CN114429592A - Automatic irrigation method and equipment based on artificial intelligence - Google Patents
Automatic irrigation method and equipment based on artificial intelligence Download PDFInfo
- Publication number
- CN114429592A CN114429592A CN202111653810.3A CN202111653810A CN114429592A CN 114429592 A CN114429592 A CN 114429592A CN 202111653810 A CN202111653810 A CN 202111653810A CN 114429592 A CN114429592 A CN 114429592A
- Authority
- CN
- China
- Prior art keywords
- irrigation
- soil
- vegetation
- hyperspectral
- water content
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
- 238000003973 irrigation Methods 0.000 title claims abstract description 206
- 230000002262 irrigation Effects 0.000 title claims abstract description 206
- 238000000034 method Methods 0.000 title claims abstract description 42
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 30
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 149
- 239000002689 soil Substances 0.000 claims abstract description 145
- 238000001514 detection method Methods 0.000 claims abstract description 33
- 238000001556 precipitation Methods 0.000 claims abstract description 31
- 238000001035 drying Methods 0.000 claims description 24
- 238000002474 experimental method Methods 0.000 claims description 24
- 230000000875 corresponding effect Effects 0.000 claims description 20
- 238000004422 calculation algorithm Methods 0.000 claims description 18
- 238000012417 linear regression Methods 0.000 claims description 12
- 238000010248 power generation Methods 0.000 claims description 12
- 238000012549 training Methods 0.000 claims description 12
- 238000001228 spectrum Methods 0.000 claims description 10
- 230000002159 abnormal effect Effects 0.000 claims description 9
- 210000004027 cell Anatomy 0.000 claims description 9
- 238000012544 monitoring process Methods 0.000 claims description 8
- 230000005540 biological transmission Effects 0.000 claims description 6
- 238000000701 chemical imaging Methods 0.000 claims description 6
- 230000001276 controlling effect Effects 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 6
- 238000003860 storage Methods 0.000 claims description 5
- 210000004460 N cell Anatomy 0.000 claims description 3
- 238000012937 correction Methods 0.000 claims description 3
- 238000005520 cutting process Methods 0.000 claims description 3
- 230000006870 function Effects 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 claims description 3
- 238000011161 development Methods 0.000 description 6
- 230000018109 developmental process Effects 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 4
- 238000007689 inspection Methods 0.000 description 4
- 238000012271 agricultural production Methods 0.000 description 3
- 230000006872 improvement Effects 0.000 description 2
- 239000003621 irrigation water Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000010238 partial least squares regression Methods 0.000 description 2
- 230000008447 perception Effects 0.000 description 2
- 244000061456 Solanum tuberosum Species 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000013505 freshwater Substances 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
- 238000005303 weighing Methods 0.000 description 1
Images
Classifications
-
- A—HUMAN NECESSITIES
- A01—AGRICULTURE; FORESTRY; ANIMAL HUSBANDRY; HUNTING; TRAPPING; FISHING
- A01G—HORTICULTURE; CULTIVATION OF VEGETABLES, FLOWERS, RICE, FRUIT, VINES, HOPS OR SEAWEED; FORESTRY; WATERING
- A01G25/00—Watering gardens, fields, sports grounds or the like
- A01G25/16—Control of watering
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
Landscapes
- Life Sciences & Earth Sciences (AREA)
- Physics & Mathematics (AREA)
- Biochemistry (AREA)
- Health & Medical Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Spectroscopy & Molecular Physics (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Engineering & Computer Science (AREA)
- Water Supply & Treatment (AREA)
- Environmental Sciences (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention discloses an automatic irrigation method and equipment based on artificial intelligence, belongs to the technical field of artificial intelligence, and is used for solving the technical problems of low utilization rate of water resources and low automation degree of irrigation areas in the prior farmland irrigation. The method comprises the following steps: regularly patrolling the irrigation area by an unmanned aerial vehicle carrying a hyperspectral camera, and collecting a plurality of hyperspectral images; segmenting vegetation pixels and soil pixels in the hyperspectral images to obtain vegetation hyperspectral images and soil hyperspectral images; identifying the hyperspectral image of the vegetation through the trained vegetation water detection model to obtain the vegetation water content; recognizing the hyperspectral image of the vegetation through the trained soil moisture detection model to obtain the soil moisture content; obtaining the predicted precipitation in the preset time of an irrigation area; and sending the vegetation water content, the soil water content and the predicted precipitation to an irrigation decision system to judge whether to irrigate the irrigated area.
Description
Technical Field
The application relates to the field of artificial intelligence, in particular to an automatic irrigation method and equipment based on artificial intelligence.
Background
Agricultural production plays an important role in the development process of social economy, and is one of basic industries supporting the development of national economy. In recent years, with the comprehensive development of science and technology, a foundation is laid for the generation and the improvement of an artificial intelligence technology to a certain extent, the artificial intelligence technology is used in all industries, particularly in the field of agricultural production, the artificial intelligence technology is widely applied to agricultural production, and the agricultural high-quality development is facilitated.
The farmland irrigation water accounts for a large proportion in agricultural water, and the pressure of fresh water resources is continuously increased along with the development of the economic society. But the prior farmland irrigation has the defects of low water resource utilization rate and low informatization degree of irrigation areas. The imperfect irrigation mode is time-consuming, water-consuming and labor-consuming, and is not beneficial to agricultural development and water resource saving and utilization.
