Method, device, equipment and storage medium for predicting farmland health status based on remote sensing images

Abstract

The embodiments of the present invention disclose a method, apparatus, device, and storage medium for predicting the health status of farmland based on remote sensing images. The method includes: obtaining health status annotation data and satellite remote sensing image data of crops in a preset area during the same preset period; the satellite remote sensing image data includes multiple pixels, and each pixel corresponds to a remote sensing indicator status data; obtaining historical vegetation indicator data of the preset area, and determining the vegetation indicator state vector corresponding to each pixel in combination with the current satellite remote sensing image data; training the health status annotation data and the remote sensing indicator state data corresponding to each pixel through a hidden Markov model to obtain a parameter set corresponding to each pixel; and determining the health status of the crops corresponding to each pixel in the current preset area based on the parameter set corresponding to each pixel and the vegetation indicator state vector. This method realizes intelligent prediction of the health status of farmland crops in a certain area, improving the accuracy and efficiency of the prediction.

Description

Technical Field

[0001] The embodiments of the present invention relate to the technical field of agricultural pest control management, and in particular to a method, apparatus, device and storage medium for predicting the health status of farmland based on remote sensing images. Background Art

[0002] In agricultural production, the health of farmland directly impacts the success of agricultural production. For example, once crops are affected by pests and diseases, their growth and yields are directly impacted, which in turn affects farmers' income. Therefore, predicting the health of crops is crucial to ensure timely intervention when problems arise, preventing further deterioration that could lead to reduced production and income.

[0003] Existing prediction methods mainly rely on manual investigation, experiments and experience prediction, which have long prediction cycles, high labor costs and cannot guarantee the accuracy of predictions. Especially when large-scale crop planting areas are affected by pests and diseases for a long time and a wide range of impact, the disadvantages of on-site investigations, such as long time consumption and high manpower consumption, are particularly obvious. Summary of the Invention

[0004] The present invention provides a method, device, equipment and storage medium for predicting the health status of farmland based on remote sensing images, so as to realize intelligent prediction of the health status of farmland and crops and improve the efficiency and accuracy of prediction.

[0005] In a first aspect, an embodiment of the present invention provides a method for predicting farmland health status based on remote sensing images. The method for predicting farmland health status based on remote sensing images includes:

[0006] Acquire crop health status annotation data and satellite remote sensing image data within a preset area during the same preset period; wherein the satellite remote sensing image data includes a plurality of pixels, and each pixel corresponds to a remote sensing indicator status data;

[0007] Acquire historical vegetation index data of the preset area, and determine the vegetation index state vector corresponding to each pixel in combination with current satellite remote sensing image data;

[0008] The health status label data and the remote sensing indicator status data corresponding to each pixel are trained through a hidden Markov model to obtain a parameter set corresponding to each pixel;

[0009] The health status of crops corresponding to each pixel in the current preset area is determined according to the parameter set and vegetation indicator state vector corresponding to each pixel.

[0010] Optionally, the parameter set includes: a probability distribution of the initial health state of crops in the preset area, a probability distribution of transitions between health states, and a probability distribution of the health state as a remote sensing indicator state.

[0011] Optionally, determining the health status of crops corresponding to each pixel in the current preset area according to the parameter set and vegetation indicator state vector corresponding to each pixel includes:

[0012] The maximum likelihood estimation method is used to determine the health status of the crops corresponding to each pixel based on the parameter set and vegetation indicator state vector corresponding to each pixel.

[0013] Optionally, after determining the health status of crops corresponding to each pixel in the current preset area according to the parameter set and vegetation indicator state vector corresponding to each pixel, the method further includes:

[0014] The risk level of the farmland in the preset area is determined according to the health status of the crops corresponding to each pixel.

[0015] Optionally, determining the risk level of the farmland in the preset area according to the health status of the crops corresponding to each pixel includes:

[0016] According to the current health status of the crops corresponding to each pixel, statistical parameters of the health status of the crops corresponding to each pixel are calculated, and the risk level of the farmland in the preset area is determined according to the statistical parameters.

[0017] Optionally, the statistical parameters include at least an expected value and a maximum likelihood estimate.

[0018] Optionally, the historical vegetation index data includes at least a normalized vegetation index and a ratio vegetation index.

