Tobacco growth quality prediction method and program product

Abstract

The invention provides a tobacco growth quality prediction method and a program product. The tobacco growth quality prediction method comprises the following steps: acquiring video data of tobacco in a growth process; acquiring first soil data and second soil data of soil in a tobacco planting area where the tobacco is located; the first soil data comprises at least one of soil temperature, soil humidity and near infrared spectrum data; the second soil data comprises at least one of nitrogen content, phosphorus content, potassium content, conductivity and pH value; through the quality prediction model, the growth quality information of the tobacco is determined based on the video data, the first soil data and the second soil data, prediction of the growth quality of the tobacco is achieved, the accuracy and efficiency of prediction of the growth quality of the tobacco are improved, and data support is provided for subsequent growth of the tobacco.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method and program product for predicting tobacco growth quality. Background Technology

[0002] As the area under tobacco cultivation increases, predicting the growth quality of tobacco becomes increasingly important.

[0003] In existing technologies, agricultural technicians conduct manual field inspections of large tobacco-growing areas, observing each tobacco leaf individually and recording its growth quality.

[0004] Manually inspecting tobacco fields in large-scale tobacco-growing areas and recording the growth quality of the leaves is time-consuming and inefficient. Furthermore, the growth quality of tobacco leaves is easily influenced by the subjective experience of agricultural technicians, leading to low accuracy in predicting the quality of tobacco leaves. Summary of the Invention

[0005] This invention provides a method and program product for predicting tobacco growth quality, so as to improve the accuracy and efficiency of tobacco growth quality prediction.

[0006] According to one aspect of the present invention, a method for predicting tobacco growth quality is provided, the method comprising:

[0007] Acquire video data of tobacco during its growth process;

[0008] Obtain first and second soil data of the soil in the tobacco-growing area; the first soil data includes at least one of soil temperature, soil moisture and near-infrared spectral data; the second soil data includes at least one of nitrogen content, phosphorus content, potassium content, electrical conductivity and pH value.

[0009] The quality prediction model is used to determine the growth quality information of tobacco based on video data, first soil data, and second soil data.

[0010] According to another aspect of the present invention, a tobacco growth quality prediction device is provided, the device comprising:

[0011] The video data acquisition module is used to acquire video data of tobacco during its growth process;

[0012] The soil data acquisition module is used to acquire first soil data and second soil data of the soil in the tobacco planting area where the tobacco is located; the first soil data includes at least one of soil temperature, soil moisture and near-infrared spectral data; the second soil data includes at least one of nitrogen content, phosphorus content, potassium content, electrical conductivity and pH value.

[0013] The growth quality information determination module is used to determine the growth quality information of tobacco based on video data, first soil data, and second soil data through a quality prediction model.

[0014] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0015] At least one processor; and

[0016] A memory that is communicatively connected to at least one processor; wherein,

[0017] The memory stores a computer program that can be executed by at least one processor, such that the at least one processor is able to perform a method for predicting tobacco growth quality provided in any embodiment of the present invention.

[0018] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement a tobacco growth quality prediction method provided in any embodiment of the present invention.

[0019] According to another aspect of the present invention, a computer program product is provided, comprising a computer program, characterized in that the computer program, when executed by a processor, implements a tobacco growth quality prediction method provided in any embodiment of the present invention.

[0020] The technical solution of this invention achieves video data acquisition by obtaining video data of tobacco during its growth process; it also acquires first and second soil data of the soil in the tobacco-growing area, where the first soil data includes at least one of soil temperature, soil moisture, and near-infrared spectral data, and the second soil data includes at least one of nitrogen content, phosphorus content, potassium content, electrical conductivity, and pH value, thus achieving the acquisition of both first and second soil data and providing data support for subsequent analysis and processing; and through a quality prediction model, it determines the growth quality information of tobacco based on the video data, first soil data, and second soil data, achieving the prediction of tobacco growth quality, solving the problems of low accuracy and low efficiency in the prediction of tobacco growth quality in the prior art, and improving the efficiency and accuracy of tobacco growth quality prediction.

[0021] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a flowchart of a method for predicting tobacco growth quality provided in an embodiment of the present invention;

[0024] Figure 2 This is a flowchart of another method for predicting tobacco growth quality provided in an embodiment of the present invention;

[0025] Figure 3 This is a flowchart of another method for predicting tobacco growth quality provided in an embodiment of the present invention;

[0026] Figure 4 This is a schematic diagram of the structure of a tobacco growth quality prediction device provided in an embodiment of the present invention;

[0027] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0028] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0029] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0030] Figure 1This is a flowchart of a tobacco growth quality prediction method provided in an embodiment of the present invention. This embodiment is applicable to situations where the growth quality of tobacco needs to be predicted. The method can be executed by a tobacco growth quality prediction device, which can be implemented in hardware and / or software. This device can be configured in an electronic device provided in this embodiment of the present invention. The electronic device can be a server, computer, or mobile terminal, such as a mobile phone or tablet computer. Figure 1 As shown, the method includes:

[0031] S110. Acquire video data of tobacco during its growth process.

[0032] The video data consists of image data recording the growth status of tobacco. The video data includes multiple video frames. The growth status of tobacco can be recorded using video capture devices to obtain video data of the tobacco during its growth process. Video capture devices include, but are not limited to, high frame rate webcams. These high frame rate webcams support the High Efficiency Video Coding (HEVC) 2.6 video compression standard and a video capture capability of 60 frames per second. Video data of the tobacco growth process can also be obtained through a video database. The video database can store multiple video data sets, and the data can be matched using a unique identifier to obtain the video data of the tobacco during its growth process. Optionally, the video data can be acquired using video capture devices deployed at designated locations within the tobacco planting area. The number of designated locations in the tobacco planting area can be three or five, depending on the requirements. Taking six designated locations as an example, with a rectangular shape for the tobacco planting area, the video capture devices can be deployed at the entrance, center, and four corners of the tobacco planting area. By deploying video acquisition devices at designated locations in tobacco-growing areas to collect video data on the tobacco growth process, we can obtain video data from all angles during the tobacco growth process, providing comprehensive data support for subsequent analysis and processing.