Disclosure of Invention
The embodiment of the application provides an automatic irrigation method and equipment based on artificial intelligence, which are used for solving the following technical problems: at present, the utilization rate of farmland irrigation water resources is not high, and the automation degree of irrigation areas is low.
The embodiment of the application adopts the following technical scheme:
on one hand, the embodiment of the application provides a method for regularly patrolling an irrigation area through an unmanned aerial vehicle carrying a hyperspectral camera and acquiring a plurality of hyperspectral images; segmenting vegetation pixels and soil pixels in the hyperspectral images to obtain vegetation hyperspectral images and soil hyperspectral images; identifying the vegetation hyperspectral image through a trained vegetation moisture detection model to obtain the vegetation moisture content; identifying the vegetation hyperspectral image through a trained soil moisture detection model to obtain the soil moisture content; obtaining the predicted precipitation in the preset time of the irrigation area; sending the vegetation water content, the soil water content and the predicted precipitation to an irrigation decision system to judge whether the irrigation area is irrigated or not; sending the irrigation strategy output by the irrigation decision system to a remote monitoring platform for storage and display; the irrigation strategy comprises whether the irrigation area is irrigated or not, and if the irrigation area needs to be irrigated, the irrigation strategy also comprises irrigation time and irrigation quantity.
In a feasible implementation manner, the vegetation pixels and the soil pixels in the hyperspectral images are segmented to obtain a vegetation hyperspectral image and a soil hyperspectral image, and the method specifically includes the following steps: converting the collected hyperspectral image into a gray image; setting the gray value interval of the gray image as [0, a](ii) a With [0, a ]]Dividing the gray image into [0, t ] by taking each gray value t in the interval as a division threshold]And [ t +1, a]Two partial images; calculating a ratio alpha of the number of pixels of each of the two partial images0、α1And the average gray value β of each part0、β1(ii) a According to
Obtaining the overall average gray value of the gray image; according to gamma2＝α0(β0-β)2+α1(β1-β)2＝α0α1(β0-β1)2Obtaining the inter-class variance of the two parts of images; determining the corresponding threshold T when the inter-class variance is maximum as the optimal segmentation threshold, and segmenting the gray level image into [0, T]And [ T +1, a]Two partial images; according to [0, T ] in the gray level image]Determining vegetation pixel points in the hyperspectral image by using pixel points in the interval; according to [ T +1, a ] in the gray level image]Determining soil pixel points in the hyperspectral image by pixel points in the interval; keeping the pixel value of a vegetation pixel point in the hyperspectral image unchanged, and setting the pixel value of a soil pixel point to be 0 or 255 to obtain the vegetation hyperspectral image; keeping the pixel value of a soil pixel point in the hyperspectral image unchanged, and vegetation pixel point pixelsAnd setting the value to be 0 or 255, and obtaining the soil hyperspectral image.
In a possible implementation manner, before the trained vegetation moisture detection model identifies the vegetation hyperspectral image to obtain the vegetation moisture content, the method further includes: constructing a first linear regression model according to a partial minimum second-product regression algorithm; collecting a plurality of leaf samples of vegetation in the irrigation area, and obtaining a plurality of groups of leaf samples with different water contents through a plurality of groups of drying experiments; wherein, the group of drying experiments comprises a plurality of drying experiments; performing hyperspectral image acquisition on each group of blade samples with different water contents by a hyperspectral imaging device, and performing black-and-white board correction on the acquired hyperspectral images; calculating the average value of the reflection spectrum of each group of blade samples according to the corrected hyperspectral images to obtain hyperspectral data of each group of blade samples; wherein the hyperspectral data comprises 256 characteristic wavelengths; obtaining a plurality of leaf water content matrixes according to the water content of each group of leaf samples; obtaining a plurality of blade hyperspectral data matrixes according to the hyperspectral data of each group of blade samples; and training the first linear regression model by taking the plurality of blade water content matrixes and the plurality of blade hyperspectral data matrixes as training sets to obtain the vegetation water detection model.
In a feasible implementation, through the trained vegetation moisture detection model, it is right a plurality of high spectrum images are discerned, obtain the vegetation moisture content, specifically include: dividing the vegetation hyperspectral image into N cells, and respectively calculating the average value of the reflection spectrum in each cell to obtain N vegetation hyperspectral data matrixes; inputting the N vegetation hyperspectral data matrixes into the vegetation moisture detection model to obtain the moisture content corresponding to each cell; according to
Obtaining a moisture content F1 corresponding to the vegetation hyperspectral image; wherein f is1，f2，……，fNThe water content corresponding to each cell is obtained; according to
Obtaining vegetation water content G1 in the irrigation area; where M is the number of acquired high-spectrum images.
In a feasible implementation mode, through a trained soil moisture detection model, the vegetation hyperspectral image is identified to obtain the soil moisture content, and the method specifically comprises the following steps: constructing a second linear regression model according to a partial least square regression algorithm; collecting a plurality of soil samples in the irrigation area, and obtaining a plurality of groups of soil samples with different water contents through a plurality of groups of drying experiments; wherein, the group of drying experiments comprises a plurality of times of drying experiments; performing hyperspectral image acquisition on each group of soil samples with different water contents by a hyperspectral imaging device; calculating the full-wave-band albedo of each group of soil samples according to the acquired hyperspectral images, and measuring the day and night temperature of the irrigation area; obtaining the apparent thermal inertia of each group of soil samples based on the full-wave albedo and the day and night temperature; obtaining a plurality of soil moisture content matrixes according to the moisture content of each group of soil samples; obtaining a plurality of soil apparent thermal inertia matrixes according to the apparent thermal inertia of each group of soil samples; taking the plurality of soil moisture content matrixes and the plurality of apparent thermal inertia matrixes as a training set, and training the second linear regression model to obtain the soil moisture detection model; and inputting the soil hyperspectral image into the soil moisture detection model to obtain the soil moisture content.