[0019] In a second aspect, an embodiment of the present invention further provides a device for predicting farmland health status based on remote sensing images, the device comprising:

[0020] A module for acquiring health status annotation data and satellite remote sensing image data, configured to acquire health status annotation data and satellite remote sensing image data of crops within a preset area during a preset period; wherein the satellite remote sensing image data includes a plurality of pixels, and each pixel corresponds to a remote sensing indicator status data;

[0021] A historical vegetation index data acquisition module, used to acquire historical vegetation index data of the preset area;

[0022] A vegetation index state vector determination module is used to determine the vegetation index state vector corresponding to each pixel based on the historical vegetation index data of the preset area and in combination with the current satellite remote sensing image data;

[0023] A parameter set acquisition module is used to train the health status annotation data and the remote sensing indicator status data corresponding to each pixel through a hidden Markov model to obtain a parameter set corresponding to each pixel;

[0024] The health status determination module is used to determine the health status of crops corresponding to each pixel in the current preset area according to the parameter set and vegetation indicator state vector corresponding to each pixel.

[0025] In a third aspect, an embodiment of the present invention further provides a prediction device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. The computer device further comprises: a remote sensing satellite for collecting remote sensing image data of crops within a preset area. When the processor executes the program, the method for predicting farmland health status based on remote sensing images as described in the first aspect is implemented.

[0026] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the farmland health status prediction method based on remote sensing images as described in the first aspect.

[0027] The present invention provides a method, apparatus, device, and storage medium for predicting farmland health based on remote sensing images. The method comprises: obtaining crop health status annotation data and satellite remote sensing image data within a preset area during the same preset period; wherein the satellite remote sensing image data includes multiple pixels, and each pixel corresponds to a remote sensing indicator state data; obtaining historical vegetation indicator data for the preset area and combining it with current satellite remote sensing image data to determine a vegetation indicator state vector corresponding to each pixel; training the health status annotation data and the remote sensing indicator state data corresponding to each pixel through a hidden Markov model to obtain a parameter set corresponding to each pixel; and determining the crop health status corresponding to each pixel within the current preset area based on the parameter set corresponding to each pixel and the vegetation indicator state vector. This method can achieve: intelligent prediction of the health status of farmland crops within a certain area, reduce labor costs, shorten the prediction cycle, and improve the accuracy of the parameter set through model training, thereby improving the accuracy and efficiency of the prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 This is a flowchart of a method for predicting farmland health status based on remote sensing images in Example 1 of the present invention;

[0029] Figure 2 Schematic diagram of the state transition of the hidden Markov model in the first embodiment of the present invention;

[0030] Figure 3 This is a flowchart of a method for predicting farmland health status based on remote sensing images in Example 2 of the present invention;

[0031] Figure 4 This is a structural block diagram of a device for predicting farmland health status based on remote sensing images in the third embodiment of the present invention;

[0032] Figure 5 This is a structural diagram of a device in embodiment 4 of the present invention. DETAILED DESCRIPTION

[0033] The present invention will be further described in detail below with reference to the accompanying drawings and examples. It will be understood that the specific embodiments described herein are intended only to illustrate the present invention and are not intended to limit the present invention. It should also be noted that, for ease of description, the accompanying drawings only illustrate portions relevant to the present invention, not all structures.

[0034] Example 1

[0035] Figure 1 This is a flowchart of a method for predicting farmland health status based on remote sensing images provided in Example 1 of the present invention. Figure 2 This is a schematic diagram of the state transition of the hidden Markov model provided in the first embodiment of the present invention. This embodiment can be applied to an agricultural pest control management platform to implement a method for intelligently and efficiently predicting the health status of farmland. This method can be executed by a farmland health status prediction device based on remote sensing images. The device can be implemented in software and / or hardware. The device can be configured in the server of the management platform. Figure 1 , specifically including the following steps:

[0036] Step 110: Acquire crop health status labeling data and satellite remote sensing image data within a preset area during the same preset period;

[0037] The satellite remote sensing image data includes a plurality of pixels, and each pixel corresponds to a remote sensing indicator status data;

[0038] Specifically, experts can be invited to mark the health status of crops in a preset area within a preset period or time, thereby obtaining time-sequential crop health status labeling data, that is, the implicit state change vector; satellite remote sensing image data of a preset area in the same preset period can be obtained through Internet download or other methods, wherein the satellite remote sensing image data contains multiple pixels, such as n pixels; after each pixel is processed by projection transformation, remapping, calculation of multiple vegetation assessment indices, etc., the time sequence corresponding to each pixel and the crop remote sensing indicator status data matching the coordinates are obtained, that is, the observable state change vector.

