Crop automatic monitoring method and system

Abstract

The invention relates to a crop automatic monitoring method and system, and the method comprises the steps: collecting multi-dimensional time series data containing environment and crop growth information through an Internet of Things sensor disposed in a crop planting region, and obtaining original monitoring data; decomposing the original monitoring data by using filter parameters dynamically adjusted according to the spatial position of a monitoring point and a crop type to obtain a plurality of data components; based on the data component and a preset crop growth model, calculating and determining a local cut-off frequency used for separating effective crop information and interference noise under different time scales; screening and reconstructing the data component by using the local cut-off frequency to obtain feature data representing the key state of the crop; and according to the feature data, analyzing crop growth conditions, predicting environmental stress risks, and generating a crop supervision state report.

Description

Technical Field

[0001] This application relates to the field of the Internet of Things, and specifically to a method and system for automatically monitoring crops. Background Art

[0002] Smart agriculture has become a key direction for the transformation and upgrading of modern agriculture. Using Internet of Things sensors to continuously and real-time monitor the growth environment of crops and the physiological state of the crops themselves is the basis for achieving precise management. By deploying environmental sensors such as soil temperature and humidity, air temperature and humidity, light intensity, and carbon dioxide concentration in the field, as well as crop body sensors such as leaf surface temperature, stem micro-changes, and canopy images, a large amount of multi-dimensional time series data can be collected. These data contain rich information about crop growth trends, resource utilization efficiency, pest and disease stress, and the impact of environmental changes. However, the original monitoring data is often complex and non-linear, usually mixed with various components of different time scales, such as slow-changing seasonal trends, daily environmental cycles, the growth rhythm of the crops themselves, and sudden environmental stress events or equipment noise. This requires effectively extracting characteristic information from these mixed data that can accurately reflect the key state of the crops. Traditional data processing methods, such as Fourier transform or wavelet transform, have limitations in processing such agricultural monitoring data with non-stationary and non-linear characteristics, and are difficult to fully adapt to the characteristics of the data frequency changing dynamically with time, which may lead to data energy leakage or the inability to accurately separate data components from different sources.

[0003] Moreover, within the same farmland, there may be significant differences in soil conditions, microclimate, and crop growth at different locations, which means that the inherent characteristics and information focus of the monitoring data from different spatial positions may also be different. If a unified filtering parameter or frequency division standard is used to process the data of all monitoring points, it may not be able to fully adapt to this spatial difference, resulting in inaccurate information extraction in some areas or key features being misjudged as noise, thereby affecting the accuracy of crop supervision. Summary of the Invention

[0004] In view of the above problems, in the first aspect of this application, a method for automatically monitoring crops is provided, including the following steps: Collect multi-dimensional time series data containing environmental and crop growth information through Internet of Things sensors deployed in the crop planting area to obtain original monitoring data; Use filter parameters dynamically adjusted according to the spatial position of the monitoring point and the crop type to decompose the original monitoring data to obtain multiple data components; based on the data components and a preset crop growth model, calculate and determine a local cut-off frequency for separating effective crop information and interference noise at different time scales; Filter and reconstruct the data component using the local cut-off frequency to obtain characteristic data representing the key state of the crop; analyze the crop growth status based on the characteristic data and predict the environmental stress risk to generate a crop supervision status report.

[0005] Preferably, the filter parameters dynamically adjusted according to the spatial position of the monitoring point and the crop type are specifically: Pre-establish a spatial position - crop type - filter parameter mapping relationship database, and the database stores the optimal initial center frequency and bandwidth range of the filter under different geographical regions and corresponding planted crop types; When processing the original monitoring data of a specific monitoring point, according to the GPS coordinates of the monitoring point and the recorded crop type, retrieve the corresponding parameter range from the database, and combine with the frequency distribution characteristics of the original monitoring data within the current time window to fine-tune and determine the filter parameters finally used for data decomposition.

[0006] Preferably, it is characterized in that the fine-tuning to determine the filter parameters finally used for data decomposition is specifically: Calculate the power spectral density or short-time Fourier transform of the original monitoring data within the current analysis time window to identify the main frequency band where the data energy is concentrated; According to the deviation between the identified main frequency band and the initial center frequency retrieved from the database, and the distribution width of the data energy within the frequency band, adjust the initial center frequency and bandwidth, so as to determine the filter parameters finally used for data decomposition.

