分辨率为0.50mm。
基于多尺度生境信息的苗期作物水肥检测和控制装置，其特征在于，包括施肥机构、信息采集系统和控制系统组成。
所述施肥机构主要由进水主管道、过滤器、进水泵、进水电磁阀、出肥管道、施肥电磁阀、搅拌电机、混肥灌、施肥泵、出肥管组成；其中进水主管道连接水源，进水主管道的另一端连接过滤器，过滤器的出水口与进水泵相连接，用于提供水肥供给系统的基础水源，进水泵的出口管路和混肥管路之间连接有进水电磁阀，以控制水源管路的通断；施肥泵通过管路与混肥罐上部相连、两者之间有施肥电磁阀，用于实现施肥管路的通断控制；施肥罐顶部安装有搅拌电机，搅拌电机输出轴末端有搅拌叶片，通过电机驱 动，实现对固体肥料颗粒的混合及均匀搅拌作业；
所述信息采集系统包括环境温度变送器、环境湿度变送器、环境光照变送器、基质含水率变送器、EC传感器、两个pH传感器、液位传感器、压力变送器、作物信息检测系统，以及A/D转换模拟量输入模块；其中环境温度变送器、环境湿度变送器、环境光照变送器安装在温室内部、基质含水率变送器安装在作物盆栽的基质中，环境温度变送器、环境湿度变送器、环境光照变送器、基质含水率变送器分别与模拟量模块相连；EC传感器连接在施肥机管路中，用于检测营养液的浓度，EC传感器的输出端与模拟量输入模块的输入端相连；两个pH传感器分别安装在距离混肥罐液面10cm处和下部距离混肥罐底部20cm处，并分别与模拟量模块的输入端相连，通过两个pH传感器的差值比较，判断肥料搅拌的均匀程度，判定搅拌电机所需的转速，并自启动和停止搅拌电机和搅拌装置；液位传感器放置在混肥罐的底部，液位传感器的信号输出端也与模拟量模块的输入端相连，基于水下压力不同的原理来进行液位判断；压力变送器连接在主管路中部，以判断主管道压力，其输出端与模拟量模块的输入端相连；所述作物信息检测系统包括micro-CT扫描系统、偏振-高光谱成像系统和三维激光扫描系统；
所述控制系统包括PLC控制器、触摸屏、变频器和执行机构，其中触摸屏与PLC控制器相连接，用于进行人机交互，输入控制模式和作物生长状态信息；其中变频器的输入端与PLC控制器的输出端相连接，变频器的输出端与施肥泵相连接，模拟量模块与PLC控制器相连接，用于实现PLC控制器对信息采集系统的多路信息采集控制；
系统根据作物的水肥需求结合环境信息，可以通过PLC控制器实现对变频器的频率控制，进而实现对施肥泵转速的控制，以调节施肥管道的流量，在恒压主管路中，最终实现对主管道施肥灌溉量的精确控制。
进一步地，所述micro-CT扫描系统中，为旋转样本托架，通过底座上的螺钉固定在检测样本仓的底部，旋转样本托架底座几何中心安装有旋转轴，旋转轴末端安装固定有圆形样本托架，检测时旋转轴带动样本托架360°旋转，同时固定在发射仓中部的X射线发射器通过俯仰动作实现对样本的断层切片扫描过程。
进一步地，其中所述偏振-高光谱图像检测系统，包括控制系统、双坐标样本台、图像采集系统、光源系统；
其中所述图像采集系统包括两个偏振-高光谱成像系统、图像采集器、立臂和悬臂；所述立臂由第一底座、带丝杠的立杆和第一滑块组成，第一底座通过螺钉固定在光箱的底部的左侧，第一底座上部通过铰链与立杆连接，立杆能够以铰链为中心做左右摆动， 完成成像设备的空间位姿调整；立杆上安装有第一滑块；第一偏振-高光谱成像系统安装在第一滑块上，第一滑块能够由丝杠驱动沿立杆上下移动，带动第一偏振-高光谱成像系统寻找最佳检测位，实现主视方向的偏振高光谱图像信息的采集；
所述悬臂由第二底座、带丝杠的横杆和第二滑块组成，第二底座通过螺钉固定在光箱的右侧板的上部，第二底座通过铰链与横杆连接，横杆能够以铰链为中心做上下摆动，完成成像设备的空间位姿调整；横杆上安装有第二滑块，第二偏振-高光谱成像系统安装在第二滑块上，第二滑块能够由丝杠驱动沿横杆沿水平方向左右移动，带动第二偏振-高光谱成像系统寻找最佳检测位，实现俯视方向的偏振高光谱图像信息的采集；
其中所述光源系统由可见光-近红外光源、云台组成，在立杆的底端和顶端、立杆的右端和左端分别安装一个云台，每个云台上分别安装可见光-近红外光源，可见光-近红外光源能够通过云台进行仰俯角设置，实现对植株的清晰匀光成像；
所述双坐标样本台固定在光箱的底平面的几何中心位置，垂直丝杠顶端安装有样本托架，用于安放待测样本，通过水平丝杠和垂直丝杠的运动，可以带动样本托架实现水平和垂直方向的匀速位移，可以配合图像采集控制系统实现推帚式的第一偏振-高光谱成像系统和第二偏振-高光谱成像系统的扫描成像；
其中所述偏振-高光谱成像系统由前到后分别由前置偏振片、偏振片驱动装置、前置滤光片、滤光片切换装置、摄谱仪和成像系统组成，偏振片在整个系统的最前端，由偏振驱动装置驱动360°旋转，可实现对任意偏振角的设定，摄谱仪和成像系统可实现偏振角的设定和步序偏振信息的采集；偏振片后为560nm和1450nm窄带滤光片，滤光片采用转轮切换的方式，配合摄谱仪和成像系统实现对作物样本主视和俯视高光谱营养和水分胁迫特征图像的采集；
所述控制系统包括控制计算机、光源控制器、图像采集器和运动控制器；
其中光源控制器连接可见光-近红外光源，实现不同光强和光质的光源控制；
图像采集器连接两个偏振-高光谱成像系统和控制计算机，由控制计算机发出指令，实现对主视和俯视偏振-高光谱成像系统的成像信息采集和分析；
运动控制器连接双坐标样本台、立臂、悬臂和云台；同时，运动控制器与控制计算机相连，由控制计算机发出指令，实现对双坐标样本台的升降和水平位移控制，对立臂、悬臂的滑块驱动控制，以及云台的仰俯角控制。
本发明是针对目前温室作物苗期营养液动态配比的精确施肥灌溉，其技术方案涉及：①营养液精确调配技术：通过控制母液的流入量，实现营养液浓度的精确调配。建立在 不同营养液配比下营养液浓度与变频控制参数的关系模型，给出营养液浓度和灌施量的调控方法。②灌溉量和施肥量决策：通过对不同光照强度、光积累、温湿度等温室内环境下作物生长状态测试研究，结合水肥对作物生长信息的影响；制定出在温室条件下的灌溉量和施肥量的决策优化方案。③针对设施生产在配肥和水肥灌施过程的营养液和杂质沉积导致的管路堵塞和配比误差等问题，设计和加装母液混肥搅拌和管路定时冲洗等配套装置，并通过研究不同营养液配比条件下肥料固体颗粒沉积的规律，确定搅拌时机，降低固体颗粒的沉积，提高营养液的利用效率。并在此基础上开发了适合日光温室、钢架大棚适用的低成本肥水一体化调控装备。