In one possible embodiment, the irrigation decision system includes a sensing layer, a network layer, and an application layer; the sensing layer comprises a soil moisture content sensor so as to obtain the actual water content of soil in the field irrigation process; the network layer is a wireless data transmission channel and is used for realizing the data transmission function between the sensing layer and the application layer; the application layer comprises an irrigation decision algorithm, the transmitted data are processed through the irrigation decision algorithm, and a corresponding irrigation strategy is output; the application layer is also used for controlling the solar power generation device and the water pump to execute corresponding actions, so that intelligent irrigation and remote monitoring management are realized; sending the vegetation water content, the soil water content and the forecast precipitation to an irrigation decision system, and judging whether to irrigate the irrigation area, wherein the method specifically comprises the following steps: acquiring the actual moisture content of the soil in the irrigation area through a soil moisture content sensor connected with the irrigation decision system; and inputting the actual water content of the soil, the vegetation water content, the soil water content and the predicted precipitation into the irrigation decision algorithm in the application layer to obtain the irrigation strategy of the irrigation area.
In a possible implementation manner, inputting the actual water content of the soil, the vegetation water content, the soil water content, and the predicted precipitation into the irrigation decision algorithm in the application layer to obtain the irrigation strategy of the irrigation area, which specifically includes: under the condition that the actual water content value of the soil is abnormal, the irrigation area is not irrigated, and first alarm information is sent out; wherein the content of the first warning information indicates that the soil moisture sensor is abnormal; under the condition that the actual water content value of the soil is normal, comparing the actual water content of the soil with the water content of the soil, if the difference value of the actual water content of the soil and the water content of the soil is larger than a first preset threshold value, not irrigating the irrigation area, and sending out second alarm information; the content of the second warning information indicates that the data collected by the unmanned aerial vehicle or the moisture detection model is abnormal; if the difference value between the two values is smaller than the first preset threshold value and the predicted precipitation is smaller than a second preset threshold value, irrigating the irrigation area; and if the difference value of the two values is smaller than the first preset threshold value and the predicted precipitation is larger than or equal to a second preset threshold value, not irrigating the irrigation area.
In a possible embodiment, after the irrigation area is irrigated if the difference between the first preset threshold and the second preset threshold is smaller than the first preset threshold and the predicted precipitation is smaller than the second preset threshold, the method further includes: acquiring the theoretical water demand X of the vegetation in the current growth stage; according to H ═ 1-omegai) X) Y, and obtaining the current water demand H of all vegetation in the irrigation area; wherein Y is the vegetation number in the irrigated area; omegaiIs as followsCurrent water demand of i vegetation, and 0<i is less than or equal to Y; obtaining irrigation quantity I of the irrigation area according to the I-H-K J-Z; wherein K is the soil water content, J is the area of the irrigation area, and Z is the predicted precipitation; and controlling an irrigation device to irrigate the irrigation area according to the irrigation quantity.
In a possible embodiment, the control of the irrigation device to irrigate the irrigation area according to the irrigation quantity specifically comprises: connecting a passage between a solar power generation device and a water pump of the irrigation device so that the solar power generation device provides electric energy for the water pump of the irrigation device; calculating the water yield of a single water outlet through a water flow sensor arranged at each water outlet, and calculating the total water yield of all the water outlets; and cutting off a passage between the solar power generation device and the water pump after the total water yield reaches the irrigation quantity.
On the other hand, this application embodiment still provides an automatic irrigation equipment based on artificial intelligence, includes: at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to cause the at least one processor to perform a method for artificial intelligence based automatic irrigation according to any of the embodiments described above.
The embodiment of the application provides an automatic irrigation method and equipment based on artificial intelligence, through regularly patrolling irrigated area to through two moisture detection models, the moisture content of vegetation and the moisture content of soil in the accurate detection irrigated area, thereby whether need water in the accurate judgement irrigated area, how much water need be watered, thereby make full use of water resource avoids causing the unnecessary waste. The full-automatic irrigation process saves manpower and physical force and lightens the workload of irrigation workers.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments described in the present application, and other drawings can be obtained by those skilled in the art without creative efforts. In the drawings:
FIG. 1 is a flow chart of an artificial intelligence based automatic irrigation method provided by an embodiment of the present application;
fig. 2 is a schematic structural diagram of an automatic irrigation device based on artificial intelligence according to an embodiment of the present application.
Detailed Description
In order to make those skilled in the art better understand the technical solutions in the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present disclosure without making any creative effort, shall fall within the protection scope of the present disclosure.
The embodiment of the application provides an automatic irrigation method based on artificial intelligence, and as shown in fig. 1, the automatic irrigation method based on artificial intelligence specifically comprises steps S101-S106:
s101, carrying out regular patrol on an irrigation area by an unmanned aerial vehicle with a hyperspectral camera, and collecting a plurality of hyperspectral images. And segmenting vegetation pixels and soil pixels in the hyperspectral images to obtain vegetation hyperspectral images and soil hyperspectral images.
Specifically, a hyperspectral camera is carried on the unmanned aerial vehicle, and then the inspection time and the inspection route of the unmanned aerial vehicle are specified through an automatic control program, so that the unmanned aerial vehicle can automatically execute the inspection work of the irrigation area at regular time. A plurality of hyperspectral images of the irrigated area are collected in each inspection.