[0007] Preferably, the calculation and determination of the local cut-off frequency for separating the effective crop information and interference noise at different time scales based on the data component and the preset crop growth model are specifically: Extract the instantaneous frequency of each data component IMF; Obtain the normal fluctuation frequency range of the key physiological indicators corresponding to the current crop growth stage or the characteristic frequency of the response to specific environmental stress from the preset crop growth model; Analyze whether the instantaneous frequency of each IMF component falls into the effective frequency interval; Based on the analysis results, set one or more frequency thresholds as the local cut-off frequency.

[0008] Preferably, the filtering and reconstructing the data component using the local cut-off frequency to obtain the characteristic data representing the key state of the crop is specifically: Traverse each data component IMF obtained by decomposition, and compare its instantaneous frequency at each time point with the local cut-off frequency; if the instantaneous frequency of the IMF component meets the effective information standard, it is determined as an effective component and retained; discard or set to zero the IMF components determined as interference or non-related information. Perform time-domain superposition on all retained valid IMF components to form characteristic data representing the key states of crops.

[0009] In the second aspect of the present application, a crop automatic monitoring system is provided, including the following modules: A data acquisition module for collecting multi-dimensional time series data containing environmental and crop growth information through Internet of Things sensors deployed in the crop planting area to obtain original monitoring data; A parameter determination module for decomposing the original monitoring data by using filter parameters dynamically adjusted according to the spatial position of the monitoring point and the crop type to obtain a plurality of data components; based on the data components and a preset crop growth model, calculating and determining a local cut-off frequency for separating valid crop information and interference noise at different time scales; An analysis and supervision module for screening and reconstructing the data components by using the local cut-off frequency to obtain characteristic data representing the key states of crops; analyzing the crop growth condition and predicting the environmental stress risk according to the characteristic data, and generating a crop supervision status report.

[0010] Preferably, the filter parameters dynamically adjusted according to the spatial position of the monitoring point and the crop type are specifically: Pre-establish a spatial position - crop type - filter parameter mapping relationship database, and the database stores the optimal initial center frequency and bandwidth range of the filter under different geographical regions and corresponding planted crop types; When processing the original monitoring data of a specific monitoring point, according to the GPS coordinates of the monitoring point and the recorded crop type, retrieve the corresponding parameter range from the database, and combine the frequency distribution characteristics of the original monitoring data within the current time window to finely adjust and determine the filter parameters finally used for data decomposition.

[0011] Preferably, the fine adjustment to determine the filter parameters finally used for data decomposition is specifically: Calculate the power spectral density or short-time Fourier transform of the original monitoring data within the current analysis time window to identify the main frequency bands where the data energy is concentrated; According to the deviation between the identified main frequency bands and the initial center frequency retrieved from the database, and the distribution width of the data energy within the frequency bands, adjust the initial center frequency and bandwidth, so as to determine the filter parameters finally used for data decomposition.

[0012] Preferably, the calculation and determination of the local cut-off frequency for separating valid crop information and interference noise at different time scales based on the data components and a preset crop growth model is specifically: Extract the instantaneous frequencies of the IMFs of each data component; Obtain the normal fluctuation frequency range of the key physiological indicators corresponding to the current crop growth stage or the characteristic frequency in response to specific environmental stresses from the preset crop growth model; Analyze whether the instantaneous frequency of each IMF component falls within the effective frequency interval; Based on the analysis results, set one or more frequency thresholds as local cut-off frequencies.

[0013] Preferably, the data components are screened and reconstructed using the local cut-off frequency to obtain characteristic data representing the key state of the crop. Specifically: Traverse each data component IMF obtained by decomposition, and compare its instantaneous frequency at each time point with the local cut-off frequency; if the instantaneous frequency of the IMF component meets the effective information standard, it is determined as an effective component and retained; the IMF components determined as interference or irrelevant information are discarded or set to zero; Perform time-domain superposition on all retained effective IMF components to form characteristic data representing the key state of the crop.