与现有技术相比，本发明具有以下有益效果。
1、本发明将作物信息探测技术、温室环境信息探测技术、液肥浓度精确调配技术与灌溉施肥量控制决策控制技术相融合，提出能依据温室温度、湿度、光照和基质含水率等主要环境信息，结合多信息融合的作物养分信息，精确控制灌溉时机、水肥灌溉量和配比的作物施肥灌溉控制策略。提出了一种基于多尺度生境信息的苗期作物水肥检测和控制方法，解决按照作物需求进行水肥动态控制的难题，克服了目前只依据环境因子信息进行施肥控制的局限性，由于实现了按需施肥，故而大幅降低了肥料的用量，降低了人力成本，提高了经济效益。
2、本发明的施肥装置利用变频器控制施肥泵转速来调节液肥管路流量的大小，并将液肥送到恒压主管道中，实现了液肥浓度配比和灌溉量的实时动态精确控制。采用基于作物生境信息反馈的变频技术，实现了温室水肥的自动配比和变量灌施作业，目前温室蔬菜生产中尚未见到基于作物生境信息反馈的恒压变频水肥配比控制技术。
3、固体肥搅拌液肥灌施模式下，液肥沉积和管路赌塞的主要原因是混合灌中搅拌后的液肥不均匀和粘性大，长时间作业和存放必然会导致较多的沉淀、管道壁堆积，造成系统的作业故障和寿命降低，通过增设基于上下废液不均匀性反馈的混肥搅拌装置以提高液肥搅拌的均匀性，大幅降低了固体颗粒的沉积率，结合管道过滤装置，不仅降低了人工搅拌的劳动强度，同时大幅减少设备故障，提高了施肥机的作业效率，降低了成本，提高了经济效益。
4、本发明采用micro-CT扫描作物营养和水分胁迫下气孔、海绵体、栅栏组织、纤毛、维管束、根系体积、主根和根毛密度等特征变量的微观形态特征；利用双位偏振-高光谱成像系统，获取水分胁迫的冠幅、株高、叶倾角等宏观形态特征；以及560nm、1450nm高光谱图像的营养和水分敏感波长下的叶脉分布、平均灰度、叶缘阴影面积和0°、45°、 90°、135°、180°特征偏振角下560nm、1450nm特征图像的偏振态、stock向量、穆勒矩阵变量、三维激光扫描系统获取的植株体积、叶面积、茎粗、株高等宏观长势形态特征，通过对植株内外结构、地上地下、宏观微观形态特征与水分胁迫下特征波长图像、偏振态的相互融合，优势互补，实现对作物苗期营养和水分胁迫特征的全面精确提取和精确定量分析，为设施水肥一体化的科学管理提供了科学依据。
附图说明
图1是本发明所述苗期作物水肥检测和控制装置的结构示意图。
图2是本发明所采用的micro-CT装置的结构示意图。
图3是本发明所采用的偏振-高光谱成像系统的结构示意图。
图4是本发明所采用的三维激光扫描系统结构示意图。
图5是本发明所述基于多尺度生境信息的苗期作物水肥检测和控制方法流程图。
图6生菜三维空间网格模型。
图7生菜茎粗坐标。
图中：
1.进水主管道；2.过滤器；3.进水泵；4.进水电磁阀；5.出肥管道；6.施肥电磁阀；7.EC传感器；8.压力数字显示仪表；9.变频器；10.控制柜；11.接触器；12.环境温度变送器；13.环境湿度变送器；14.环境光照变送器；15.基质含水率变送器；16.模拟量模块；17.电源空气开关；18.PLC控制器；19.pH传感器；20.液位传感器；21.触摸屏；22.搅拌电机；23.混肥灌；24.施肥泵；25.压力变送器；26.施肥管道；27.旋转样本托架；29.X射线发射器；30.计算机；32.双坐标样本台；33.立臂，33-1第一底座；33-2立杆；33-3第一滑块；34.悬臂；34-1第二底座；34-2悬杆；34-3第二滑块；35.偏振-高光谱成像系统，35-1第一偏振-高光谱成像仪；35-2.第二偏振-高光谱成像仪；36.云台；37.可见光-近红外光源；38.运动控制器；39.图像采集器；40.光源控制器；41.控制计算机；42.光箱；43.PC机；44.FireWire适配器；45.FireWire电缆；46.手持三维扫描头；47.电源模块。
具体实施方式
下面结合附图以及具体实施例对本发明作进一步的说明，但本发明的保护范围并不限于此。
图1所示为本发明所述的基于多尺度生境信息的苗期作物水肥检测和控制装置，该装置包括施肥机构、信息采集系统和控制系统组成。所述施肥机构主要由进水主管道1、过滤器2、进水泵3、进水电磁阀4、出肥管道5、施肥电磁阀6、搅拌电机22、混肥灌 23、施肥泵24、出肥管26组成；其中进水主管道1连接水源，进水主管道1的另一端连接过滤器2，过滤器2的出水口与进水泵3相连接，用于提供水肥供给系统的基础水源，进水泵3的出口管路和混肥管路之间连接有进水电磁阀4，以控制水源管路的通断；施肥泵24通过管路与混肥罐23上部相连、两者之间有施肥电磁阀6，用于实现施肥管路的通断控制；施肥罐23顶部安装有搅拌电机22，搅拌电机22输出轴末端有搅拌叶片，通过电机驱动，可以实现对固体肥料颗粒的混合及均匀搅拌作业。
所述信息采集系统包括环境温度变送器12、环境湿度变送器13、环境光照变送器14、基质含水率变送器15、EC传感器7、两个pH传感器19-1和19-2、液位传感器21、压力变送器25、作物信息检测系统，以及9通道的A/D转换模拟量输入模块16。其中环境温度变送器12、环境湿度变送器13、环境光照变送器14安装在温室内部，基质含水率变送器15安装在作物珍珠岩盆栽的基质中；环境温度变送器12、环境湿度变送器13、环境光照变送器14、基质含水率变送器15分别与模拟量模块16相连。
EC传感器7连接在施肥机管路中，用于检测营养液的浓度，EC传感器7的输出端与模拟量输入模块16的输入端相连，两个pH传感器19-1和19-2分别安装在距离混肥罐23液面10cm处和下部距离混肥罐底部20cm处，混肥罐23内上下布置的两个pH传感器19-1和19-2分别与模拟量模块16的输入端相连；通过上下两个pH传感器19-1和19-2的差值比较，判断肥料搅拌的均匀程度，判定搅拌电机所需的转速，并自启动和停止搅拌电机和搅拌装置。
液位传感器20放置在混肥罐23的底部，基于水下压力不同的原理来进行液位判断，其4-20mA信号输出端与模拟量模块16的输入端相连，压力变送器25连接在主管路中部，以判断主管道压力，压力变送器25的输出端与模拟量模块16的输入端相连。
