Dual-axis interactive method, device, equipment and medium based on crop growth

By constructing a growth trend estimate model, adjusting the time span according to changes in crop species and time axis, generating crop growth trends, solving the problem of cumbersome operations in the existing technology, and improving the interactive experience and efficiency of crop growth cycle query.

CN114239716BActive Publication Date: 2024-11-08PING AN TECH (SHENZHEN) CO LTD
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202111536680.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-12-15
Publication Date
2024-11-08
Estimated Expiration
2041-12-15

AI Technical Summary

Technical Problem

The existing crop growth data interaction methods are mainly limited to the NDVI value of the timeline or a single growth cycle. It is impossible to observe the NDVI value of the crop growth cycle throughout the year at the same time. The operation steps are cumbersome, which increases the user's behavioral burden.

Method used

By constructing a growth trend estimate model, the time spans of the growth, harvest and sowing period are adjusted according to the types and timeline changes of the crops to be tested, the final growth trend is generated, and the differences are displayed in a chart manner to simplify the operation process.

Benefits of technology

It has achieved rapid simulation of crop growth cycles and trends, reduced operation steps, improved the interactive experience of crop growth cycle query, and met the needs of different users.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114239716B_ABST
    Figure CN114239716B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of artificial intelligence, and proposes a two-axis interactive method, device, equipment and medium based on the growth of crops. The method comprises: obtaining the type of crops to be tested; using a preset growth trend prediction model to obtain the initial growth trend of the crops to be tested corresponding to the growth period, harvest period and sowing period within a preset time period according to the type of crops to be tested; obtaining the time span on the time axis corresponding to each growth cycle of the crops to be tested; adjusting the time period corresponding to the growth period, harvest period and sowing period of the crops to be tested on the time axis according to the time span of the crops to be tested, updating the initial growth trend corresponding to the crops to be tested in each growth cycle according to the time period to generate a final growth trend, and displaying the difference in a graphical manner. The present invention is not only simple to operate, but also can meet the needs of various different users, and greatly improves the interactive experience of determining the growth of crops.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence, and provides a dual-axis interaction method, device, equipment and medium based on the growth of crops. Background Art

[0002] With the development of remote sensing technology, the use of remote sensing images to conduct large-scale typical crop research and quickly and accurately obtain the spatial distribution characteristics of farmland and crop planting categories has extremely important practical significance for promoting agricultural modernization and maintaining regional ecological stability and sustainable development. In the field of remote sensing applications, the vegetation index NDVI value has been widely used to qualitatively and quantitatively evaluate vegetation coverage and its growth vitality, because the vegetation spectrum is a complex mixed reaction of vegetation, soil brightness, environmental impact, shadow, soil color and humidity, and is affected by atmospheric space and temporal changes.

[0003] However, existing methods for interacting with crop growth data are mostly limited to the timeline or the NDVI value of a single crop growth cycle. It is impossible to simultaneously observe the NDVI values ​​throughout the year and the entire crop growth cycle from a single chart, nor is it possible to directly view relevant information on the crop growth period, sowing period, and maturity period within a certain time span. The only option is to select the time span and time period for matching separately. This operation step is relatively cumbersome and increases the behavioral burden on users. Summary of the invention

[0004] The present invention provides a dual-axis interactive method, device, equipment and medium based on the growth of crops. Its main purpose is to construct a growth trend prediction model, adjust the time span of various crops to be tested according to the trend prediction model, and use the linkage effect between the time axis and the growth cycle to obtain the growth trend corresponding to the crops to be tested, so as to quickly simulate the growth cycle and growth trend of each crop to be tested, reduce cumbersome operations, and optimize and improve the interactive experience of crop growth cycle query.

[0005] To achieve the above object, the present invention provides a dual-axis interaction method based on crop growth, the method comprising:

[0006] Obtain the types of crops to be tested;

[0007] Receiving the type of the crop to be tested, and using a preset growth trend prediction model to obtain the initial growth trend of the crop to be tested in each growth cycle within a preset time period according to the type of the crop to be tested, wherein the growth cycle includes at least one of a growing period, a harvest period, and a sowing period;

[0008] Obtaining the time span on the time axis corresponding to each growth cycle of the crop to be tested, wherein the time span is determined by the change of the time axis of the crop to be tested;

[0009] When it is monitored that the time span corresponding to at least any one of the growth period, harvest period and sowing period of the crop to be tested changes, the time period corresponding to the growth period, harvest period and sowing period of the crop to be tested on the time axis is adjusted according to the time span of the crop to be tested, and the initial growth trend corresponding to the crop to be tested in each growth cycle is updated according to the time period to generate the final growth trend, and the differences are displayed in a graphical form.

[0010] Optionally, the growth trend is determined by:

[0011] Determine the vegetation cover index based on the invisible bands in satellite remote sensing images;

[0012] Calculate the average and standard deviation of the actual vegetation coverage index of several sampling points in each area during the current period;

[0013] Determine the regional difference between the average vegetation cover index and the actual vegetation cover index in each region;

[0014] Determine a growth reference factor according to the regional difference and the standard deviation, wherein the growth reference factor includes the quotient of the standard deviation and the regional difference;

[0015] The growth trend is determined according to the growth potential reference factor and the preset growth potential mapping relationship.

[0016] Optionally, when it is monitored that the time span corresponding to at least any one of the growth period, harvest period and sowing period of the tested crop changes, the time period corresponding to the growth period, harvest period and sowing period of the tested crop on the time axis is adjusted according to the time span of the tested crop, the initial growth trend corresponding to the tested crop in each growth cycle is updated according to the time period to generate the final growth trend, and the steps of displaying the difference in a graphical form include:

[0017] Acquire the growth cycle corresponding to the crop to be tested in the current display interface, where the growth cycle is at least one of a growing period, a harvest period, and a sowing period;

[0018] Determine the time span of the change on the time axis according to the growth cycle corresponding to the current crop to be tested;

[0019] When it is monitored that the time span corresponding to at least any one of the growth period, harvest period and sowing period of the tested crop changes, the time span corresponding to each growth cycle and the growth influence data of the current crop caused by the time difference are determined by using the changed time span, and the growth influence data at least includes the light, temperature, rainfall and pest change factors corresponding to the tested crop in the current landform caused by the time difference change;

[0020] According to the change factors of the growth of the tested crops in the time period corresponding to each growth cycle, the average value and standard deviation of the actual vegetation coverage index of several sampling points in each area are calculated respectively;

[0021] Determine the regional difference between the average vegetation cover index and the actual vegetation cover index in each region;

[0022] The linkage change between the time axis and the growth cycle is determined according to the regional difference and the standard deviation, and the growth trend corresponding to the tested crop in each growth cycle is determined and displayed in a graphical manner.