[0014] By introducing filter parameters that are dynamically adjusted according to the spatial position of the monitoring point and the crop type, the present application can adaptively process the monitoring data from different farmland areas, improve the adaptability of the data decomposition process to spatial heterogeneity, and avoid local information distortion that may be caused by using unified parameters. Moreover, by introducing the preset crop growth model into the calculation process of the local cut-off frequency, the frequency is no longer divided solely based on the mathematical characteristics of the data itself, but combined with the characteristic frequency range of the physiological response of the crop at a specific growth stage, making the target of frequency separation clearer and enabling more accurate screening of the characteristic data components that truly reflect the key state of the crop from complex data, while suppressing environmental background noise and irrelevant interference. The present application improves the accuracy of crop supervision. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 Is the flowchart of Embodiment 1; Figure 2 Is the schematic diagram of the local cut-off frequency; Figure 3 Is the schematic diagram of the reconstructed data; Figure 4 Is the structural diagram of Embodiment 2. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0016] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application. For a better understanding of the technical solutions of the present application, the following will be described in detail in conjunction with the accompanying drawings of the specification and specific embodiments.

[0017] Embodiment 1, in this embodiment, a method for automatic monitoring of crops is provided, such asFigure 1 As shown, it includes the following steps: S1. Collect multi-dimensional time series data containing environmental and crop growth information through Internet of Things sensors deployed in the crop planting area to obtain original monitoring data; A variety of sensors are arranged in the fields of the crop planting area, including but not limited to soil moisture meters, thermometers, and light sensors. In one embodiment, the sensors also include cameras, and the cameras are directed at the crop leaves to capture changes in leaf color, morphology, etc. The sensors record and report data at regular intervals, for example, every 1 minute. These data constitute multi-dimensional time series data. Exemplarily, at time point 1: [humidity = 65%, temperature = 25°C, light = 800 lux, leaf = feature]; among them, the leaf color extracts the features of the pictures taken by the camera through methods such as convolution. These original data are sent to a remote server through a wireless network for storage and subsequent processing.

[0018] S2. Use filter parameters dynamically adjusted according to the spatial position of the monitoring point and the crop type to decompose the original monitoring data to obtain multiple data components; based on the data components and a preset crop growth model, calculate and determine local cut-off frequencies for separating effective crop information and interference noise at different time scales; Use dynamically adjusted filter parameters for decomposition to ensure that the decomposition method can adapt to the characteristics of different plots and different crops. After decomposing the data components, calculate one or more local cut-off frequencies through the crop model as the boundary for distinguishing normal growth, water-deficient data, and other data such as noise. In one embodiment, for different data sources, the local medium frequencies are different. For example, even for data collected at the same time and the same location for humidity and temperature, their local cut-off frequencies are different. Figure 2 Shows the local cut-off frequency. Decompose the original data, such as continuous soil moisture readings, into several IMF components, and each part represents components with different change speeds, such as slow seasonal dry-wet changes, faster day-night transpiration effects, and very fast sensor noise.

[0019] In one embodiment, according to the filter parameters, namely the center frequency and the bandwidth, the coverage range of the band - pass filter is determined. Preferably, the coverage range of the band - pass filter is from the center frequency - bandwidth / 2 to the center frequency + bandwidth / 2. The current residual signal is passed through this band - pass filter to preliminarily separate the oscillatory components with frequencies within this target range. The envelope mean calculation similar to the standard EMD is performed on the filtered signal to more precisely extract the IMF component at the current scale. The extracted IMF component is subtracted from the current residual signal to obtain a new residual signal. The above steps are repeated until the residual signal meets the stop condition, such as becoming a monotonic function or having an energy lower than the threshold. Finally, a series of IMF components and the final residual component are obtained, where each IMF represents the oscillatory mode of the original signal at different characteristic time scales, and its extraction process is guided by the spatial position and crop type information.

[0020] S3. Using the local cut - off frequency to screen and reconstruct the data components to obtain the characteristic data representing the key state of the crop; analyzing the crop growth status and predicting the environmental stress risk based on the characteristic data, and generating a crop supervision status report.