作物信息检测系统包括micro-CT扫描系统、偏振-高光谱成像系统和三维激光扫描系统，用于检测表征植株受氮素和水肥胁迫程度的微尺度、叶片尺度和冠层尺度的作物信息。
如图2所示，所述micro-CT扫描系统中，27为旋转样本托架，通过底座上的螺钉固定在检测样本仓的底部，旋转样本托架底座几何中心安装有旋转轴，旋转轴末端安装固定有圆形样本托架，检测时旋转轴带动样本托架360°旋转，同时固定在发射仓中部的X射线发射器29通过俯仰动作实现对样本的断层切片扫描过程。
偏振-高光谱图像检测系统如图3所示，包括控制系统、双坐标样本台32、图像采集系统、光源系统；其中所述图像采集系统包括两个偏振-高光谱成像系统35、图像采集器 38、立臂33和悬臂34；所述立臂33由第一底座33-1、带丝杠的立杆33-2和第一滑块33-3组成，第一底座33-1通过螺钉固定在光箱42的底部的左侧，第一底座33-1上部通过铰链与立杆33-2连接，立杆33-2能够以铰链为中心做左右摆动，完成成像设备的空间位姿调整；立杆33-2上安装有第一滑块33-3；第一偏振-高光谱成像系统35-1安装在第一滑块33-3上，第一滑块33-3能够由丝杠驱动沿立杆33-2上下移动，带动第一偏振-高光谱成像系统35-1寻找最佳检测位，实现主视方向的偏振高光谱图像信息的采集。
所述悬臂34由第二底座34-1、带丝杠的横杆34-2和第二滑块34-3组成，第二底座34-1通过螺钉固定在光箱42的右侧板的上部，第二底座34-1通过铰链与横杆34-2连接，横杆34-2能够以铰链为中心做上下摆动，完成成像设备的空间位姿调整；横杆34-2上安装有第二滑块34-3，第二偏振-高光谱成像系统35-2安装在第二滑块34-3上，第二滑块34-3能够由丝杠驱动沿横杆34-2沿水平方向左右移动，带动第二偏振-高光谱成像系统35-2寻找最佳检测位，实现俯视方向的偏振高光谱图像信息的采集。
其中所述光源系统由可见光-近红外光源37、云台36组成，在立杆33-2的底端和顶端、立杆34-2的右端和左端分别安装一个云台36，每个云台上分别安装可见光-近红外光源37，可见光-近红外光源37能够通过云台36进行仰俯角设置，实现对植株的清晰匀光成像。
所述双坐标样本台32固定在光箱42的底平面的几何中心位置，其中32-1为水平丝杠，32-2位垂直丝杠，垂直丝杠32-2顶端安装有样本托架，用于安放待测样本，通过水平丝杠32-1和垂直丝杠32-2的运动，可以带动样本托架实现水平和垂直方向的匀速位移，可以配合图像采集控制系统实现推帚式的偏振-高光谱成像系统35-1和偏振-高光谱成像系统35-2的扫描成像。
其中所述偏振-高光谱成像系统35由前到后分别由前置偏振片、偏振片驱动装置、前置滤光片、滤光片切换装置、摄谱仪和成像系统组成，偏振片在整个系统的最前端，由偏振驱动装置驱动360°旋转，可实现对任意偏振角的设定，摄谱仪和成像系统可实现偏振角的设定和步序偏振信息的采集；偏振片后为560nm和1450nm窄带滤光片，滤光片采用转轮切换的方式，配合摄谱仪和成像系统实现对作物样本主视和俯视高光谱营养和水分胁迫特征图像的采集；所述控制系统包括控制计算机41、光源控制器40、图像采集器39和运动控制器38；其中光源控制器40连接可见光-近红外光源37，实现不同光强和光质的光源控制；图像采集器39连接两个偏振-高光谱成像系统35和控制计算机41，由控制计算机41发出指令，实现对主视和俯视偏振-高光谱成像系统35的成像信息 采集和分析；运动控制器38连接双坐标样本台32、立臂33、悬臂34和云台36；同时，运动控制器38与控制计算机41相连，由控制计算机41发出指令，实现对双坐标样本台32的升降和水平位移控制，对立臂33、悬臂34的滑块驱动控制，以及云台36的仰俯角控制。
三维扫描成像系统如图4所示，由PC机43、FireWire适配器44、FireWire电缆45、手持三维扫描头46和电源模块47组成，其中手持三维扫描头46通过FireWire电缆45连接FireWire适配器44，并通过FireWire适配器44与PC机相连，通过PC机软件实现对手持三维扫描头46的三维扫描控制和信息采集，电源模块47通过与FireWire适配器44相连为PC机提供电源；电源模块47与手持三维扫描头46相连，为手持三维扫描头46提供电源。
所述控制系统包括PLC控制器18、触摸屏21、变频器9和执行机构，其中触摸屏21与PLC控制器18的422通讯端口相连接，用于进行人机交互，输入控制模式和作物生长状态信息。其中变频器9的输入端与PLC控制器18的输出端相连接，变频器9的输出端与施肥泵24相连接，模拟量模块16与PLC控制器18相连接，用于实现PLC控制器18对信息采集系统的多路信息采集。
PLC控制器18中内置有多特征融合氮素和水分胁迫精确定量检测模型、施肥灌溉量模型施肥泵转速与施肥流量、以及施肥泵转速与施肥流量之间的关系模型，PLC控制器18根据多特征融合氮素和水分胁迫精确定量检测模型判定植株当前的水分胁迫、氮素胁迫状态和程度；再基于施肥灌溉量模型，结合施肥泵转速与施肥流量、以及施肥泵转速与施肥流量之间的关系模型，利用PLC控制变频器输出频率，进而控制施肥泵的转速，实现施肥配比和灌溉量的精确节控制。
如图5所示，本发明所述的基于多尺度生境信息的苗期作物水肥检测和控制方法，主要包括如下步骤：
步骤1，采取标准营养液配方，利用珍珠岩作为培养基质，采用无土栽培定植温室蔬菜作物，采用温室标准化管理方式进行管理，保证苗期作物营养元素和水分的正常供给。
步骤2：选取定植一周后的植株分别进行水分胁迫、氮素胁迫样本培育；
(1)定植一周后，在营养元素不变的条件下，将水分供应量按照标准供应量的100％、80％、60％、40％和20％分成5个不同的水平，每个水平20个样本，共计100个样本，进行水分胁迫样本培育；
(2)定植一周后，在保持充足水分供应的条件下，将氮素供应量按照标准供应量的200％、100％、50％和25％分成4个不同的水平，每个水平20个样本，共计80个样本，进行氮素胁迫样本培育.
