A fitness calculation method and device based on peanut trait information classification
By classifying and annotating peanut trait information, combining planting yield and area, a more accurate fitness is calculated, which solves the problem of low calculation accuracy of planting fitness in the prior art and achieves higher planting accuracy.
Patent Information
- Application Number
- CN202311360608.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-10-19
- Publication Date
- 2024-12-17
- Estimated Expiration
- 2043-10-19
AI Technical Summary
In the peanut planting, the calculation adaptability of the trait information of different species is similar in peanut planting, and the calculation accuracy is low.
The fitness calculation method based on peanut trait information classification is adopted. By collecting and labeling trait information, it is converted into label fields for classification, and the fitness is calculated based on planting yield and area, and the calculation accuracy is improved.
The accuracy of peanut planting fitness calculation can be improved, and the adaptability of peanuts of different species can be more accurately determined, avoiding planting errors.
Smart Images
Figure CN117453965B_ABST
Abstract
Description
Background Art
[0002] Peanuts are known as "vegetable meat" because of their high oil content, excellent quality and fragrant smell. In addition to being edible, they are also used in printing and dyeing, papermaking industry and medicine. Therefore, they have important cultivation value. In the process of cultivating peanuts, it is necessary to record the different characteristics of different parts such as seed coat, cotyledon and embryo, including: color, size, grain length, yield, etc., to facilitate subsequent breeding management.
[0003] Since the trait information corresponding to different species of peanuts contains multiple data, in order to determine the individual adaptability of different species of peanuts, the commonly used method is to determine the trait characteristic data of peanuts, use the trait characteristics to assign weights, and define the fitness function based on each weight, and calculate its fitness based on the fitness function.
[0004] However, the commonly used methods currently have the following technical problems: Since the plant morphology of peanuts of different species is roughly similar or close, and the plant morphology may be affected by different factors (soil, sunlight, fertilization, etc.), the fitness calculated based on the weights assigned to the plant morphology and traits may be similar for different species and different trait information, thereby confusing peanuts of different species and resulting in low calculation accuracy. Summary of the invention
[0005] The present invention proposes a fitness calculation method and device based on the classification of peanut trait information. The method can collect the annotation information added by the user to the data contained in each trait information, convert the annotation information into label fields of trait information of different species of peanuts, classify multiple label fields and perform fitness calculation based on the classification results and the planting yield and planting area of peanuts to improve the calculation accuracy.
[0006] A first aspect of an embodiment of the present invention provides a method for calculating fitness based on peanut trait information classification, the method comprising:
[0007] After collecting trait information of multiple peanuts of different species, classifying them based on fields corresponding to the trait information to obtain trait classification results;
[0008] After displaying the trait classification result, obtaining a user's fitness analysis request;
[0009] Based on the peanut species corresponding to the fitness analysis request, extracting a plurality of target trait information from the trait classification result;
[0010] The planting adaptability is calculated based on the planting yield and planting area of the plurality of trait information.
[0011] In a possible implementation manner of the first aspect, the classifying based on the field corresponding to the trait information to obtain the trait classification result includes:
[0012] Acquire annotation information added by a user, where the annotation information is annotation content added by the user to each data included in each of the property information;
[0013] Merging the multiple annotation information corresponding to each of the trait information into a field to obtain multiple label fields, wherein each of the trait information corresponds to one label field;
[0014] Classify multiple label fields to obtain trait classification results.
[0015] In a possible implementation manner of the first aspect, merging the multiple annotation information of each trait information into a field includes:
[0016] Convert the plurality of annotation information corresponding to each of the trait information into fields respectively to obtain a plurality of annotation fields;
[0017] Extracting a characteristic field of each of the marked fields, wherein the characteristic field is a field with a preset number of digits in each of the marked fields;
[0018] The plurality of annotation fields are sorted based on the numerical values of the plurality of characteristic fields, and the plurality of annotation fields are sequentially concatenated according to the sorting result to obtain a label field.
[0019] In a possible implementation manner of the first aspect, classifying the multiple tag fields to obtain a trait classification result includes:
[0020] If the classification operation is the first classification, the preset first neural network is called to aggregate and classify the multiple label fields according to the matching degree of the fields to obtain a plurality of individual trait classification results, each of which contains at least one label field.
[0021] In a possible implementation manner of the first aspect, classifying the multiple tag fields to obtain a trait classification result includes:
[0022] If the classification operation is not the first classification, a preset classification list is obtained, wherein the classification list is composed of a plurality of classification categories, and each of the classification categories corresponds to a category field;
[0023] Calling a preset second neural network to calculate the field matching degree between each of the label fields and the category field of each of the classification categories;
[0024] The field matching degree with the largest value is selected from the multiple field matching degrees corresponding to each of the label fields, and the classification category corresponding to the field matching degree with the largest selected value is used as the trait classification result of the label field.
[0025] In a possible implementation manner of the first aspect, calculating the planting adaptability based on the planting yield and planting area of the plurality of trait information includes:
[0026] Calculating the ratio of the planting yield to the planting area of each trait information to obtain the yield-area ratio;
[0027] Based on the yield-area ratio, a corresponding calculation weight value is assigned;
[0028] A weighted average is performed on a plurality of the calculated weight values to obtain the planting fitness.
[0029] In a possible implementation manner of the first aspect, after the step of obtaining the annotation information added by the user, the method further includes:
[0030] Each of the annotation information is formatted to obtain conversion information, and a plurality of the conversion information are used to perform classification operations.
[0031] In a possible implementation manner of the first aspect, after the step of calculating the planting adaptability based on the planting yield and planting area of the plurality of trait information, the method further includes:
[0032] An analysis recommendation table is generated using the planting yield and planting area of the plurality of trait information and the planting fitness, and the analysis recommendation table is sent to a user corresponding to the fitness analysis request for user reference.
[0033] A second aspect of an embodiment of the present invention provides a fitness calculation device based on peanut trait information classification, the device comprising:
[0034] A classification module, for collecting trait information of multiple peanuts of different species, and classifying them based on fields corresponding to the trait information to obtain trait classification results;
[0035] An acquisition request module, used to acquire a user's fitness analysis request after displaying the trait classification result;
[0036] An information extraction module, used for extracting a plurality of target trait information from the trait classification result based on the peanut species corresponding to the fitness analysis request;
[0037] The fitness calculation module is used to calculate the planting fitness based on the planting yield and planting area of the plurality of trait information.
[0038] Compared with the prior art, the fitness calculation method and device based on the classification of peanut trait information provided by the embodiment of the present invention have the following beneficial effects: the present invention can obtain the user's fitness analysis request after displaying the trait information of peanuts of different species that have been classified; extract several target trait information based on the peanut species corresponding to the fitness analysis request; and then calculate the planting fitness according to the planting yield and planting area of multiple trait information, thereby improving the calculation accuracy, and determining whether the corresponding species of peanuts are suitable for planting in the corresponding land according to the planting fitness. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] Figure 1 It is a flowchart of a method for calculating fitness based on peanut trait information classification provided by an embodiment of the present invention;
[0040] Figure 2 This is an operation flow chart of a method for calculating fitness based on peanut trait information classification provided by one embodiment of the present invention;
[0041] Figure 3 It is a structural schematic diagram of a fitness calculation device based on peanut trait information classification provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0042] 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.
[0043] In order to solve the above problems, a fitness calculation method based on peanut trait information classification provided in an embodiment of the present application will be introduced and explained in detail through the following specific embodiments.
[0044] Reference Figure 1 , showing a flow chart of a method for calculating fitness based on peanut trait information classification provided by an embodiment of the present invention.
[0045] In one embodiment, the method is applicable to an information processing platform, which can be installed in an intelligent terminal or a computer for users to call and record data and perform various data analysis, processing, and calculation operations.
[0046] Wherein, as an example, the fitness calculation method based on peanut trait information classification may include:
[0047] S11. After collecting trait information of multiple peanuts of different species, classification is performed based on fields corresponding to the trait information to obtain trait classification results.
[0048] In one embodiment, each trait information may be a trait characteristic of different species of peanuts, for example, the name, planting year, planting place, planting soil, yield, kernel length, color, etc. Since the trait information includes a plurality of different data.
[0049] The field corresponding to the trait information may be a characteristic field of the trait information. Since each trait information contains specific trait characteristics for classification, the category corresponding to the peanut species is determined to obtain a trait classification result.
[0050] In order to accurately determine the trait categories of peanuts of different species, step S11 may include the following sub-steps as an example:
[0051] S111. Obtaining annotation information added by a user, wherein the annotation information is annotation content added by the user to each data included in each property information.
[0052] In one embodiment, each trait information may be a trait characteristic of different species of peanuts, for example, the name, planting year, planting place, planting soil, yield, grain length, color, etc. Since the trait information contains multiple different data. In order to facilitate the management of different data, the information processing platform may first display the trait information of each genus of peanuts on the screen, and may obtain the annotation information added by the user, which is the annotation content added to each data contained in the trait information after the user views the trait information of each genus of peanuts. The annotation content may be a string of characters, a digital code, a user-defined identifier, a user's rating value, etc.
[0053] Referring to the table below, several data included in the property information in one embodiment are shown.
[0054] Field Name Field meaning project_name Project Name Project_year Experimental year Project_season Experimental Season Project_file_path Project storage path Filename Project data file name Analyse_time Analysis time Analyse_code Analyze status codes
[0055] Referring to the following table, various annotation information added by the user in an optional embodiment is shown. When in use, the user can create a data table, which can include seed_manage, seed_describe, seed_area, and edit and manage data for each item.
[0056] Among them, seed_manage is shown in the following table:
[0057] Field Name Field meaning Seed_name Peanut variety name Character_name Trait name Character_value Values of quantitative traits Character_label Values of quality traits Test_area Pilot
[0058] The specific details of seed_describe are shown in the following table:
[0059] Field Name Field meaning Seed_name Peanut variety name seed_describe Variety description
[0060] The specific seed_area is shown in the following table:
[0061] Field Name Field meaning Seed_name Peanut variety name area Pilot Areas Suitable Is it suitable for planting?
[0062] In yet another optional embodiment, the user may input corresponding values according to application requirements, and the information management platform may synchronously update the above data table.
[0063] Optionally, the information management platform can use the entered trait information of each variety to output a corresponding Excel table, which can be viewed according to the corresponding traits and supports filtering and other operations.
[0064] In another optional embodiment, after the information management platform creates the data tables warehouse_manage, seed_enter, and seed_out, it can manage them. The information management platform can add corresponding storage warehouses for different inventory units and manage the collected seed resources.
[0065] Optionally, ordinary users can add and manage the library corresponding to the user, and add annotation data to the trait information of different species of peanuts in the library, where the annotation data includes warehouse_manage, seed_enter and seed_out.
[0066] Among them, warehouse_manage is shown in the following table:
[0067] Field Name Field meaning Mechanism_unit Organizational Unit warehouse_code Warehouse Number Warehouse_type Warehouse Type Max_capacity Maximum storage capacity of warehouse Storage_time Storage time (days) Now_capacity Existing capacity
[0068] The specific seed_enter is shown in the following table:
[0069]
[0070]
[0071] The specific seed_enter is shown in the following table:
[0072] Field Name Field meaning Seed_name Germplasm name Mechanism_unit Storage Unit Warehouse_code Storage Warehouse capacity Outgoing quantity Apply_member Applicants for outbound Apply_unit Applicant Out_time Shipping time Apply_code Application Status
[0073] In another embodiment, the user can also create data tables Character_manage, Character_classify, and Character_classify_manage to manage the characteristics, and classify the peanuts according to their different economic characteristics. The last digit of each classification is added with a classification code, which is used as the label information.
[0074] Character_manage is shown in the following table:
[0075]
[0076]
[0077] Character_classify is shown in the following table:
[0078] Field Name Field meaning Classify_name Category label name Classify_code Classification code such as (A, B, C)
[0079] Character_classify_manage is shown in the following table:
[0080] Field Name Field meaning Character_name Shape Name Classify_name Category Tags
[0081] After viewing the trait information, users can make judgments or evaluations on it. The annotations can be used as the user's evaluation of the peanut species, and then the peanuts can be classified in the future based on the user's evaluation. On the one hand, comprehensive classification can be performed to avoid using the single factor of plant morphology. On the other hand, comprehensive and detailed classification can be performed based on the annotations added by users to improve the accuracy of classification.
[0082] S112: Merge the multiple annotation information corresponding to each of the property information into a field to obtain multiple label fields, where each of the property information corresponds to one label field.
[0083] Since the trait information added by the user may contain various contents, the input users may be different, and the input contents and methods may be different, in order to unify the various trait information, it is convenient to use unified information for subsequent classification processing.
[0084] In one embodiment, the annotation information of each data included in the trait information may be first converted into a field according to a preset format, and then multiple fields may be concatenated or merged to form a field that can be used as a label or identifier for the trait information, thereby obtaining a label field.
[0085] In an optional embodiment, step S112 may include the following sub-steps:
[0086] S1121. Convert the multiple annotation information corresponding to each of the property information into fields respectively to obtain multiple annotation fields.
[0087] Specifically, the multiple annotation information corresponding to each trait information may be converted into fields of corresponding formats respectively, so as to obtain the multiple annotation fields corresponding to each trait information.
[0088] For example, there are five trait information, A, B, C, D, and E, and each trait information corresponds to 10 annotation information. Each annotation information can be converted into a field of the corresponding format, thereby converting a total of 50 annotation fields, among which each trait information corresponds to 10 annotation fields.
[0089] The corresponding format here can be a corresponding number of bits, for example, 8-bit, 16-bit or 32-bit format. For example, one of the annotation information of the A trait information is converted into the annotation field as 69813491. If the number of bits is insufficient, 0 can be added. The specific format can be adjusted according to actual needs.
[0090] S1122: Extract a characteristic field from each of the marked fields, where the characteristic field is a field with a preset number of digits in each of the marked fields.
[0091] After the field conversion, the corresponding characteristic field can be extracted from each annotation field, and the characteristic field is a field with a specific number of digits or a preset number of digits in the annotation field. After the conversion, each field with a specific number of digits corresponds to the annotation content of a specific user.
[0092] For example, if one of the annotation fields is 609080103040901, the first four digits "6090" can be used as the characteristic field, the middle four digits "0103" can be used as the characteristic field, or the last four digits "0901" can be used as the characteristic field.
[0093] The specific details can be adjusted according to actual needs.
[0094] S1123: Sort the multiple annotation fields based on the numerical values of the multiple characteristic fields, and sequentially concatenate the multiple annotation fields according to the sorting result to obtain a label field.
[0095] In an optional embodiment, the characteristic field may be converted into a numerical value of a specific base, or the numerical size of the characteristic field may be directly used as its numerical value.
[0096] Then, the numerical values are compared, and the multiple annotation fields corresponding to each property information are sorted according to the numerical values. Among them, the specific field can be the main consideration for the user to prioritize or classify the property information when adding annotation information.
[0097] For example, trait A corresponds to 10 annotation information. When adding annotation information, the user can add a rating value after each annotation. After being converted into an annotation field, the rating value is the field corresponding to a specific number of digits, which is the feature field. Then, the feature field is extracted to obtain the user's rating of the annotation field when adding annotation information, that is, the user rating value of the trait information data. If the rating value is higher, the corresponding feature field value is also larger, indicating that the user also pays more attention to or is the main annotation of the classification during classification.
[0098] After sorting by numerical value, multiple label fields can be sorted according to the sorting result. Finally, multiple label fields can be concatenated according to the sorting result to form the final label field.
[0099] For example, trait information A corresponds to 10 annotation information. These 10 annotation fields are sorted, and then the first and second fields are concatenated in sequence.
[0100] S113. Classify multiple label fields to obtain trait classification results.
[0101] After the splicing is completed, the multiple spliced label fields can be classified to determine which fields are similar, and then the classification result is used as the classification result of the trait information, and multiple different trait information are stored in the classification.
[0102] For example, there are five trait information, A, B, C, D, and E, and five corresponding label fields. The five label fields can be aggregated and classified to obtain classification results. For example, the label fields of A and B are one category, the label fields of C and D are one category, and the label field of E is one category, resulting in three categories in total, which are the trait classification results.
[0103] Optionally, after the classification is completed, the trait information of A and B can be stored as one category, the trait information of C and D can be stored as one category, and the trait information of E can be stored as one category according to the above classification results.
[0104] In one embodiment, this classification operation may be the first classification of the trait information of peanuts of different species, wherein, as an example, step S113 may include the following sub-steps:
[0105] S21. If the classification operation is the first classification, call a preset first neural network to aggregate and classify the multiple label fields according to the matching degree of the fields to obtain a plurality of classification results, each of which contains at least one label field.
[0106] Specifically, the information processing platform can determine whether there is already classified and stored trait information in the database. If not, it means that this classification operation is the first classification.
[0107] A preset first neural network may be called to calculate the matching degree between multiple label fields, and the multiple label fields may be aggregated and classified based on the matching degree.
[0108] The first neural network may be a classification model obtained by training a CNN neural network using multiple different fields after the user pre-edits multiple different fields. The specific training method may adopt conventional technical means in the art.
[0109] In the above example, there are five trait information, A, B, C, D, and E, and five label fields correspondingly. The five label fields can be aggregated and classified to obtain classification results, for example, the label fields of A and B are one category, the label fields of C and D are one category, and the label field of E is one category, for a total of three categories.
[0110] In one embodiment, since this classification operation is the first classification performed, after the classification is completed, the label field of this operation can be used as a training field to train the neural network to improve the classification accuracy of the neural network.
[0111] In one embodiment, this classification operation may not be the first classification of the trait information of peanuts of different species, wherein, as an example, step S113 may include the following sub-steps:
[0112] S31. If the classification operation is not the first classification, a preset classification list is obtained, wherein the classification list is composed of a plurality of classification categories, and each of the classification categories corresponds to a category field.
[0113] Specifically, the information processing platform can determine whether there is already classified and stored trait information in the database. If so, it means that this classification operation is not the first classification.
[0114] If the classification operation is not the first classification, the information processing platform can search for the classification result of the previous classification, which can be a classification list, to obtain a preset classification list. This preset classification list can be composed of multiple classification categories, each classification category corresponds to a category field, and also corresponds to a category of trait information. The category field can be a label field corresponding to one of the trait information of this category.