[0023] Optionally, also include:

[0024] Obtaining a query request for the crop to be tested, the query request at least including the name of the crop to be tested, location information of the crop to be tested, and consulting content describing the query;

[0025] Extracting consulting information of the consulting content using the professional knowledge base;

[0026] Acquire several guide directories from the professional knowledge base according to the consulting information, determine the sorting relationship of the guide directories according to the knowledge graph corresponding to the consulting information, and use the leading guide directory in the sorting relationship as the query directory;

[0027] The query catalog is used for consultation guidance, and the preset growth trend prediction model responds according to the name of the crop to be tested determined by the query catalog and the time span corresponding to each growth cycle of the crop to be tested on the time axis. According to the response results, the growth trends of the crop to be tested in each growth cycle are displayed differently in a graphical form.

[0028] Optionally, the training method of the preset growth trend prediction model includes:

[0029] Dividing the sample data set into a training set and a validation set;

[0030] The preset basic model is trained by the training set, and the preset basic model after training is verified by the verification set, until the accuracy of the output result of the preset basic model after training reaches the preset accuracy, thereby obtaining a growth trend prediction model;

[0031] Wherein, dividing the sample data set into a training set and a validation set includes any one of the following:

[0032] Dividing the sample data set into a training set and a validation set according to a preset ratio;

[0033] The sample data set is divided into a training group and a validation group, the validation group is divided into at least two validation sub-groups, at least one of the validation sub-groups is used as a validation set, and the remaining validation sub-groups and the training group are used as training sets, to generate several pairs of training sets and validation sets, wherein the validation sets of each pair of training sets and validation sets are different.

[0034] Optionally, before training the preset basic model, the method further includes:

[0035] Obtaining preset growth impact data corresponding to the machine learning model;

[0036] The minimum value of the preset growth influence data and the total recognition accuracy is used as the credibility factor of the machine learning model;

[0037] The growth potential recognition accuracy, the credibility factor, the growth potential recognition accuracy of the machine learning model, and the preset growth potential influence data corresponding to the machine learning model are processed by a preset calculation formula to obtain credibility;

[0038] A credible model is determined according to the credibility corresponding to the machine learning model, and the target growth trend output by the credible model is used as the final growth trend, and the credibility of the credible model is used as the final growth trend credibility.

[0039] Optionally, the preset basic model includes at least one of a naive Bayes model and a nonlinear support vector machine model;

[0040] The Naive Bayes model includes,

[0041]

[0042] Among them, xi is the growth potential, yk is the growth potential influencing data, and n is the number of growth potential classifications;

[0043] The nonlinear support vector machine model includes:

[0044]

[0045] Among them, w is the normal vector of the hyperplane, C is the penalty coefficient, which indicates the degree of penalty for the model's incorrect prediction of growth potential, and δ i is the penalty term, which indicates the penalty for incorrectly predicting the growth potential of each sample, N is the total number of sample data in the training set for training the nonlinear support vector machine, Xi is the growth potential influence vector input to the nonlinear support vector machine, and y i is the actual growth trend, and b is the adjustment coefficient.

[0046] In addition, to achieve the above-mentioned purpose, the present invention also provides a two-axis interactive device based on the growth of crops, the device comprising:

[0047] An acquisition module, used to acquire the types of crops to be tested;

[0048] A growth trend determination module, for receiving the type of the crop to be tested, and using a preset growth trend prediction model to obtain the initial growth trend of the crop to be tested in each growth cycle within a preset time period according to the type of the crop to be tested, wherein the growth cycle includes at least one of a growing period, a harvest period, and a sowing period;

[0049] A time axis acquisition module, used to acquire the time span on the time axis corresponding to each growth cycle of the crop to be tested, wherein the time span is determined by the change of the time axis of the crop to be tested;

[0050] The dual-axis interaction module is used to adjust the time periods corresponding to the growth period, harvest period and sowing period of the crop to be tested on the time axis according to the time span of the crop to be tested, when it is monitored that the time span corresponding to at least any one of the growth period, harvest period and sowing period of the crop to be tested changes, and update the initial growth trend corresponding to the crop to be tested in each growth cycle according to the time period to generate the final growth trend, and display the difference in a graphical form.

[0051] In addition, to achieve the above objectives, the present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method described in any one of the above embodiments when executing the computer program.

[0052] In addition, to achieve the above-mentioned purpose, the present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the method described in any of the above embodiments are implemented.

[0053] The present invention proposes a two-axis interactive method, device, equipment and medium based on the growth of crops. The method performs two-axis interaction with each growth cycle of crops through a time axis. After receiving the type of the crop to be tested, a preset growth trend prediction model can be used to obtain the initial growth trend of the crop to be tested in the growth period, harvest period and sowing period within a preset time period according to the type of the crop to be tested. Then, the time span is determined according to the different time periods of the crop to be tested in the growth period, sowing period and maturity period. The initial growth trend corresponding to the crop to be tested in each growth cycle is updated by using linkage changes to generate the final growth trend, and the differences are displayed in a graphical form. A graph is used to display the growth trend change graph of different growth cycles. Not only is the operation simple, but it can also meet the needs of various different users, greatly improving the interactive experience of determining the growth of crops. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 A schematic flow chart of a dual-axis interaction method based on crop growth provided in one embodiment of the present invention;

[0055] Figure 2 Another schematic diagram of a process flow in a dual-axis interaction method based on crop growth provided in one embodiment of the present invention;

[0056] Figure 3 A schematic diagram of a growth trend prediction model training process in a dual-axis interactive method based on crop growth provided in one embodiment of the present invention;

[0057] Figure 4 A schematic diagram of a growth trend prediction model determination process in a dual-axis interactive method based on crop growth provided in one embodiment of the present invention;

[0058] Figure 5 A schematic diagram of a complete structure of a dual-axis interactive device based on crop growth provided in one embodiment of the present invention;

[0059] Figure 6 A schematic diagram of the structure of a computer device provided in one embodiment of the invention.

[0060] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0061] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0062] In one embodiment, a two-axis interaction method based on crop growth is provided, referring to Figure 1 As shown, the method comprises the following steps:

[0063] S101, obtaining the type of the crop to be tested;

[0064] Among them, the crops to be tested can be divided into two categories: food crops and cash crops. Food crops include rice, corn, beans, potatoes, barley, broad beans, wheat, etc.; cash crops are further divided into oil crops, vegetable crops, and flowers, grasses, and trees. Common varieties include apples, peaches, apricots, cabbage, celery, leeks, mustard, peanuts, sesame, etc. The types of crops to be tested can be selected according to user needs and are not further limited here.