[0021] Use the local cut - off frequency determined in S2 as the screening criterion to filter the data components decomposed in S2. Only retain those components whose frequency characteristics meet the effective information criterion, and recombine them to obtain a characteristic data that can better reflect a specific state, as Figure 3 shown. Analyze the purified characteristic data. For example, if the amplitude of this data continues to increase, it may indicate that the water stress is intensifying. Based on this analysis, combined with other information, generate a report such as "The corn at monitoring point P001 is suspected of being moderately drought - stressed, with a high risk level. It is recommended to pay attention and irrigate in a timely manner."

[0022] In an alternative embodiment, the filter parameters dynamically adjusted according to the spatial position of the monitoring point and the crop type are specifically: Pre - establish a spatial position - crop type - filter parameter mapping relationship database, which stores the optimal initial center frequency and bandwidth range of the filter for different geographical regions and corresponding planted crop types; When processing the original monitoring data of a specific monitoring point, according to the GPS coordinates of the monitoring point and the recorded crop type, retrieve the corresponding parameter range from the database, and fine - tune to determine the final filter parameters for data decomposition in combination with the frequency distribution characteristics of the original monitoring data within the current time window.

[0023] The database can be a simple table or key-value store, such as { "Region A - Corn": { "Initial center frequency": 0.5Hz, "Initial bandwidth": 0.2Hz}, "Region B - Rice": { "Initial center frequency": 0.8Hz, "Initial bandwidth": 0.3Hz}}. When processing data from a certain sensor in Region A where corn is planted (GPS coordinates fall within Region A), the parameters { "Initial center frequency": 0.5Hz, "Initial bandwidth": 0.2Hz} will be retrieved as a reference for subsequent fine-tuning.

[0024] In an alternative embodiment, the fine-tuning determines the filter parameters finally used for data decomposition, specifically: Calculate the power spectral density or short-time Fourier transform of the original monitoring data within the current analysis time window, and identify the main frequency band where the data energy is concentrated; Based on the deviation between the identified main frequency band and the initial center frequency retrieved from the database, as well as the distribution width of the data energy within the frequency band, adjust the initial center frequency and bandwidth, thereby determining the filter parameters finally used for data decomposition.

[0025] Perform real-time data-driven correction on the initial parameters provided by the database. For example, for the sensor data of corn in Region A above, assume the retrieved initial center frequency is 0.5Hz. Take the data of a current short period, such as the data in the past 1 hour, and calculate its power spectral density (PSD). If it is found that the data energy is actually mainly concentrated around 0.6Hz, and the width of the energy distribution, such as the full width at half maximum, corresponds to a bandwidth of 0.25Hz, then adjust the initial parameters provided by the database to parameters closer to the actual data characteristics, such as a center frequency of 0.6Hz and a bandwidth of 0.25Hz. The adjusted parameters will be used for subsequent data decomposition steps, making the decomposition more adaptable to the actual situation of the current data.

[0026] In an alternative embodiment, find local maxima, that is, peaks, on the calculated power spectrum, and use the frequency corresponding to the peak or the maximum peak as the main frequency band, or use the frequency range of a preset length including the frequency corresponding to the peak or the maximum peak as the main frequency band.

[0027] In an alternative embodiment, based on the data components and a preset crop growth model, calculate and determine the local cut-off frequency for separating effective crop information and interference noise at different time scales, specifically: Extract the instantaneous frequencies of each data component IMF; Obtain the normal fluctuation frequency range of the key physiological indicators corresponding to the current crop growth stage or the characteristic frequencies in response to specific environmental stresses from the preset crop growth model; Analyze whether the instantaneous frequencies of each IMF component fall within the effective frequency interval; Based on the analysis results, set one or more frequency thresholds as the local cut-off frequencies.

