Agricultural multimodal data fusion system and method based on intelligent technology

Abstract

The present invention discloses an agricultural multimodal data fusion system and method based on intelligent technology, which relates to the field of data processing technology, including: an acquisition and collection module, which is used to acquire sub-regions and collect agricultural data of sub-regions; a feature extraction module, which is used to acquire first data, second data and third data; an index calculation module, which is used to acquire the first decision index corresponding to the sub-region, and acquire the second decision index corresponding to the sub-region, and acquire the third decision index corresponding to the sub-region; a judgment and fusion module, which is used to acquire the real-time correlation coefficients of adjacent sub-regions, and judge whether all of the multiple real-time correlation coefficients are greater than a preset threshold value, and if so, acquire the fusion decision index corresponding to the sub-region based on the fusion model, the first decision index, the second decision index and the third decision index. The present invention has the advantages of comprehensive and accurate data fusion, reliable processing and high efficiency.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to an agricultural multimodal data fusion system and method based on intelligent technology. Background Art

[0002] In recent years, with the rapid development of Internet of Things technology, big data analysis and artificial intelligence technology, the application of multimodal data fusion technology in the agricultural field has gradually attracted attention. Multimodal data fusion refers to the integration of information from different sensors or data sources to obtain a more complete and accurate description of the agricultural environment.

[0003] However, existing agricultural data fusion methods often lack refined division and separate monitoring of agricultural environmental characteristics, resulting in incomplete data acquisition; at the same time, in the feature extraction and indicator calculation stages, data processing efficiency is low, the fusion strategy is single, and the calculation of decision indicators is inaccurate; in addition, existing fusion strategies often ignore the mutual influence between adjacent agricultural areas, resulting in a lack of spatial coherence in the calculation of decision indicators, and a lack of verification mechanism for abnormal areas, which further affects the reliability of agricultural multimodal data fusion results. Summary of the invention

[0004] In view of the defects in the prior art, the present invention provides an agricultural multimodal data fusion system and method based on intelligent technology, aiming to achieve efficient fusion of multimodal data and accurate decision-making in the agricultural environment through intelligent data collection, feature extraction, indicator calculation and decision fusion process.

[0005] A method for fusion of agricultural multimodal data based on intelligent technology comprises: an acquisition and collection module, which is used to acquire a monitoring area and divide the monitoring area into a plurality of sequentially adjacent sub-areas, and collect agricultural data of the sub-areas; a feature extraction module, which is used to acquire first data under a temperature monitoring type, second data under a humidity monitoring type, and third data under a soil moisture monitoring type in the sub-area according to the agricultural data; an index calculation module, which is used to acquire a first decision index corresponding to the sub-area according to the first data and a temperature decision model, and acquire a second decision index corresponding to the sub-area according to the second data and a humidity decision model, and acquire a third decision index corresponding to the sub-area according to the third data and a soil moisture decision model; a judgment and fusion module, which is used to acquire a real-time correlation coefficient of an adjacent sub-area according to the first decision index, the second decision index, and the third decision index of the adjacent sub-area, and judge whether all the multiple real-time correlation coefficients are greater than a preset threshold value, and if so, acquire a fusion decision index corresponding to the sub-area based on the fusion model, the first decision index, the second decision index, and the third decision index.

[0006] Optionally, the temperature decision model in the indicator calculation module is expressed as: ;in, is the first decision indicator, is the first data under the temperature monitoring type at time t, is the target temperature.

[0007] Optionally, the humidity decision model in the indicator calculation module is expressed as: ;in, is the second decision indicator, is the second data under the humidity monitoring type at time t, is the target humidity.

[0008] Optionally, the soil moisture decision model in the indicator calculation module is expressed as: ;in, is the third decision indicator, is the third data under the soil moisture monitoring type at time t, is the target soil moisture content.

[0009] Optionally, the judgment and fusion module is also used to: obtain a first difference based on a first decision indicator of an adjacent sub-area, obtain a second difference based on a second decision indicator of an adjacent sub-area, and obtain a second difference based on the second decision indicator of an adjacent sub-area; obtain a real-time correlation coefficient based on the association model, the first difference, the second difference and the third difference.

[0010] Optionally, the association model in the judgment and fusion module is expressed as: ;in, is the correlation coefficient, is the first difference, is the second difference, is the third difference, is the adjustment coefficient.

[0011] Optionally, the judgment and fusion module is also used to: normalize the first decision indicator, the second decision indicator and the third decision indicator respectively based on the fusion model and generate a first character string, a second character string and a third character string; extract the characteristic character of the first character string as the first character, and extract the characteristic character of the second character string as the second character, and extract the characteristic character of the third character string as the third character; concatenate the first character, the second character and the third character in sequence to form a fusion decision indicator.

[0012] Optionally, the judgment and fusion module is also used to: trigger a fusion exception strategy if multiple real-time correlation coefficients are not all greater than a preset threshold.

[0013] A method for agricultural multimodal data fusion based on intelligent technology is also provided, the method comprising: obtaining a monitoring area and dividing the monitoring area into a plurality of sequentially adjacent sub-areas, and collecting agricultural data of the sub-areas; obtaining first data under a temperature monitoring type, second data under a humidity monitoring type, and third data under a soil moisture monitoring type in the sub-areas according to the agricultural data; obtaining a first decision indicator corresponding to the sub-area according to the first data and a temperature decision model, and obtaining a second decision indicator corresponding to the sub-area according to the second data and a humidity decision model, and obtaining a third decision indicator corresponding to the sub-area according to the third data and a soil moisture decision model; obtaining a real-time correlation coefficient of an adjacent sub-area according to the first decision indicator, the second decision indicator, and the third decision indicator of the adjacent sub-area, and judging whether all the multiple real-time correlation coefficients are greater than a preset threshold value, and if so, obtaining a fusion decision indicator corresponding to the sub-area based on the fusion model, the first decision indicator, the second decision indicator, and the third decision indicator.

[0014] Optionally, obtaining a fusion decision indicator corresponding to a sub-region based on a fusion model, a first decision indicator, a second decision indicator and a third decision indicator includes: normalizing the first decision indicator, the second decision indicator and the third decision indicator respectively based on the fusion model and generating a first character string, a second character string and a third character string; extracting a characteristic character of the first character string as the first character, extracting a characteristic character of the second character string as the second character, and extracting a characteristic character of the third character string as the third character; concatenating the first character, the second character and the third character in sequence to form a fusion decision indicator.

[0015] The beneficial effects of the present invention are embodied in:

[0016] In the entire agricultural multimodal data fusion method based on intelligent technology, the word frequency index of each category of words is first calculated based on the newly added data set, and the word frequency sequence is formed by sorting. Then, the significant fluctuation points in the word frequency sequence are identified through the fluctuation calculation model, and the word frequency index to be processed is screened out accordingly; further, the importance of the word frequency index to be processed is further subdivided by the second threshold, so as to formulate a more reasonable expansion strategy, improve the comprehensiveness and accuracy of data expansion, avoid the introduction of redundant and irrelevant data, and enable the newly added data set to be integrated into the relevant categories in the knowledge graph more quickly. At the same time, it ensures that these categories are arranged in order according to the correlation with the newly added data set, thereby improving agricultural production efficiency, promoting agricultural scientific and technological innovation, and providing strong support for agricultural knowledge management and services. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for the specific embodiments or the description of the prior art. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn according to the actual scale.

[0018] Figure 1 It is a schematic diagram of the composition of the agricultural multimodal data fusion system based on intelligent technology of the present invention;

[0019] Figure 2 A schematic diagram of the steps of the agricultural multimodal data fusion method based on intelligent technology of the present invention;

[0020] Figure 3 It is a schematic diagram of some steps of S4 in the agricultural multimodal data fusion method based on intelligent technology of the present invention.

[0021] Reference numerals:

[0022] 100-acquisition and collection module, 200-feature extraction module, 300-indicator calculation module, 400-judgment and fusion module. DETAILED DESCRIPTION

[0023] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be described clearly and completely in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0024] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. 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.

[0025] It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in the subsequent drawings. In addition, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0026] like Figure 1 As shown, an agricultural multimodal data fusion system based on intelligent technology is provided, including:

[0027] The acquisition and collection module 100 is used to acquire the monitoring area and divide the monitoring area into a plurality of sequentially adjacent sub-areas, and collect agricultural data of the sub-areas;

[0028] A feature extraction module 200, for acquiring first data of a temperature monitoring type, second data of a humidity monitoring type, and third data of a soil moisture monitoring type in a sub-region according to the agricultural data;

[0029] An index calculation module 300, used to obtain a first decision index corresponding to the sub-region according to the first data and the temperature decision model, obtain a second decision index corresponding to the sub-region according to the second data and the humidity decision model, and obtain a third decision index corresponding to the sub-region according to the third data and the soil moisture decision model;

[0030] The judgment and fusion module 400 is used to obtain the real-time correlation coefficients of adjacent sub-regions based on the first decision indicator, the second decision indicator and the third decision indicator of the adjacent sub-regions, and to judge whether all multiple real-time correlation coefficients are greater than a preset threshold. If so, the fusion decision indicator corresponding to the sub-region is obtained based on the fusion model, the first decision indicator, the second decision indicator and the third decision indicator.

[0031] In this embodiment, it should be noted that the entire system is able to achieve refined monitoring of the agricultural environment, efficiently extract key features and accurately calculate decision-making indicators, and fully consider the mutual influence between adjacent agricultural areas, thereby providing more scientific and accurate agricultural management decision support.

[0032] Furthermore, when the acquisition and collection module 100 is executed, the monitoring area is divided and agricultural data of each sub-area is collected. When the acquisition and collection module is in operation, first, through high-precision geographic information technology, the system can automatically identify and obtain the agricultural area to be monitored. This step ensures the accuracy and pertinence of subsequent data processing; then the area is divided, such as based on the location of the multimodal data collection terminal, agricultural terrain, crop type, soil type and other information. It is only necessary to accurately divide the entire monitoring area into a plurality of adjacent sub-areas, and each sub-area includes at least one group of multimodal data collection terminals. This division not only helps to realize the refined collection of data, but also provides more detailed basic data for subsequent feature extraction and indicator calculation. For example, in an actual agricultural monitoring scenario, the acquisition and collection module first uses satellite remote sensing images and drone aerial photography data to accurately identify a vast wheat field and determine its boundaries. Subsequently, the module divides the wheat field into several sub-areas based on factors such as the terrain, crop growth conditions, and soil type distribution in the wheat field. Each sub-area is equipped with temperature, humidity, and soil moisture sensors, which can collect agricultural data in their respective sub-areas in real time, such as key information such as temperature, humidity, and soil moisture content.

[0033] Furthermore, when the feature extraction module 200 is executed, the data to be fused is obtained from the agricultural data, laying the foundation for subsequent indicator calculation and fusion decision-making. After receiving the collected agricultural data of the sub-area, the module will perform detailed data preprocessing and obtain agricultural data of different monitoring types, such as temperature, humidity and soil moisture, respectively; specifically, the feature extraction module will extract the temperature data at the current moment as the first data, the humidity data at the current moment as the second data, and the soil moisture data at the current moment as the third data. Through these data extractions, the feature extraction module can provide accurate and comprehensive data support for subsequent indicator calculation and fusion decision-making.

[0034] Furthermore, when the index calculation module 300 is executed, the raw data from the feature extraction module is converted into a decision index with practical decision guidance. The module first applies the temperature decision model, the humidity decision model and the soil moisture decision model according to different types of agricultural data such as temperature, humidity and soil moisture. The target of these decision models is single-mode data, which can accurately reflect the relationship between decision and environmental data under a single mode. Specifically, in the temperature decision model, the optimal temperature is obtained according to the influence of different temperatures on the growth rate of crops, the probability of occurrence of pests and diseases, and the quality of crops, and then the first decision index is calculated based on the optimal temperature and the current temperature data; in the humidity decision model, the optimal humidity is calculated in combination with the demand of crops for different humidity, and then the second decision index is calculated based on the optimal humidity and the current humidity data; in the soil moisture decision model, the optimal soil moisture content is obtained according to the influence of different soil moisture contents on the root growth and water use efficiency of crops, and then the third decision index is calculated based on the optimal soil moisture content and the current soil moisture data.

[0035] For example, in an actual agricultural monitoring scenario, the indicator calculation module receives the temperature, humidity and soil moisture data of a sub-area from the feature extraction module. For the first data, the module inputs it into the temperature decision model, which analyzes that the first data is 1.5°C lower than the optimal growth temperature of the crop, and then calculates the first decision indicator based on this. For humidity data, the module uses the humidity decision model, combined with the crop type and growth stage in the area, to obtain the optimal growth humidity of 85%, and then calculates the second decision indicator based on the current second data and the optimal growth humidity. For soil moisture data, the module applies the soil moisture decision model to analyze that the current soil moisture content is close to the optimal level for crop growth, and obtains the optimal soil moisture content of 0.2, and then calculates the third decision indicator based on the current third data and the optimal soil moisture content. These detailed indicator calculations are helpful for subsequent multimodal data fusion.

[0036] Furthermore, when the judgment and fusion module 400 is executed, the real-time correlation coefficients between adjacent sub-regions and the decision indicators of each sub-region are comprehensively considered to make accurate data fusion decisions. In this module, the real-time correlation coefficients between adjacent sub-regions are first calculated based on the first decision indicator (temperature related), the second decision indicator (humidity related) and the third decision indicator (soil moisture related) of the adjacent sub-regions. These correlation coefficients reflect the similarity and mutual influence of adjacent sub-regions in agricultural environmental parameters, and are the key to ensuring that the data fusion results have spatial coherence. When calculating the real-time correlation coefficient, methods such as Pearson correlation coefficient or Spearman rank correlation coefficient calculation can be used, or simple numerical difference calculation and threshold comparison can be used. Furthermore, the preset threshold is a pre-set standard or limit, and the preset threshold is usually determined according to the actual application scenario and data characteristics to ensure the accuracy and reliability of the fusion result. If the real-time correlation coefficients of all adjacent sub-regions are not greater than the preset threshold, it involves marking and feedback of abnormal sub-regions, and various fusion abnormal warnings and notifications will be issued to implement an abnormal verification mechanism. If the real-time correlation coefficients of all adjacent sub-regions are greater than the preset threshold, this indicates that the agricultural environmental parameters between the sub-regions are relatively stable and interrelated, and data fusion can be performed. At this time, the fusion model will be started, and the first decision indicator, the second decision indicator, and the third decision indicator of each sub-region will be comprehensively processed to calculate the corresponding fusion decision indicator of each sub-region. Each fusion decision indicator can reflect the agricultural environmental status of the monitored area, and fully consider the mutual influence between adjacent regions, providing a more scientific and accurate basis for agricultural management decisions.

[0037] In summary, in the entire agricultural multimodal data fusion system based on intelligent technology, the monitoring area is finely divided through the acquisition and collection module, and the high-precision geographic information technology and multimodal data collection terminal are used to achieve comprehensive and accurate monitoring of the agricultural environment, which provides more detailed and accurate basic data for subsequent feature extraction and indicator calculation; further, in the feature extraction module, agricultural data of different monitoring types can be processed in detail to extract data such as temperature, humidity and soil moisture, laying the foundation for subsequent data fusion and decision-making; further, through the indicator calculation module, the temperature decision model, humidity decision model and soil moisture decision model are used to convert the original data into decision indicators with practical decision-making guidance significance. These indicators can accurately reflect the crop growth environment. The state of the agricultural environment provides a scientific basis for agricultural management; further, the judgment and fusion module fully considers the mutual influence between adjacent agricultural areas, and ensures the spatial coherence of the data fusion results by calculating the real-time correlation coefficient. This module not only uses advanced algorithms to calculate the correlation coefficient, but also introduces a fusion model to comprehensively process the first decision-making index, the second decision-making index and the third decision-making index of each sub-area, so as to obtain the final fusion decision-making index. These fusion decision-making indicators not only comprehensively reflect the agricultural environmental conditions in the monitored area, but also fully consider the mutual influence between regions, providing agricultural managers with more comprehensive and in-depth decision-making support; in addition, the system also has a verification mechanism for abnormal areas, which can detect and handle abnormal situations in time, further improving the reliability and accuracy of data fusion.

[0038] In one embodiment, the temperature decision model in the indicator calculation module is expressed as:

[0039] ;in,

[0040] is the first decision indicator, is the first data under the temperature monitoring type at time t, is the target temperature.

[0041] In this embodiment, it should be noted that: Represents the first decision indicator at time t, which reflects the current temperature With target temperature the degree of difference between is the temperature data monitored by the temperature sensor at time t, that is, the first data under the temperature monitoring type; and It is the target temperature set based on optimal conditions for crop growth or specific management goals.

[0042] Specifically, the expression The square of the difference between the current temperature and the target temperature is calculated, which can amplify the difference between the two, especially when the difference is large, it can be reflected more significantly. The square operation ensures that no matter Is it higher or lower? , the calculated values ​​are all positive. By adding a constant of 1 and taking the natural logarithm, the degree of difference is smoothed, and large differences can be reduced and small differences can be amplified. This is very suitable for processing temperature data because it allows the system to be sensitive to small fluctuations in temperature while avoiding overreaction to extreme temperature differences. Due to the operation of square sum plus 1, the result of the expression is always positive, and it avoids the situation when = When the denominator is zero or the result is zero.

[0043] In one embodiment, the humidity decision model in the index calculation module is expressed as:

[0044] ;in,

[0045] is the second decision indicator, is the second data under the humidity monitoring type at time t, is the target humidity.

[0046] In this embodiment, it should be noted that the humidity decision model is similar to the temperature decision model. Represents the second decision indicator at time t, which is used to reflect the current humidity With target humidity the degree of difference between is the humidity data monitored by the humidity sensor at time t, i.e., the second data under the humidity monitoring type; and It is the target humidity set based on optimal conditions for crop growth or specific management goals.

[0047] Similar to the temperature decision model, here The square of the difference between the current humidity and the target humidity is calculated to amplify the difference between the two, ensuring that no matter Is it higher or lower? , the difference values ​​are all positive. By adding a constant of 1 and taking the natural logarithm, the expression achieves smoothing of the humidity difference, which helps to reduce the large changes in decision indicators caused by small fluctuations in humidity data, thereby improving the robustness of the decision. At the same time, when the humidity difference is very large, the growth rate of the decision indicator will slow down, thereby avoiding the excessive impact of extreme values ​​on the decision results.

[0048] In one embodiment, the soil moisture decision model in the index calculation module is expressed as:

[0049] ;in,

[0050] is the third decision indicator, is the third data under the soil moisture monitoring type at time t, is the target soil moisture content.

[0051] In this embodiment, it should be noted that the expression Partial calculation of current soil moisture content and target soil moisture content The absolute value ensures that no matter Is it higher or lower? , the difference values ​​are all positive. , this expression realizes the exponential amplification of the difference in soil moisture, which means that when the difference between soil moisture and the target value increases, the decision indicator The growth rate of will be very fast, so it can more sensitively reflect the extreme changes in soil moisture, that is, it is also highly sensitive to small changes in soil moisture, and can issue early warnings in time to help agricultural managers take quick measures to adjust irrigation plans. Compared with other modal decision indicators (such as the first decision indicator of temperature, the second decision indicator of humidity, etc.), the soil moisture decision indicator has unique sensitivity and amplification characteristics, which enables it to play an important role in multimodal data fusion.

[0052] In one embodiment, the judgment and fusion module is further used to:

[0053] Obtain a first difference value according to a first decision indicator of adjacent sub-regions, obtain a second difference value according to a second decision indicator of adjacent sub-regions, and obtain a third difference value according to the second decision indicator of adjacent sub-regions;

[0054] A real-time correlation coefficient is obtained based on the correlation model, the first difference, the second difference and the third difference.

[0055] In this embodiment, it should be noted that when the judgment and fusion module is executed, the first difference, the second difference and the third difference are calculated first; assuming that there are two adjacent sub-areas A and B, their first decision indicators are respectively and , then the first difference can be expressed as | ∣, their second decision indicators are and , then the first difference can be expressed as | ∣, their third decision indicators are and , then the first difference can be expressed as | ∣.

[0056] After calculating the first difference, the second difference, and the third difference, the judgment and fusion module will use a predefined association model to calculate the real-time association coefficient between adjacent sub-areas. This association model may be a complex function that takes into account the impact of different decision indicator differences on the association coefficient, as well as the interaction between these impacts, reflecting the similarity or correlation between adjacent sub-areas in terms of temperature, humidity, and soil moisture. The real-time association coefficient can be used in the subsequent decision fusion process to help the system more accurately understand the environmental conditions of the entire farmland and make corresponding management decisions.

[0057] In one embodiment, the association model in the judgment and fusion module is expressed as:

[0058] ;in,

[0059] is the correlation coefficient, is the first difference, is the second difference, is the third difference, is the adjustment coefficient.

[0060] In this embodiment, it should be noted that, where R is the correlation coefficient, , and They represent the first difference (the difference of the first decision indicator), the second difference (the difference of the second decision indicator) and the third difference (the difference of the third decision indicator) respectively, and a is an adjustment coefficient.

[0061] By dividing the three differences , and This expression comprehensively considers the differences in temperature, humidity and soil moisture between adjacent sub-regions. This comprehensive consideration helps to more comprehensively evaluate the environmental similarities and correlations between sub-regions. To process the sum of the differences, when the difference increases, the value of the correlation coefficient R will decrease rapidly. This exponential decay characteristic can reflect that when the environmental differences between sub-regions increase, the correlation between them will be significantly weakened. The adjustment coefficient a is introduced to adjust the calculation of the correlation coefficient. This coefficient can be set according to the specific farmland environment and crop growth requirements to ensure that the association model can more accurately reflect the actual situation. The value of coefficient a can rely on expert opinions. Specifically, experts or scholars in the agricultural field are invited to set a suitable adjustment coefficient a for specific farmland environment and crop growth requirements based on their professional knowledge and experience; it can also be obtained based on historical data, and the performance of the association model under different adjustment coefficients can be analyzed using historical data, and the best performing coefficient is selected as the value of a.

[0062] In one embodiment, the judgment and fusion module is further used to:

[0063] Based on the fusion model, the first decision index, the second decision index and the third decision index are respectively normalized and a first character string, a second character string and a third character string are generated;

[0064] Extracting a characteristic character of the first character string as the first character, extracting a characteristic character of the second character string as the second character, and extracting a characteristic character of the third character string as the third character;

[0065] The first character, the second character and the third character are sequentially concatenated to form a fusion decision index.

[0066] In this embodiment, it should be noted that after multiple real-time correlation coefficients are all greater than the preset threshold, fusion processing is performed. First, the first decision indicator, the second decision indicator and the third decision indicator are normalized and corresponding character strings are generated. Subsequently, feature characters are extracted from these character strings and concatenated to form a fusion decision indicator.

[0067] Specifically, normalization is a common data preprocessing technology used to convert data of different dimensions or value ranges into a unified standard form for comparison and analysis. In this embodiment, the judgment and fusion module normalizes the first decision indicator, the second decision indicator, and the third decision indicator, including the following steps: respectively determine the theoretical maximum and theoretical minimum values ​​of the decision indicators under the temperature monitoring type, the humidity monitoring type, and the soil moisture monitoring type, which are generally obtained based on relevant records and literature and domain expert experience, and then use the normalization formula, the first string = (first decision indicator - theoretical minimum value of the decision indicator under the temperature monitoring type) / (theoretical maximum value of the decision indicator under the temperature monitoring type - theoretical minimum value of the decision indicator under the temperature monitoring type); the second string = (second decision indicator - theoretical minimum value of the decision indicator under the humidity monitoring type) / (theoretical maximum value of the decision indicator under the humidity monitoring type - theoretical minimum value of the decision indicator under the humidity monitoring type); the third string = (third decision indicator - theoretical minimum value of the decision indicator under the soil moisture monitoring type) / (theoretical maximum value of the decision indicator under the soil moisture monitoring type - theoretical minimum value of the decision indicator under the soil moisture monitoring type). This way, all normalized values ​​will fall between 0 and 1 (or adjust the range as needed).

[0068] Then, a string is generated. For each generated string, the judgment and fusion module extracts its characteristic character. This characteristic character is generally the first character after the decimal point, which can represent the maximum value of the string. For example, if the first string is 0.3, then the first character is 3.

[0069] Finally, the judgment and fusion module concatenates the extracted characteristic characters (the first character, the second character, and the third character) in sequence to form a new string, namely the fusion decision index. This fusion decision index combines information on temperature, humidity, and soil moisture, and can be used for subsequent decision analysis or data fusion processes.

[0070] In one embodiment, the judgment and fusion module is further used to:

[0071] If multiple real-time correlation coefficients are not all greater than the preset threshold, the fusion exception strategy is triggered.

[0072] In this embodiment, it should be noted that when the multiple real-time correlation coefficients are not all greater than the preset threshold, the judgment and fusion module will trigger the fusion exception strategy.

[0073] This strategy may include the following countermeasures. Alarm prompts: Alarm prompts are issued to users through sound, light, text messages, etc. to inform them of fusion anomalies. Data verification: Re-verify the original data to ensure the accuracy and completeness of the data. If there are problems with the data, corrections or re-collection are required. Model adjustment: According to the situation of fusion anomalies, the associated model is adjusted or optimized to improve the accuracy and adaptability of the model. Manual intervention: When necessary, introduce manual experts to intervene and guide and adjust the fusion process based on professional knowledge and experience. Record logs: Record the fusion anomalies and related handling measures in the log for subsequent analysis and improvement.

[0074] like Figure 2 As shown, a method for agricultural multimodal data fusion based on intelligent technology is also provided, and the method includes:

[0075] S1. Obtain a monitoring area and divide the monitoring area into a plurality of adjacent sub-areas, and collect agricultural data of the sub-areas;

[0076] S2. Acquire first data of the temperature monitoring type, second data of the humidity monitoring type, and third data of the soil moisture monitoring type in the sub-area according to the agricultural data;

[0077] S3, obtaining a first decision indicator corresponding to the sub-area according to the first data and the temperature decision model, obtaining a second decision indicator corresponding to the sub-area according to the second data and the humidity decision model, and obtaining a third decision indicator corresponding to the sub-area according to the third data and the soil moisture decision model;

[0078] S4. Obtain the real-time correlation coefficient of the adjacent sub-region according to the first decision indicator, the second decision indicator and the third decision indicator of the adjacent sub-region, and determine whether all the multiple real-time correlation coefficients are greater than the preset threshold. If so, obtain the fusion decision indicator corresponding to the sub-region based on the fusion model, the first decision indicator, the second decision indicator and the third decision indicator.

[0079] like Figure 3 As shown, in one embodiment, obtaining the fusion decision indicator corresponding to the sub-region based on the fusion model, the first decision indicator, the second decision indicator and the third decision indicator includes:

[0080] S41, normalizing the first decision indicator, the second decision indicator, and the third decision indicator respectively based on the fusion model and generating a first character string, a second character string, and a third character string;

[0081] S42, extracting a characteristic character of the first character string as the first character, extracting a characteristic character of the second character string as the second character, and extracting a characteristic character of the third character string as the third character;

[0082] S43, sequentially concatenate the first character, the second character, and the third character to form a fusion decision index.

[0083] In this embodiment, it should be noted that, regarding the above-mentioned agricultural multimodal data fusion method based on intelligent technology, the specific steps have been described in detail in the implementation of the agricultural multimodal data fusion system based on intelligent technology, and will not be elaborated here.

[0084] The preferred embodiments of the present disclosure are described in detail above in conjunction with the accompanying drawings; however, the present disclosure is not limited to the specific details in the above embodiments. Within the technical concept of the present disclosure, a variety of simple modifications can be made to the technical solution of the present disclosure, and these simple modifications all fall within the protection scope of the present disclosure.

[0085] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, the present disclosure will not further describe various possible combinations.

[0086] In addition, various embodiments of the present disclosure may be arbitrarily combined, and as long as they do not violate the concept of the present disclosure, they should also be regarded as the contents disclosed by the present disclosure.

[0087] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. An agricultural multimodal data fusion system based on intelligent technology, characterized in that: include: The acquisition and collection module is used to acquire the monitoring area and divide the monitoring area into a plurality of sequentially adjacent sub-areas, and collect agricultural data of the sub-areas; A feature extraction module, used for acquiring first data under a temperature monitoring type, second data under a humidity monitoring type, and third data under a soil moisture monitoring type in a sub-area according to the agricultural data; The index calculation module is used to obtain the first decision index corresponding to the sub-area according to the first data and the temperature decision model, obtain the second decision index corresponding to the sub-area according to the second data and the humidity decision model, and obtain the third decision index corresponding to the sub-area according to the third data and the soil moisture decision model; the temperature decision model in the index calculation module is expressed as: ;in, is the first decision indicator, is the first data under the temperature monitoring type at time t, is the target temperature; the humidity decision model in the index calculation module is expressed as: ;in, is the second decision indicator, is the second data under the humidity monitoring type at time t, is the target humidity; the soil moisture decision model in the index calculation module is expressed as: ;in, is the third decision indicator, is the third data under the soil moisture monitoring type at time t, is the target soil moisture content; The judgment and fusion module is used to obtain the real-time correlation coefficient of the adjacent sub-region according to the first decision indicator, the second decision indicator and the third decision indicator of the adjacent sub-region, and judge whether all the multiple real-time correlation coefficients are greater than the preset threshold value. If so, the fusion decision indicator corresponding to the sub-region is obtained based on the fusion model, the first decision indicator, the second decision indicator and the third decision indicator; the judgment and fusion module is also used to: obtain the first difference according to the first decision indicator of the adjacent sub-region, obtain the second difference according to the second decision indicator of the adjacent sub-region, and obtain the second difference according to the second decision indicator of the adjacent sub-region; obtain the real-time correlation coefficient based on the correlation model, the first difference, the second difference and the third difference; the correlation model in the judgment and fusion module is expressed as: ;in, is the real-time correlation coefficient, is the first difference, is the second difference, is the third difference, is the adjustment coefficient; The judgment and fusion module is also used to: normalize the first decision indicator, the second decision indicator and the third decision indicator respectively based on the fusion model and generate a first character string, a second character string and a third character string; extract the characteristic character of the first character string as the first character, extract the characteristic character of the second character string as the second character, and extract the characteristic character of the third character string as the third character; splice the first character, the second character and the third character in sequence to form a fusion decision indicator.

2. The agricultural multimodal data fusion system based on intelligent technology according to claim 1 is characterized in that: The judgment and fusion module is also used for: If multiple real-time correlation coefficients are not all greater than the preset threshold, the fusion exception strategy is triggered.

3. An agricultural multimodal data fusion method based on intelligent technology, characterized in that: The method comprises: Acquire a monitoring area and divide the monitoring area into a plurality of sequentially adjacent sub-areas, and collect agricultural data of the sub-areas; Acquire first data under the temperature monitoring type, second data under the humidity monitoring type, and third data under the soil moisture monitoring type in the sub-area according to the agricultural data; The first decision index corresponding to the sub-area is obtained according to the first data and the temperature decision model, the second decision index corresponding to the sub-area is obtained according to the second data and the humidity decision model, and the third decision index corresponding to the sub-area is obtained according to the third data and the soil moisture decision model; the temperature decision model in the index calculation module is expressed as: ;in, is the first decision indicator, is the first data under the temperature monitoring type at time t, is the target temperature; the humidity decision model in the index calculation module is expressed as: ;in, is the second decision indicator, is the second data under the humidity monitoring type at time t, is the target humidity; the soil moisture decision model in the index calculation module is expressed as: ;in, is the third decision indicator, is the third data under the soil moisture monitoring type at time t, is the target soil moisture content; Obtaining a real-time correlation coefficient of an adjacent sub-region according to the first decision indicator, the second decision indicator, and the third decision indicator of the adjacent sub-region, and judging whether all the multiple real-time correlation coefficients are greater than a preset threshold value, and if so, obtaining a fusion decision indicator corresponding to the sub-region based on the fusion model, the first decision indicator, the second decision indicator, and the third decision indicator; Wherein, obtaining the real-time correlation coefficient of the adjacent sub-region according to the first decision indicator, the second decision indicator and the third decision indicator of the adjacent sub-region includes: obtaining the first difference according to the first decision indicator of the adjacent sub-region, obtaining the second difference according to the second decision indicator of the adjacent sub-region, and obtaining the second difference according to the second decision indicator of the adjacent sub-region; obtaining the real-time correlation coefficient based on the correlation model, the first difference, the second difference and the third difference; wherein the correlation model in the judgment and fusion module is expressed as: ;in, is the real-time correlation coefficient, is the first difference, is the second difference, is the third difference, is the adjustment coefficient; Among them, the method of obtaining the fusion decision indicator corresponding to the sub-region based on the fusion model, the first decision indicator, the second decision indicator and the third decision indicator includes: normalizing the first decision indicator, the second decision indicator and the third decision indicator respectively based on the fusion model and generating a first character string, a second character string and a third character string; extracting the characteristic character of the first character string as the first character, extracting the characteristic character of the second character string as the second character, and extracting the characteristic character of the third character string as the third character; splicing the first character, the second character and the third character in sequence to form a fusion decision indicator.

Images