[0065] S102, receiving the type of the crop to be tested, and using a preset growth trend prediction model to obtain the initial growth trend of the crop to be tested in each growth cycle within a preset time period according to the type of the crop to be tested, wherein the growth cycle includes at least one of a growing period, a harvest period, and a sowing period;

[0066] Among them, the growing period refers to the time in a year that is suitable for crop growth in a region, generally measured by the number of days with a synoptic average temperature ≥ 10°C. For example, the growing period of crops in the south of the Yangtze River is relatively long, generally 8-10 months, 3-5 months in the north of the Huaihe River, and about 4 months in the northeast region.

[0067] Harvest season: refers to the time suitable for harvesting crops; sowing season: refers to the time suitable for sowing crops.

[0068] Growth cycle: refers to the time required for a crop from sowing to harvesting. For example, corn is generally 95-130 days. Depending on the type of crop to be tested, the corresponding growth period, harvest period, sowing period and growth cycle are different, which will not be listed here.

[0069] Specifically, the preset time period is set by the user according to needs.

[0070] S103, obtaining the time span on the time axis corresponding to each growth cycle of the crop to be tested, wherein the time span is determined by the change of the time axis of the crop to be tested;

[0071] Specifically, the time span graph is a period within a year, and the change accuracy of the time span graph on the time axis can be accurate to days.

[0072] For example, the time span graph may be the time corresponding to the growing period, the harvesting period, and the sowing period respectively, or the time span graph may be the time corresponding to the entire growth cycle.

[0073] It should be noted that, depending on the type of crops to be tested, the time periods corresponding to the growth period, harvest period and sowing period are also different. The difference is related to the characteristics of the current crops to be tested. For example, rice and corn are spring and summer crops, and their corresponding time periods should also be spring and summer time periods. Similarly, potatoes, cabbage, etc. are autumn and winter crops, and their corresponding time periods should also be autumn and winter time periods. I will not go into details here.

[0074] S104, when it is monitored that the time span corresponding to at least any one of the growth period, harvest period and sowing period of the crop to be tested changes, the time period corresponding to the growth period, harvest period and sowing period of the crop to be tested on the time axis is adjusted according to the time span of the crop to be tested, and the initial growth trend corresponding to the crop to be tested in each growth cycle is updated according to the time period to generate the final growth trend, and the difference is displayed in a graphical form.

[0075] Among them, due to the linkage effect between the time axis and the growth cycle, when it is monitored that the time span corresponding to at least any one stage of the growth period, harvest period and sowing period of the crop to be tested changes, for example, the time period corresponding to the growth period, harvest period and sowing period of the crop to be tested on the time axis is adjusted according to the time span of the crop to be tested, and the initial growth trend corresponding to the crop to be tested in each growth cycle is updated according to the time period to generate the final growth trend, and the linkage reflection display is used to display the initial growth trend corresponding to each growth cycle in a graphical form to generate the final growth trend. For example, corn is selected as an example of the crop to be tested, and the corn is input into the growth trend prediction model. The NDVI value of a preset time period such as one year in the current display interface is obtained. The growth trend includes the vegetation coverage index within the preset time period, or the growth trend is determined according to the growth reference factor and the preset growth mapping relationship, thereby obtaining the growth trend of the corn; when it is detected that the time span corresponding to each growth cycle of the corn changes, the time period of each growth cycle of the corn changes due to the change in the time span is linked to the change, and the initial growth trend corresponding to the corn in each growth cycle is updated to form a final growth trend, and the initial growth trend corresponding to each growth cycle is displayed in a graphical form through linkage reflection to form a final growth trend.

[0076] Through the above method, users can select different types of crops to be tested, stretch and change the time span as needed, and move this time span within the time axis to achieve selection of different months throughout the year. The data of the growth trend prediction model for its growing period, harvest period, and sowing period can be displayed, making the entire information display clearer and the user operation simpler; at the same time, the real-time updated growth trend caused by the change in time span can also be intuitively seen.

[0077] In addition, the growth trend graph of the crops to be tested in the current interface can be quickly obtained, and the growth trends of the growth cycles in different periods of time can be seen in the same graph, which is convenient for monitoring or simulating various crops to be tested.

[0078] In some embodiments, the growth trend is determined by:

[0079] Determine the vegetation cover index based on the invisible bands in satellite remote sensing images;

[0080] Calculate the average and standard deviation of the actual vegetation coverage index of several sampling points in each area during the current period;

[0081] Determine the regional difference between the average vegetation cover index and the actual vegetation cover index in each region;

[0082] Determine a growth reference factor according to the regional difference and the standard deviation, wherein the growth reference factor includes the quotient of the standard deviation and the regional difference;

[0083] The growth trend is determined according to the growth potential reference factor and the preset growth potential mapping relationship.

[0084] Among them, the method of acquiring satellite remote sensing images can be implemented in a manner known to those skilled in the art, and the method of determining the actual vegetation cover index (NDVI) based on the invisible bands in the satellite remote sensing images can also be implemented in a manner known to those skilled in the art, which is not limited here.

[0085] The total average value is determined by adding up all the actual vegetation cover indices collected and obtaining the average value as the total average value.

[0086] The regional average value is the average value obtained by summing up the actual vegetation cover index corresponding to each sampling point in a certain area.

[0087] The standard deviation is determined as follows:

[0088]

[0089] Among them, σ is the standard deviation, n is the total number of regions collected, xi is the regional average of region i, and x is the total average.

[0090] For example, determine the regional difference between the regional mean and the overall rating mean;

[0091] The regional difference is used as a growth potential reference factor;

[0092] The actual growth potential is determined according to the growth potential reference factor and the preset growth potential mapping relationship.

[0093] Of course, the actual growth trend can also be determined by variance, and its specific implementation method is similar to the above-mentioned method of using standard deviation, which will not be repeated here.

[0094] An exemplary preset growth potential mapping relationship is as follows:

[0095] 10—Growth potential reference factor is less than -3*standard deviation;

[0096] 9—Growth potential reference factor is between -3*standard deviation and -2*standard deviation;

[0097] 8—Growth potential reference factor is between -2*standard deviation and -1*standard deviation;

[0098] 7—Growth potential reference factor is between -1*standard deviation and -0.5*standard deviation;

[0099] 6—Growth potential reference factor is between 0.5*standard deviation and 0;

[0100] 5—Growth potential reference factor is between 0 and 0.5*standard deviation;

[0101] 4—The growth potential reference factor is between 0.5*standard deviation and 1*standard deviation;

[0102] 3—The growth potential reference factor is between 1*standard deviation and 2*standard deviation;

[0103] 2—The growth potential reference factor is between 2*standard deviation and 3*standard deviation;

[0104] 1—Growth reference factor is greater than 3*standard deviation.