[0028] For each IMF obtained by data decomposition, such as IMF1, IMF2, etc., calculate its instantaneous frequency varying with time, preferably implemented using the Hilbert transform. Then query the crop model, for example, according to the model, "the frequency characteristics of the early response data under mild drought stress during the maize germination stage are usually in the range of 0.1Hz - 0.3Hz". Next, analyze the instantaneous frequency sequence of each IMF. Suppose most of the instantaneous frequency values of IMF2 fall between 0.1Hz - 0.3Hz, while the frequency of IMF1 is much higher than 0.3Hz (which may be equipment noise), and the frequency of IMF3 is much lower than 0.1Hz (which may be very slow sunlight changes). Based on the above analysis, set a local cut-off frequency, for example, use 0.1Hz and 0.3Hz as the thresholds to determine which IMF components are the effective information that needs attention.

[0029] Among them, the crop growth model can, according to the input environmental data and the characteristics of the crop itself, predict which growth stage the crop should be in, and within that stage, the normal change range or characteristic pattern of its key physiological indicators. In an alternative embodiment, the crop growth model divides the growth stages based on the accumulated temperature and associates a set of empirical characteristic frequency ranges with each stage.

[0030] In an alternative embodiment, the step of using the local cut-off frequency to screen and reconstruct the data components to obtain the characteristic data representing the key state of the crop is specifically as follows: Traverse each data component IMF obtained by decomposition, and compare its instantaneous frequency at each time point with the local cut-off frequency; if the instantaneous frequency of the IMF component meets the effective information standard, then determine it as an effective component and retain it; discard or set to zero the IMF components determined to be interference or irrelevant information. Perform time-domain superposition on all the retained effective IMF components to form the characteristic data representing the key state of the crop.

[0031] Check each IMF. If the frequency of IMF1 is higher than 0.3Hz most of the time, which does not meet the standard, determine it as interference and discard it; if the frequency of IMF2 is between 0.1Hz - 0.3Hz most of the time, which meets the drought stress response characteristics, determine it as an effective component and retain it; if the frequency of IMF3 is lower than 0.1Hz most of the time, which does not meet the stress response characteristics, determine it as irrelevant slow-varying information and discard it. Finally, directly add all the retained effective components on the time axis, and the obtained data is the finally extracted characteristic data that can represent the key state that the current crop may be experiencing mild drought stress. This characteristic data can be subsequently used to generate early warnings or guide irrigation decisions.

[0032] Figure 4 Shows the structural diagram of the second embodiment. The second embodiment provides an automatic crop monitoring system, including the following modules: The data acquisition module is used to collect multi-dimensional time series data containing environmental and crop growth information through Internet of Things sensors deployed in the crop planting area, and obtain the original monitoring data; The parameter determination module is used to decompose the original monitoring data by using filter parameters dynamically adjusted according to the spatial position of the monitoring point and the crop type, and obtain multiple data components; based on the data components and a preset crop growth model, calculate and determine the local cut-off frequency for separating effective crop information and interference noise at different time scales; The analysis and supervision module is used to screen and reconstruct the data components by using the local cut-off frequency to obtain the characteristic data representing the key state of the crop; analyze the crop growth status according to the characteristic data, predict the environmental stress risk, and generate a crop supervision status report.

[0033] Preferably, the filter parameters dynamically adjusted according to the spatial position of the monitoring point and the crop type are specifically: Pre-establish a spatial position - crop type - filter parameter mapping relationship database, and the database stores the optimal initial center frequency and bandwidth range of the filter under different geographical regions and corresponding planted crop types; When processing the original monitoring data of a specific monitoring point, according to the GPS coordinates of the monitoring point and the recorded crop type, retrieve the corresponding parameter range from the database, and combine the frequency distribution characteristics of the original monitoring data within the current time window to fine-tune and determine the filter parameters finally used for data decomposition.

[0034] Preferably, the fine-tuning to determine the filter parameters finally used for data decomposition is specifically: Calculate the power spectral density or short-time Fourier transform of the original monitoring data within the current analysis time window, and identify the main frequency band where the data energy is concentrated; According to the deviation between the identified main frequency band and the initial center frequency retrieved from the database, and the distribution width of the data energy within the frequency band, adjust the initial center frequency and bandwidth, so as to determine the filter parameters finally used for data decomposition.