[0115] Because in the previous classification operation, a category may contain one or more trait information, and each trait information has its corresponding label field, and the method for generating this label field is as above. Then a label field of a trait information can be randomly selected as the category field of this classification category.
[0116] S32. Calling a preset second neural network to calculate the field matching degree between each of the label fields and the category field of each of the classification categories.
[0117] Next, a preset second neural network may be called to calculate the matching degree between each label field and the category field of each classification category to obtain multiple field matching degrees.
[0118] The preset second neural network may also be a classification model obtained by training the CNN neural network using multiple different fields after the user pre-edits multiple different fields. The specific training method may adopt conventional technical means in the art.
[0119] S33: Filter the field matching degree with the largest value among the multiple field matching degrees corresponding to each of the label fields, and use the classification category corresponding to the filtered field matching degree with the largest value as the classification result of the label field.
[0120] In an actual operation, since there are multiple classification categories, there are also multiple corresponding category fields. After calculating the matching degree between the label field and each category field, multiple field matching degrees corresponding to each label field can be calculated. The field matching degree with the largest value can be filtered from multiple field matching degrees, and the classification category corresponding to the field matching degree with the largest value can be used as the classification category of this label field. When each label field has a classification category, the classification result is obtained.
[0121] It should be noted that, in an optional embodiment, the label field may not match the category field of each classification category.
[0122] In order to avoid the above situation, after filtering and obtaining the field matching degree with the largest filtering value, it can be determined whether the field matching degree is greater than the preset matching degree. If so, the classification category corresponding to the field matching degree with the largest value can be used as the classification category of this label field. If not, it means that this label field does not match the category field of each classification category, and a category can be reassigned to the trait information of this label field, and this label field can be used as the category field of this category.
[0123] In one embodiment, different users may add annotation information, and different users may add different contents and even different formats. For example, some users add English, some users add Chinese, or numbers, etc. In order to unify the format of information and reduce the deviation of subsequent operations, as an example, step S11 may include the following sub-steps:
[0124] S41, converting the format of each of the annotation information to obtain conversion information, and performing classification operations using the plurality of conversion information.
[0125] Specifically, each annotation information may be converted into a format to obtain conversion information, and the format conversion may be converted into a number of specific digits or a specific base. If it is in Chinese or English, it may be converted into text first, and then the text may be converted into numbers.
[0126] After the conversion is completed, the classification operation is performed according to the content disclosed in the above embodiment.
[0127] The present invention can first collect the annotation information added by the user to the data contained in each trait information, then convert the annotation information into label fields of trait information of different species of peanuts, and finally classify multiple label fields and store different trait information according to the classification results, thereby improving the classification accuracy.
[0128] S12. After displaying the trait classification result, obtaining a user's fitness analysis request.
[0129] Specifically, the trait information of multiple peanuts of different genera or types can be classified according to the operation method of the above embodiment, and after the classification is completed, the multiple trait information is stored in the database of the information processing platform.
[0130] The trait information of different classification results can be displayed, or the trait information of all categories can be displayed to the user in a table form, and the user can view and select one or more trait information of any genus or multiple trait information of different genera. The selected trait information is used to calculate the fitness.
[0131] In an actual operation mode, after displaying various information, the user's fitness analysis request can be obtained, and the request can include the soil data that the user wants to plant, and can also include the trait information selected by the user.
[0132] S13. Based on the peanut species corresponding to the fitness analysis request, extract several target trait information from the trait classification result.
[0133] In a specific implementation, a fitness analysis request may be responded to, the trait information selected by the user may be determined according to the fitness analysis request, and then a number of target trait information for fitness calculation may be extracted from the database.
[0134] In one embodiment, the prior property information may include soil data for the crops, namely, the location, type of soil, pH value, fertilizer content, etc.
[0135] In an actual operation mode, after extracting the target trait information, it can be determined whether the soil data contained in the target trait information to be selected is the same or similar to the soil data to be planted corresponding to the fitness analysis request, and the above-mentioned land type, pH value, and fertilizer content can be used as comparison criteria. For example, if the difference between the two values is less than a preset threshold, it means that the two data are similar.
[0136] If they are similar, the target trait information can be directly used for subsequent calculations. If they are not similar, it means that the peanut species corresponding to the target trait information has not been planted in this soil or has been planted but its data has not been recorded. At this time, it may not be appropriate to use the target trait information for subsequent calculations, and a prompt message can be fed back to the user so that the user can reselect the target trait information.
[0137] S14, calculating the planting adaptability based on the planting yield and planting area of the plurality of trait information.