[0105] Through the above method, on the one hand, the growth trend can be accurately determined, which is convenient for users to understand the status of the crops to be tested. On the other hand, combined with the dual-axis interaction method of the time axis and the growth cycle of the crops to be tested in this embodiment, the growth trend of the optimal time period can be quickly determined, which helps users to plant according to this growth trend and form effective reference experience.

[0106] See also Figure 2 , is another flow chart of a dual-axis interactive method based on crop growth provided in one embodiment of the present invention, comprising:

[0107] When it is monitored that the time span corresponding to at least any one of the growth period, harvest period and sowing period of the tested crop changes, the time period corresponding to the growth period, harvest period and sowing period of the tested crop on the time axis is adjusted according to the time span of the tested crop, the initial growth trend corresponding to the tested crop in each growth cycle is updated according to the time period to generate the final growth trend, and the steps of displaying the difference in a chart are described in detail as follows:

[0108] Step S201, obtaining a growth cycle corresponding to the crop to be tested in the current display interface, wherein the growth cycle is at least one of a growing period, a harvest period, and a sowing period;

[0109] Step S202, determining the time span of the change on the time axis according to the growth cycle corresponding to the current crop to be tested;

[0110] Step S203, when it is monitored that the time span corresponding to at least any one of the growing period, harvest period and sowing period of the tested crop changes, the changed time span is used to determine the time period corresponding to each growth cycle and the growth influence data of the current crop caused by the time difference, and the growth influence data at least includes the light, temperature, rainfall and pest change factors corresponding to the tested crop in the current landform caused by the time difference change;

[0111] Step S204, calculating the average value and standard deviation of the actual vegetation coverage index of several sampling points in each current region according to the change factors of the growth influencing data of the tested crop in the time period corresponding to each growth cycle;

[0112] Step S205, determining the regional difference between the average vegetation coverage index of each region and the actual vegetation coverage index;

[0113] Step S206, determining the linkage change between the time axis and the growth cycle according to the regional difference and the standard deviation, determining the growth trend of the tested crop in each growth cycle and displaying the difference in a chart.

[0114] It should be noted that, by determining the time span of the change on the time axis, the growth trend (the final growth trend) corresponding to each growth cycle can be directly calculated according to the above-mentioned change factor.

[0115] Specifically, the time span of a certain section or the entire growth cycle of the crop to be tested on the time axis is changed by clicking, dragging, pulling, etc.; when it is detected that the time span corresponding to the change in a certain section or the entire growth cycle of the crop to be tested changes, the growth impact data caused by the time factor is determined according to the difference in the time span.

[0116] Specifically, the crops to be tested are crops of the same type in a certain area, and the growth data affected by time changes in the area are collected. For example, the growth data affected by time changes in a certain area include light, temperature, climate (rainfall), etc.

[0117] For example, if the growth influencing data changes, the accuracy of the growth trend prediction model can be optimized by adjusting the growth influencing data within the growth trend prediction model.

[0118] Among them, the methods of determining growth trends based on regional differences and standard deviations include:

[0119] Determine a growth reference factor according to the regional difference and the standard deviation, wherein the growth reference factor includes the quotient of the standard deviation and the regional difference;

[0120] The growth trend is determined according to the growth potential reference factor and the preset growth potential mapping relationship.

[0121] For example, of course, the determination of growth potential can also be carried out by means of variance, and its specific implementation method is similar to the above-mentioned method of using standard deviation, which will not be repeated here.

[0122] Through the above method, the time span of the time change on the time axis is used to adjust the time period corresponding to the tested crop in the growth cycle, and according to the change factor of the growth influence data in the time period, the average value and standard deviation of the actual vegetation coverage index of several sampling points in each current area are calculated respectively, and the regional difference between the average vegetation coverage index and the actual vegetation coverage index of each area is determined; according to the regional difference and the standard deviation, the linkage change between the time axis and the growth cycle is determined, and the growth trend corresponding to the tested crop in each growth cycle is determined and displayed in a graphical manner, thereby meeting the needs of various users, greatly simplifying the operation process, being able to quickly determine the growth trend corresponding to the tested crop in each growth cycle, and significantly improving the user experience of determining the growth of crops.

[0123] In some other embodiments, it also includes:

[0124] Obtaining a query request for the crop to be tested, the query request at least including the name of the crop to be tested, location information of the crop to be tested, and consulting content describing the query;

[0125] Extracting consulting information of the consulting content using the professional knowledge base;

[0126] Acquire several guide directories from the professional knowledge base according to the consulting information, determine the sorting relationship of the guide directories according to the knowledge graph corresponding to the consulting information, and use the leading guide directory in the sorting relationship as the query directory;

[0127] The query catalog is used for consultation guidance, and the preset growth trend prediction model responds according to the name of the crop to be tested determined by the query catalog and the time span corresponding to each growth cycle of the crop to be tested on the time axis. According to the response results, the growth trends of the crop to be tested in each growth cycle are displayed differently in a graphical form.

[0128] The query request may be a query statement manually input by a user, or may be a query statement collected by voice, which is not limited here.

[0129] Specifically, professional corpus is used to perform semantic segmentation on key information to obtain training keywords and test keywords; the keywords and valid questions are trained to generate a keyword-question logic model; the trained keyword-question logic model is tested with the test keywords, and the keyword-question logic model is corrected according to the test results to obtain a keyword-question logic model that meets the requirements; the key information is imported into the keyword-question logic model to automatically combine to form a number of questions; and a guide directory related to each of the questions is generated to obtain a number of guide directories.

[0130] For example, by segmenting the sentence to be processed, a plurality of individual words are obtained; the entities in the plurality of individual words are identified, and the entities are grouped into entity pairs; for each entity pair, the sentence vector of the corresponding sentence is obtained, and the sentence vector includes the word-position vectors of the plurality of individual words, and the word-position vector of the individual word includes the word vector of the individual word, and the relative position vector between the individual word and the entity pair; the representation features of multiple aspects of the sentence are extracted according to the sentence vector; a fusion feature is obtained according to the representation features of the plurality of aspects, and the fusion feature includes no less than one feature vector; the entity relationship of the entity pair is predicted according to the fusion feature; and a knowledge graph is constructed according to the entity pair and the corresponding entity relationship. Among them, in order to fully reflect the correlation between the phrase and the entity pair, the features between the entity relationship in the sentence and the sentence can be mined from multiple aspects and angles. Through fusion means, the entity relationship between the entity pairs can be predicted more comprehensively, and thus the knowledge graph can be constructed more completely.