Further, the collected hyperspectral image is converted into a gray level image, and the gray level interval of the gray level image is set as 0, a]. Then with [0, a ]]Dividing the gray image into [0, t ] by using each gray value t in the interval as a dividing threshold]And [ t +1，a]Two partial images. Calculating the ratio alpha of the number of pixels of each of the two partial images0、α1And the average gray value β of each part0、β1。
Further in accordance with
The overall average gray value of the gray image is obtained. Then according to gamma2＝α0(β0-β)2+α1(β1-β)2＝α0α1(β0-β1)2And obtaining the inter-class variance of the two partial images. Determining the threshold T corresponding to the maximum inter-class variance as the optimal segmentation threshold, and segmenting the grayscale image into [0, T]And [ T +1, a]Two partial images.
Further, determining vegetation pixel points in the hyperspectral image according to pixel points in the [0, T ] interval in the grayscale image. And determining soil pixel points in the hyperspectral image according to the pixel points in the [ T +1, a ] interval in the grayscale image. And keeping the pixel value of the vegetation pixel point in the hyperspectral image unchanged, and setting the pixel value of the soil pixel point to be 0 or 255 to obtain the vegetation hyperspectral image. And keeping the pixel value of the soil pixel point in the hyperspectral image unchanged, and setting the pixel value of the vegetation pixel point to be 0 or 255 to obtain the soil hyperspectral image.
As a feasible implementation manner, in the original image of the hyperspectral image, the pixel points at the positions corresponding to the pixel points in the [0, T ] interval in the grayscale image are found, that is, the vegetation pixel points. Similarly, in the original image of the hyperspectral image, the pixel points at the positions corresponding to the pixel points in the [ T +1, a ] interval in the grayscale image are found, and the pixel points are the soil pixel points.
S102, identifying the hyperspectral image of the vegetation through the trained vegetation water detection model to obtain the vegetation water content.
Specifically, a first linear regression model is constructed according to a partial least squares regression algorithm. Collecting a plurality of leaf samples of vegetation in an irrigation area, and obtaining a plurality of groups of leaf samples with different water contents through a plurality of groups of drying experiments; wherein, a set of drying experiments comprises a plurality of drying experiments. Then, hyperspectral image acquisition is carried out on the blade samples with different water contents in each group through a hyperspectral imaging device, and black and white board correction is carried out on the acquired hyperspectral images.
As a possible embodiment, potato plants are grown, a critical period of water management (the emergence period) is selected, and 60 leaf samples are taken from vegetation planted in the irrigated area. The leaf samples were taken back to the laboratory, each leaf sample weighed and the mass recorded as M0The hyperspectral image of each leaf sample is then scanned. Then putting the leaf sample into an oven at 40 ℃, baking for 40min, taking out and weighing again, wherein the mass is recorded as M1. And again scan the hyperspectral image of each leaf sample. Repeating the measurement for 3 times, wherein each time of drying is 40min to obtain the mass M2、M3、M4. Then, the leaves are put into an oven at 80 ℃, dried to constant weight, and the mass is recorded as Md. After 4 times of drying, the leaf samples which are not dried are added, which is equal to 300 leaf samples with different water contents, and the number of the obtained hyperspectral images is 300. Then, a water content calculation formula of the leaves is used:
and calculating the water content in the blade sample after the nth drying experiment. Wherein M isnThe mass of the blade sample after the nth drying experiment is in a value range of 0-4. The 4 drying experiments are a group of drying experiments, and the leaves of the vegetation can be collected at other places to carry out multiple groups of experiments to obtain more samples.
Further, according to the corrected hyperspectral images, the average value of the reflection spectrum of each group of leaf samples is calculated, and hyperspectral data of each group of leaf samples are obtained. The hyperspectral data comprises 256 characteristic wavelengths. And obtaining a plurality of leaf water content matrixes according to the water content of each group of leaf samples. And obtaining a plurality of blade hyperspectral data matrixes according to the hyperspectral data of each group of blade samples.
As a possible implementation, the water cut of each group of leaf samples is represented by a water cut matrix of the number of samples × 1. And then opening the corrected hyperspectral image data of the blade sample in each group of experiments in software ENVl5.1, and calculating the average value of the reflection spectrum of the whole blade as the spectrum data of the sample. If the number of the blade samples in each group is 300, a plurality of 300 × 256 wavelength spectral data matrixes are finally obtained.
And further, training the first linear regression model by taking the plurality of blade water content matrixes and the plurality of blade hyperspectral data matrixes as training sets to obtain a vegetation water detection model.
And further, dividing the vegetation hyperspectral image into N cells, and respectively calculating the average value of the reflection spectrums in each cell to obtain N vegetation hyperspectral data matrixes. And inputting the N vegetation hyperspectral data matrixes into a vegetation moisture detection model to obtain the moisture content corresponding to each cell.
Further in accordance with
Obtaining a moisture content F1 corresponding to the vegetation hyperspectral image; wherein f is1，f2，……，fNThe corresponding water content of each cell. According to
Obtaining vegetation water content G1 in the irrigation area; and M is the number of the collected hyperspectral images.
S103, recognizing the vegetation hyperspectral image through the trained soil moisture detection model to obtain the soil moisture content.