步骤3，苗期样本水分胁迫3天后，营养胁迫分别在胁迫后的7天、15天和21天进行样本微尺度的micro-CT、叶片尺度的偏振-高光谱成像扫描和冠层尺度的植株冠层三维扫描成像的多信息采集，同时获取温室作物生长环境的温度、湿度、光照强度和栽培基质的含水率信息。上述信息的获取按照以下步骤进行：
1)作物生长的环境信息采集步骤：采用温度变送器、湿度变送器、光照度变送器、含水率变送器获取作物生长环境的温度、湿度、光照以及基质含水率信息。
2)微尺度的micro-CT检测步骤：
①将不同营养和水分胁迫水平的样本，依次放入Micro-CT扫描系统的样本仓的旋转样本托架27上，通过控制计算机30启动Micro-CT扫描系统，顺序进行扫描，获得各个样本的CT剖面；
②利用IPL软件，进行样本断层图片的边界和轮廓选取；
③选取不同断层切面进行图像分析，根据断层图片中目标灰阶的不同，调节高、低阈值，选定目标阈值范围，并二值化样本目标的断层图像；
④利用IPL软件结合目标图像分析，获取植株叶片和茎秆的气孔密度、海绵体厚度、栅栏组织、纤毛密度、维管束的剖面结构和断层图像灰度等特征参数；
⑤基于选择的边界和阈值，剥离珍珠岩基质，生成根系三维图像，执行IPL语言导出根系体积和主根、根毛密度及分布参数。
3)叶片尺度的偏振-高光谱扫描成像检测步骤：
获取micro-CT的扫描图像并完成特征提取后，依次取出样本进行偏振-高光谱图像的扫描成像，并按照下述步骤进行：
①将样本置于偏振-高光谱成像系统的双坐标样本台32上，设置可见光-近红外光源系统37的波长范围为300-2200nm，设置光强范围为6500lux，调整成像系统的几何中心与位移台水平和垂直轴线XZ轴几何中心一致；
②利用具有前置偏振滤光片组的两台高光谱成像系统35-1和35-2，设置偏振片的采样偏振角分别为：0°、45°、90°、135°、180°；高光谱前置滤光片透过波长为560nm、1450nm，分别在水平方向和垂直方向上，进行推帚式的偏振-高光谱扫描成像，获取主视和俯视方向的偏振-高光谱特征图像；
③通过坐标匹配和主视/俯视特征图像融合，提取营养和水分胁迫样本的主视和俯视视场下的高光谱特征图像，并提取植株冠幅、株高、叶倾角图像；
④基于前置560nm和1450nm滤光片，提取特征波长下的冠层高光谱特征图像，提取叶面的560nm和1450nm高光谱营养水分敏感波长下的叶脉分布，平均灰度、叶缘阴影面积等特征参数；
⑤基于获取的0°、45°、90°、135°、180°特征偏振角的560nm、1450nm的偏振高光谱图像，提取氮素和水分胁迫植株样本的偏振态、斯托克向量、穆勒矩阵变量。
4)冠层尺度的三维激光扫描成像检测步骤：
获取micro-CT的扫描图像并完成特征提取后，依次取出样本进行偏振-高光谱图像的扫描成像：在三维激光扫描数据采集前，需预先确定扫描仪传感器的激光功率、快门时间和采集软件的分辨率以保证三维模型的清晰。经过分析比较，设定激光功率为65％，快门时间为7.2ms，分辨率为0.50mm，设定参数下所采集的植株的三维形态。
①首先在所要扫描的作物叶片和栽植器皿上方粘贴直径为6mm的黑色轮廓高反射目标点，由于叶片表面弯曲，在贴反射目标点时，两目标点之间的最短距离控制在20mm；
②运行扫描系统，用三维激光扫描仪测量校准板，以纠正传感器参数，确保数据采集精度；
③最后通过手持扫描的方式，依次获取所有作物样本的三维数据。
步骤4、进行常规理化检测：称量样本的干湿重，确定植株的含水率真值；使用英国SEAL公司生产的AutoAnalyzer3型连续流动分析仪，通过凯氏定氮营养检测测定样品全氮含量；采用扫描电镜和超景深3D显微成像技术，获取植株的气孔和纤毛密度，以及海绵体和栅栏组织厚度，维管束分布密度和管径等实测值；
步骤5、对步骤3提取的micro-CT的特征变量、偏振-高光谱图像特征变量和三维扫描成像特征变量进行归一化处理，使其特征值范围统一在0-1之间；
步骤6、对步骤5提取的归一化特征参数，利用主成分分析，结合分段逐步回归法进行特征降维和优化，以相关性和独立性为原则，在显著性水平α＝0.005时，当变量进入时模型的F>4.35则保留，变量回判时模型的F<2.95则剔除，同时保证R 2>0.9，以获取相关性最大，多重共线性最小，相对检测误差最小为优化原则，进行特征寻优，获取最优化的特征变量作为植株水分胁迫、氮素胁迫诊断的特征变量；
步骤7、利用SVR支持向量机回归法进行特征层融合，建立基于micro-CT系统获取的气孔、海绵体、栅栏组织、纤毛、维管束、根系体积、主根和根毛密度等特征变量， 以及基于偏振-高光谱图像系统获取的冠幅、株高、叶倾角、560nm和1450nm高光谱水分敏感波长下的叶脉分布、平均灰度、叶缘阴影面积；0°、45°、90°、135°、180°特征偏振角下，560nm和1450nm特征图像的偏振态、stock向量、穆勒矩阵变量，以及三维激光扫描的植株体积、叶面积、茎粗等特征变量多特征融合氮素和水分胁迫精确定量检测模型，并将多特征融合氮素和水分胁迫精确定量检测模型输入PLC控制系统；
步骤8、利用步骤3获取的作物生长环境的温度、湿度、光照和基质含水率信息，统计分析计算自定植以来的温度累计、光照累计值，结合基质含水率和环境温度、湿度、光照信息，计算植株的蒸腾量，在此基础上，基于采集的作物micro-CT图像、偏振高光谱图像特征、三维激光扫描成像特征，进行作物营养和水分胁迫的跟踪连续检测，得出植株的施肥灌溉量模型，并将该模型输入PLC控制系统；
步骤9、在主管路恒压的条件下，建立施肥流量与施肥量、施肥泵转速与施肥流量之间的关系模型，并将施肥流量与施肥量、施肥泵转速与施肥流量之间的模型输入PLC控制系统；
步骤10、根据步骤(3)来获取表征植株的生长状态的micro-CT图像、偏振高光谱图像特征、三维激光扫描成像特征参数，PLC控制系统根据多特征融合氮素和水分胁迫精确定量检测模型判定植株当前的水分胁迫、氮素胁迫状态和程度；再基于施肥灌溉量模型，结合施肥泵转速与施肥流量、以及施肥泵转速与施肥流量之间的关系模型，利用PLC控制变频器输出频率，进而控制施肥泵的转速，实现施肥配比和灌溉量的精确节控制。
所述实施例为本发明的优选的实施方式，但本发明并不限于上述实施方式，在不背离本发明的实质内容的情况下，本领域技术人员能够做出的任何显而易见的改进、替换或变型均属于本发明的保护范围。