[0131] Specifically, several guide directories are sorted through the knowledge graph constructed in the professional knowledge base, wherein each of the knowledge graphs is sorted in a descending order of relevance, and the guide directories are sorted by sorting the knowledge graphs. The query directory input by the current user is obtained according to the sorting relationship of the guide directories, thereby improving the accuracy of the query directory.

[0132] Specifically, through the location information of the crops to be tested, not only can the meteorological data corresponding to the geographical location be obtained, but also by calling the sensors of the environment in which the geographical location is located, the topographic information, rainfall, temperature, fertilization conditions, pest conditions and other factors that may affect the growth of the crops to be tested can be obtained.

[0133] Through the above method, during information interaction, the consulting content of the user's query request can be accurately obtained, and the consulting information in the consulting content is used to match professional information answers in the agricultural information knowledge base, which not only improves the user interaction experience, but also improves the information of the crops to be tested.

[0134] In some embodiments, see Figure 3 , is a schematic diagram of the training process of the growth trend prediction model in the dual-axis interactive method based on the growth of crops provided in one embodiment of the present invention, which is described in detail as follows:

[0135] Step S301, dividing the sample data set into a training set and a validation set;

[0136] Step S302, training the preset basic model through the training set, and verifying the trained preset basic model through the verification set, until the accuracy of the output result of the trained preset basic model reaches a preset accuracy, thereby obtaining a growth trend prediction model;

[0137] Wherein, dividing the sample data set into a training set and a validation set includes any one of the following:

[0138] Dividing the sample data set into a training set and a validation set according to a preset ratio;

[0139] The sample data set is divided into a training group and a validation group, the validation group is divided into at least two validation sub-groups, at least one of the validation sub-groups is used as a validation set, and the remaining validation sub-groups and the training group are used as training sets, to generate several pairs of training sets and validation sets, wherein the validation sets of each pair of training sets and validation sets are different.

[0140] The preset ratio can be set by those skilled in the art as needed, and the division of the training set and the validation set according to the preset ratio can be achieved by random allocation. The division of the sample data set into the training group and the validation group can also be achieved by random allocation.

[0141] The following is an example to specifically illustrate the generation methods of several pairs of training sets and validation sets. The data in the sample data set is divided into a training group and a validation group, the training group is divided into several training sub-groups, the validation group is split into a newly added training group and a remaining validation group for multiple times, the training set is generated based on the newly added training group and the training sub-group, the validation set is generated based on the remaining validation group, and the remaining validation sub-group obtained by each initial validation sub-group split is different. For another example, the data in the sample data set is divided into a training group and a validation group, the training group is divided into M initial training sub-groups, the validation group is divided into N initial validation sub-groups, the initial validation sub-groups are split in turn, the newly added training sub-groups and the remaining validation sub-groups are obtained, the training set is generated based on the newly added training sub-groups and the initial training sub-groups, the validation set is generated based on the remaining validation sub-groups, and the remaining validation sub-groups are different each time the initial validation sub-groups are split, wherein the newly added training sub-groups and the remaining validation sub-groups are split in turn, including taking (N-1) validation sub-groups as the (M+1)th training sub-group and the remaining one remaining validation sub-group as the validation set. Through the above implementation, multiple pairs of training sets and validation sets with certain differences can be obtained, and the preset basic model can be trained more times to improve the accuracy of the machine learning model obtained by training.

[0142] In some embodiments, see Figure 4 , is a schematic diagram of a flow chart of determining a growth trend prediction model in a two-axis interactive method based on crop growth provided in an embodiment of the present invention, and before training the preset basic model, it also includes:

[0143] Step S401, obtaining preset growth influence data corresponding to the machine learning model;

[0144] Step S402, taking the minimum value of the preset growth influence data and the total recognition accuracy as the credibility factor of the machine learning model;

[0145] Step S403, processing the growth potential recognition accuracy, the credibility factor, the growth potential recognition accuracy of the machine learning model, and the preset growth potential influence data corresponding to the machine learning model through a preset calculation formula to obtain credibility;

[0146] Step S404, determining a credible model according to the credibility corresponding to the machine learning model, and taking the target growth trend output by the credible model as the final growth trend, and taking the credibility of the credible model as the final growth trend credibility.

[0147] For example, the preset impact factor can be set by those skilled in the art according to different types of preset basic models. The preset impact factor is greater than 0 and less than 1.

[0148] Optionally, the credibility determination method (preset calculation formula) includes:

[0149] Ai=Si*wi / max(Sb*wb);

[0150] Among them, Ai is the credibility, Si is the growth recognition accuracy of the target predicted growth output by the i-th machine learning model, w1 is the preset influence factor of the i-th machine learning model, Sb is the growth recognition accuracy under the target predicted growth of the b-th machine learning model, wb is the preset influence factor of the b-th machine learning model, i∈[1,n], b={1……n}, n is the number of machine learning models, n≥2.

[0151] Among them, max(Sb*wb) is the largest value of the product of the preset influence factor and the growth potential recognition accuracy of each machine learning model. The credible model is the machine learning model with the highest credibility.

[0152] Selecting machine learning models through credibility not only helps to improve the output target growth trend, but also improves the prediction accuracy of the final growth trend. At the same time, by comparing with the past growth trends of the crops to be tested in the current target area, it is not only possible to predict whether the growth trend is abnormal, but also to issue early warning information to remind the target area that there is a risk of abnormal growth trend, so that the staff of the relevant crops to be tested can prepare in advance.

[0153] In some embodiments, the preset basic model includes at least one of a naive Bayes model and a nonlinear support vector machine model;

[0154] The Naive Bayes model includes,

[0155]

[0156] Among them, xi is the growth potential, yk is the growth potential influencing data, and n is the number of growth potential classifications;

[0157] The nonlinear support vector machine model includes:

[0158]

[0159] Among them, w is the normal vector of the hyperplane, C is the penalty coefficient, which indicates the degree of penalty for the model's incorrect prediction of growth potential, and δ iis the penalty term, which indicates the penalty for incorrectly predicting the growth potential of each sample, N is the total number of sample data in the training set for training the nonlinear support vector machine, Xi is the growth potential influence vector input to the nonlinear support vector machine, and y i is the actual growth trend, and b is the adjustment coefficient.