Specifically, a second linear regression model is constructed according to a partial least squares regression algorithm. A plurality of soil samples in the irrigation area are collected, and a plurality of groups of soil samples with different water contents are obtained through a plurality of groups of drying experiments. Wherein, a set of drying experiments comprises a plurality of drying experiments. Then, hyperspectral image acquisition is carried out on the soil samples with different water contents in each group through a hyperspectral imaging device.
Further, according to the hyperspectral images of the collected soil samples, the full-wave-section albedo of each group of soil samples is calculated, and the day and night temperature of the irrigation area is measured. And obtaining the apparent thermal inertia of each group of soil samples based on the full-wave-band albedo and the day-night temperature.
As a possible embodiment, according to
And obtaining the apparent thermal inertia ATI of each group of soil samples. Wherein ABE is the full-wave-band albedo, T, of each group of soil samplesdIs the daytime temperature of the irrigated area, TnIs the temperature of the irrigated area at night.
And further, obtaining a plurality of soil moisture content matrixes according to the moisture content of each group of soil samples. And obtaining a plurality of soil apparent thermal inertia matrixes according to the apparent thermal inertia of each group of soil samples. And training the second linear regression model by taking the plurality of soil moisture content matrixes and the plurality of apparent thermal inertia matrixes as a training set to obtain a soil moisture detection model.
And further, inputting the hyperspectral image of the soil into a soil moisture detection model to obtain the moisture content of the soil.
And S104, obtaining the predicted precipitation in the preset time of the irrigation area.
As a feasible implementation mode, the predicted precipitation of the region where the irrigation area is located within the preset time can be crawled in the Internet through a script library in python.
And S105, sending the vegetation water content, the soil water content and the forecast precipitation to an irrigation decision system, and judging whether to irrigate the irrigated area.
Specifically, the irrigation decision system comprises a perception layer, a network layer and an application layer. The perception layer includes the soil moisture content sensor to obtain the actual moisture content of soil among the farmland irrigation process. The network layer is a wireless data transmission channel and is used for realizing a data transmission function between the sensing layer and the application layer. The application layer comprises an irrigation decision algorithm, the transmitted data are processed through the irrigation decision algorithm, and a corresponding irrigation strategy is output. The irrigation strategy comprises whether irrigation is carried out on the irrigation area or not, and if the irrigation area needs to be irrigated, the irrigation strategy also comprises irrigation time and irrigation quantity. The application layer is also used for controlling the solar power generation device and the water pump to execute corresponding actions, and intelligent irrigation and remote monitoring management are achieved.
Furthermore, the actual moisture content of the soil in the irrigation area is collected through a soil moisture content sensor connected with an irrigation decision system. A plurality of soil moisture content sensors are installed in the soil of the irrigation area. And inputting the actual water content of the soil, the vegetation water content, the soil water content and the predicted precipitation into an irrigation decision algorithm in the application layer to obtain an irrigation strategy of the irrigation area.
As a possible implementation, the irrigation decision algorithm comprises the following steps: under the condition that the actual water content value of the soil is abnormal, the irrigation area is not irrigated, and first alarm information is sent out; wherein the content of the first warning information indicates that the soil moisture content sensor is abnormal. Under the condition that the actual water content value of the soil is normal, comparing the actual water content of the soil with the water content of the soil, if the difference value of the actual water content of the soil and the water content of the soil is larger than a first preset threshold value, not irrigating the irrigation area, and sending second alarm information; and the content of the second alarm information indicates that the data acquired by the unmanned aerial vehicle or the moisture detection model is abnormal. And if the difference value of the two values is smaller than a first preset threshold value and the predicted precipitation is smaller than a second preset threshold value, irrigating the irrigation area. And if the difference value of the two is smaller than a first preset threshold value and the predicted precipitation is larger than or equal to a second preset threshold value, not irrigating the irrigation area.
In one embodiment, if the difference between the actual water content of the soil and the water content of the soil is smaller than a first preset threshold value, precipitation is obtained within the last half hour, and the precipitation is larger than the difference, the irrigation area does not need to be irrigated.
Further, if the irrigation area needs to be irrigated, the irrigation decision system obtains the theoretical water demand X of the vegetation in the irrigation area at the current growth stage. Then according to H ═ 1-omegai) X) Y, obtaining the current water demand H of all vegetation in the irrigation area. Wherein Y is the vegetation number in the irrigated area；ωiCurrent water demand for the ith vegetation, and 0<i is less than or equal to Y. Obtaining irrigation quantity I of the irrigation area according to the I-H-K J-Z; wherein K is the water content of the soil, J is the area of the irrigation area, and Z is the predicted precipitation. And the irrigation decision system controls the irrigation device to irrigate the irrigation area according to the irrigation quantity.
Further, after the irrigation decision-making system determines the irrigation quantity, a switch between the solar power generation device and a water pump of the irrigation device is controlled to be closed, so that the solar power generation device provides electric energy for the water pump of the irrigation device. And the water flow sensor arranged at each water outlet is used for calculating the water yield of a single water outlet and calculating the total water yield of all the water outlets. And after the total water yield reaches the irrigation quantity, cutting off the passage between the solar power generation device and the water pump so as to save electric energy. Solar power system still can charge for the soil moisture content sensor, charge for unmanned aerial vehicle and for the automatic irrigation equipment that this application provided based on artificial intelligence provides the electric energy.
And S106, sending the irrigation strategy output by the irrigation decision system to a remote monitoring platform for storage and display.
Specifically, the irrigation strategy output by the irrigation decision system every time is sent to the remote monitoring platform for storage and visual display, so that irrigation workers in the irrigation area can check the irrigation condition of the irrigation area in real time.