[0033] Optionally, the triggering conditions for video data acquisition include time-based triggering conditions.

[0034] The acquisition trigger condition refers to the conditions that trigger the video acquisition device to collect video data. The time trigger condition is a pre-set condition based on the time dimension that triggers video data acquisition. When the time trigger condition is met, the video acquisition device collects video data of the tobacco. For example, video acquisition can be triggered at fixed time periods each day, with the time trigger condition being the set acquisition time interval. By setting the time trigger condition, the video acquisition device collects video data of the tobacco when the time trigger condition is met, thus achieving automatic acquisition of tobacco data.

[0035] Optionally, the video data acquisition trigger conditions include image color difference change conditions.

[0036] This system allows for real-time or timed video / image monitoring of tobacco to determine the numerical value of color difference changes in tobacco images. Specifically, it acquires monitoring videos / images of tobacco and determines the numerical value of the color difference change between the monitored video / image and the previously acquired video data. The image color difference change condition is triggered automatically when the numerical value of the color difference change between the monitored video / image and the previously acquired video data exceeds a preset color difference change threshold. For example, by processing the monitored video / image using image analysis algorithms to determine the color difference change rate, the video acquisition device acquires video data of the tobacco when the color difference change rate exceeds 15%. By setting the image color difference change condition, the video acquisition device can accurately capture the color changes of the tobacco, providing accurate data support for subsequent analysis.

[0037] Optionally, the video data acquisition trigger conditions include time-based trigger conditions and image color difference change conditions.

[0038] Specifically, when both the time-triggered condition and the image color difference change condition are met simultaneously, the video acquisition device can accurately capture the color change of the tobacco by acquiring video data, thus reducing the waste of resources.

[0039] Optionally, there may be multiple tobacco growing areas, each with its own regional coding information.

[0040] There are multiple tobacco growing areas, and a single tobacco growing area may include the same variety of tobacco or different varieties. The area code information serves as a unique identifier for each tobacco growing area.

[0041] Specifically, a Geographic Information System (GIS) can be used to divide the total tobacco-growing area into multiple tobacco-growing zones based on preset planting areas and spacing. The GIS stores boundary data, land use data, and spatial analysis algorithms for the total tobacco-growing area. These algorithms, along with the corresponding regional codes for each zone, allow for precise delineation of tobacco-growing areas, facilitating zoned management.

[0042] Optionally, there are multiple tobacco growing areas, each with corresponding location information.

[0043] The location information refers to the specific geographical location of each tobacco-growing area. For example, the location information could be the latitude and longitude of the center of the tobacco-growing area.

[0044] Optionally, there may be multiple tobacco growing areas, each with corresponding area coding information and location information.

[0045] S120. Obtain first soil data and second soil data of the soil in the tobacco planting area where the tobacco is located; the first soil data includes at least one of soil temperature, soil moisture and near-infrared spectral data; the second soil data includes at least one of nitrogen content, phosphorus content, potassium content, electrical conductivity and pH value.

[0046] The first soil data refers to parameters reflecting the physical state and basic spectral characteristics of the soil, including at least one of soil temperature, soil moisture, and near-infrared spectral data. Optionally, the first soil data is obtained based on detection by pre-deployed sensors in the tobacco growing area. Different types of first soil data can be obtained using different types of sensors. For example, soil temperature can be detected by temperature sensors deployed in the tobacco growing area; soil moisture can be detected by moisture sensors deployed in the tobacco growing area; and near-infrared (NIR) spectral data can be obtained by measuring moisture sensors deployed in the tobacco growing area using NIR sensors. Obtaining the first soil data through pre-deployed sensors in the tobacco growing area achieves accurate determination of the first soil data, providing accurate sensor data for subsequent analysis and processing.

[0047] Based on the above embodiments, after multiple detections by the same type of sensor, multiple detection data are obtained. The average value of the multiple detection data is then processed by Kalman filtering to obtain the first soil data.

[0048] The second soil data comprises parameters reflecting the chemical and physicochemical properties of the soil, including at least one of nitrogen content, phosphorus content, potassium content, electrical conductivity, and pH value. Optionally, the second soil data is obtained by analyzing soil samples periodically collected from the tobacco-growing area using soil sampling and analysis equipment. This equipment includes, but is not limited to, atomic absorption spectrometry. Optionally, the second soil data can be generated by at least one of the following methods: determining the nitrogen and phosphorus content of the soil in the tobacco-growing area using atomic absorption spectrometry; determining the potassium content of the soil in the tobacco-growing area using colorimetry; and determining the electrical conductivity and pH value of the soil in the tobacco-growing area using the standard electrode method.

[0049] Atomic absorption spectrometry (AAS) is an instrument that determines the elemental content of a substance based on the absorption characteristics of gaseous atoms to light of specific wavelengths. The process of determining the nitrogen content of soil in a tobacco-growing area using AAS is as follows: soil sampling and pretreatment, nitrogen conversion, and content calculation. In the soil sampling and pretreatment process, soil is collected from the tobacco-growing area, and impurities such as stones and plant debris are removed. The soil is placed in a digestion tube, a mixed digestion solution is added, and the mixture is heated on a hot plate until the solution becomes clear. After cooling, the digestion solution is transferred to a volumetric flask with deionized water, diluted to the mark, and allowed to stand. The supernatant is then used as the test solution. In the nitrogen conversion process, the nitrogen in the soil sample is converted into a form that can combine with metal ions. First, a 40% formaldehyde solution is added to the test solution, followed by the addition of sodium hydroxide solution to neutralize it, resulting in a mixed solution. In the content calculation process, the mixed solution is irradiated with an atomic absorption spectrometer to obtain the absorbance of the remaining sodium ions in the mixed solution. The concentration of the remaining sodium ions is calculated according to the standard curve. Based on the stoichiometric relationship, the concentration of hydrogen ions is derived from the amount of sodium ions consumed, and then the concentration of ammonium ions in the test solution is calculated. The concentration of ammonium ions is then converted into the nitrogen content of the soil. The process of determining the phosphorus content of soil in tobacco-growing areas using atomic absorption spectrometry is similar to that of determining the nitrogen content of soil in tobacco-growing areas, and will not be described in detail here.