[0160] Optionally, in order to improve the growth trend of the tested crops,

[0161] The data in the sample data set are divided into several growth trend groups according to the actual growth trend, and the growth influence data in the growth trend groups are clustered to obtain several cluster groups, including the growth influence data in N cluster groups with a large amount of data as the relevant growth influence data corresponding to each growth trend, and the target growth trend corresponding to the target growth trend influence vector is obtained. The ideal growth trend is determined according to the target growth trend and the final predicted growth trend, and the ideal influence data corresponding to the ideal growth trend is determined from the relevant growth influence data. The disaster relief plan is determined according to the target growth influence data and the ideal growth influence data.

[0162] Among them, the ideal growth trend is the better growth trend between the target growth and the final predicted growth.

[0163] The growth impact data include topographic information, crop types, rainfall, temperature, fertilization, and pest conditions. The disaster relief plan is determined based on the target growth impact data and the ideal growth impact data, including:

[0164] If the rainfall in the target growth impact data is higher than the rainfall in the ideal growth impact data, add drainage channels;

[0165] If the rainfall in the target growth impact data is lower than the rainfall in the ideal growth impact data, artificial rainfall or enhanced irrigation shall be carried out;

[0166] If the fertilizer amount in the target growth impact data is lower than the fertilizer amount in the ideal growth impact data, topdressing is performed;

[0167] If the amount of insect pests in the target growth impact data is higher than the amount of fertilizer in the ideal growth impact data, spray pesticides or biological insect repellents.

[0168] After dividing the data in the sample data set into a plurality of growth trend groups according to the actual growth trend, and before clustering the growth influence data in the growth trend groups respectively, the method further includes:

[0169] Obtaining attribute information of the growth-affecting data, the attribute information can be set in advance, the attribute information includes type and adjustability, and the adjustability includes adjustable and non-adjustable;

[0170] The attribute information including the adjustable growth influence data is clustered according to the type.

[0171] For example, the types of growth data may include landform information, crop types, rainfall, temperature, fertilization conditions, pest conditions, etc., among which the data of landform information and crop types are not adjustable, while the data of rainfall, temperature, fertilization conditions, and pest conditions are adjustable.

[0172] By summarizing the main influencing factors of each growth trend, we can know that if the predicted growth is worse than the target growth trend of the previous period, we can make corresponding adjustments based on the main influencing factors to reduce the impact of the disaster.

[0173] The present embodiment provides a two-axis interaction method based on the growth of crops. The method performs two-axis interaction with each growth cycle of crops through a time axis. After receiving the type of the crop to be tested, the method can use a preset growth trend prediction model to obtain the initial growth trend of the crop to be tested in the growth period, harvest period and sowing period within a preset time period according to the type of the crop to be tested. Then, the time span is determined according to the different time periods of the crop to be tested in the growth period, sowing period and maturity period. The initial growth trend corresponding to the crop to be tested in each growth cycle is updated by using linkage changes to generate a final growth trend, and the differences are displayed in a graphical form. Using one graph to display the growth trend change graph of different growth cycles is not only simple to operate, but also can meet the needs of various different users, greatly improving the interactive experience of determining the growth of crops.

[0174] The following is a specific example to illustrate the above-mentioned two-axis interaction method based on the growth of crops. The specific two-axis interaction method based on the growth of crops includes:

[0175] Pull down to select the type of crop to be tested.

[0176] In the initial state, the growing period, harvesting period, and sowing period of crops throughout the year are displayed according to different categories of crops to be tested; at the same time, the average NDVI value of the selected crops to be tested in the past five years and the NDVI values ​​of 2021 and 2020 are displayed throughout the year; for example, when a certain type of crop to be tested is selected, the relevant data information is called to display the relevant data values ​​(NDVI values, i.e., vegetation index) throughout the year in the initial state.

[0177] The time span is selected by stretching the time axis. When the growth cycle of the crop to be tested changes in linkage, the display of the crop growth period, sowing period or harvest period can be realized. For example, when the harvest period of the selected crop to be tested is in progress, the time range corresponding to the current harvest period is determined by stretching the time span of the time axis. If the selected crop to be tested is corn, the corresponding time period on the time axis is from June 1, 2021 to December 31, 2021.

[0178] By default, when the time span does not move, the chart can display the sowing period, maturity period or harvest period of crops according to the time span. It also displays the average NDVI value of crops in the past five years within a fixed time span and the NDVI values ​​of 2021 and 2020, displayed with different color curves.

[0179] By moving the time span left and right on the time axis, the time span corresponding to each growth cycle can be adjusted and determined; according to the movement of the time span, the time of the crop growth period, harvest period or sowing period and the average NDVI value of the crops in the past five years and the NDVI values ​​in 2021 and 2020 are displayed.

[0180] Through the above method, using the dual-axis interaction method corresponding to the time axis and the growth cycle of the crops to be tested, different crops to be tested can be selected according to user needs. At the same time, the time span can be associated with the corresponding month to quickly query the NDVI value of the crops to be tested.

[0181] In addition, to achieve the above purpose, the present invention also provides a two-axis interactive device 500 based on crop growth, see Figure 5 , the device comprises:

[0182] An acquisition module 501 is used to acquire the type of the crop to be tested;

[0183] The growth tendency determination module 502 is used to receive the type of the crop to be tested, and obtain the initial growth trend of the crop to be tested in each growth cycle within a preset time period according to the type of the crop to be tested using a preset growth tendency estimation model, wherein the growth cycle includes at least one of a growing period, a harvest period, and a sowing period;

[0184] The time axis acquisition module 503 is used to acquire the time span on the time axis corresponding to each growth cycle of the crop to be tested, wherein the time span is determined by the change of the time axis of the crop to be tested;

[0185] The dual-axis interaction module 504 is used to adjust the time periods corresponding to the growth period, harvest period and sowing period of the crop to be tested on the time axis according to the time span of the crop to be tested, when it is monitored that the time span corresponding to at least any one of the growth period, harvest period and sowing period of the crop to be tested changes, and update the initial growth trend corresponding to the crop to be tested in each growth cycle according to the time period to generate the final growth trend, and display the difference in a graphical form.

[0186] Specifically, in this embodiment, the growth trend is determined by:

[0187] Determine the vegetation cover index based on the invisible bands in satellite remote sensing images;

[0188] Calculate the average and standard deviation of the actual vegetation coverage index of several sampling points in each area during the current period;

[0189] Determine the regional difference between the average vegetation cover index and the actual vegetation cover index in each region;

[0190] Determine a growth reference factor according to the regional difference and the standard deviation, wherein the growth reference factor includes the quotient of the standard deviation and the regional difference;

[0191] The growth trend is determined according to the growth potential reference factor and the preset growth potential mapping relationship.