In addition, this application embodiment still provides an automatic irrigation equipment based on artificial intelligence, as shown in fig. 2, automatic irrigation equipment based on artificial intelligence specifically includes: at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor to cause the at least one processor to:
regularly patrolling the irrigation area by an unmanned aerial vehicle carrying a hyperspectral camera, and collecting a plurality of hyperspectral images;
segmenting vegetation pixels and soil pixels in the hyperspectral images to obtain vegetation hyperspectral images and soil hyperspectral images;
identifying the hyperspectral image of the vegetation through the trained vegetation water detection model to obtain the vegetation water content; recognizing the hyperspectral image of the vegetation through the trained soil moisture detection model to obtain the soil moisture content;
obtaining the predicted precipitation in the preset time of an irrigation area;
sending the vegetation water content, the soil water content and the predicted precipitation into an irrigation decision system, and judging whether to irrigate the irrigated area;
sending the irrigation strategy output by the irrigation decision system to a remote monitoring platform for storage and display; the irrigation strategy comprises whether irrigation is carried out on the irrigation area or not, and if the irrigation area needs to be irrigated, the irrigation strategy also comprises irrigation time and irrigation quantity.
The embodiments in the present application are described in a progressive manner, and the same and similar parts among the embodiments can be referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the apparatus embodiment, since it is substantially similar to the method embodiment, the description is relatively simple, and for the relevant points, reference may be made to the partial description of the method embodiment.
The foregoing description of specific embodiments of the present application has been presented. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
The above description is only an example of the present application and is not intended to limit the present application. Various modifications and alterations to the embodiments of the present application will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement or the like made within the spirit and principle of the embodiments of the present application shall be included in the scope of the claims of the present application.
Claims (10)
1. An artificial intelligence based automatic irrigation method, characterized in that the method comprises:
regularly patrolling the irrigation area by an unmanned aerial vehicle carrying a hyperspectral camera, and collecting a plurality of hyperspectral images;
segmenting vegetation pixels and soil pixels in the hyperspectral images to obtain vegetation hyperspectral images and soil hyperspectral images;
identifying the vegetation hyperspectral image through a trained vegetation moisture detection model to obtain the vegetation moisture content; identifying the vegetation hyperspectral image through a trained soil moisture detection model to obtain the soil moisture content;
obtaining the predicted precipitation in the preset time of the irrigation area;
sending the vegetation water content, the soil water content and the predicted precipitation to an irrigation decision system to judge whether the irrigation area is irrigated or not;
sending the irrigation strategy output by the irrigation decision system to a remote monitoring platform for storage and display; the irrigation strategy comprises whether the irrigation area is irrigated or not, and if the irrigation area needs to be irrigated, the irrigation strategy also comprises irrigation time and irrigation quantity.
2. The artificial intelligence based automatic irrigation method according to claim 1, wherein the vegetation hyperspectral image and the soil hyperspectral image are obtained by segmenting vegetation pixels and soil pixels in the hyperspectral images, and the method specifically comprises the following steps:
converting the collected hyperspectral image into a gray image;
setting a gray value interval of the gray image as [0, a ];
dividing the gray image into [0, t ] and [ t +1, a ] two-part images by taking each gray value t in the [0, a ] interval as a division threshold;
calculating the two partial imagesThe ratio of the number of pixels of each part of alpha0、α1And the average gray value β of each part0、β1；
According to
Obtaining the overall average gray value of the gray image;
according to gamma2＝α0(β0-β)2+α1(β1-β)2＝α0α1(β0-β1)2Obtaining the inter-class variance of the two parts of images;
determining a threshold value T corresponding to the maximum inter-class variance as an optimal segmentation threshold value, and segmenting the gray-scale image into two parts of images, namely [0, T ] and [ T +1, a ];
determining vegetation pixel points in the hyperspectral image according to pixel points in the [0, T ] interval in the grayscale image;
determining soil pixel points in the hyperspectral image according to pixel points in a [ T +1, a ] interval in the grayscale image;
keeping the pixel value of a vegetation pixel point in the hyperspectral image unchanged, and setting the pixel value of a soil pixel point to be 0 or 255 to obtain the vegetation hyperspectral image;
and keeping the pixel value of a soil pixel point in the hyperspectral image unchanged, and setting the pixel value of a vegetation pixel point to be 0 or 255 to obtain the soil hyperspectral image.
3. The artificial intelligence based automatic irrigation method according to claim 1, wherein before the vegetation hyperspectral image is identified through a trained vegetation water detection model to obtain vegetation water content, the method further comprises:
constructing a first linear regression model according to a partial least square regression algorithm;
collecting a plurality of leaf samples of vegetation in the irrigation area, and obtaining a plurality of groups of leaf samples with different water contents through a plurality of groups of drying experiments; wherein, the group of drying experiments comprises a plurality of times of drying experiments;
performing hyperspectral image acquisition on each group of blade samples with different water contents through a hyperspectral imaging device, and performing black-and-white board correction on the acquired hyperspectral images;
calculating the average value of the reflection spectrum of each group of blade samples according to the corrected hyperspectral images to obtain hyperspectral data of each group of blade samples; wherein the hyperspectral data comprises 256 characteristic wavelengths;
obtaining a plurality of leaf water content matrixes according to the water content of each group of leaf samples;
obtaining a plurality of blade hyperspectral data matrixes according to the hyperspectral data of each group of blade samples;
and taking the plurality of blade water content matrixes and the plurality of blade hyperspectral data matrixes as training sets, and training the first linear regression model to obtain the vegetation water detection model.