[0050] The potassium content of soil in tobacco-growing areas was determined using a colorimetric method. The specific process is as follows: soil pretreatment and potassium extraction, colorimetric reaction, standard curve plotting, and soil absorbance measurement and potassium content calculation. In the soil pretreatment and potassium extraction process, soil was collected from tobacco-growing areas, air-dried, ground, and sieved. The soil was then placed in an Erlenmeyer flask, and the extraction reagent was added. The mixture was then shaken on a shaker, and after shaking, it was filtered through qualitative filter paper to obtain the test solution. In the colorimetric reaction process, a certain amount of the test solution was pipetted into a volumetric flask, phenolphthalein indicator was added, and sodium hydroxide solution was added dropwise until the solution turned slightly pink. Equal volumes of ethylenediaminetetraacetic acid solution and formaldehyde solution were added, and the mixture was shaken well to obtain the colorimetric solution to be tested. In the process of constructing the standard curve, equal volumes of potassium standard solutions of 0, 10, 20, 30, and 40 μg / mL were taken and processed according to the colorimetric reaction procedure to obtain multiple standard solutions. Using a spectrophotometer, the absorbance of each standard solution was measured at a wavelength of 420 nm, with the 0 μg / mL standard solution as a blank control. A standard curve was constructed with potassium concentration on the x-axis and the corresponding absorbance on the y-axis. In the process of soil absorbance determination and potassium content calculation, the colorimetric solution to be tested was poured into a cuvette, and the absorbance was measured at a wavelength of 420 nm. Based on the soil absorbance, the corresponding potassium concentration was determined according to the standard curve. The potassium content of the soil was then determined based on the conversion relationship.

[0051] The electrical conductivity of soil in tobacco-growing areas was determined using the standard electrode method. The specific process is as follows: soil pretreatment and extract preparation, conductivity meter calibration, and conductivity measurement. In the soil pretreatment and extract preparation process, soil was collected from the tobacco-growing area, air-dried, ground, and sieved. The soil was then placed in a beaker, deionized water was added, and the mixture was stirred to ensure thorough mixing. The mixture was then shaken on a shaker to ensure all soluble salts were completely dissolved in the water. The extract was then filtered through qualitative filter paper to obtain the soil extract. During conductivity meter calibration, the conductivity electrodes were rinsed with deionized water and dried with filter paper. An appropriate amount of 0.01 mol / L potassium chloride standard solution was placed in a beaker, and the electrodes were immersed in the solution. The conductivity meter was then turned on and calibrated according to the instrument instructions. During conductivity measurement, a certain amount of soil extract was placed in a beaker, and the calibrated electrodes were immersed in the extract. Once the instrument reading stabilized, the conductivity value was recorded. The process of determining the pH value of soil in tobacco-growing areas using the standard electrode method is similar to the process of determining the electrical conductivity of soil in tobacco-growing areas using the standard electrode method, and will not be described in detail here.

[0052] Specifically, the nitrogen and phosphorus contents of the soil in the tobacco-growing area were determined by atomic absorption spectrometry, the potassium content of the soil in the tobacco-growing area was determined by colorimetry, and the electrical conductivity and pH value of the soil in the tobacco-growing area were determined by the standard electrode method. This achieved accurate measurement of the nitrogen, phosphorus, electrical conductivity, and pH value of the soil, providing accurate data support for subsequent analysis and processing.

[0053] Based on the above embodiments, during the soil collection process in tobacco-growing areas, an S-shaped path can be used to collect soil at different depths. For example, an S-shaped path can be used to collect soil at two depths: 0–20 cm and 20–40 cm, obtaining soil samples at varying depths, which helps improve the authenticity and accuracy of the second soil data.

[0054] Optionally, the method further includes: receiving processing requirement information, the processing requirement information including target area encoding information; and obtaining video data corresponding to tobacco in the tobacco planting area, as well as first soil information and second soil information of the soil in the tobacco planting area, based on the target area encoding information.

[0055] The processing requirement information is a directive used to obtain data related to tobacco-growing areas. This information includes target area coding information, which serves as a unique identifier for the tobacco-growing area used for data retrieval. For example, users can input processing requirement information into the GIS through an interactive interface. The GIS can then match this information against a regional coding information database based on the target area coding information, obtaining the corresponding video data, first soil information, and second soil information, thus achieving accurate acquisition of these data.

[0056] Optionally, the method further includes: receiving processing demand information, the processing demand information including target location information; and acquiring video data corresponding to tobacco in the tobacco planting area, as well as first soil information and second soil information of the soil in the tobacco planting area, based on the target location information.

[0057] The target location information refers to the location information used for data querying, which includes, but is not limited to, latitude and longitude information. For example, users can input processing requirements into the GIS through an interactive interface. The GIS can then match the target location information in the processing requirements against the location information database to obtain the corresponding video data, first soil information, and second soil information, thus achieving accurate acquisition of these data.

[0058] Optionally, the method further includes: receiving processing requirement information, the processing requirement information including target area coding information and target location information; and obtaining video data corresponding to tobacco in the tobacco planting area, as well as first soil information and second soil information of the soil in the tobacco planting area, based on the target area coding information and target location information.

[0059] Specifically, users can input processing requirements into the GIS through an interactive interface. The GIS can then match the target area coding information and target location information in the area coding information-location information database to obtain the video data, first soil information, and second soil information corresponding to the matched target area coding information and target location information, thus achieving accurate acquisition of video data, first soil information, and second soil information.

[0060] S130. Based on video data, first soil data, and second soil data, determine the growth quality information of tobacco using a quality prediction model.