Claims (9)

  1. 一种基于多尺度生境信息的苗期作物水肥检测和控制方法，其特征在于，包括如下步骤：
    (1)采取标准营养液配方，利用珍珠岩作为培养基质，采用无土栽培定植温室蔬菜作物，采用温室标准化管理方式进行管理，保证苗期作物营养元素和水分的正常供给；
    (2)选取定植一周后的植株分别进行水分胁迫、氮素胁迫样本培育；
    (3)苗期样本水分胁迫、营养胁迫分别在胁迫后通过微尺度的micro-CT、叶片尺度的偏振-高光谱成像扫描、冠层尺度的植株冠层三维扫描成像进行植株生长状态的连续检测与信息采集，同时获取植株生长环境信息；
    (4)进行常规理化检测：称量样本的干湿重，确定植株的含水率真值；用凯氏定氮法测定样品全氮含量；采用扫描电镜和超景深3D显微成像技术，获取植株的气孔和纤毛密度，以及海绵体和栅栏组织厚度，纤维管束分布密度和管径实测值；
    (5)对步骤(3)提取的micro-CT的特征变量、偏振-高光谱图像特征变量和三维扫描成像特征变量进行归一化处理，使其特征值范围统一在0-1之间；
    (6)对步骤(5)提取的归一化特征参数，利用主成分分析，结合分段逐步回归法进行特征降维和优化，以相关性和独立性为原则，在显著性水平α＝0.005时，当变量进入时模型的F>4.35则保留，变量回判时模型的F<2.95则剔除，同时保证R 2>0.9，以获取相关性最大、多重共线性最小、相对检测误差最小为优化原则，进行特征寻优，获取最优化的特征变量作为植株水分胁迫、氮素胁迫诊断的特征变量；
    (7)利用SVR支持向量机回归法进行特征层融合，建立基于micro-CT系统获取的气孔、海绵体、栅栏组织、纤毛、维管束、根系体积、主根和根毛密度特征变量，以及基于偏振-高光谱图像系统获取的冠幅、株高、叶倾角、560nm和1450nm高光谱水分敏感波长下的叶脉分布、平均灰度、叶缘阴影面积；0°、45°、90°、135°、180°特征偏振角下，560nm和1450nm特征图像的偏振态、stock向量、穆勒矩阵变量，以及三维激光扫描的植株体积、叶面积、茎粗特征变量多特征融合氮素和水分胁迫精确定量检测模型，并将多特征融合氮素和水分胁迫精确定量检测模型输入PLC控制系统；
    (8)利用步骤(3)获取的作物生长环境的温度、湿度、光照和基质含水率信息，统计分析计算自定植以来的温度累计、光照累计值，结合基质含水率和环境温度、湿度、光照信息，计算植株的蒸腾量，在此基础上，基于采集的作物micro-CT图像、偏振高光谱图像特征、三维激光扫描成像的最优化的特征变量，进行作物营养和水分胁迫的跟踪连续 检测，得出植株的施肥灌溉量模型，并将植株的施肥灌溉量模型输入PLC控制系统；
    (9)在主管路恒压的条件下，建立施肥流量与施肥量、施肥泵转速与施肥流量之间的关系模型，并将施肥流量与施肥量、施肥泵转速与施肥流量之间的模型输入PLC控制系统；
    (10)根据步骤(3)来获取表征植株的生长状态的micro-CT图像、偏振高光谱图像特征、三维激光扫描成像特征参数，PLC控制系统根据多特征融合氮素和水分胁迫精确定量检测模型判定植株当前的水分胁迫、氮素胁迫状态和程度；再基于施肥灌溉量模型，结合施肥泵转速与施肥流量、以及施肥泵转速与施肥流量之间的关系模型，利用PLC控制变频器输出频率，进而控制施肥泵的转速，实现施肥配比和灌溉量的精确节控制。
  2. 根据权利要求1所述的基于多尺度生境信息的苗期作物水肥检测和控制方法，其特征在于，所述生长环境信息包括温室作物生长环境的温度、湿度、光照强度和栽培基质的含水率信息，采用温度变送器、湿度变送器、光照度变送器、含水率变送器来获取。
  3. 根据权利要求1所述的基于多尺度生境信息的苗期作物水肥检测和控制方法，其特征在于，所述微尺度的micro-CT检测步骤为：
    ①将植株样本(28)，放入Micro-CT扫描系统的样本仓的旋转样本托架(27)上，启动Micro-CT扫描系统，进行扫描，获得样本的CT剖面；
    ②利用IPL软件，进行样本断层图片的边界和轮廓选取；
    ③选取不同断层切面进行图像分析，根据断层图片中目标灰阶的不同，调节高、低阈值，选定目标阈值范围，并二值化样本目标的断层图像；
    ④利用IPL软件结合目标图像分析，获取植株叶片和茎秆的气孔密度、海绵体厚度、栅栏组织、纤毛密度、维管束的剖面结构和断层图像灰度等特征参数；
    ⑤基于选择的边界和阈值，剥离珍珠岩基质，生成根系三维图像，执行IPL语言导出根系体积和主根、根毛密度及分布参数。
  4. 根据权利要求1所述的基于多尺度生境信息的苗期作物水肥检测和控制方法，其特征在于，叶片尺度的偏振-高光谱扫描成像检测步骤：
    ①将样本(28)置于偏振-高光谱成像系统的双坐标样本台(32)上，设置可见光-近红外光源系统(37)的波长范围为300-2200nm，设置光强范围为6500lux，调整成像系统的几何中心与位移台水平和垂直轴线XZ轴几何中心一致，进行偏振-高光谱图像的扫描成像；
    ②利用具有前置偏振滤光片组的两台高光谱成像系统，设置偏振片的采样偏振角分别为0°、45°、90°、135°、180°；高光谱前置滤光片透过波长为560nm、1450nm， 分别在水平方向和垂直方向上，进行推帚式的偏振-高光谱扫描成像，获取主视和俯视方向的偏振-高光谱特征图像；
    ③通过坐标匹配和主视/俯视特征图像融合，提取营养和水分胁迫样本的主视和俯视视场下的高光谱特征图像，并提取植株冠幅、株高、叶倾角图像；
    ④基于前置560nm和1450nm滤光片，提取特征波长下的冠层高光谱特征图像，提取叶面的560nm和1450nm高光谱营养水分敏感波长下的叶脉分布，平均灰度、叶缘阴影面积等特征参数；
    ⑤基于获取的0°、45°、90°、135°、180°特征偏振角的560nm、1450nm的偏振高光谱图像，提取氮素和水分胁迫植株样本的偏振态、斯托克向量、穆勒矩阵变量。
  5. 根据权利要求1所述的基于多尺度生境信息的苗期作物水肥检测和控制方法，其特征在于，冠层尺度的三维激光扫描成像检测步骤：
    ①首先在所要扫描的作物叶片和栽植器皿上方粘贴直径为6mm的黑色轮廓高反射目标点，由于叶片表面弯曲，在贴反射目标点时，两目标点之间的最短距离控制在20mm；
    ②运行扫描系统，用三维激光扫描仪测量校准板，以纠正传感器参数，确保数据采集精度；
    ③最后通过手持扫描的方式，依次获取所有作物样本的三维数据。
  6. 根据权利要求5所述的基于多尺度生境信息的苗期作物水肥检测和控制方法，其特征在于，在三维激光扫描成像检测时，激光功率为65％，快门时间为7.2ms，分辨率为0.50mm。
  7. 基于多尺度生境信息的苗期作物水肥检测和控制装置，其特征在于，包括施肥机构、信息采集系统和控制系统组成。
    所述施肥机构主要由进水主管道(1)、过滤器(2)、进水泵(3)、进水电磁阀(4)、出肥管道(5)、施肥电磁阀(6)、搅拌电机(22)、混肥灌(23)、施肥泵(24)、出肥管(26)组成；其中进水主管道(1)连接水源，进水主管道(1)的另一端连接过滤器(2)，过滤器(2)的出水口与进水泵(3)相连接，用于提供水肥供给系统的基础水源，进水泵(3)的出口管路和混肥管路之间连接有进水电磁阀(4)，以控制水源管路的通断；施肥泵(24)通过管路与混肥罐(23)上部相连、两者之间有施肥电磁阀(6)，用于实现施肥管路的通断控制；施肥罐(23)顶部安装有搅拌电机(22)，搅拌电机(22)输出轴末端有搅拌叶片，通过电机驱动，实现对固体肥料颗粒的混合及均匀搅拌作业；