[0192] Specifically, in this embodiment, when it is monitored that the time span corresponding to at least any one of the growth period, harvest period and sowing period of the tested crop changes, the time period corresponding to the growth period, harvest period and sowing period of the tested crop on the time axis is adjusted according to the time span of the tested crop, the initial growth trend corresponding to the tested crop in each growth cycle is updated according to the time period to generate the final growth trend, and the steps of displaying the difference in a graphical form include:

[0193] Acquire the growth cycle corresponding to the crop to be tested in the current display interface, where the growth cycle is at least one of a growing period, a harvest period, and a sowing period;

[0194] Determine the time span of the change on the time axis according to the growth cycle corresponding to the current crop to be tested;

[0195] When it is monitored that the time span corresponding to at least any one of the growth period, harvest period and sowing period of the tested crop changes, the time span corresponding to each growth cycle and the growth influence data of the current crop caused by the time difference are determined by using the changed time span, and the growth influence data at least includes the light, temperature, rainfall and pest change factors corresponding to the tested crop in the current landform caused by the time difference change;

[0196] According to the change factors of the growth of the tested crops in the time period corresponding to each growth cycle, the average value and standard deviation of the actual vegetation coverage index of several sampling points in each area are calculated respectively;

[0197] Determine the regional difference between the average vegetation cover index and the actual vegetation cover index in each region;

[0198] The linkage change between the time axis and the growth cycle is determined according to the regional difference and the standard deviation, and the growth trend corresponding to the tested crop in each growth cycle is determined and displayed in a graphical manner.

[0199] Specifically, in this embodiment, it also includes:

[0200] Obtaining a query request for the crop to be tested, the query request at least including the name of the crop to be tested, location information of the crop to be tested, and consulting content describing the query;

[0201] Extracting consulting information of the consulting content using the professional knowledge base;

[0202] Acquire several guide directories from the professional knowledge base according to the consulting information, determine the sorting relationship of the guide directories according to the knowledge graph corresponding to the consulting information, and use the leading guide directory in the sorting relationship as the query directory;

[0203] The query catalog is used for consultation guidance, and the preset growth trend prediction model responds according to the name of the crop to be tested determined by the query catalog and the time span corresponding to each growth cycle of the crop to be tested on the time axis. According to the response results, the growth trends of the crop to be tested in each growth cycle are displayed differently in a graphical form.

[0204] Specifically, in this embodiment, the training method of the preset growth trend prediction model includes:

[0205] Dividing the sample data set into a training set and a validation set;

[0206] The preset basic model is trained by the training set, and the preset basic model after training is verified by the verification set, until the accuracy of the output result of the preset basic model after training reaches the preset accuracy, thereby obtaining a growth trend prediction model;

[0207] Wherein, dividing the sample data set into a training set and a validation set includes any one of the following:

[0208] Dividing the sample data set into a training set and a validation set according to a preset ratio;

[0209] The sample data set is divided into a training group and a validation group, the validation group is divided into at least two validation sub-groups, at least one of the validation sub-groups is used as a validation set, and the remaining validation sub-groups and the training group are used as training sets, to generate several pairs of training sets and validation sets, wherein the validation sets of each pair of training sets and validation sets are different.

[0210] Specifically, in this embodiment, before training the preset basic model, the method further includes:

[0211] Obtaining preset growth impact data corresponding to the machine learning model;

[0212] The minimum value of the preset growth influence data and the total recognition accuracy is used as the credibility factor of the machine learning model;

[0213] The growth potential recognition accuracy, the credibility factor, the growth potential recognition accuracy of the machine learning model, and the preset growth potential influence data corresponding to the machine learning model are processed by a preset calculation formula to obtain credibility;

[0214] A credible model is determined according to the credibility corresponding to the machine learning model, and the target growth trend output by the credible model is used as the final growth trend, and the credibility of the credible model is used as the final growth trend credibility.

[0215] Specifically, in this embodiment, the preset basic model includes at least one of a naive Bayes model and a nonlinear support vector machine model;

[0216] The Naive Bayes model includes,

[0217]

[0218] Among them, xi is the growth potential i, yk is the growth potential impact data, and n is the number of growth potential classifications;

[0219] The nonlinear support vector machine model includes:

[0220]

[0221] Among them, w is the normal vector of the hyperplane, C is the penalty coefficient, which indicates the degree of penalty for the model's incorrect prediction of growth potential, and δ i is the penalty term, which indicates the penalty for incorrectly predicting the growth potential of each sample, N is the total number of sample data in the training set for training the nonlinear support vector machine, Xi is the growth potential influence vector input to the nonlinear support vector machine, and y i is the actual growth trend, and b is the adjustment coefficient.

[0222] The present embodiment provides a two-axis interactive device based on the growth of crops. The device performs two-axis interaction with each growth cycle of crops through a time axis. After receiving the type of the crop to be tested, the device can use a preset growth trend prediction model to obtain the initial growth trend of the crop to be tested in the growth period, harvest period and sowing period within a preset time period according to the type of the crop to be tested. Then, the time span is determined according to the different time periods of the crop to be tested in the growth period, sowing period and maturity period. The initial growth trend corresponding to the crop to be tested in each growth cycle is updated by using linkage changes to generate a final growth trend, and the differences are displayed in a graphical form. Using one graph to display the growth trend change graph of different growth cycles is not only simple to operate, but also can meet the needs of various different users, greatly improving the interactive experience of determining the growth of crops.

[0223] It should be known that the above-mentioned dual-axis interactive device system based on crop growth is essentially equipped with multiple modules to execute the dual-axis interactive method based on crop growth in any of the above-mentioned embodiments. The specific functions and technical effects can be referred to the above-mentioned embodiments and will not be repeated here.

[0224] In one embodiment, see Figure 6 This embodiment also provides a computer device 600, including a memory 601, a processor 602, and a computer program stored in the memory and executable on the processor, wherein the processor 602 implements the steps of the method described in any of the above embodiments when executing the computer program.

[0225] In one embodiment, a computer-readable storage medium is further provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method described in any of the above embodiments are implemented.

[0226] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0227] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the existence of other identical elements in the process, device, article or method including the element.