[0061] The quality prediction model is used to analyze and process video data, first soil data, and second soil data. The quality prediction model includes, but is not limited to, neural network models, such as convolutional neural network models or machine learning models, which can be selected according to user needs and are not limited here. Optionally, the quality prediction model can be a Transformer model based on a multi-head attention mechanism. The Transformer model based on the multi-head attention mechanism includes a three-layer Transformer encoder and two fully connected layers. Each Transformer encoder layer includes five self-attention heads and a 512-dimensional hidden layer.

[0062] The growth quality information is output by a quality prediction model, reflecting the quality status of tobacco during its growth process. Optionally, the tobacco growth quality information includes at least one of the following: tobacco quality grade information, soil condition information, and pest and disease information. The tobacco quality grade information can include four grades: excellent, good, medium, and inferior, as well as the probability data corresponding to each quality grade. The tobacco quality grade information can be set according to needs and is not limited here. The soil condition information reflects the physical, chemical, and biological characteristics of the soil in the tobacco growing area. For example, soil condition information can include the deviations between the nitrogen content, phosphorus content, potassium content, electrical conductivity, and pH value in the soil and their corresponding expected values. The pest and disease information includes data on pathogens (such as fungi, bacteria, and viruses) and pests (such as aphids and tobacco budworms) that infest the tobacco during its growth.

[0063] Specifically, by using a quality prediction model, video data, primary soil data, and secondary soil data are analyzed and processed to obtain tobacco growth quality information, thereby enabling the prediction of tobacco growth quality and improving the efficiency of tobacco growth quality prediction.

[0064] Optionally, the method further includes: generating a decision work order based on the growth quality information when the tobacco quality grade information meets the first preset condition; wherein the decision work order includes at least one of the following: fertilizer ratio parameters, irrigation parameters, pest and disease control parameters, and recommended execution time.

[0065] The first preset condition refers to the conditions that the tobacco quality grade information must meet based on the tobacco planting objectives. For example, the first preset condition could be that the probability of the intermediate grade in the tobacco quality grade information is greater than or equal to 40%. The decision work order is a document used to guide tobacco planting, and it can be generated based on growth quality information. The decision work order may include one of the following: fertilization ratio parameters, irrigation parameters, pest and disease control parameters, and recommended execution time; it may also include multiple of these parameters. The fertilization ratio parameters are the proportions and specific application rates of major nutrients and micronutrients set to meet the nutrient requirements of tobacco growth. For example, the fertilization ratio parameters may include recommended application rates of nitrogen, phosphorus, and potassium. Irrigation parameters are the data corresponding to irrigating tobacco. For example, irrigation parameters may include irrigation timing and irrigation water volume. Pest and disease control parameters are the data used to control tobacco pests and diseases. For example, pest and disease control parameters may include the type and dosage of pesticides. The recommended implementation time can include the recommended implementation time for fertilizer ratio parameters, irrigation parameters, and pest and disease control parameters.

[0066] Specifically, when the probability of the intermediate grade in the tobacco quality grade information is greater than or equal to 40%, a decision work order is generated based on the growth quality information, providing a management basis for subsequent tobacco planting.

[0067] Optionally, the method further includes: generating a decision work order based on growth quality information when the soil condition information meets the second preset condition; wherein the decision work order includes at least one of fertilizer ratio parameters, irrigation parameters, pest and disease control parameters, and recommended execution time.

[0068] The second preset condition is a condition that the soil condition information must meet based on the tobacco planting goals. For example, the second preset condition could be that the soil electrical conductivity fluctuates by more than ±15% compared to the desired electrical conductivity.

[0069] Specifically, when the soil electrical conductivity fluctuates by more than ±15% compared to the expected electrical conductivity, a decision work order is generated based on the growth quality information, providing a management basis for subsequent tobacco planting.

[0070] Optionally, the method further includes: generating a decision work order based on the growth quality information when the tobacco quality grade information meets the first preset condition and the soil condition information meets the second preset condition; wherein the decision work order includes at least one of the following: fertilizer ratio parameters, irrigation parameters, pest and disease control parameters, and recommended execution time.

[0071] Specifically, when the probability of the intermediate grade in the tobacco quality grade information is greater than or equal to 40%, and the soil electrical conductivity fluctuates by more than ±15% compared to the expected electrical conductivity, a decision work order is generated based on the growth quality information, providing a management basis for subsequent tobacco planting.

[0072] Based on the above embodiments, the method further includes: sending the generated decision work order to the target terminal, so that agricultural technicians can view the decision work order through the target terminal and manage the tobacco according to the decision work order.

[0073] The technical solution of this embodiment achieves video data acquisition by obtaining video data of tobacco during its growth process; it also acquires first and second soil data of the soil in the tobacco-growing area, where the first soil data includes at least one of soil temperature, soil moisture, and near-infrared spectral data, and the second soil data includes at least one of nitrogen content, phosphorus content, potassium content, electrical conductivity, and pH value, thus achieving the acquisition of both first and second soil data and providing data support for subsequent analysis and processing; and through a quality prediction model, it determines the growth quality information of tobacco based on the video data, first soil data, and second soil data, thereby achieving the prediction of tobacco growth quality and improving the efficiency and accuracy of tobacco growth quality prediction.

[0074] Figure 2 This is a flowchart of another method for predicting tobacco growth quality provided by an embodiment of the present invention. This embodiment is a refinement of the above embodiments. Based on the foregoing embodiments, it provides a detailed explanation of how to determine tobacco growth quality information using a quality prediction model based on video data, first soil data, and second soil data. For specific implementation details, please refer to the technical solution of this embodiment. Technical terms that are the same as or corresponding to those in the above embodiments will not be repeated here. Figure 2 As shown, the method includes:

[0075] S310. Acquire video data of tobacco during its growth process.

[0076] S320. Obtain first soil data and second soil data of the soil in the tobacco planting area where the tobacco is located; the first soil data includes at least one of soil temperature, soil moisture and near-infrared spectral data; the second soil data includes at least one of nitrogen content, phosphorus content, potassium content, electrical conductivity and pH value.

[0077] To improve the accuracy of the first and second soil data, dynamic calibration can be performed on both.