    所述信息采集系统包括环境温度变送器(12)、环境湿度变送器(13)、环境光照变送器(14)、基质含水率变送器(15)、EC传感器(7)、两个pH传感器(19-1、19-2)、液 位传感器(21)、压力变送器(25)、作物信息检测系统，以及A/D转换模拟量输入模块(16)；其中环境温度变送器(12)、环境湿度变送器(13)、环境光照变送器(14)安装在温室内部、基质含水率变送器(15)安装在作物盆栽的基质中，环境温度变送器(12)、环境湿度变送器(13)、环境光照变送器(14)、基质含水率变送器(15)分别与模拟量模块(16)相连；EC传感器(7)连接在施肥机管路中，用于检测营养液的浓度，EC传感器(7)的输出端与模拟量输入模块(16)的输入端相连；两个pH传感器(19-1、19-2)分别安装在距离混肥罐(23)液面10cm处和下部距离混肥罐底部20cm处，并分别与模拟量模块(16)的输入端相连，通过两个pH传感器(19-1、19-2)的差值比较，判断肥料搅拌的均匀程度，判定搅拌电机所需的转速，并自启动和停止搅拌电机和搅拌装置；液位传感器(20)放置在混肥罐(23)的底部，液位传感器(20)的信号输出端也与模拟量模块(16)的输入端相连，基于水下压力不同的原理来进行液位判断；压力变送器(25)连接在主管路中部，以判断主管道压力，其输出端与模拟量模块(16)的输入端相连；所述作物信息检测系统包括micro-CT扫描系统、偏振-高光谱成像系统和三维激光扫描系统；
    所述控制系统包括PLC控制器(18)、触摸屏(21)、变频器(9)和执行机构，其中触摸屏(21)与PLC控制器(18)相连接，用于进行人机交互，输入控制模式和作物生长状态信息；其中变频器(9)的输入端与PLC控制器(18)的输出端相连接，变频器(9)的输出端与施肥泵(24)相连接，模拟量模块(16)与PLC控制器(18)相连接，用于实现PLC控制器(18)对信息采集系统的多路信息采集控制；
    系统根据作物的水肥需求结合环境信息，可以通过PLC控制器(18)实现对变频器(9)的频率控制，进而实现对施肥泵(24)转速的控制，以调节施肥管道的流量，在恒压主管路中，最终实现对主管道施肥灌溉量的精确控制。
  8. 根据权利要求7所述基于多尺度生境信息的苗期作物水肥检测和控制装置，其特征在于，所述micro-CT扫描系统中，(27)为旋转样本托架，通过底座上的螺钉固定在检测样本仓的底部，旋转样本托架底座几何中心安装有旋转轴，旋转轴末端安装固定有圆形样本托架，检测时旋转轴带动样本托架360°旋转，同时固定在发射仓中部的X射线发射器(29)通过俯仰动作实现对样本的断层切片扫描过程。
  9. 根据权利要求7所述基于多尺度生境信息的苗期作物水肥检测和控制装置，其特征在于，其中所述偏振-高光谱图像检测系统，包括控制系统、双坐标样本台(32)、图像采集系统、光源系统；
    其中所述图像采集系统包括两个偏振-高光谱成像系统(35)、图像采集器(38)、立 臂(33)和悬臂(34)；所述立臂(33)由第一底座(33-1)、带丝杠的立杆(33-2)和第一滑块(33-3)组成，第一底座(33-1)通过螺钉固定在光箱(42)的底部的左侧，第一底座(33-1)上部通过铰链与立杆(33-2)连接，立杆(33-2)能够以铰链为中心做左右摆动，完成成像设备的空间位姿调整；立杆(33-2)上安装有第一滑块(33-3)；第一偏振-高光谱成像系统(35-1)安装在第一滑块(33-3)上，第一滑块(33-3)能够由丝杠驱动沿立杆(33-2)上下移动，带动第一偏振-高光谱成像系统(35-1)寻找最佳检测位，实现主视方向的偏振高光谱图像信息的采集；
    所述悬臂(34)由第二底座(34-1)、带丝杠的横杆(34-2)和第二滑块(34-3)组成，第二底座(34-1)通过螺钉固定在光箱(42)的右侧板的上部，第二底座(34-1)通过铰链与横杆(34-2)连接，横杆(34-2)能够以铰链为中心做上下摆动，完成成像设备的空间位姿调整；横杆(34-2)上安装有第二滑块(34-3)，第二偏振-高光谱成像系统(35-2)安装在第二滑块(34-3)上，第二滑块(34-3)能够由丝杠驱动沿横杆(34-2)沿水平方向左右移动，带动第二偏振-高光谱成像系统(35-2)寻找最佳检测位，实现俯视方向的偏振高光谱图像信息的采集；
    其中所述光源系统由可见光-近红外光源(37)、云台(36)组成，在立杆(33-2)的底端和顶端、立杆(34-2)的右端和左端分别安装一个云台(36)，每个云台上分别安装可见光-近红外光源(37)，可见光-近红外光源(37)能够通过云台(36)进行仰俯角设置，实现对植株的清晰匀光成像；
    所述双坐标样本台(32)固定在光箱(42)的底平面的几何中心位置，垂直丝杠(32-2)顶端安装有样本托架，用于安放待测样本，通过水平丝杠(32-1)和垂直丝杠(32-2)的运动，可以带动样本托架实现水平和垂直方向的匀速位移，可以配合图像采集控制系统实现推帚式的第一偏振-高光谱成像系统(35-1)和第二偏振-高光谱成像系统(35-2)的扫描成像；
    其中所述偏振-高光谱成像系统(35)由前到后分别由前置偏振片、偏振片驱动装置、前置滤光片、滤光片切换装置、摄谱仪和成像系统组成，偏振片在整个系统的最前端，由偏振驱动装置驱动360°旋转，可实现对任意偏振角的设定，摄谱仪和成像系统可实现偏振角的设定和步序偏振信息的采集；偏振片后为560nm和1450nm窄带滤光片，滤光片采用转轮切换的方式，配合摄谱仪和成像系统实现对作物样本主视和俯视高光谱营养和水分胁迫特征图像的采集；
    所述控制系统包括控制计算机(41)、光源控制器(40)、图像采集器(39)和运动控制器(38)；
    其中光源控制器(40)连接可见光-近红外光源(37)，实现不同光强和光质的光源控制；
    图像采集器(39)连接两个偏振-高光谱成像系统(35和控制计算机(41)，由控制计算机(41)发出指令，实现对主视和俯视偏振-高光谱成像系统(35)的成像信息采集和分析；
    运动控制器(38)连接双坐标样本台(32)、立臂(33)、悬臂(34)和云台(36)；同时，运动控制器(38)与控制计算机(41)相连，由控制计算机(41发出指令，实现对双坐标样本台(32)的升降和水平位移控制，对立臂(33)、悬臂(34)的滑块驱动控制，以及云台(36)的仰俯角控制。
PCT/CN2017/117190 2017-12-05 2017-12-19 一种基于多尺度生境信息的苗期作物水肥检测和控制方法及装置 WO2019109384A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US16/646,369 US11406057B2 (en) 2017-12-05 2017-12-19 Multi-scale habitat information-based method and device for detecting and controlling water and fertilizer for crops in seedling stage