[0228] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages and disadvantages of the embodiments. Through the description of the above implementation modes, those skilled in the art can clearly understand that the above embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course, by hardware, but in many cases the former is a better implementation mode. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes a number of instructions for enabling a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0229] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A two-axis interactive method based on crop growth, characterized in that: The method comprises: Obtain the types of crops to be tested; Receiving the type of the crop to be tested, and using a preset growth trend prediction model to obtain the initial growth trend of the crop to be tested in each growth cycle within a preset time period according to the type of the crop to be tested, wherein the growth cycle includes at least one of a growing period, a harvest period, and a sowing period; Obtaining the time span on the time axis corresponding to each growth cycle of the crop to be tested, wherein the time span is determined by the change of the time axis of the crop to be tested; When it is monitored that the time span corresponding to at least any one of the growth period, harvest period and sowing period of the tested crop changes, the time period corresponding to the growth period, harvest period and sowing period of the tested crop on the time axis is adjusted according to the time span of the tested crop, and the initial growth trend corresponding to the tested crop in each growth cycle is updated according to the time period to generate the final growth trend, and the difference is displayed in a chart; Among them, the method for determining the growth trend includes: determining the vegetation coverage index based on the invisible bands in the satellite remote sensing images; calculating the average value and standard deviation of the actual vegetation coverage index of several sampling points in each region at the current time; determining the regional difference between the average vegetation coverage index and the actual vegetation coverage index of each region; determining the growth reference factor according to the regional difference and the standard deviation, the growth reference factor including the quotient of the standard deviation and the regional difference; determining the growth trend according to the growth reference factor and a preset growth mapping relationship.

2. The two-axis interaction method based on crop growth as claimed in claim 1, characterized in that: Also includes: Obtaining a query request for the crop to be tested, the query request at least including the name of the crop to be tested, location information of the crop to be tested, and consulting content describing the query; Extracting consulting information of the consulting content using the professional knowledge base; Acquire several guide directories from the professional knowledge base according to the consulting information, determine the sorting relationship of the guide directories according to the knowledge graph corresponding to the consulting information, and use the leading guide directory in the sorting relationship as the query directory; The query catalog is used for consultation guidance, and the preset growth trend prediction model responds according to the name of the crop to be tested determined by the query catalog and the time span corresponding to each growth cycle of the crop to be tested on the time axis. According to the response results, the growth trends of the crop to be tested in each growth cycle are displayed differently in a graphical form.

3. The two-axis interaction method based on crop growth as claimed in claim 1, characterized in that: The training method of the preset growth trend prediction model includes: Dividing the sample data set into a training set and a validation set; The preset basic model is trained by the training set, and the preset basic model after training is verified by the verification set, until the accuracy of the output result of the preset basic model after training reaches the preset accuracy, thereby obtaining a growth trend prediction model; Wherein, dividing the sample data set into a training set and a validation set includes any one of the following: Dividing the sample data set into a training set and a validation set according to a preset ratio; The sample data set is divided into a training group and a validation group, the validation group is divided into at least two validation sub-groups, at least one of the validation sub-groups is used as a validation set, and the remaining validation sub-groups and the training group are used as training sets, to generate several pairs of training sets and validation sets, wherein the validation sets of each pair of training sets and validation sets are different.

4. The two-axis interaction method based on crop growth as claimed in claim 3, characterized in that: Before training the preset basic model, the method further includes: Obtaining preset growth impact data corresponding to the machine learning model; The minimum value of the preset growth influence data and the total recognition accuracy is used as the credibility factor of the machine learning model; The growth potential recognition accuracy, the credibility factor, the growth potential recognition accuracy of the machine learning model, and the preset growth potential influence data corresponding to the machine learning model are processed by a preset calculation formula to obtain credibility; A credible model is determined according to the credibility corresponding to the machine learning model, and the target growth trend output by the credible model is used as the final growth trend, and the credibility of the credible model is used as the final growth trend credibility.

5. The two-axis interactive method based on crop growth potential according to claim 3 or 4, characterized in that: The preset basic model includes at least one of a naive Bayes model and a nonlinear support vector machine model; The Naive Bayes model includes, Among them, xi is the growth potential, yk is the growth potential influencing data, and n is the number of growth potential classifications; The nonlinear support vector machine model includes: Among them, w is the normal vector of the hyperplane, C is the penalty coefficient, which indicates the degree of penalty imposed by the model on incorrectly predicted growth potential. is the penalty term, which indicates the penalty for incorrectly predicting the growth potential of each sample. N is the total number of sample data in the training set for training the nonlinear support vector machine. is the growth influence vector input to the nonlinear support vector machine, is the actual growth trend, and b is the adjustment coefficient.

6. A two-axis interactive device based on crop growth, characterized in that: The device comprises: An acquisition module, used to acquire the types of crops to be tested; A growth trend determination module, for receiving the type of the crop to be tested, and using a preset growth trend prediction model to obtain the initial growth trend of the crop to be tested in each growth cycle within a preset time period according to the type of the crop to be tested, wherein the growth cycle includes at least one of a growing period, a harvest period, and a sowing period; A time axis acquisition module, used to acquire the time span on the time axis corresponding to each growth cycle of the crop to be tested, wherein the time span is determined by the change of the time axis of the crop to be tested; A dual-axis interaction module is used for adjusting the time periods corresponding to the growth period, harvest period and sowing period of the crop to be tested on the time axis according to the time span of the crop to be tested, when it is monitored that the time span corresponding to at least any one of the growth period, harvest period and sowing period of the crop to be tested changes, and updating the initial growth trend corresponding to the crop to be tested in each growth cycle according to the time period to generate a final growth trend, and displaying the difference in a graphical form; wherein the growth trend is determined in the following manner: determining the vegetation coverage index based on the invisible band in the satellite remote sensing image; calculating the average value and standard deviation of the actual vegetation coverage index of several sampling points in each area of ​​the current period; determining the regional difference between the average vegetation coverage index and the actual vegetation coverage index of each area; determining the growth reference factor according to the regional difference and the standard deviation, the growth reference factor including the quotient of the standard deviation and the regional difference; determining the growth trend according to the growth reference factor and a preset growth mapping relationship.

7. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 5 are implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Citation Information

Patent Citations

  • Remote sensing extraction method of agricultural disaster information based on vegetation index time-space statistical characteristics

    CN106780091A

  • Method for monitoring soil moisture status in whole growth period of large-area crops on basis of modified NDVI (normalized difference vegetation index) time sequence

    CN106908415A

  • Crop growth condition analysis method and device and storage medium

    CN112581464A

  • Crop intelligent identification and growth prediction method and system

    CN113343783A

  • Wide-area agricultural monitoring and prediction

    US20120101784A1