4. The artificial intelligence-based automatic irrigation method according to claim 1, wherein the hyperspectral images are identified through a trained vegetation water detection model to obtain vegetation water content, and the method specifically comprises the following steps:
dividing the vegetation hyperspectral image into N cells, and respectively calculating the average value of the reflection spectrum in each cell to obtain N vegetation hyperspectral data matrixes;
inputting the N vegetation hyperspectral data matrixes into the vegetation moisture detection model to obtain the moisture content corresponding to each cell;
according to
Obtaining a moisture content F1 corresponding to the vegetation hyperspectral image; wherein, f1，f2，……，fNThe water content corresponding to each cell is obtained;
according to
Obtaining vegetation water content G1 in the irrigation area; and M is the number of the collected hyperspectral images.
5. The artificial intelligence-based automatic irrigation method according to claim 1, characterized in that the vegetation hyperspectral image is identified through a trained soil moisture detection model to obtain the soil moisture content, and the method specifically comprises the following steps:
constructing a second linear regression model according to a partial least square regression algorithm;
collecting a plurality of soil samples in the irrigation area, and obtaining a plurality of groups of soil samples with different water contents through a plurality of groups of drying experiments; wherein, the group of drying experiments comprises a plurality of times of drying experiments;
performing hyperspectral image acquisition on each group of soil samples with different water contents through a hyperspectral imaging device;
calculating the full-wave-band albedo of each group of soil samples according to the collected hyperspectral images, and measuring the day and night temperature of the irrigation area;
obtaining the apparent thermal inertia of each group of soil samples based on the full-wave albedo and the day and night temperature;
obtaining a plurality of soil moisture content matrixes according to the moisture content of each group of soil samples;
obtaining a plurality of soil apparent thermal inertia matrixes according to the apparent thermal inertia of each group of soil samples;
taking the plurality of soil moisture content matrixes and the plurality of apparent thermal inertia matrixes as a training set, and training the second linear regression model to obtain the soil moisture detection model;
and inputting the soil hyperspectral image into the soil moisture detection model to obtain the soil moisture content.
6. The artificial intelligence based automatic irrigation method according to claim 1, wherein the irrigation decision system comprises a sensing layer, a network layer and an application layer;
the sensing layer comprises a soil moisture content sensor so as to obtain the actual water content of soil in the field irrigation process;
the network layer is a wireless data transmission channel and is used for realizing a data transmission function between the sensing layer and the application layer;
the application layer comprises an irrigation decision algorithm, the transmitted data are processed through the irrigation decision algorithm, and a corresponding irrigation strategy is output;
the application layer is also used for controlling the solar power generation device and the water pump to execute corresponding actions so as to realize intelligent irrigation and remote monitoring management;
sending the vegetation water content, the soil water content and the forecast precipitation to an irrigation decision system, and judging whether the irrigation area is irrigated or not, wherein the method specifically comprises the following steps:
acquiring the actual moisture content of the soil in the irrigation area through a soil moisture content sensor connected with the irrigation decision system;
and inputting the actual water content of the soil, the vegetation water content, the soil water content and the predicted precipitation into the irrigation decision algorithm in the application layer to obtain the irrigation strategy of the irrigation area.
7. The artificial intelligence based automatic irrigation method according to claim 6, wherein the step of inputting the actual water content of the soil, the water content of the vegetation, the water content of the soil and the predicted precipitation into the irrigation decision algorithm in the application layer to obtain the irrigation strategy of the irrigation area specifically comprises the steps of:
under the condition that the actual water content value of the soil is abnormal, the irrigation area is not irrigated, and first alarm information is sent out; wherein the content of the first warning information indicates that the soil moisture sensor is abnormal;
under the condition that the actual water content value of the soil is normal, comparing the actual water content of the soil with the water content of the soil, if the difference value of the actual water content of the soil and the water content of the soil is larger than a first preset threshold value, not irrigating the irrigation area, and sending second alarm information; the content of the second warning information indicates that the data acquired by the unmanned aerial vehicle or the moisture detection model is abnormal;
if the difference value between the two values is smaller than the first preset threshold value and the predicted precipitation is smaller than a second preset threshold value, irrigating the irrigation area;
and if the difference value of the two values is smaller than the first preset threshold value and the predicted precipitation is larger than or equal to a second preset threshold value, not irrigating the irrigation area.
8. The method according to claim 7, wherein after the irrigation area is irrigated if the difference between the first and second predetermined threshold values is less than the first predetermined threshold value and the predicted precipitation is less than a second predetermined threshold value, the method further comprises:
acquiring the theoretical water demand X of the vegetation in the current growth stage;
according to H ═ 1-omegai) X) Y, and obtaining the current water demand H of all vegetation in the irrigation area; wherein Y is the vegetation number in the irrigated area; omegaiCurrent water demand for the ith vegetation, and 0<i≤Y；
Obtaining irrigation quantity I of the irrigation area according to the I-H-K J-Z; wherein K is the soil water content, J is the area of the irrigation area, and Z is the predicted precipitation;
and controlling an irrigation device to irrigate the irrigation area according to the irrigation quantity.
9. The artificial intelligence based automatic irrigation method according to claim 8, wherein controlling an irrigation device to irrigate the irrigation area according to the irrigation quantity comprises:
connecting a passage between a solar power generation device and a water pump of the irrigation device so that the solar power generation device provides electric energy for the water pump of the irrigation device;
calculating the water yield of a single water outlet through a water flow sensor arranged at each water outlet, and calculating the total water yield of all the water outlets;
and cutting off the passage between the solar power generation device and the water pump after the total water yield reaches the irrigation quantity.