[0138] Whether a plant is suitable can be intuitively evaluated by its yield, and the yield can be used as an auxiliary evaluation based on the planting area. The planting adaptability can be calculated based on the relationship between the planting yield and the planting area.
[0139] Wherein, as an example, step S14 may include the following sub-steps:
[0140] S141, calculating the ratio of the planting yield to the planting area of each trait information to obtain the yield-to-area ratio.
[0141] S142. Allocate corresponding calculation weight values based on the yield-area ratio.
[0142] S143, performing weighted averaging on the plurality of calculated weight values to obtain planting adaptability.
[0143] Specifically, each trait information will include the planting yield and planting area when planted in a certain soil.
[0144] The planting yield is a specific yield value, for example, 50 jin or 200 jin, and the planting area is the actual planting area of the peanuts, for example, one mu or three mu, or 30 square meters or 60 square meters.
[0145] Next, the ratio of crop yield to crop area can be calculated, for example, 0.1, 0.5, 6.4, 8.9, and so on.
[0146] Then, it can be determined whether the yield-area ratio is greater than the preset threshold. If the yield-area ratio is greater than the preset threshold, the yield-area ratio can be added to the preset first constant to obtain the calculated weight value; if the yield-area ratio is less than the preset threshold, the yield-area ratio can be subtracted from the preset second constant to obtain the calculated weight value.
[0147] For another example, if the yield-area ratio is greater than a preset threshold, the yield-area ratio can be multiplied by a preset first constant to obtain a calculated weight value; if the yield-area ratio is less than the preset threshold, the yield-area ratio can be multiplied by a preset second constant to obtain a calculated weight value.
[0148] The specific method can be adjusted according to actual needs.
[0149] Finally, after calculating the calculation weight value corresponding to each trait information, a plurality of calculation weight values can be weighted averaged to finally obtain the planting fitness.
[0150] Users can determine whether the peanut variety corresponding to the target trait information is suitable for planting based on the planting adaptability.
[0151] Because when extracting the target trait information, each peanut with target trait information has been planted in this soil, therefore, combining multiple target trait information for calculation can effectively improve the precision and accuracy of the technology.
[0152] In an optional embodiment, if there is only one target trait information, its calculated weight can be directly used as the planting suitability.
[0153] It should be noted that the planting adaptability can be an evaluation value of a certain genus or variety of peanuts planted in a certain place or soil. The higher the planting adaptability, the higher the evaluation value, which means it is more suitable for planting in this place or soil. Conversely, if the planting adaptability is lower, the evaluation value is also lower, which means it is less suitable for planting in this place or soil.
[0154] In order to facilitate the user to view the above calculation results, as an example, the method may further include:
[0155] S15, generating an analysis recommendation table using the planting yield and planting area of the plurality of trait information and the planting fitness, and sending the analysis recommendation table to the user corresponding to the fitness analysis request for the user's reference.
[0156] In one implementation, the planting yield and planting area of each target trait information and the planting fitness can be added to a preset table to generate an analysis recommendation table, and then the information processing platform can directly send the analysis recommendation table to the user corresponding to the fitness analysis request for reference.
[0157] In one embodiment, the information processing platform can also test the average per-acre yield and kernel yield of different varieties of peanuts in the test data against the reference varieties. Based on the above results, the yield information of different varieties is described in text and output to a storage file. Alternatively, the information processing platform performs a visual analysis of the yield of each experimental variety in different pilots and outputs it to a corresponding file.
[0158] Optionally, the information processing platform can also use pandas to visualize the characteristics of different varieties and output the results to a file. Based on the above results, a script program is written to generate a descriptive evaluation text report for each variety. All output results are summarized and sorted, and the results are output to the front-end page of the web page. Users can directly download the sorted report file and the analysis result data file generated by the above steps.
[0159] Alternatively, if the user needs to test the regional adaptability of the variety, the information processing platform can read the data table of the variety uploaded by the expert and the corresponding regional planting conditions to verify the results. According to the different regions of the user's input file, the recommended output table is given for the user's reference.
[0160] In this embodiment, an embodiment of the present invention provides a method for calculating the fitness of a planting area, and its beneficial effects are: the present invention can obtain a user's fitness analysis request after displaying the trait information of different species of peanuts that have been classified; extract a number of target trait information based on the peanut species corresponding to the fitness analysis request; and then calculate the planting fitness based on the planting yield and planting area of multiple trait information to improve the calculation accuracy, and determine whether the corresponding species of peanuts are suitable for planting in the corresponding land based on the planting fitness.
[0161] Reference Figure 2 , shows an operation flow chart of a fitness calculation method based on peanut trait information classification provided by an embodiment of the present invention.
[0162] Specifically, the operation of the above method may include the following steps:
[0163] The first step is user information management.
[0164] The second step is user project analysis data entry management.
[0165] The third step is to enter and display variety information.
[0166] Step 4: The super administrator adds the corresponding storage warehouse.
[0167] Step 5: The system initializes the commonly used settings by default.
[0168] Step 6: Users add the required public traits.
[0169] Step 7: Classification of trait information.
[0170] Step 8: Summarize and organize based on the output results.
[0171] Step 9. Download the main report file and analysis data result file.
[0172] Step 10: Read the data table to verify the results.
[0173] Step 11. Output reference table.
[0174] The embodiment of the present invention also provides a fitness calculation device based on peanut trait information classification, see Figure 3 , showing a structural schematic diagram of a fitness calculation device based on peanut trait information classification provided by an embodiment of the present invention.
[0175] Wherein, as an example, the fitness calculation device based on peanut trait information classification may include:
[0176] The classification module 301 is used to collect the trait information of multiple peanuts of different species and classify them based on the fields corresponding to the trait information to obtain the trait classification results;
[0177] An acquisition request module 302 is used to acquire a user's fitness analysis request after displaying the trait classification result;
[0178] An information extraction module 303 is used to extract a number of target trait information from the trait classification result based on the peanut species corresponding to the fitness analysis request;
[0179] The fitness calculation module 304 is used to calculate the planting fitness based on the planting yield and planting area of the plurality of trait information.
[0180] Optionally, the classification module is further used to:
[0181] Acquire annotation information added by a user, where the annotation information is annotation content added by the user to each data included in each of the property information;
[0182] Merging the multiple annotation information corresponding to each of the trait information into a field to obtain multiple label fields, wherein each of the trait information corresponds to one label field;
[0183] Classify multiple label fields to obtain trait classification results.
[0184] Optionally, the classification module is further used to:
[0185] Convert the plurality of annotation information corresponding to each of the trait information into fields respectively to obtain a plurality of annotation fields;
[0186] Extracting a characteristic field of each of the marked fields, wherein the characteristic field is a field with a preset number of digits in each of the marked fields;
[0187] The plurality of annotation fields are sorted based on the numerical values of the plurality of characteristic fields, and the plurality of annotation fields are sequentially concatenated according to the sorting result to obtain a label field.
[0188] Optionally, the classification module is further used to:
[0189] If the classification operation is the first classification, the preset first neural network is called to aggregate and classify the multiple label fields according to the matching degree of the fields to obtain a plurality of individual trait classification results, each of which contains at least one label field.
[0190] Optionally, the classification module is further used to:
[0191] If the classification operation is not the first classification, a preset classification list is obtained, wherein the classification list is composed of a plurality of classification categories, and each of the classification categories corresponds to a category field;
[0192] Calling a preset second neural network to calculate the field matching degree between each of the label fields and the category field of each of the classification categories;
[0193] The field matching degree with the largest value is selected from the multiple field matching degrees corresponding to each of the label fields, and the classification category corresponding to the field matching degree with the largest selected value is used as the trait classification result of the label field.
[0194] Optionally, the fitness calculation module is further used to:
[0195] Calculating the ratio of the planting yield to the planting area of each trait information to obtain the yield-area ratio;
[0196] Based on the yield-area ratio, a corresponding calculation weight value is assigned;
[0197] A weighted average is performed on a plurality of the calculated weight values to obtain the planting fitness.
[0198] Optionally, the device further comprises:
[0199] The format conversion module is used to convert the format of each of the annotation information to obtain conversion information, and perform classification operations using multiple conversion information.
[0200] Optionally, the device further comprises:
[0201] The analysis table making module is used to generate an analysis recommendation table using the planting yield and planting area of multiple trait information and the planting fitness, and send the analysis recommendation table to the user corresponding to the fitness analysis request for user reference.
[0202] Those skilled in the art can clearly understand that, for the sake of convenience and brevity of description, the specific working process of the device described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.
[0203] Furthermore, an embodiment of the present application also provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the fitness calculation method based on peanut trait information classification as described in the above embodiment is implemented.
[0204] Furthermore, an embodiment of the present application also provides a computer-readable storage medium, which stores a computer-executable program, and the computer-executable program is used to enable a computer to execute the fitness calculation method based on peanut trait information classification as described in the above embodiment.
[0205] The above is a preferred embodiment of the present invention. It should be pointed out that a person skilled in the art can make several improvements and modifications without departing from the principle of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.
Claims
1. A fitness calculation method based on peanut trait information classification, characterized in that: The method comprises: After collecting the trait information of multiple peanuts of different species, obtaining the annotation information added by the user, wherein the annotation information is the annotation content added by the user to each data included in each of the trait information; The plurality of annotation information corresponding to each of the trait information are respectively converted into fields to obtain a plurality of annotation fields; a characteristic field of each of the annotation fields is extracted, wherein the characteristic field is a field with a preset number of digits in each of the annotation fields; the plurality of annotation fields are sorted based on the numerical values of the plurality of characteristic fields, and the plurality of annotation fields are sequentially spliced according to the sorting results to obtain a plurality of label fields, wherein each of the trait information corresponds to one label field; If the classification operation is the first classification, calling a preset first neural network to aggregate and classify the plurality of label fields according to the matching degree of the fields, and obtaining a plurality of individual trait classification results, each of which contains at least one label field; If the classification operation is not the first classification, a preset classification list is obtained, the classification list is composed of several classification categories, and each classification category corresponds to a category field; a preset second neural network is called to calculate the field matching degree of each of the label fields and the category field of each of the classification categories; the field matching degree with the largest value is selected from the multiple field matching degrees corresponding to each of the label fields, and the classification category corresponding to the field matching degree with the largest value is used as the trait classification result of the label field; After displaying the trait classification result, obtaining a user's fitness analysis request; Based on the peanut species corresponding to the fitness analysis request, extracting a plurality of target trait information from the trait classification result; Calculating planting adaptability based on the planting yield and planting area of the plurality of trait information; The step of calculating the planting adaptability based on the planting yield and planting area of the plurality of trait information includes: Calculating the ratio of the planting yield to the planting area of each trait information to obtain the yield-area ratio; Based on the yield-area ratio, a corresponding calculation weight value is assigned; A weighted average is performed on a plurality of the calculated weight values to obtain the planting fitness.
2. The fitness calculation method based on peanut trait information classification according to claim 1, characterized in that: After the step of obtaining the annotation information added by the user, the method further includes: Each of the annotation information is formatted to obtain conversion information, and a plurality of the conversion information are used to perform classification operations.
3. The fitness calculation method based on peanut trait information classification according to claim 1, characterized in that: After the step of calculating the planting adaptability based on the planting yield and planting area of the plurality of trait information, the method further comprises: An analysis recommendation table is generated using the planting yield and planting area of the plurality of trait information and the planting fitness, and the analysis recommendation table is sent to a user corresponding to the fitness analysis request for user reference.
4. A fitness calculation device based on peanut trait information classification, characterized in that: The device comprises: A classification module, for collecting trait information of multiple peanuts of different species, and classifying them based on fields corresponding to the trait information to obtain trait classification results; An acquisition request module, used to acquire a user's fitness analysis request after displaying the trait classification result; An information extraction module, used for extracting a plurality of target trait information from the trait classification result based on the peanut species corresponding to the fitness analysis request; A fitness calculation module, used for calculating the planting fitness based on the planting yield and planting area of the plurality of trait information; The step of calculating the planting adaptability based on the planting yield and planting area of the plurality of trait information includes: Calculating the ratio of the planting yield to the planting area of each trait information to obtain the yield-area ratio; Based on the yield-area ratio, a corresponding calculation weight value is assigned; Taking a weighted average of the plurality of calculated weight values to obtain a planting adaptability; The classification module is used to classify based on the fields corresponding to the trait information to obtain the trait classification result, including: obtaining the annotation information added by the user, the annotation information is the annotation content added by the user to each data included in each trait information; The plurality of annotation information corresponding to each of the trait information are respectively converted into fields to obtain a plurality of annotation fields; a characteristic field of each of the annotation fields is extracted, wherein the characteristic field is a field with a preset number of digits in each of the annotation fields; the plurality of annotation fields are sorted based on the numerical values of the plurality of characteristic fields, and the plurality of annotation fields are sequentially spliced according to the sorting results to obtain a plurality of label fields, wherein each of the trait information corresponds to one label field; If the classification operation is the first classification, calling a preset first neural network to aggregate and classify the plurality of label fields according to the matching degree of the fields, and obtaining a plurality of individual trait classification results, each of which contains at least one label field; If the classification operation is not the first classification, obtain a preset classification list, which consists of several classification categories, and each classification category corresponds to a category field; call a preset second neural network to calculate the field matching degree between each of the label fields and the category field of each of the classification categories; select the field matching degree with the largest value among the multiple field matching degrees corresponding to each of the label fields, and use the classification category corresponding to the field matching degree with the largest selected value as the trait classification result of the label field.
5. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer-executable program, and the computer-executable program is used to enable a computer to execute the fitness calculation method based on peanut trait information classification as described in any one of claims 1 to 3.