[0078] Optionally, the method further includes: dynamically calibrating the first soil data based on the first real data, and dynamically calibrating the second soil data based on the second real data, to obtain the dynamically calibrated first soil data and second soil data; wherein the first real data and the second real data are respectively obtained based on laboratory analysis of soil in tobacco growing areas.

[0079] The first set of real data consists of soil samples from tobacco-growing areas obtained through standard laboratory analysis methods. This data is used for dynamic calibration of the first soil data. The first set of real data corresponds to the first soil data. The first set of real data can be obtained through laboratory analysis of the soil in tobacco-growing areas. The second set of real data also corresponds to the second soil data. The second set of real data can be obtained through laboratory analysis of the soil in tobacco-growing areas.

[0080] Specifically, the first soil data is dynamically calibrated based on the first set of real data, and the second soil data is dynamically calibrated based on the second set of real data, resulting in dynamically calibrated first and second soil data. This dynamic calibration of the first and second soil data improves their accuracy and, consequently, the accuracy of tobacco growth quality prediction.

[0081] Optionally, the first soil data is dynamically calibrated based on the first real data to obtain the dynamically calibrated first soil data, including: determining the first calibration coefficient corresponding to the first soil data based on the first real data; and dynamically calibrating the first soil data based on the first calibration coefficient corresponding to the first soil data to obtain the dynamically calibrated first soil data.

[0082] The first calibration coefficient is a correction parameter calculated using statistical methods based on a comparative analysis of the first real data and the first soil data for the same soil sample. The first calibration coefficient is used to correct for systematic biases in the first soil parameters. For example, the deviations between multiple sets of first real data and first soil data are calculated, and the deviations between these sets are fitted to obtain the first calibration coefficient.

[0083] Specifically, by calculating the deviation between multiple sets of first real data and first soil data, the deviation between multiple sets of first real data and first soil data is fitted to obtain the first calibration coefficient. Based on the first calibration coefficient corresponding to the first soil data, the first soil data is dynamically calibrated to obtain the dynamically calibrated first soil data. This realizes the dynamic calibration of the first soil data and helps to improve the accuracy of the first soil data.

[0084] Optionally, the second soil data is dynamically calibrated based on the second real data to obtain dynamically calibrated second soil data, including: determining the second calibration coefficient corresponding to the second soil data based on the second real data; and dynamically calibrating the second soil data based on the second calibration coefficient corresponding to the second soil data to obtain dynamically calibrated second soil data.

[0085] The second calibration coefficient is a correction parameter calculated using statistical methods through a comparative analysis of the second true data and the second soil data for the same soil sample. For example, the deviation between multiple sets of second true data and second soil data is calculated, and the deviation between the multiple sets of second true data and second soil data is fitted to obtain the second calibration coefficient.

[0086] Specifically, by calculating the deviation between multiple sets of second real data and second soil data, the deviation between the multiple sets of second real data and second soil data is fitted to obtain a second calibration coefficient. Based on the second calibration coefficient corresponding to the second soil data, the second soil data is dynamically calibrated to obtain dynamically calibrated second soil data. This achieves dynamic calibration of the second soil data, which is beneficial to improving the accuracy of the second soil data.

[0087] Optionally, determining the first calibration coefficient corresponding to the first soil data based on the first real data includes: obtaining the first initial calibration coefficient of the first soil data; calibrating the first real data and the first soil data using a weighted regression-robust regression model to obtain a first target loss function; and adjusting the first initial calibration coefficient based on the target loss function to obtain the first calibration coefficient corresponding to the first soil data.

[0088] The first initial calibration coefficient is the initial calibration coefficient of the first soil data. For sensors that have undergone multiple dynamic calibrations, the first initial calibration coefficient can be the calibration coefficient of the previous first soil data; for sensors that have not undergone dynamic calibration, the first initial calibration coefficient can be the calibration parameter indicated in the manufacturer's manual. The weighted regression-robust regression model is a hybrid statistical model that combines the advantages of weighted regression and robust regression, used to analyze the relationship between the first real data and the first soil data and to solve for the first calibration coefficient. The weighted regression-robust regression model uses a combination of weighted regression and robust regression to calibrate the first soil data. The first objective loss function is an indicator that measures the difference between the first real data and the first soil data. Optionally, the first objective loss function can be the Huber loss function. When the first objective loss function is less than or equal to the first loss threshold, the corresponding calibration coefficient is used as the first calibration coefficient corresponding to the first soil data.

[0089] Specifically, by obtaining the first initial calibration coefficient of the first soil data, the relationship between the first real data and the first soil data is analyzed and the first calibration coefficient is solved through a weighted regression-robust regression model. The first soil data is then calibrated based on the first calibration coefficient to obtain the first objective loss function. When the first objective loss function is less than or equal to the first loss threshold, the corresponding calibration coefficient is used as the first calibration coefficient corresponding to the first soil data. This achieves dynamic calibration of the first soil data, which is beneficial to improving the accuracy of the first soil data.

[0090] Optionally, determining the second calibration coefficient corresponding to the second soil data based on the second real data includes: obtaining the second initial calibration coefficient of the second soil data; calibrating the second real data and the second soil data using a weighted regression-robust regression model to obtain a second objective loss function; and adjusting the second initial calibration coefficient based on the second objective loss function to obtain the second calibration coefficient corresponding to the second soil data.

[0091] The second initial calibration coefficient is the initial calibration coefficient for the second soil data. For soil sampling and analysis equipment that has undergone multiple dynamic calibrations, the second initial calibration coefficient can be the calibration coefficient of the previous second soil data; for soil sampling and analysis equipment that has not undergone dynamic calibration, the second initial calibration coefficient can be the calibration parameter indicated in the manufacturer's manual of the soil sampling and analysis equipment. The second target loss function is an indicator that measures the difference between the second real data and the second soil data. Optionally, the second target loss function can be the Huber loss function. When the second target loss function is less than or equal to the second loss threshold, the corresponding calibration coefficient is used as the second calibration coefficient corresponding to the second soil data. The first target loss function and the second target loss function can be the same or different, and the first target loss function and the second target loss function can be set according to requirements.