10. An artificial intelligence based automatic irrigation apparatus, characterized in that said apparatus comprises:
at least one processor; and the number of the first and second groups,
a memory communicatively coupled to the at least one processor; wherein,
the memory stores instructions executable by the at least one processor to cause the at least one processor to perform a method for artificial intelligence based automatic irrigation according to any one of claims 1-9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111653810.3A CN114429592A (en) | 2021-12-30 | 2021-12-30 | Automatic irrigation method and equipment based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111653810.3A CN114429592A (en) | 2021-12-30 | 2021-12-30 | Automatic irrigation method and equipment based on artificial intelligence |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114429592A true CN114429592A (en) | 2022-05-03 |
Family
ID=81311517
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111653810.3A Pending CN114429592A (en) | 2021-12-30 | 2021-12-30 | Automatic irrigation method and equipment based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114429592A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115240126A (en) * | 2022-09-23 | 2022-10-25 | 江苏景瑞农业科技发展有限公司 | Intelligent drip irrigation method based on artificial intelligence |
| CN115443889A (en) * | 2022-08-24 | 2022-12-09 | 中国农业大学 | Accurate irrigation method and device for crops |
| CN116050679A (en) * | 2023-04-03 | 2023-05-02 | 中化现代农业有限公司 | Irrigation decision-making method and device and electronic equipment |
| CN116403154A (en) * | 2023-03-06 | 2023-07-07 | 广州市林业和园林科学研究院 | Intelligent monitoring method for combined three-dimensional greening abnormality of landscape architecture |
| CN118235688A (en) * | 2024-05-28 | 2024-06-25 | 江苏龙祥给排水设备有限公司 | Integrated intelligent pump station |
-
2021
- 2021-12-30 CN CN202111653810.3A patent/CN114429592A/en active Pending
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115443889A (en) * | 2022-08-24 | 2022-12-09 | 中国农业大学 | Accurate irrigation method and device for crops |
| CN115240126A (en) * | 2022-09-23 | 2022-10-25 | 江苏景瑞农业科技发展有限公司 | Intelligent drip irrigation method based on artificial intelligence |
| CN115240126B (en) * | 2022-09-23 | 2022-12-20 | 江苏景瑞农业科技发展有限公司 | Intelligent drip irrigation method based on artificial intelligence |
| CN116403154A (en) * | 2023-03-06 | 2023-07-07 | 广州市林业和园林科学研究院 | Intelligent monitoring method for combined three-dimensional greening abnormality of landscape architecture |
| CN116403154B (en) * | 2023-03-06 | 2023-11-03 | 广州市林业和园林科学研究院 | Intelligent monitoring method and system for landscape combined type three-dimensional greening abnormality |
| CN116050679A (en) * | 2023-04-03 | 2023-05-02 | 中化现代农业有限公司 | Irrigation decision-making method and device and electronic equipment |
| CN118235688A (en) * | 2024-05-28 | 2024-06-25 | 江苏龙祥给排水设备有限公司 | Integrated intelligent pump station |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114429592A (en) | Automatic irrigation method and equipment based on artificial intelligence | |
| CN113040034B (en) | Water-saving irrigation control system and control method | |
| CN109583663B (en) | Night water dissolved oxygen amount prediction method suitable for aquaculture pond | |
| Li et al. | Prediction of plant transpiration from environmental parameters and relative leaf area index using the random forest regression algorithm | |
| CN116108318B (en) | Rape nitrogen fertilizer recommended dressing amount calculation method based on unmanned aerial vehicle multispectral image | |
| CN115345076B (en) | Wind speed correction processing method and device | |
| CN113273449A (en) | Digital twin body construction method for precise monitoring of sunlight greenhouse | |
| KR20190136774A (en) | Prediction system for harvesting time of crop and the method thereof | |
| CN117391482B (en) | Greenhouse temperature intelligent early warning method and system based on big data monitoring | |
| Djalilov et al. | DEVELOPMENT OF AUTOMATED INTELLIGENT DRIP IRRIGATION SYSTEM | |
| CN112493100B (en) | Cotton moisture monitoring drip irrigation control method and system based on soil water potential | |
| CN111159640A (en) | Small rain emptying method, system, electronic equipment and storage medium suitable for grid forecast | |
| Yaseen et al. | Smart green farm | |
| CN116452358B (en) | Intelligent agriculture management system based on Internet of things | |
| CN116746463A (en) | Intelligent irrigation system based on weather forecast and soil monitoring | |
| CN116508635A (en) | Agricultural thing networking intelligent water-saving irrigation system | |
| CN113723670B (en) | Photovoltaic power generation power short-term prediction method with variable time window | |
| CN205450843U (en) | Intelligent agriculture remote monitoring device | |
| CN215074427U (en) | Lawn irrigation system | |
| CN116076339B (en) | Intelligent irrigation system for fields | |
| CN116195495B (en) | Method and system for water-saving irrigation based on Internet of things technology | |
| CN110781602A (en) | Method for obtaining space-time continuous soil water based on characteristic space method | |
| CN116883221B (en) | River basin ecology monitoring method | |
| AU2021106558A4 (en) | Solar Moisture and Fertilizer Measurement Recommendation System | |
| Lavanya et al. | Agriculture improvement using IoT |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |