Intelligent decision-making method and system for growth of interplanting of fructus forsythiae and konjak based on image analysis

By using drones equipped with high-definition and multispectral cameras to acquire crop images and combining them with image analysis models, the problem of inaccurate data caused by leaf occlusion has been solved, enabling precise monitoring and management of crop growth status and improving agricultural production efficiency and product quality.

CN118966520AInactive Publication Date: 2024-11-15SHAANXI FENGRUN AGRICULTURE & FORESTRY TECHNOLOGY CO LTD +2
View PDF 0 Cites 3 Cited by

Patent Information

Application Number
CN202410932540.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-11
Publication Date
2024-11-15
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Existing technologies are easily affected by leaf occlusion when acquiring crop images, leading to inaccurate data acquisition and affecting the judgment of crop water and nutrient requirements.

Method used

The system uses drones equipped with high-definition and multispectral cameras to acquire crop images. Combined with image analysis models, it analyzes soil conditions and crop growth to generate irrigation and fertilization instructions.

Benefits of technology

It improves the accuracy of crop data collection, enables timely feedback on the status of planting areas, provides precise agricultural production decision support, and improves production efficiency and agricultural product quality.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118966520A_ABST
    Figure CN118966520A_ABST
Patent Text Reader

Abstract

The invention discloses a forsythia suspensa interplanting konjak growth intelligent decision-making method and system based on image analysis, and particularly relates to the technical field of agricultural planting.The method comprises the steps that soil monitoring data of crops in the nth planting subarea are obtained, and the soil monitoring data comprise a soil humidity evaluation coefficient and a soil fertilizer evaluation coefficient; analyzing the soil monitoring data of the crops in the nth planting sub-region to judge whether a soil state is abnormal or not; if the soil state is abnormal, generating a soil abnormity instruction; if the soil exception instruction is generated, controlling the unmanned aerial vehicle to obtain a crop image set of the planting sub-region corresponding to the soil exception instruction; based on soil monitoring data and crop images, the growth condition of crops is comprehensively analyzed, whether soil state abnormity and crop growth abnormity exist or not is judged, and an abnormity instruction is generated in time; accurate decision support can be provided for agricultural production, and the production efficiency and the quality of agricultural products are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of agricultural planting technology, and more specifically, to an intelligent decision-making method and system for the growth of forsythia interplanted with konjac based on image analysis. Background Art

[0002] Forsythia is a deep-rooted crop that is highly adaptable, light-loving, drought-resistant and barren. It is suitable for growing in warm and humid climates and chooses fertile, loose, well-drained sandy soil. Konjac has a shallow root system and prefers semi-shade and semi-sun. It is suitable for growing in sandy soil like Forsythia. With the development of the agricultural field, agricultural planting methods often face problems such as low utilization of land resources, deterioration of soil quality, and frequent occurrence of diseases and insect pests. In order to solve these problems, people began to explore various innovative planting techniques, including intercropping. Intercropping of konjac and Forsythia refers to planting two different plants on the same piece of land at the same time, with Forsythia as the main crop and konjac as the secondary crop, thereby improving land utilization efficiency, reducing the occurrence of diseases and insect pests, and improving soil quality.

[0003] Existing field management methods can monitor and analyze crop varieties, crop growth status, planting environment parameters, pest and disease conditions, nutrient and water deficiency conditions in planting areas through image-based acquisition. For example, Chinese patent publication number CN115424151A discloses an agricultural intelligent platform based on image processing. Although the above method can improve the efficiency of planting area management and improve the growth quality of konjac and forsythia, the inventors have studied and applied the above method and the prior art and found that when obtaining crop images, a comprehensive photo of the crops is easily affected by the occlusion between the crop leaves, especially when the crop leaves grow densely, resulting in partial obstruction of some leaves and inability to obtain complete image information, which in turn leads to inaccurate data acquisition and affects the judgment of the moisture and nutrient content requirements of the crops.

[0004] To this end, the present invention provides an intelligent decision-making method and system for the growth of forsythia interplanted with konjac based on image analysis. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art, the present invention provides an intelligent decision-making method and system for the growth of forsythia interplanted with konjac based on image analysis to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solution: an intelligent decision-making method for the growth of forsythia interplanted with konjac based on image analysis, comprising:

[0007] Step 1: Acquire soil monitoring data of crops in the nth planting sub-area, wherein the soil monitoring data includes a soil moisture assessment coefficient and a soil fertilizer assessment coefficient;

[0008] Step 2: Analyze the soil monitoring data of the crops in the nth planting sub-area to determine whether there is an abnormal soil state; if there is an abnormal soil state, generate a soil abnormality instruction;

[0009] Step 3: If a soil abnormality instruction is generated, the UAV is controlled to obtain a crop image set of the planting sub-area corresponding to the soil abnormality instruction; the crop image set includes V crop images A and M crop images B, where V and M are both integer sets greater than zero;

[0010] Step 4: Analyze the V crop images A in the crop image set to determine whether to generate irrigation instructions;

[0011] Step 5: Analyze the M crop images B in the crop image set to determine whether to generate a fertilization instruction.

[0012] Furthermore, soil monitoring data of crops in the nth planting sub-area is obtained, including:

[0013] Step a1: within a set monitoring period, soil moisture data of crops are obtained; a humidity fluctuation graph is constructed with the time in the soil moisture data as the horizontal axis and the soil moisture value in the soil moisture data as the vertical axis; the set monitoring period may be hours, days or may be determined by a person skilled in the art according to circumstances;

[0014] Step a2: dividing the humidity fluctuation graph into equal parts according to E humidity intervals to obtain E actual humidity line graphs, wherein the E actual humidity line graphs are combined into an actual humidity line graph set, where E is an integer greater than zero;

[0015] Step a3: extract the e-th actual humidity line graph in the actual humidity line graph set, e∈E, and the initial value of e is 1;

[0016] Step a4: Obtain the humidity range of the crop, extract the standard humidity line graph associated with the humidity range, calculate the similarity between the actual humidity line graph and the standard humidity line graph, if the similarity between the actual humidity line graph and the standard humidity line graph is greater than or equal to the preset humidity similarity threshold, jump to step a5; if the similarity between the actual humidity line graph and the standard humidity line graph is less than the preset humidity similarity threshold, mark the actual humidity line graph as an abnormal humidity line graph, and jump to step a5;

[0017] Step a5: set e=e+1 and jump back to step a3;

[0018] Step a6: repeat the above steps a3 to a5 until e=E, then end the cycle and obtain multiple abnormal humidity line graphs;

[0019] Step a7: The percentage of the number of abnormal humidity line graphs to the total number of actual humidity line graphs is used as the soil humidity assessment coefficient.

[0020] Further, obtaining soil monitoring data of crops in the nth planting sub-area also includes:

[0021] Step b1: within a set monitoring period, obtaining soil fertilizer data of crops; constructing a fertilizer content fluctuation graph with time in the soil fertilizer data as the horizontal axis and soil fertilizer content in the soil fertilizer data as the vertical axis; the soil fertilizer content includes but is not limited to one of nutrient content, pH value and salt content;

[0022] Step b2: dividing the fertilizer content fluctuation graph into equal parts according to F fertilizer content intervals to obtain F actual fertilizer content line graphs, wherein the F actual fertilizer content line graphs are combined into an actual fertilizer content line graph set, where F is an integer greater than zero;

[0023] Step b3: extract the fth actual fertilizer content line graph in the actual fertilizer content line graph set, f∈F, and the initial value of f is 1;

[0024] Step b4: Obtain the fertilizer content range of the crop, extract the standard fertilizer content line graph associated with the fertilizer content range, calculate the similarity between the actual fertilizer content line graph and the standard fertilizer content line graph, if the similarity between the actual fertilizer content line graph and the standard fertilizer content line graph is greater than or equal to the preset fertilizer similarity threshold, jump to step b5; if the similarity between the actual fertilizer content line graph and the standard fertilizer content line graph is less than the preset fertilizer similarity threshold, mark the actual fertilizer content line graph as an abnormal fertilizer content line graph, and jump to step b5;

[0025] Step b5: set f=f+1, and jump back to step b3;

[0026] Step b6: repeat the above steps b3 to b5 until f=F, then end the cycle and obtain multiple abnormal fertilizer content line graphs;

[0027] Step b7: The percentage of the number of abnormal fertilizer content line graphs to the total number of actual fertilizer content line graphs is used as the soil fertilizer evaluation coefficient.

[0028] Furthermore, the soil monitoring data of crops in the nth planting sub-area is analyzed, including:

[0029] Obtain the soil moisture assessment coefficient and soil fertilizer assessment coefficient from the soil monitoring data of crops in the nth planting sub-area, and mark them as Ts respectively. n and Tf n ;

[0030] The soil moisture evaluation coefficient Tsn and soil fertilizer evaluation coefficient Tf n A formula calculation is performed to obtain a soil comprehensive assessment coefficient; the calculation formula of the soil comprehensive assessment coefficient is as follows:

[0031]

[0032] Where, Tp n is the comprehensive soil assessment coefficient of crops in the nth planting sub-area, θ1 is the weight factor of the soil moisture assessment coefficient, and θ2 is the weight factor of the soil fertilizer assessment coefficient.

[0033] Furthermore, judging whether there is abnormal soil condition includes:

[0034] In the nth planting sub-area, there are drones and Q crop monitoring devices to monitor the growth of crops and obtain soil data;

[0035] Preset soil assessment thresholds and compare the soil comprehensive assessment coefficient with the soil assessment threshold range;

[0036] If the soil comprehensive assessment coefficient is within the soil assessment threshold range, no soil abnormality instruction is generated, indicating that there is no soil abnormality in the growth stage of the crop;

[0037] If the soil comprehensive assessment coefficient is outside the soil assessment threshold range and is not equal to 0, a soil abnormality instruction is generated, indicating that there is an abnormal soil state during the growth stage of the crop;

[0038] If the soil comprehensive evaluation coefficient is 0, the system's automatic control system controls the crop monitoring device to start and calculate the soil comprehensive evaluation coefficient again.

[0039] Furthermore, the drone is controlled to obtain a set of crop images of the planting sub-area corresponding to the soil abnormality instruction, including:

[0040] Step c1: In the corresponding planting sub-area, the starting point coordinates and the end point coordinates of the monitored crops and the coordinates of W hovering shooting points are obtained by the drone, where W is an integer set greater than zero;

[0041] Step c2: Control the drone to fly from the starting point coordinates to the hovering shooting point coordinates;

[0042] Step c3: When arriving at the hovering shooting point, the drone is controlled to hover, and at time T, a crop image A is acquired by using a first camera device carried by the drone, where T is an integer set greater than zero;

[0043] Step c4: analyzing the crop image A based on a preconfigured preset image analysis model to determine whether there is an abnormality in the crop image A;

[0044] If there is no abnormality, the crop image B is collected by using the second camera device carried by the drone, and T=T+i is set, and the process returns to step c3, where i is an integer greater than zero;

[0045] If there is an abnormality, the drone is controlled to move to the coordinates of the next hovering shooting point and returns to step c3;

[0046] Step c5: repeat the above steps c3 to c4 until the coordinates of the hovering shooting point are the end coordinates, then end the loop and obtain a crop image set.

[0047] The drone is equipped with a first camera device and a second camera device. The first camera device is a high-definition camera that shoots crops at a 45-degree angle, and the second camera device is a multi-spectral camera that shoots crops at a straight-on angle.

[0048] Furthermore, the crop image A is analyzed based on the preconfigured preset image analysis model, including:

[0049] A historical training image set is obtained, wherein the historical training image set includes abnormal crop images and normal crop images, and labels are annotated on the historical training image set, wherein the label annotation logic is: annotating abnormal crop images with "1" and annotating normal crop images with "0"; the historical training image set is divided into an image training set and an image test set, and a classifier is constructed, wherein the abnormal crop images and normal crop images in the image training set are used as inputs of the classifier, and the annotations in the image training set are used as outputs of the classifier, and the classifier is trained to obtain an initial classifier, and the initial classifier is tested using the image test set, and the initial classifier that meets a preset accuracy is output as an image analysis model; a crop image A is input into the image analysis model to learn whether there is an abnormality in the crop image A; and the classifier is specifically one of a decision tree classification model, a random forest classification model, a support vector machine classification model, a logistic regression network model, or a neural network model.

[0050] Furthermore, the V crop images A in the crop image set are analyzed, including:

[0051] Step d1: extract the vth crop image A from the crop image set, v∈V, and the initial value of v is 1;

[0052] Step d2: obtaining a first sample image corresponding to the crop image A, calculating the similarity between the vth crop image A and the first sample image, if the similarity between the vth crop image A and the first sample image is less than a preset image similarity threshold, jumping to step d3; if the similarity between the vth crop image A and the first sample image is greater than or equal to the preset image similarity threshold, marking the vth crop image A as a first abnormal image, and jumping to step d3;

[0053] Step d3: set v=v+1 and jump back to step d1;

[0054] Step d4: repeat the above steps d1 to d3 until v=V, then end the loop and obtain multiple first abnormal images;

[0055] Step d5: taking the percentage of the number of first abnormal images to the total number of crop images A as the first image evaluation coefficient;

[0056] The first sample image is an image of a crop whose leaves are dry, curled and yellowed due to lack of water, collected by a first camera device and pre-stored in a system database;

[0057] Determine whether to generate an irrigation instruction, including:

[0058] Presetting a first coefficient threshold, and comparing the first image evaluation coefficient with the first coefficient threshold;

[0059] If the first image evaluation coefficient is greater than or equal to a first coefficient threshold, generating an irrigation instruction;

[0060] If the first image evaluation coefficient is less than the first coefficient threshold, no irrigation instruction is generated.

[0061] Furthermore, the M crop images B in the crop image set are analyzed, including:

[0062] Step s1: extract the mth crop image B from the crop image set, m∈M, and the initial value of m is 1;

[0063] Step s2: obtaining a second sample image corresponding to the crop image B, calculating the similarity between the m-th crop image B and the second sample image, if the similarity between the m-th crop image B and the second sample image is less than a preset image similarity threshold, jumping to step s3; if the similarity between the m-th crop image B and the second sample image is greater than or equal to the preset image similarity threshold, marking the m-th crop image B as a second abnormal image, and jumping to step s3;

[0064] Step s3: set m=m+1 and jump back to step s1;

[0065] Step s4: repeat the above steps s1 to s3 until m=M, then end the loop and obtain multiple second abnormal images;

[0066] Step s5: taking the percentage of the number of the second abnormal images to the total number of the crop images B as the second image evaluation coefficient;

[0067] The second sample image is an image of a crop whose stem or fruit is eaten by insect pests, which is collected by a second camera device and is pre-stored in a system database;

[0068] Determine whether to generate a fertilization instruction, including:

[0069] Presetting a second coefficient threshold, and comparing the second image evaluation coefficient with the second coefficient threshold;

[0070] If the second image evaluation coefficient is greater than or equal to the second coefficient threshold, generating a fertilization instruction;

[0071] If the second image evaluation coefficient is less than the second coefficient threshold, no fertilization instruction is generated.

[0072] In a second aspect, the present invention provides an intelligent decision-making system for the growth of forsythia interplanted with konjac based on image analysis, comprising:

[0073] A soil data acquisition module, used to obtain soil monitoring data of crops in the nth planting sub-area, wherein the soil monitoring data includes a soil moisture assessment coefficient and a soil fertilizer assessment coefficient;

[0074] A judgment module is used to analyze the soil monitoring data of the crops in the nth planting sub-area to determine whether there is an abnormal soil state; if there is an abnormal soil state, a soil abnormality instruction is generated;

[0075] An image acquisition module is used to control the UAV to acquire a crop image set of the planting sub-area corresponding to the soil abnormality instruction if a soil abnormality instruction is generated; the crop image set includes V crop images A and M crop images B, where V and M are both integer sets greater than zero;

[0076] The first image analysis module is used to analyze V crop images A in the crop image set to determine whether to generate an irrigation instruction;

[0077] The second image analysis module is used to analyze the M crop images B in the crop image set to determine whether to generate a fertilization instruction.

[0078] Technical effects and advantages of the present invention:

[0079] 1. The present invention first obtains soil monitoring data of crops in the nth planting sub-area, wherein the soil monitoring data includes a soil moisture assessment coefficient and a soil fertilizer assessment coefficient; then the soil monitoring data of crops in the nth planting sub-area is analyzed to determine whether there is an abnormal soil state; if there is an abnormal soil state, a soil abnormality instruction is generated; if the soil abnormality instruction is generated, the drone is controlled to obtain a crop image set of the planting sub-area corresponding to the soil abnormality instruction; the crop image set includes V crop images A and M crop images B, then the V crop images A in the crop image set are analyzed to determine whether an irrigation instruction is generated; finally, the M crop images B in the crop image set are analyzed. The method comprises the following steps: performing analysis to determine whether to generate a fertilization instruction; based on the above steps, the present invention collects soil monitoring data and crop images in the crop planting area through the cooperation between the drone and the crop monitoring device, and can monitor soil moisture, fertilizer content and crop growth in real time, and timely feedback the status of the crop planting area, thereby improving the accuracy of crop data collection; based on the soil monitoring data and crop images, comprehensively analyzing the growth of crops, determining whether there are abnormal soil conditions and abnormal crop growth, and generating abnormal instructions in time; further analyzing the crop images, and generating irrigation instructions and fertilization instructions according to the analysis results, thereby providing accurate decision support for agricultural production and improving production efficiency and the quality of agricultural products.

[0080] 2. The present invention is equipped with a first camera device and a second camera device through a drone. The first camera device uses a high-definition camera to shoot crops at a 45-degree angle, making it easier to capture images of the bottom of crop leaves. The second camera device uses a multi-spectral camera to shoot crops at a straight-on angle, making it easier to capture images of the crop surface. It can also detect pests hiding in hidden areas. By comprehensively analyzing crop images, the growth, pest and disease conditions, and nutrient requirements of crops can be more accurately evaluated, thereby guiding agricultural production management and decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] Figure 1 is a flow chart of the method of Example 1;

[0082] Figure 2 This is a schematic diagram of the planting layout of Example 1;

[0083] Figure 3 is a system schematic diagram of Example 2;

[0084] Figure 4 This is a schematic diagram of an electronic device according to Embodiment 3;

[0085] Figure 5 A schematic diagram of a computer-readable storage medium of Example 4. DETAILED DESCRIPTION

[0086] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0087] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0088] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, a first unit may be referred to as a second unit, and a similar second unit may be referred to as a first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0089] Example 1

[0090] See also Figure 1 As shown, this embodiment discloses an intelligent decision-making method for the growth of forsythia interplanted with konjac based on image analysis:

[0091] Step 1: Acquire soil monitoring data of crops in the nth planting sub-area, wherein the soil monitoring data includes a soil moisture assessment coefficient and a soil fertilizer assessment coefficient;

[0092] See also Figure 2As shown, it should be noted that: the nth planting sub-area is obtained from the target planting area according to a specific planting layout; specifically, in the nth planting sub-area, the planting density of Forsythia suspensa is 3 meters in row spacing and 2.5 meters in plant spacing; an earth embankment (gray area in the image) is formed between the Forsythia suspensa planting rows, and commercial taro konjac and breeding seed taro are planted alternately in the earth embankment; in this planting layout, konjac and Forsythia suspensa with complementary characteristics are interplanted, and intelligent field management is realized in combination with information technology, and land resources are fully utilized to improve land utilization efficiency; it should be supplemented that: commercial taro konjac refers to the main konjac variety used for commercial purposes, while breeding seed taro refers to the variety used to breed and plant the next season's konjac, and they play different roles in the interplanting of Forsythia suspensa and konjac, and jointly promote the maximum utilization of farmland resources and the increase of konjac yield.

[0093] It should be understood that: a drone and Q crop monitoring devices are provided in the nth planting sub-area for monitoring the growth of crops and obtaining soil data; each planting sub-area is serially labeled, named and sorted, and a flight route is set for each sorted planting sub-area, so that the drone collects data along a predetermined route to improve the data collection efficiency of the drone; each crop monitoring device includes a plurality of sensors, including but not limited to humidity sensors and soil testers, etc. The type and monitoring position of the crop monitoring device are determined by those skilled in the art according to the situation, and are not limited to too much here. The crop in this embodiment refers to one of konjac or forsythia;

[0094] Specifically, the soil monitoring data of crops in the nth planting sub-area is obtained, including:

[0095] Step a1: within a set monitoring period, soil moisture data of crops are obtained; a humidity fluctuation graph is constructed with the time in the soil moisture data as the horizontal axis and the soil moisture value in the soil moisture data as the vertical axis; the set monitoring period may be hours, days or may be determined by a person skilled in the art according to circumstances;

[0096] Step a2: dividing the humidity fluctuation graph into equal parts according to E humidity intervals to obtain E actual humidity line graphs, wherein the E actual humidity line graphs are combined into an actual humidity line graph set, where E is an integer greater than zero;

[0097] Step a3: extract the e-th actual humidity line graph in the actual humidity line graph set, e∈E, and the initial value of e is 1;

[0098] Step a4: Obtain the humidity range of the crop, extract the standard humidity line graph associated with the humidity range, calculate the similarity between the actual humidity line graph and the standard humidity line graph, if the similarity between the actual humidity line graph and the standard humidity line graph is greater than or equal to the preset humidity similarity threshold, jump to step a5; if the similarity between the actual humidity line graph and the standard humidity line graph is less than the preset humidity similarity threshold, mark the actual humidity line graph as an abnormal humidity line graph, and jump to step a5;

[0099] Step a5: set e=e+1 and jump back to step a3;

[0100] Step a6: repeat the above steps a3 to a5 until e=E, then end the cycle and obtain multiple abnormal humidity line graphs;

[0101] Step a7: taking the percentage of the number of abnormal humidity line graphs to the total number of actual humidity line graphs as the soil humidity assessment coefficient;

[0102] Obtain soil monitoring data for crops in the nth planting sub-area, including:

[0103] Step b1: within a set monitoring period, obtaining soil fertilizer data of crops; constructing a fertilizer content fluctuation graph with time in the soil fertilizer data as the horizontal axis and soil fertilizer content in the soil fertilizer data as the vertical axis; the soil fertilizer content includes but is not limited to one of nutrient content, pH value and salt content;

[0104] Step b2: dividing the fertilizer content fluctuation graph into equal parts according to F fertilizer content intervals to obtain F actual fertilizer content line graphs, wherein the F actual fertilizer content line graphs are combined into an actual fertilizer content line graph set, where F is an integer greater than zero;

[0105] Step b3: extract the fth actual fertilizer content line graph in the actual fertilizer content line graph set, f∈F, and the initial value of f is 1;

[0106] Step b4: Obtain the fertilizer content range of the crop, extract the standard fertilizer content line graph associated with the fertilizer content range, calculate the similarity between the actual fertilizer content line graph and the standard fertilizer content line graph, if the similarity between the actual fertilizer content line graph and the standard fertilizer content line graph is greater than or equal to the preset fertilizer similarity threshold, jump to step b5; if the similarity between the actual fertilizer content line graph and the standard fertilizer content line graph is less than the preset fertilizer similarity threshold, mark the actual fertilizer content line graph as an abnormal fertilizer content line graph, and jump to step b5;

[0107] Step b5: set f=f+1, and jump back to step b3;

[0108] Step b6: repeat the above steps b3 to b5 until f=F, then end the cycle and obtain multiple abnormal fertilizer content line graphs;

[0109] Step b7: taking the percentage of the number of abnormal fertilizer content line graphs to the total number of actual fertilizer content line graphs as the soil fertilizer evaluation coefficient;

[0110] Step 2: Analyze the soil monitoring data of the crops in the nth planting sub-area to determine whether there is an abnormal soil state; if there is an abnormal soil state, generate a soil abnormality instruction;

[0111] Analyze the soil monitoring data of crops in the nth planting sub-area, including:

[0112] Obtain the soil moisture assessment coefficient and soil fertilizer assessment coefficient from the soil monitoring data of crops in the nth planting sub-area, and mark them as Ts respectively. n and Tf n ;

[0113] The soil moisture evaluation coefficient Ts n and soil fertilizer evaluation coefficient Tf n A formula calculation is performed to obtain a soil comprehensive assessment coefficient; the calculation formula of the soil comprehensive assessment coefficient is as follows:

[0114]

[0115] Where, Tp n is the soil comprehensive assessment coefficient of crops in the nth planting sub-area, θ1 is the weight factor of the soil moisture assessment coefficient, and θ2 is the weight factor of the soil fertilizer assessment coefficient;

[0116] Determine whether there is any abnormal soil condition, including:

[0117] Preset soil assessment thresholds and compare the soil comprehensive assessment coefficient with the soil assessment threshold range;

[0118] If the soil comprehensive assessment coefficient is within the soil assessment threshold range, no soil abnormality instruction is generated, indicating that there is no soil abnormality in the growth stage of the crop;

[0119] If the soil comprehensive assessment coefficient is outside the soil assessment threshold range and is not equal to 0, a soil abnormality instruction is generated, indicating that there is an abnormal soil state during the growth stage of the crop;

[0120] If the soil comprehensive evaluation coefficient is 0, it means that the crop monitoring device is not turned on, and the system's automatic control system controls the crop monitoring device to turn on and calculate the soil comprehensive evaluation coefficient again;

[0121] It should be noted that: the soil assessment threshold range is that the staff collected the crop monitoring device data multiple times during the historical normal working stage of the crop monitoring device, wherein multiple groups of crop monitoring device data were collected at one time, the average of the multiple groups of crop monitoring device data was calculated, the soil comprehensive assessment coefficient corresponding to the average of the crop monitoring device data was calculated, the multiple soil comprehensive assessment coefficients were sorted, the maximum value of the multiple soil comprehensive assessment coefficients was taken as the maximum value of the soil assessment threshold, and the minimum value of the multiple soil comprehensive assessment coefficients was taken as the minimum value of the soil assessment threshold; the range between the minimum value of the soil assessment threshold and the maximum value of the soil assessment threshold is the soil assessment threshold range;

[0122] It is further explained that it is not accurate enough to judge whether there is abnormal soil state of crops only through soil monitoring data. The reason is that due to the influence of sensor failure or external signal interference in the collection process of soil monitoring data, the collected soil monitoring data is abnormal, which leads to inaccurate judgment; therefore, other operations are required to accurately judge whether there is abnormal growth state in the growth stage of crops;

[0123] Step 3: If a soil abnormality instruction is generated, the UAV is controlled to obtain a crop image set of the planting sub-area corresponding to the soil abnormality instruction;

[0124] Control the drone to obtain the crop image set of the planting sub-area corresponding to the soil abnormality instruction, including:

[0125] Step c1: In the corresponding planting sub-area, the starting point coordinates and the end point coordinates of the monitored crops and the coordinates of W hovering shooting points are obtained by the drone, where W is an integer set greater than zero;

[0126] It should be noted that the starting point coordinates, the end point coordinates and the hovering shooting point coordinates of the monitored crops are actually calibrated by those skilled in the art according to the conditions of the monitored crops, and no further limitation is made thereto;

[0127] Step c2: Control the drone to fly from the starting point coordinates to the hovering shooting point coordinates;

[0128] Step c3: When arriving at the hovering shooting point, the drone is controlled to hover, and at time T, a crop image A is acquired by using a first camera device carried by the drone, where T is an integer set greater than zero;

[0129] Step c4: analyzing the crop image A based on a preconfigured preset image analysis model to determine whether there is an abnormality in the crop image A;

[0130] If there is no abnormality, the crop image B is collected by using the second camera device carried by the drone, and T=T+i is set, and the process returns to step c3, where i is an integer greater than zero;

[0131] If there is an abnormality, the drone is controlled to move to the coordinates of the next hovering shooting point and returns to step c3;

[0132] It should be noted that: the drone is equipped with a first camera device and a second camera device, the first camera device is a high-definition camera that shoots crops at a 45-degree angle, and the second camera device is a multi-spectral camera that shoots crops at a straight-on angle. Further explanation is that since the high-definition camera shoots at a 45-degree angle, it is easier to capture images of the bottom of crop leaves for preliminary detection of whether there are any abnormalities in the crops, such as leaf damage or leaf discoloration. If there are no abnormalities, the crop image B is collected by the second camera device. This is because pests may hide in more concealed areas, such as the junction of stems and leaves. Therefore, this embodiment uses an algorithm to determine and control the timing of collecting crop image B, which is conducive to avoiding useless crop image data;

[0133] It should also be noted that the crop image A is analyzed based on the preconfigured preset image analysis model, including:

[0134] Acquire a historical training image set, wherein the historical training image set includes abnormal crop images and normal crop images, and label the historical training image set, wherein the label labeling logic is: label the abnormal crop images as "1", and label the normal crop images as "0", "1" indicates that there is an abnormality in the image, that is, it indicates that the crop corresponding to the crop image A has leaf damage or leaf discoloration, and "0" indicates that there is no abnormality in the image, that is, it indicates that the crop corresponding to the crop image A has no leaf damage or leaf discoloration; divide the historical training image set into an image training set and an image test set, construct a classifier, use the abnormal crop images and normal crop images in the image training set as inputs of the classifier, use the labels in the image training set as outputs of the classifier, train the classifier to obtain an initial classifier, test the initial classifier using the image test set, and output the initial classifier that meets a preset accuracy as an image analysis model; input the crop image A into the image analysis model to learn whether there is an abnormality in the crop image A; the classifier is specifically one of a decision tree classification model, a random forest classification model, a support vector machine classification model, a logistic regression network model, or a neural network model;

[0135] Step c5: repeating the above steps c3 to c4 until the coordinates of the hovering shooting point are the end coordinates, ending the loop and obtaining a crop image set; the crop image set includes V crop images A and M crop images B, where V and M are both integer sets greater than zero;

[0136] Step 4: Analyze the V crop images A in the crop image set to determine whether to generate irrigation instructions;

[0137] Specifically, analyzing V crop images A in the crop image set includes:

[0138] Step d1: extract the vth crop image A from the crop image set, v∈V, and the initial value of v is 1;

[0139] Step d2: obtaining a first sample image corresponding to the crop image A, calculating the similarity between the vth crop image A and the first sample image, if the similarity between the vth crop image A and the first sample image is less than a preset image similarity threshold, jumping to step d3; if the similarity between the vth crop image A and the first sample image is greater than or equal to the preset image similarity threshold, marking the vth crop image A as a first abnormal image, and jumping to step d3;

[0140] Step d3: set v=v+1 and jump back to step d1;

[0141] Step d4: repeat the above steps d1 to d3 until v=V, then end the loop and obtain multiple first abnormal images;

[0142] Step d5: taking the percentage of the number of first abnormal images to the total number of crop images A as the first image evaluation coefficient;

[0143] It should be noted that the first sample image is an image of a crop whose leaves are dry, curled and yellowed due to lack of water, which is collected by a first camera device and is pre-stored in a system database; it should also be noted that before analyzing the V crop images A in the crop image set, the crop images A need to be pre-processed, and the pre-processing includes but is not limited to flipping, scaling, cropping and image denoising, etc., so that they are in the same dimension as the first sample image;

[0144] Specifically, determining whether to generate an irrigation instruction includes:

[0145] Presetting a first coefficient threshold, and comparing the first image evaluation coefficient with the first coefficient threshold;

[0146] If the first image evaluation coefficient is greater than or equal to the first coefficient threshold, an irrigation instruction is generated, indicating that the crops in the corresponding planting sub-area need to be irrigated;

[0147] If the first image evaluation coefficient is less than the first coefficient threshold, no irrigation instruction is generated, indicating that the crops in the corresponding planting sub-area do not need to be irrigated temporarily, and the crops continue to be monitored.

[0148] Step 5: Analyze the M crop images B in the crop image set to determine whether to generate a fertilization instruction;

[0149] Specifically, the M crop images B in the crop image set are analyzed, including:

[0150] Step s1: extract the mth crop image B from the crop image set, m∈M, and the initial value of m is 1;

[0151] Step s2: obtaining a second sample image corresponding to the crop image B, calculating the similarity between the m-th crop image B and the second sample image, if the similarity between the m-th crop image B and the second sample image is less than a preset image similarity threshold, jumping to step s3; if the similarity between the m-th crop image B and the second sample image is greater than or equal to the preset image similarity threshold, marking the m-th crop image B as a second abnormal image, and jumping to step s3;

[0152] Step s3: set m=m+1 and jump back to step s1;

[0153] Step s4: repeat the above steps s1 to s3 until m=M, then end the loop and obtain multiple second abnormal images;

[0154] Step s5: taking the percentage of the number of the second abnormal images to the total number of the crop images B as the second image evaluation coefficient;

[0155] It should be noted that: the second sample image is an image of a crop in which the stem or fruit of the crop is eaten by pests, which is collected by the second camera device and is pre-stored in the system database; it should also be noted that before analyzing the M crop images B in the crop image set, the crop images B need to be pre-processed, and the pre-processing includes but is not limited to flipping, scaling, cropping and image denoising, etc., so that they are in the same dimension as the second sample image;

[0156] Specifically, determining whether to generate a fertilization instruction includes:

[0157] Presetting a second coefficient threshold, and comparing the second image evaluation coefficient with the second coefficient threshold;

[0158] If the second image evaluation coefficient is greater than or equal to the second coefficient threshold, a fertilization instruction is generated, indicating that the crops in the corresponding planting sub-area need to be fertilized;

[0159] If the second image evaluation coefficient is less than the second coefficient threshold, no fertilization instruction is generated, indicating that the crops in the corresponding planting sub-area do not need to be fertilized temporarily, and the crops continue to be monitored.

[0160] The formulas involved in the above are all calculated by removing dimensions and taking their numerical values. They are a formula that is closest to the actual situation obtained by collecting a large amount of data and performing software simulation. The weight factors in the formula and the preset thresholds in the analysis process are set by technicians in this field according to actual conditions or obtained by simulating a large amount of data; the size of the weight factor is to quantify each parameter to obtain a specific value for subsequent comparison. The size of the weight factor depends on the amount of sample data and the corresponding processing coefficient initially set by technicians in this field for each group of sample data; as long as it does not affect the proportional relationship between the parameter and the quantized value.

[0161] This embodiment first obtains soil monitoring data of crops in the nth planting sub-area, wherein the soil monitoring data includes a soil moisture assessment coefficient and a soil fertilizer assessment coefficient; then, the soil monitoring data of crops in the nth planting sub-area is analyzed to determine whether there is an abnormal soil state; if there is an abnormal soil state, a soil abnormality instruction is generated; if the soil abnormality instruction is generated, the drone is controlled to obtain a crop image set of the planting sub-area corresponding to the soil abnormality instruction; the crop image set includes V crop images A and M crop images B, and then the V crop images A in the crop image set are analyzed to determine whether to generate an irrigation instruction; finally, the M crop images B in the crop image set are analyzed. Analysis, determine whether to generate a fertilization instruction; based on the above steps, this embodiment cooperates with the drone and the crop monitoring device to collect soil monitoring data and crop images in the crop planting area, which can monitor soil moisture, fertilizer content and crop growth in real time, and timely feedback the status of the crop planting area, thereby improving the accuracy of crop data collection; based on the soil monitoring data and crop images, comprehensively analyze the growth of crops, determine whether there are abnormal soil conditions and abnormal crop growth, and generate abnormal instructions in time; further analyze the crop images, and generate irrigation instructions and fertilization instructions according to the analysis results, which can provide accurate decision support for agricultural production and improve production efficiency and the quality of agricultural products.

[0162] In this embodiment, the drone is equipped with a first camera device and a second camera device. The first camera device uses a high-definition camera to shoot crops at a 45-degree angle, making it easier to capture images of the bottom of crop leaves. The second camera device uses a multi-spectral camera to shoot crops at a straight-on angle, making it easier to capture images of the crop surface. It can also detect pests hiding in hidden areas. By comprehensively analyzing crop images, the growth, pest and disease conditions, and nutrient requirements of crops can be more accurately evaluated, thereby guiding agricultural production management and decision-making.

[0163] Example 2

[0164] See also Figure 3As shown, this embodiment provides an intelligent decision-making system for the growth of forsythia interplanted with konjac based on image analysis, including:

[0165] A soil data acquisition module, used to obtain soil monitoring data of crops in the nth planting sub-area, wherein the soil monitoring data includes a soil moisture assessment coefficient and a soil fertilizer assessment coefficient;

[0166] A judgment module is used to analyze the soil monitoring data of the crops in the nth planting sub-area to determine whether there is an abnormal soil state; if there is an abnormal soil state, a soil abnormality instruction is generated;

[0167] An image acquisition module is used to control the UAV to acquire a crop image set of the planting sub-area corresponding to the soil abnormality instruction if a soil abnormality instruction is generated; the crop image set includes V crop images A and M crop images B, where V and M are both integer sets greater than zero;

[0168] The first image analysis module is used to analyze V crop images A in the crop image set to determine whether to generate an irrigation instruction;

[0169] The second image analysis module is used to analyze the M crop images B in the crop image set to determine whether to generate a fertilization instruction.

[0170] Example 3

[0171] See also Figure 4 As shown, this embodiment provides an electronic device, including: a processor and a memory, wherein the memory stores a computer program that can be called by the processor;

[0172] The processor executes the intelligent decision-making method for interplanting konjac with Forsythia suspensa based on image analysis of Example 1 by calling the computer program stored in the memory.

[0173] Example 4

[0174] See also Figure 5 As shown, this embodiment provides a computer-readable storage medium storing instructions. When the instructions are executed on a computer, the computer executes the intelligent decision-making method for the growth of forsythia interplanted with konjac based on image analysis of Example 1.

[0175] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0176] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An intelligent decision-making method for the growth of forsythia interplanted with konjac based on image analysis, characterized in that: include: Step 1: Acquire soil monitoring data of crops in the nth planting sub-area, wherein the soil monitoring data includes a soil moisture assessment coefficient and a soil fertilizer assessment coefficient; Step 2: Analyze the soil monitoring data of the crops in the nth planting sub-area to determine whether there is an abnormal soil state; if there is an abnormal soil state, generate a soil abnormality instruction; Step 3: If a soil abnormality instruction is generated, the UAV is controlled to obtain a crop image set of the planting sub-area corresponding to the soil abnormality instruction; the crop image set includes V crop images A and M crop images B, where V and M are both integer sets greater than zero; Step 4: Analyze the V crop images A in the crop image set to determine whether to generate irrigation instructions; Step 5: Analyze the M crop images B in the crop image set to determine whether to generate a fertilization instruction.

2. The intelligent decision-making method for the growth of forsythia interplanted with konjac based on image analysis according to claim 1, characterized in that, The step of obtaining soil monitoring data of crops in the nth planting sub-area includes: Step a1: within a set monitoring period, soil moisture data of crops is obtained; a humidity fluctuation graph is constructed with the time in the soil moisture data as the horizontal axis and the soil moisture value in the soil moisture data as the vertical axis; the set monitoring period is hours or days; Step a2: dividing the humidity fluctuation graph into equal parts according to E humidity intervals to obtain E actual humidity line graphs, wherein the E actual humidity line graphs are combined into an actual humidity line graph set, where E is an integer greater than zero; Step a3: extract the e-th actual humidity line graph in the actual humidity line graph set, e∈E, and the initial value of e is 1; Step a4: Obtain the humidity range of the crop, extract the standard humidity line graph associated with the humidity range, calculate the similarity between the actual humidity line graph and the standard humidity line graph, if the similarity between the actual humidity line graph and the standard humidity line graph is greater than or equal to the preset humidity similarity threshold, jump to step a5; if the similarity between the actual humidity line graph and the standard humidity line graph is less than the preset humidity similarity threshold, mark the actual humidity line graph as an abnormal humidity line graph, and jump to step a5; Step a5: set e=e+1 and jump back to step a3; Step a6: repeat the above steps a3 to a5 until e=E, then end the cycle and obtain multiple abnormal humidity line graphs; Step a7: The percentage of the number of abnormal humidity line graphs to the total number of actual humidity line graphs is used as the soil humidity assessment coefficient.

3. The intelligent decision-making method for the growth of forsythia interplanted with konjac based on image analysis according to claim 2, characterized in that, Obtain soil monitoring data for crops in the nth planting sub-area, including: Step b1: within a set monitoring period, obtaining soil fertilizer data of crops; constructing a fertilizer content fluctuation graph with time in the soil fertilizer data as the horizontal axis and soil fertilizer content in the soil fertilizer data as the vertical axis; the soil fertilizer content includes but is not limited to one of nutrient content, pH value and salt content; Step b2: dividing the fertilizer content fluctuation graph into equal parts according to F fertilizer content intervals to obtain F actual fertilizer content line graphs, wherein the F actual fertilizer content line graphs are combined into an actual fertilizer content line graph set, where F is an integer greater than zero; Step b3: extract the fth actual fertilizer content line graph in the actual fertilizer content line graph set, f∈F, and the initial value of f is 1; Step b4: Obtain the fertilizer content range of the crop, extract the standard fertilizer content line graph associated with the fertilizer content range, calculate the similarity between the actual fertilizer content line graph and the standard fertilizer content line graph, if the similarity between the actual fertilizer content line graph and the standard fertilizer content line graph is greater than or equal to the preset fertilizer similarity threshold, jump to step b5; if the similarity between the actual fertilizer content line graph and the standard fertilizer content line graph is less than the preset fertilizer similarity threshold, mark the actual fertilizer content line graph as an abnormal fertilizer content line graph, and jump to step b5; Step b5: set f=f+1, and jump back to step b3; Step b6: repeat the above steps b3 to b5 until f=F, then end the cycle and obtain multiple abnormal fertilizer content line graphs; Step b7: The percentage of the number of abnormal fertilizer content line graphs to the total number of actual fertilizer content line graphs is used as the soil fertilizer evaluation coefficient.

4. The intelligent decision-making method for the growth of forsythia interplanted with konjac based on image analysis according to claim 3, characterized in that, Analyze the soil monitoring data of crops in the nth planting sub-area, including: Obtain the soil moisture assessment coefficient and soil fertilizer assessment coefficient from the soil monitoring data of crops in the nth planting sub-area, and mark them as Ts respectively. n and Tf n ; The soil moisture evaluation coefficient Ts n and soil fertilizer evaluation coefficient Tf n A formula calculation is performed to obtain a soil comprehensive assessment coefficient; the calculation formula of the soil comprehensive assessment coefficient is as follows: Where, Tp n is the comprehensive soil assessment coefficient of crops in the nth planting sub-area, θ1 is the weight factor of the soil moisture assessment coefficient, and θ2 is the weight factor of the soil fertilizer assessment coefficient.

5. The intelligent decision-making method for the growth of forsythia interplanted with konjac based on image analysis according to claim 4, characterized in that, Determine whether there is any abnormal soil condition, including: In the nth planting sub-area, there are drones and Q crop monitoring devices to monitor the growth of crops and obtain soil data; Preset soil assessment thresholds and compare the soil comprehensive assessment coefficient with the soil assessment threshold range; If the soil comprehensive assessment coefficient is within the soil assessment threshold range, no soil abnormality instruction is generated, indicating that there is no soil abnormality in the growth stage of the crop; If the soil comprehensive assessment coefficient is outside the soil assessment threshold range and is not equal to 0, a soil abnormality instruction is generated, indicating that there is an abnormal soil state during the growth stage of the crop; If the soil comprehensive evaluation coefficient is 0, the system's automatic control system controls the crop monitoring device to start and calculate the soil comprehensive evaluation coefficient again.