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201711269626.2 2017-12-05
CN201711269626.2A CN108323295B (zh) 2017-12-05 2017-12-05 一种基于多尺度生境信息的苗期作物水肥检测和控制方法及装置

Publications (1)

Publication Number Publication Date
WO2019109384A1 true WO2019109384A1 (zh) 2019-06-13

Family

ID=62923093

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/117190 WO2019109384A1 (zh) 2017-12-05 2017-12-19 一种基于多尺度生境信息的苗期作物水肥检测和控制方法及装置

Country Status (3)

Country Link
US (1) US11406057B2 (zh)
CN (1) CN108323295B (zh)
WO (1) WO2019109384A1 (zh)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112806142A (zh) * 2019-11-15 2021-05-18 陈优为 一种水肥一体化自动灌溉系统
CN112949445A (zh) * 2021-02-24 2021-06-11 中煤科工集团重庆智慧城市科技研究院有限公司 基于空间关系的城市管理应急联动系统及方法
CN114731821A (zh) * 2022-05-17 2022-07-12 五华县伟鑫达种植专业合作社 一种基地全套水肥智能化管控系统
CN117243090A (zh) * 2023-10-19 2023-12-19 新疆水利水电科学研究院 一种农业灌溉用水定额配给系统
CN117268476A (zh) * 2023-11-21 2023-12-22 连云港华企立方信息技术有限公司 一种基于生态农业的智慧管理方法

Families Citing this family (31)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3944144A1 (en) * 2016-10-28 2022-01-26 Beckman Coulter, Inc. Substance preparation evaluation system
CN110275036B (zh) * 2019-05-23 2023-07-18 中国人民解放军第五七一九工厂 一种离线式油液颗粒度检测系统及检测方法
CN110398465A (zh) * 2019-07-06 2019-11-01 中国海洋大学 一种基于光谱遥感影像的养殖紫菜生物量测定方法
CN110736809A (zh) * 2019-10-28 2020-01-31 沈阳农业大学 一种日光温室作物水分胁迫诊断系统
CN110749555B (zh) * 2019-10-30 2022-05-31 宜宾五粮液股份有限公司 基于高光谱技术白酒曲块内部发酵状态检测装置及方法
CN113689374B (zh) * 2020-05-18 2023-10-27 浙江大学 一种植物叶片表面粗糙度确定方法及系统
CN112314142A (zh) * 2020-11-09 2021-02-05 洛阳智能农业装备研究院有限公司 一种针对便携式水肥一体机施肥流量的控制方法、装置
CN112488467A (zh) * 2020-11-16 2021-03-12 中国科学院合肥物质科学研究院 一种基于多尺度生境信息的水培作物施肥装置
CN112690078A (zh) * 2020-12-07 2021-04-23 广州顺绿喷灌设备有限公司 一种智能水肥药一体化灌溉系统及方法
CN112715238A (zh) * 2020-12-25 2021-04-30 南京信息职业技术学院 一种小型温室培养装置
CN114764833A (zh) * 2021-01-12 2022-07-19 富泰华工业（深圳）有限公司 植物生长曲线确定方法、装置、电子设备及介质
CN112947635B (zh) * 2021-01-15 2022-04-29 江南大学 基于卡尔曼滤波器多模型麦苗生长舱最优参数预测方法
CN112956319B (zh) * 2021-01-29 2022-04-05 新疆生产建设兵团第一师农业科学研究所 一种控失肥料检测装置及检测方法
CN114185467A (zh) * 2021-10-29 2022-03-15 北京市农林科学院信息技术研究中心 一种营养液信息采集方法及装置
CN114365614A (zh) * 2021-11-22 2022-04-19 湖南大学 基于物联网的水肥精准调控方法、智能装备及系统
WO2023144680A1 (en) * 2022-01-25 2023-08-03 Cet Electronics Snc Method and apparatus for estimating the water stress of a plant
CN114600605B (zh) * 2022-04-01 2023-03-28 青岛农业大学 一种变尺度变量追肥装置
CN114991047A (zh) * 2022-08-02 2022-09-02 山东省凯麟环保设备股份有限公司 基于智能视觉和双向地面清洁度判断的清扫方法及系统
CN115349338B (zh) * 2022-08-17 2023-06-30 江苏省农业机械试验鉴定站 一种基于作物长势的变量追肥控制系统与方法
CN115530052B (zh) * 2022-09-19 2024-01-26 南京林业大学 一种可移动门架式植物表型平台及其精确作业管理方法
CN115423643A (zh) * 2022-11-04 2022-12-02 中化现代农业有限公司 水肥计算方法、装置、电子设备和存储介质
CN115841470B (zh) * 2022-12-05 2023-08-11 中国科学院合肥物质科学研究院 一种基于根际的雾培番茄水分胁迫估算方法
CN115589802B (zh) * 2022-12-15 2023-02-28 山西乐村淘网络科技有限公司 基于作物生长反馈的智慧农业自动施肥装置
CN116147950B (zh) * 2023-01-03 2023-10-20 石河子大学 供水方法及滴灌系统性能测试平台
CN116223398B (zh) * 2023-02-01 2023-11-03 广州华立学院 一种基于光谱分析的水肥药一体化浓度配比方法
CN115861827B (zh) * 2023-02-28 2023-06-02 北京市农林科学院智能装备技术研究中心 作物水肥胁迫的决策方法、装置及手机终端
CN117063688A (zh) * 2023-10-16 2023-11-17 潍坊市农业科学院(山东省农业科学院潍坊市分院) 一种用于液态肥料的施肥控制系统及方法
CN117441472B (zh) * 2023-11-14 2024-10-11 南京农业大学 基于多信息融合的设施番茄水肥养分动态按需调控系统
CN117272213B (zh) * 2023-11-21 2024-02-02 中南大学 地下污染物的地物化综合参数扫面方法、装置、设备及介质
CN117616961B (zh) * 2024-01-15 2024-04-30 中国农业科学院农业环境与可持续发展研究所 一种滴灌水肥控制系统及方法
CN117616962B (zh) * 2024-01-15 2024-05-24 中国农业科学院农业环境与可持续发展研究所 一种卷盘喷灌水肥同步控制系统及方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN203773326U (zh) * 2014-03-25 2014-08-13 南京农业大学 基于高光谱信息的变量追肥监控系统
CN104198396A (zh) * 2014-07-30 2014-12-10 江苏大学 偏振-高光谱技术诊断作物氮磷钾亏缺的方法
WO2015092799A1 (en) * 2013-12-19 2015-06-25 Phytech Ltd. Method and system for crop management
EP3135102A1 (en) * 2015-08-28 2017-03-01 Ricoh Company, Ltd. Plant cultivation supporting apparatus, plant cultivation supporting method, program, and recording medium
CN106664937A (zh) * 2017-03-20 2017-05-17 浙江省农业科学院 水肥一体化四控灌溉施肥系统
CN107173184A (zh) * 2017-05-25 2017-09-19 河南嘉禾智慧农业科技有限公司 一种智能化农业灌溉系统及方法