[0092] Specifically, by obtaining the first initial calibration coefficient of the second soil data, the relationship between the second real data and the second soil data is analyzed through a weighted regression-robust regression model, and the second calibration coefficient is solved. The second soil data is then calibrated based on the second calibration coefficient to obtain the second objective loss function. When the second objective loss function is less than or equal to the second loss threshold, the corresponding calibration coefficient is used as the second calibration coefficient corresponding to the second soil data. This achieves dynamic calibration of the second soil data, which is beneficial to improving the accuracy of the second soil data.

[0093] S330: Extract keyframes from video data to obtain multiple target images; extract features from multiple target images to obtain image features; perform dimensionality reduction on the first and second soil data to obtain dimensionality-reduced features; stitch the image features and dimensionality-reduced features to obtain stitched features; process the stitched features according to the quality prediction model to obtain tobacco growth quality information.

[0094] The target image is used for tobacco quality prediction. The target image can be determined based on video data. For example, if the video data includes multiple video frames, video frames are extracted from the video data at a preset fixed extraction interval. The extracted keyframes are then used as the target image for tobacco quality prediction.

[0095] Image features are those obtained by extracting features from multiple target images. For example, image feature extraction models can be used to extract features from target images to obtain image features. Image feature extraction models include, but are not limited to, neural network models. For example, an image feature extraction model can be a ResNet50 model. Optionally, the ResNet50 model can be optimized based on Tensor Runtime (TensorRT). The ResNet50 model can be deployed on edge servers, which helps reduce inference latency and improve throughput, thereby improving the efficiency of tobacco growth quality prediction.

[0096] Dimensionality reduction features are features obtained by reducing the dimensionality of the first and second soil data. For example, dimensionality reduction feature extraction methods can be used to reduce the dimensionality of the first and second soil data to obtain dimensionality reduction features. These methods include, but are not limited to, neural network models. For example, Principal Component Analysis (PCA) can be used. Optionally, after dimensionality reduction of the first and second soil data using PCA, a predetermined number of principal component vectors can be selected to generate dimensionality reduction features. For example, 10 principal component vectors with the largest explained variance can be selected to generate dimensionality reduction features.

[0097] The concatenated feature is a feature obtained by concatenating image features and dimensionality-reduced features. For example, image features and dimensionality-reduced features can be concatenated along the channel dimension to achieve the fusion of multimodal data. For instance, if the image features have a dimension of 2048 and the dimensionality-reduced features have a dimension of 10, concatenating the image features and dimensionality-reduced features along the channel dimension results in a concatenated feature with a dimension of 2058.

[0098] Specifically, firstly, video frames are extracted from the video data according to a preset fixed extraction interval, and the extracted keyframes are used as target images. Secondly, features are extracted from the target images using a ResNet50 model optimized based on TensorRT to obtain image features. Then, PCA is used to perform dimensionality reduction processing on the first and second soil data to obtain dimensionality-reduced features. Next, the image features and dimensionality-reduced features are stitched together in the channel dimension to obtain stitched features. Finally, the stitched features are processed by a quality prediction model to obtain tobacco growth quality information, thus achieving accurate prediction of tobacco growth quality.

[0099] Building upon the above embodiments, the target image can be preprocessed, including but not limited to local contrast enhancement, background removal, and cropping and scaling. Specifically, the Contrast Limited Adaptive Histogram Equalization (CLAHE) method can be used to enhance the local contrast of the target image. This involves partitioning the target image and performing histogram equalization within each small region, while introducing a contrast limiting factor to suppress noise interference caused by over-enhancement. This improves the clarity of tobacco leaf texture and the separability of lesion features. The U-Net model can then be used to remove the background from the locally contrast-enhanced image, retaining the image corresponding to the tobacco leaf region and removing the image corresponding to the background region. The image after background removal can be cropped and scaled to meet the input requirements of the image feature extraction model.

[0100] For example, see Figure 3 , Figure 3 This is a flowchart of another method for predicting tobacco growth quality provided in an embodiment of the present invention. The method includes:

[0101] (1) Sub-region division and metadata management are carried out through the GIS platform.

[0102] (2) Video data is collected by high-definition cameras used for video surveillance. The video data is then enhanced by CLAHE deployed on the edge computing server. Background segmentation is performed by the U-Net model, and feature extraction is performed by the ResNet50 model to obtain image features.

[0103] (3) Collect soil temperature, humidity, electrical conductivity, pH value and near-infrared spectral data through soil monitoring nodes, and perform data preprocessing such as Kalman filtering to remove noise, time series data completion and missing value interpolation.

[0104] (4) N / P / K is determined and calibration parameters are determined through laboratory testing. Least square fitting, calibration coefficient generation and parameter distribution and updating are performed through the dynamic calibration module.

[0105] (5) The image features of the 2048-dimensional vector and the soil PCA of the 10-dimensional principal components are fused to obtain a 2058-dimensional fused vector.

[0106] (6) Tobacco leaf quality is classified into four grades: Excellent, Good, Medium, and Substandard, using a Transformer deep network. The Transformer deep network consists of a 3-layer encoder, with each layer containing 8 attention heads. The Transformer deep network is optimized using the Adam optimizer and trained using the cross-entropy loss function. The Transformer deep network outputs quality prediction results and intelligent decision-making. The quality prediction results include classification probability output and anomaly warning triggers. The intelligent decision-making includes fertilization / irrigation suggestions and pest and disease control plans.

[0107] The technical solution of this embodiment achieves video data acquisition by acquiring video data of tobacco during its growth process; it also acquires first and second soil data of the soil in the tobacco-growing area, where the first soil data includes at least one of soil temperature, soil moisture, and near-infrared spectral data, and the second soil data includes at least one of nitrogen content, phosphorus content, potassium content, electrical conductivity, and pH value, thus achieving the acquisition of first and second soil data and providing data support for subsequent analysis and processing; it extracts keyframes from the video data to obtain multiple target images, extracts features from the multiple target images to obtain image features, performs dimensionality reduction processing on the first and second soil data to obtain dimensionality-reduced features, and stitches the image features and dimensionality-reduced features to obtain stitched features; and processes the stitched features according to a quality prediction model to obtain tobacco growth quality information, thus achieving the prediction of tobacco growth quality and improving the efficiency and accuracy of tobacco growth quality prediction.