[0035] Preferably, the calculation and determination of the local cut-off frequency for separating effective crop information and interference noise at different time scales based on the data components and a preset crop growth model is specifically: Extract the instantaneous frequency of each data component IMF; Obtain the normal fluctuation frequency range of the key physiological indicators corresponding to the current crop growth stage or the characteristic frequency of response to specific environmental stresses from the preset crop growth model; Analyze whether the instantaneous frequency of each IMF component falls within the effective frequency interval; Based on the analysis results, set one or more frequency thresholds as local cut-off frequencies.

[0036] Preferably, the data components are screened and reconstructed using the local cut-off frequencies to obtain characteristic data representing the key state of the crop, specifically: Traverse each data component IMF obtained by decomposition, and compare its instantaneous frequency at each time point with the local cut-off frequency; if the instantaneous frequency of the IMF component meets the effective information standard, it is determined as an effective component and retained; the IMF components determined as interference or non-related information are discarded or set to zero; Perform time-domain superposition on all the retained effective IMF components to form characteristic data representing the key state of the crop.

[0037] The computer-readable storage medium provided in this application can be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or components, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections with 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 of the above. In this embodiment, the computer-readable storage medium can be any tangible medium that contains or stores a program, and this program can be used by or in combination with an instruction execution system, device, or component. The program code contained on the computer-readable storage medium can be transmitted using any suitable medium, including but not limited to: wires, optical cables, radio frequency (RF), etc., or any suitable combination of the above.

[0038] The above computer-readable storage medium can be included in an intelligent warehousing management device; it can also exist separately and not be assembled into the intelligent warehousing management device.

[0039] The above computer-readable storage medium carries one or more programs. When the one or more programs are executed by an intelligent warehousing management device, the intelligent warehousing management device can write computer program code for performing the operations of this application in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, executed as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (for example, by using an Internet service provider to connect through the Internet).

[0040] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in the flowchart or block diagram can represent a module, a program segment, or a part of code that contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and the combination of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0041] The modules involved in the embodiments described in this application can be implemented in software or in hardware. Among them, the name of the module does not constitute a limitation to the unit itself in some cases.

[0042] The readable storage medium provided by this application is a computer-readable storage medium. The computer-readable storage medium stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned intelligent warehousing management method, which can solve the technical problem of how to improve the warehousing operation efficiency through accurate storage location allocation. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by this application are the same as those of the intelligent warehousing management method provided by the above embodiments, and will not be elaborated here.

[0043] The above are only partial embodiments of this application, and thus do not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application by using the content of the specification and drawings of this application, or direct / indirect applications in other related technical fields are included in the patent protection scope of this application.

Claims

1. An automatic monitoring method for crops, characterized in that, Including the following steps: Collect multi-dimensional time series data containing environmental and crop growth information through Internet of Things sensors deployed in crop planting areas to obtain original monitoring data; Use filter parameters dynamically adjusted according to the spatial position of the monitoring point and the crop type to decompose the original monitoring data to obtain multiple data components; Based on the data components and a preset crop growth model, calculate and determine local cut-off frequencies for separating effective crop information and interference noise at different time scales; Use the local cut-off frequencies to screen and reconstruct the data components to obtain characteristic data representing the key state of the crop; Analyze the crop growth status based on the characteristic data, predict the environmental stress risk, and generate a crop supervision status report.

2. The method according to claim 1, characterized in that The filter parameters dynamically adjusted according to the spatial position of the monitoring point and the crop type are specifically: Pre-establish a spatial position-crop type-filter parameter mapping relationship database, and the database stores the optimal initial center frequency and bandwidth range of the filter under different geographical regions and corresponding planted crop types; When processing the original monitoring data of a specific monitoring point, retrieve the corresponding parameter range from the database according to the GPS coordinates of the monitoring point and the recorded crop type, and combine the frequency distribution characteristics of the original monitoring data within the current time window to fine-tune and determine the filter parameters finally used for data decomposition.

3. The method according to claim 2, wherein The fine-tuning to determine the filter parameters finally used for data decomposition is specifically: Calculate the power spectral density or short-time Fourier transform of the original monitoring data within the current analysis time window to identify the main frequency band where the data energy is concentrated; Adjust the initial center frequency and bandwidth according to the deviation between the identified main frequency band and the initial center frequency retrieved from the database, as well as the distribution width of the data energy within the frequency band, so as to determine the filter parameters finally used for data decomposition.