Family Cites Families (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE19514223B4 (de) * 1995-04-15 2005-06-23 Claas Kgaa Mbh Verfahren zur Einsatzoptimierung von Landmaschinen
US7715013B2 (en) * 2005-09-16 2010-05-11 The United States Of America As Represented By The Administrator Of The United States Environmental Protection Agency Optical system for plant characterization
US7617057B2 (en) * 2005-12-21 2009-11-10 Inst Technology Development Expert system for controlling plant growth in a contained environment
CN103048266B (zh) * 2012-12-11 2015-06-10 江苏大学 一种设施番茄氮磷钾胁迫自动识别方法和装置
CN103018179B (zh) * 2012-12-11 2015-04-22 江苏大学 一种作物水分胁迫的近红外偏振超光谱成像检测装置和方法
US10139279B2 (en) * 2015-05-12 2018-11-27 BioSensing Systems, LLC Apparatuses and methods for bio-sensing using unmanned aerial vehicles
CN204576603U (zh) * 2015-05-21 2015-08-19 河南省华西高效农业有限公司 一种新型soa架构的精准农业管理系统
WO2017130249A1 (ja) * 2016-01-29 2017-08-03 パナソニックＩｐマネジメント株式会社 水分量観察装置、水分量観察方法及び栽培装置
CN106774069B (zh) * 2016-12-26 2023-03-31 机械工业勘察设计研究院有限公司 一种基于三维激光扫描的土方填筑监控装置及方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2015092799A1 (en) * 2013-12-19 2015-06-25 Phytech Ltd. Method and system for crop management
CN203773326U (zh) * 2014-03-25 2014-08-13 南京农业大学 基于高光谱信息的变量追肥监控系统
CN104198396A (zh) * 2014-07-30 2014-12-10 江苏大学 偏振-高光谱技术诊断作物氮磷钾亏缺的方法
EP3135102A1 (en) * 2015-08-28 2017-03-01 Ricoh Company, Ltd. Plant cultivation supporting apparatus, plant cultivation supporting method, program, and recording medium
CN106664937A (zh) * 2017-03-20 2017-05-17 浙江省农业科学院 水肥一体化四控灌溉施肥系统
CN107173184A (zh) * 2017-05-25 2017-09-19 河南嘉禾智慧农业科技有限公司 一种智能化农业灌溉系统及方法

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112806142A (zh) * 2019-11-15 2021-05-18 陈优为 一种水肥一体化自动灌溉系统
CN112949445A (zh) * 2021-02-24 2021-06-11 中煤科工集团重庆智慧城市科技研究院有限公司 基于空间关系的城市管理应急联动系统及方法
CN112949445B (zh) * 2021-02-24 2024-04-05 中煤科工集团重庆智慧城市科技研究院有限公司 基于空间关系的城市管理应急联动系统及方法
CN114731821A (zh) * 2022-05-17 2022-07-12 五华县伟鑫达种植专业合作社 一种基地全套水肥智能化管控系统
CN117243090A (zh) * 2023-10-19 2023-12-19 新疆水利水电科学研究院 一种农业灌溉用水定额配给系统
CN117243090B (zh) * 2023-10-19 2024-03-29 新疆水利水电科学研究院 一种农业灌溉用水定额配给系统
CN117268476A (zh) * 2023-11-21 2023-12-22 连云港华企立方信息技术有限公司 一种基于生态农业的智慧管理方法
CN117268476B (zh) * 2023-11-21 2024-03-22 连云港华企立方信息技术有限公司 一种基于生态农业的智慧管理方法

Also Published As

Publication number Publication date
US20210289692A1 (en) 2021-09-23
CN108323295A (zh) 2018-07-27
CN108323295B (zh) 2019-12-03
US11406057B2 (en) 2022-08-09

Similar Documents

Publication Publication Date Title
WO2019109384A1 (zh) 一种基于多尺度生境信息的苗期作物水肥检测和控制方法及装置
US11436824B2 (en) Water stress detection method for tomatoes in seedling stage based on micro-CT and polarization-hyperspectral imaging multi-feature fusion
CN108376419B (zh) 一种盆栽生菜的综合长势监测方法及装置
CN104457936B (zh) 一种盆栽作物长势动态检测装置及其检测方法
CN103959973B (zh) 一种精细化作物施肥系统及氮肥施肥方法
KR101903018B1 (ko) 에이치엠아이 기반의 자동화 고설재배 시스템
CN103329780A (zh) 一种基质栽培作物的灌溉决策方法及灌溉系统
CN108064531A (zh) 一种轻简化温室水肥灌施装置
CN111693551B (zh) 一种水稻植株及根系三维性状无损测量装置及方法
CN108830741A (zh) 一种农田环境智能监测系统
CN1789980A (zh) 基于近红外光谱的植物生长信息获取装置
WO2016074513A1 (zh) 一种旋臂式设施作物生物量多传感检测装置及方法
CN2864669Y (zh) 基于近红外光谱的植物生长信息获取装置
CN115885837A (zh) 一种水耕栽培营养液流速试验装置及方法
CN107421489B (zh) 一种实时非破环性农作物根系深度判别系统及方法
CN115152357A (zh) 植物种子培养皿、萌发检验装置、培育装置及萌发方法
CN106643521B (zh) 一种农作物冠层高度的检测方法及装置
CN105387934B (zh) 冠层内光合有效辐射自动跟踪测量装置
CN108605458A (zh) 一种基于物联网的农业种植土壤酸碱度调节装置
CN107027550A (zh) 一种基于人工气候箱的多功能植物保护装置及其使用方法
CN216117317U (zh) 一种非接触式近红外土壤墒情在线检测装置
CN206531572U (zh) 一种作物冠层内光合有效辐射测量装置
CN206389884U (zh) 一种基于人工气候箱的多功能植物保护装置
CN210298680U (zh) 一种用于高通量植物根系可视化分析的培养及测量装置
CN106092910A (zh) 一种枣树冠层氮含量的检测方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17934206

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17934206

Country of ref document: EP

Kind code of ref document: A1