[0108] Figure 4 This is a schematic diagram of a tobacco growth quality prediction device provided in an embodiment of the present invention. Figure 4 As shown, the device includes a video data acquisition module 310, a soil data acquisition module 320, and a growth quality information determination module 330.

[0109] The video data acquisition module 310 is used to acquire video data of tobacco during its growth process; the soil data acquisition module 320 is used to acquire first soil data and second soil data of the soil in the tobacco planting area; the first soil data includes at least one of soil temperature, soil moisture and near-infrared spectral data; the second soil data includes at least one of nitrogen content, phosphorus content, potassium content, electrical conductivity and pH value; and the growth quality information determination module 330 is used to determine the growth quality information of tobacco based on the video data, the first soil data and the second soil data through a quality prediction model.

[0110] The technical solution of this embodiment utilizes a video data acquisition module 310 to acquire video data of tobacco during its growth process, thus achieving video data acquisition; a soil data acquisition module 320 to acquire first and second soil data of the soil in the tobacco planting area, wherein the first soil data includes at least one of soil temperature, soil moisture, and near-infrared spectral data, and the second soil data includes at least one of nitrogen content, phosphorus content, potassium content, electrical conductivity, and pH value, thus achieving the acquisition of first and second soil data and providing data support for subsequent analysis and processing; and a growth quality information determination module 330 to determine the growth quality information of tobacco based on the video data, first soil data, and second soil data using a quality prediction model, thus achieving the prediction of tobacco growth quality and improving the efficiency and accuracy of tobacco growth quality prediction.

[0111] Based on the above embodiments, optionally, the video data is obtained from video acquisition devices deployed at designated locations in the tobacco growing area.

[0112] Optionally, the triggering conditions for video data acquisition include time-based triggering conditions.

[0113] Optionally, the video data acquisition trigger conditions include image color difference change conditions.

[0114] Optionally, the video data acquisition trigger conditions include time-based trigger conditions and image color difference change conditions.

[0115] Optionally, the first soil data is obtained based on detections by sensors deployed in tobacco-growing areas.

[0116] Optionally, the second soil data is obtained by analyzing soil samples taken periodically from tobacco growing areas using soil sampling and analysis equipment.

[0117] Optionally, the soil data acquisition module 320 is also used to generate second soil data according to at least one of the following methods: determining the nitrogen and phosphorus content of the soil in the tobacco growing area using an atomic absorption spectrometer; determining the potassium content of the soil in the tobacco growing area using a colorimetric method; and determining the electrical conductivity and pH value of the soil in the tobacco growing area using a standard electrode method.

[0118] Optionally, the device further includes dynamic calibration, used for: dynamically calibrating the first soil data based on the first real data, and dynamically calibrating the second soil data based on the second real data, to obtain the dynamically calibrated first soil data and second soil data; wherein the first real data and the second real data are respectively obtained based on laboratory analysis of soil in tobacco growing areas.

[0119] Optionally, dynamic calibration is also used for: determining a first calibration coefficient corresponding to the first soil data based on the first real data; and performing dynamic calibration on the first soil data based on the first calibration coefficient corresponding to the first soil data to obtain the dynamically calibrated first soil data.

[0120] Optionally, dynamic calibration is also used for: determining the second calibration coefficient corresponding to the second soil data based on the second real data; and performing dynamic calibration on the second soil data based on the second calibration coefficient corresponding to the second soil data to obtain the dynamically calibrated second soil data.

[0121] Optionally, dynamic calibration is also used for: obtaining the first initial calibration coefficient of the first soil data; calibrating the first real data and the first soil data using a weighted regression-robust regression model to obtain the first target loss function; and adjusting the first initial calibration coefficient based on the target loss function to obtain the first calibration coefficient corresponding to the first soil data.

[0122] Optionally, the growth quality information determination module 330 is further configured to: extract keyframes from video data to obtain multiple target images; extract features from the multiple target images to obtain image features; perform dimensionality reduction processing on the first soil data and the second soil data to obtain dimensionality reduction features; perform splicing processing on the image features and the dimensionality reduction features to obtain spliced ​​features; and process the spliced ​​features according to the quality prediction model to obtain tobacco growth quality information.

[0123] Optionally, the tobacco growth quality information includes at least one of the following: tobacco quality grade information, soil condition information, and pest and disease information.

[0124] Optionally, the device further includes a decision work order generation module, used to: generate a decision work order based on growth quality information when the tobacco quality grade information meets a first preset condition; wherein the decision work order includes at least one of fertilizer ratio parameters, irrigation parameters, pest and disease control parameters, and recommended execution time.

[0125] Optionally, the device also includes a decision work order generation module, used to: generate a decision work order based on growth quality information when the soil condition information meets the second preset condition; wherein the decision work order includes at least one of fertilizer ratio parameters, irrigation parameters, pest and disease control parameters, and recommended execution time.

[0126] Optionally, the device further includes a decision work order generation module, used to: generate a decision work order based on growth quality information when the tobacco quality grade information meets the first preset condition and the soil condition information meets the second preset condition; wherein the decision work order includes at least one of fertilizer ratio parameters, irrigation parameters, pest and disease control parameters, and recommended execution time.

[0127] Optionally, there may be multiple tobacco growing areas, and each tobacco growing area may correspond to at least one of the following: regional coding information and location information.

[0128] Optionally, the device further includes a processing demand information receiving module, used to: receive processing demand information, the processing demand information including target area coding information and / or target location information; and, based on the target area coding information and / or target location information, acquire video data corresponding to tobacco in the tobacco planting area, as well as first soil information and second soil information of the soil in the tobacco planting area.