4. The method according to claim 1, wherein The calculation and determination of the local cut-off frequencies for separating effective crop information and interference noise at different time scales based on the data components and a preset crop growth model are specifically: Extract the instantaneous frequencies of the intrinsic mode functions (IMFs) of each data component; Obtain the normal fluctuation frequency range of the key physiological indicators corresponding to the current crop growth stage or the characteristic frequencies in response to specific environmental stresses from the preset crop growth model; Analyze whether the instantaneous frequencies of each IMF component fall within the effective frequency interval; Based on the analysis results, set one or more frequency thresholds as local cut-off frequencies.

5. The method according to claim 1, characterized in that, The use of the local cut-off frequencies to screen and reconstruct the data components to obtain characteristic data representing the key state of the crop is specifically: Traverse each IMF of the decomposed data components, compare its instantaneous frequency at each time point with the local cut-off frequency; If the instantaneous frequency of the IMF component meets the effective information standard, it is determined as an effective component and retained; Discard or set to zero the IMF components determined as interference or non-related information; Perform time-domain superposition on all retained effective IMF components to form characteristic data representing the key state of the crop.

6. An automatic crop monitoring system, characterized in that, Including the following modules: A data acquisition module, which is used to collect multi-dimensional time series data containing environmental and crop growth information through Internet of Things sensors deployed in the crop planting area, so as to obtain the original monitoring data; A parameter determination module, which is used to decompose the original monitoring data by using filter parameters dynamically adjusted according to the spatial position of the monitoring point and the crop type, so as to obtain multiple data components; based on the data components and a preset crop growth model, calculate and determine the local cut-off frequency for separating effective crop information and interference noise at different time scales; An analysis and supervision module, which is used to screen and reconstruct the data components by using the local cut-off frequency to obtain characteristic data representing the key state of the crop; analyze the crop growth status according to the characteristic data, predict the environmental stress risk, and generate a crop supervision status report.

7. The system according to claim 6, characterized in that, The filter parameters dynamically adjusted according to the spatial position of the monitoring point and the crop type are specifically: A spatial position-crop type-filter parameter mapping relationship database is established in advance, and the database stores the optimal initial center frequency and bandwidth range of the filter under different geographical regions and corresponding planted crop types; When processing the original monitoring data of a specific monitoring point, according to the GPS coordinates of the monitoring point and the recorded crop type, retrieve the corresponding parameter range from the database, and combine the frequency distribution characteristics of the original monitoring data within the current time window to finely adjust and determine the filter parameters finally used for data decomposition.

8. The system according to claim 7, wherein The fine adjustment to determine the filter parameters finally used for data decomposition is specifically: Calculate the power spectral density or short-time Fourier transform of the original monitoring data within the current analysis time window, and identify the main frequency band where the data energy is concentrated; According to the deviation between the identified main frequency band and the initial center frequency retrieved from the database, and the distribution width of the data energy within the frequency band, adjust the initial center frequency and bandwidth, so as to determine the filter parameters finally used for data decomposition.

9. The system according to claim 6, wherein The calculation and determination of the local cut-off frequency for separating effective crop information and interference noise at different time scales based on the data components and a preset crop growth model is specifically: Extract the instantaneous frequency of each data component IMF; Obtain the normal fluctuation frequency range of the key physiological indexes corresponding to the current crop growth stage or the characteristic frequency of the response to specific environmental stress from the preset crop growth model; Analyze whether the instantaneous frequency of each IMF component falls within the effective frequency range; Based on the analysis results, set one or more frequency thresholds as the local cut-off frequency.

10. The system according to claim 6, wherein The screening and reconstruction of the data components by using the local cut-off frequency to obtain characteristic data representing the key state of the crop is specifically: Traverse each data component IMF obtained by decomposition, and compare its instantaneous frequency at each time point with the local cut-off frequency; if the instantaneous frequency of the IMF component meets the effective information standard, it is determined as an effective component and retained; the IMF components determined as interference or irrelevant information are discarded or set to zero; Perform time-domain superposition on all retained effective IMF components to form characteristic data representing the key state of the crop.

Images