6. The intelligent decision-making method for the growth of forsythia interplanted with konjac based on image analysis according to claim 5, characterized in that, The controlling UAV to obtain a crop image set of a planting sub-area corresponding to a soil abnormality instruction includes: Step c1: In the corresponding planting sub-area, the starting point coordinates and the end point coordinates of the monitored crops and the coordinates of W hovering shooting points are obtained by the drone, where W is an integer set greater than zero; Step c2: Control the drone to fly from the starting point coordinates to the hovering shooting point coordinates; Step c3: When arriving at the hovering shooting point, the drone is controlled to hover, and at time T, a crop image A is acquired by using a first camera device carried by the drone, where T is an integer set greater than zero; Step c4: analyzing the crop image A based on a preconfigured preset image analysis model to determine whether there is an abnormality in the crop image A; If there is no abnormality, the crop image B is collected by using the second camera device carried by the drone, and T=T+i is set, and the process returns to step c3, where i is an integer greater than zero; If there is an abnormality, the drone is controlled to move to the coordinates of the next hovering shooting point and returns to step c3; Step c5: repeating the above steps c3 to c4 until the coordinates of the hovering shooting point are the end coordinates, ending the loop and obtaining a crop image set; The drone is equipped with a first camera device and a second camera device. The first camera device is a high-definition camera that shoots crops at a 45-degree angle, and the second camera device is a multi-spectral camera that shoots crops at a straight-on angle.

7. The intelligent decision-making method for the growth of forsythia interplanted with konjac based on image analysis according to claim 6, characterized in that, The analyzing the crop image A based on the preconfigured preset image analysis model includes: A historical training image set is obtained, wherein the historical training image set includes abnormal crop images and normal crop images, and labels are annotated on the historical training image set, wherein the label annotation logic is: annotating abnormal crop images with "1" and annotating normal crop images with "0"; the historical training image set is divided into an image training set and an image test set, and a classifier is constructed, wherein the abnormal crop images and normal crop images in the image training set are used as inputs of the classifier, and the annotations in the image training set are used as outputs of the classifier, and the classifier is trained to obtain an initial classifier, and the initial classifier is tested using the image test set, and the initial classifier that meets a preset accuracy is output as an image analysis model; a crop image A is input into the image analysis model to learn whether there is an abnormality in the crop image A; and the classifier is specifically one of a decision tree classification model, a random forest classification model, a support vector machine classification model, a logistic regression network model, or a neural network model.

8. The intelligent decision-making method for the growth of forsythia interplanted with konjac based on image analysis according to claim 7, characterized in that, The analyzing V crop images A in the crop image set includes: Step d1: extract the vth crop image A from the crop image set, v∈V, and the initial value of v is 1; Step d2: obtaining a first sample image corresponding to the crop image A, calculating the similarity between the vth crop image A and the first sample image, if the similarity between the vth crop image A and the first sample image is less than a preset image similarity threshold, jumping to step d3; if the similarity between the vth crop image A and the first sample image is greater than or equal to the preset image similarity threshold, marking the vth crop image A as a first abnormal image, and jumping to step d3; Step d3: set v=v+1 and jump back to step d1; Step d4: repeat the above steps d1 to d3 until v=V, then end the loop and obtain multiple first abnormal images; Step d5: taking the percentage of the number of first abnormal images to the total number of crop images A as the first image evaluation coefficient; The first sample image is an image of a crop whose leaves are dry, curled and yellowed due to lack of water, collected by a first camera device and pre-stored in a system database; The determining whether to generate an irrigation instruction comprises: Presetting a first coefficient threshold, and comparing the first image evaluation coefficient with the first coefficient threshold; If the first image evaluation coefficient is greater than or equal to a first coefficient threshold, generating an irrigation instruction; If the first image evaluation coefficient is less than the first coefficient threshold, no irrigation instruction is generated.

9. The intelligent decision-making method for the growth of forsythia interplanted with konjac based on image analysis according to claim 8, characterized in that, The analyzing the M crop images B in the crop image set includes: Step s1: extract the mth crop image B from the crop image set, m∈M, and the initial value of m is 1; Step s2: obtaining a second sample image corresponding to the crop image B, calculating the similarity between the m-th crop image B and the second sample image, if the similarity between the m-th crop image B and the second sample image is less than a preset image similarity threshold, jumping to step s3; if the similarity between the m-th crop image B and the second sample image is greater than or equal to the preset image similarity threshold, marking the m-th crop image B as a second abnormal image, and jumping to step s3; Step s3: set m=m+1 and jump back to step s1; Step s4: repeat the above steps s1 to s3 until m=M, then end the loop and obtain multiple second abnormal images; Step s5: taking the percentage of the number of the second abnormal images to the total number of the crop images B as the second image evaluation coefficient; The second sample image is an image of a crop whose stem or fruit is eaten by insect pests, which is collected by a second camera device and is pre-stored in a system database; The determining whether to generate a fertilization instruction includes: Presetting a second coefficient threshold, and comparing the second image evaluation coefficient with the second coefficient threshold; If the second image evaluation coefficient is greater than or equal to the second coefficient threshold, generating a fertilization instruction; If the second image evaluation coefficient is less than the second coefficient threshold, no fertilization instruction is generated.

10. An intelligent decision-making system for the growth of forsythia interplanted with konjac based on image analysis, used for implementing the intelligent decision-making method for the growth of forsythia interplanted with konjac based on image analysis according to any one of claims 1 to 9, characterized in that: include: A soil data acquisition module, used to obtain soil monitoring data of crops in the nth planting sub-area, wherein the soil monitoring data includes a soil moisture assessment coefficient and a soil fertilizer assessment coefficient; A judgment module is used to analyze the soil monitoring data of crops in the nth planting sub-area to determine whether there is an abnormal soil state; If there is an abnormal soil state, a soil abnormality instruction is generated; An image acquisition module is used to control the UAV to acquire a crop image set of the planting sub-area corresponding to the soil abnormality instruction if a soil abnormality instruction is generated; the crop image set includes V crop images A and M crop images B, where V and M are both integer sets greater than zero; The first image analysis module is used to analyze V crop images A in the crop image set to determine whether to generate an irrigation instruction; The second image analysis module is used to analyze the M crop images B in the crop image set to determine whether to generate a fertilization instruction.

Citation Information

Cited By

  • Accurate data acquisition optimization method and system for farmland low-carbon production

    CN119338132A

  • Konjak continuous cropping obstacle prevention and soil improvement system based on multi-factor analysis

    CN119623716A

  • A Multifactor Analysis-Based System for Preventing Continuous Cropping Obstacles in Konjac and Improving Soil

    CN119623716B