[0129] The tobacco growth quality prediction device provided in this embodiment of the invention can execute the tobacco growth quality prediction method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0130] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. The electronic device 10 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0131] like Figure 5 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 may also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0132] Multiple components in electronic device 10 are connected to input / output (I / O) interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of monitors, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0133] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as a method for predicting tobacco growth quality.

[0134] In some embodiments, a tobacco growth quality prediction method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via read-only memory (ROM) 12 and / or communication unit 19. When the computer program is loaded into random access memory (RAM) 13 and executed by processor 11, one or more steps of the tobacco growth quality prediction method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform a tobacco growth quality prediction method by any other suitable means (e.g., by means of firmware).

[0135] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transferring data and instructions to the storage system, the at least one input device, and the at least one output device.

[0136] A computer program for implementing a method for predicting tobacco growth quality according to the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the functions / operations specified in the flowcharts and / or block diagrams are performed. The computer program can be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0137] This invention also provides a computer-readable storage medium storing computer instructions for causing a processor to execute a method for predicting tobacco growth quality, the method comprising:

[0138] Acquire video data of tobacco during its growth process; acquire first soil data and second soil data of the soil in the tobacco-growing area; the first soil data includes at least one of soil temperature, soil moisture and near-infrared spectral data; the second soil data includes at least one of nitrogen content, phosphorus content, potassium content, electrical conductivity and pH value; determine the growth quality information of tobacco based on the video data, the first soil data and the second soil data through a quality prediction model.

[0139] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0140] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0141] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0142] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0143] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication unit 19, or installed from storage unit 18, or installed from read-only memory (ROM) 12. When the computer program is executed by processor 11, it performs the functions defined in the methods of the embodiments of the present invention.

[0144] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0145] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for predicting tobacco growth quality, characterized in that, include: Acquire video data of tobacco during its growth process; Obtain first soil data and second soil data of the soil in the tobacco-growing area where the tobacco is located; the first soil data includes at least one of soil temperature, soil moisture and near-infrared spectral data; the second soil data includes at least one of nitrogen content, phosphorus content, potassium content, electrical conductivity and pH value; The growth quality information of the tobacco is determined based on the video data, the first soil data, and the second soil data using a quality prediction model.

2. The method according to claim 1, characterized in that, The video data is acquired based on video acquisition equipment deployed at designated locations in the tobacco planting area. The acquisition triggering conditions for the video data include time triggering conditions and / or image color difference change conditions. The first soil data was obtained based on detections by sensors deployed in the tobacco growing area; The second soil data is obtained by analyzing soil samples taken periodically from the tobacco planting area using soil sampling and analysis equipment.

3. The method according to claim 1, characterized in that, The second soil data is generated in a manner that includes at least one of the following: The nitrogen and phosphorus contents of the soil in the tobacco growing area were determined using atomic absorption spectrometry. The potassium content of the soil in the tobacco-growing area was determined by colorimetry. The electrical conductivity and pH value of the soil in the tobacco growing area were determined using the standard electrode method.

4. The method according to claim 1, characterized in that, The method further includes: The first soil data is dynamically calibrated based on the first real data, and the second soil data is dynamically calibrated based on the second real data to obtain the dynamically calibrated first soil data and second soil data. The first real data and the second real data are respectively obtained based on laboratory analysis of the soil in the tobacco planting area.

5. The method according to claim 4, characterized in that, The step of dynamically calibrating the first soil data based on first real data and dynamically calibrating the second soil data based on second real data to obtain dynamically calibrated first and second soil data includes: Based on the first real data, a first calibration coefficient corresponding to the first soil data is determined; based on the first calibration coefficient corresponding to the first soil data, the first soil data is dynamically calibrated to obtain the dynamically calibrated first soil data; and... Based on the second real data, a second calibration coefficient corresponding to the second soil data is determined; based on the second calibration coefficient corresponding to the second soil data, the second soil data is dynamically calibrated to obtain the dynamically calibrated second soil data.

6. The method according to claim 5, characterized in that, The step of determining the first calibration coefficient corresponding to the first soil data based on the first real data includes: Obtain the first initial calibration coefficient of the first soil data; perform calibration processing on the first real data and the first soil data through a weighted regression-robust regression model to obtain a first target loss function; adjust the first initial calibration coefficient based on the target loss function to obtain the first calibration coefficient corresponding to the first soil data; The step of determining the second calibration coefficient corresponding to the second soil data based on the second real data includes: obtaining the second initial calibration coefficient of the second soil data; calibrating the second real data and the second soil data using a weighted regression-robust regression model to obtain a second objective loss function; and adjusting the second initial calibration coefficient based on the second objective loss function to obtain the second calibration coefficient corresponding to the second soil data.

7. The method according to claim 1, characterized in that, The step of determining the growth quality information of the tobacco based on the video data, the first soil data, and the second soil data using a quality prediction model includes: The video data is processed by extracting keyframes to obtain multiple target images; Feature extraction is performed on multiple target images to obtain image features; The first soil data and the second soil data are subjected to dimensionality reduction processing to obtain dimensionality-reduced features; The image features and the dimensionality-reduced features are concatenated to obtain concatenated features; The splicing features are processed according to the quality prediction model to obtain the growth quality information of the tobacco.

8. The method according to claim 1, characterized in that, The tobacco growth quality information includes at least one of the following: tobacco quality grade information, soil condition information, and pest and disease information. The method further includes: If the tobacco quality grade information meets the first preset condition and / or the soil condition information meets the second preset condition, a decision work order is generated based on the growth quality information. The decision work order includes at least one of the following: fertilizer ratio parameters, irrigation parameters, pest and disease control parameters, and recommended execution time.

9. The method according to claim 1, characterized in that, There are multiple tobacco planting areas, and each tobacco planting area corresponds to at least one of the area coding information and location information; The method further includes: Receive processing request information, which includes target area encoding information and / or target location information; Based on the target area encoding information and / or target location information, obtain video data corresponding to tobacco in the tobacco planting area, as well as first soil information and second soil information of the soil in the tobacco planting area.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements a method for predicting tobacco growth quality as described in any one of claims 1-9.

Images