[0039] The calculation of vegetation assessment indices can be performed by obtaining multiple vegetation assessment indices, such as the Ratio Vegetation Index (RVI), based on multiple image channels storing electromagnetic wave sampling information at different wavelengths in satellite remote sensing image data. These indices are then further divided into different growth states based on preset rules, with each index representing an integer, such as 0-10, representing the relative goodness or badness of the indicator, serving as the observable state in the Hidden Markov Model. The preset rules are division rules that can be set based on actual conditions and are not specifically limited here.

[0040] Step 120: Acquire historical vegetation index data of a preset area, and determine the vegetation index state vector corresponding to each pixel in combination with current satellite remote sensing image data.

[0041] The historical vegetation index data of agricultural crops in the preset area can be obtained from the database. The satellite remote sensing image data of the current preset area can be obtained by downloading from the Internet, etc., wherein the current satellite remote sensing image data also includes a plurality of pixels, and the number of pixels is the same as the number of pixels of the satellite remote sensing image data in the preset area during the preset period in step 110 and corresponds one to one, that is, the number is also n. By analyzing the multi-band satellite image values ​​and other data, the vegetation state value of each pixel can be obtained. Then, the vegetation state value of each pixel can be added to the vegetation state value of the historical vegetation index data of the preset area to derive the vegetation index state vector corresponding to each pixel. The derived vegetation index state vector corresponding to each pixel is stored in the database to update the historical vegetation index data in the database.

[0042] For example, assuming that the historical vegetation index data is (10, 10, 9) and the current date is August 16, the method for obtaining the vegetation index state vector corresponding to a single pixel is shown in Table 1.

[0043] As shown in Table 1, the vegetation state value of the current pixel in the current period is 8, and the historical vegetation index data is (10, 10, 9), so the vegetation index state vector of the current pixel is (10, 10, 9, 8).

[0044] Table 1. Methods for obtaining the vegetation indicator state vector corresponding to a single pixel

[0045] date August 1 August 6 August 11 August 16 (current) Historical vegetation index data 10 10 9 Satellite remote sensing image data 8 Vegetation indicator status 10 10 9 8

[0046] Step 130: The health status labeling data and the remote sensing indicator status data corresponding to each pixel are trained through a hidden Markov model to obtain a parameter set corresponding to each pixel.

[0047] Among them, the hidden Markov model is a random process probability model, a type of Markov chain, which includes elements such as observable state, hidden state, transition probability between hidden states, and transition probability from hidden state to observable state. The elements in the model sometimes contain unknown items, which need to be obtained by inputting a certain amount of training data. For example, the state transition of the hidden Markov model can be referred to Figure 2 , where T i,j represents the transition probability from implicit state i to j, P i,k represents the transition probability from implicit state i to observable state k.

[0048] Among them, the health status of crops refers to the degree to which crops are affected by pests and diseases, wherein the pest and disease situation is used as a hidden state of the model, for example, represented by an integer 0-10. In agricultural production, pest and disease disasters have a strong temporal correlation, and different pest and disease situations affect the different growth states of crops to a great extent. These realistic conditions are very consistent with the use conditions of the hidden Markov model. Therefore, the present invention uses the health status, i.e., the degree of influence of pests and diseases, as a hidden state, and the growth state, i.e., the quality of vegetation indicators, as its corresponding observable state. The transition probability between implicit states and from implicit states to observable states is set as an unknown quantity, and a hidden Markov model is constructed, so that a parameter set composed of unknown quantities can be obtained by training with labeled vegetation indicator data, and the degree of influence of pests and diseases corresponding to the new vegetation indicator changes can be evaluated accordingly.

[0049] Specifically, the health status of crops within a preset area at a preset time period is labeled as a latent state vector. Satellite remote sensing imagery of crops within the same preset area at the same time period is used to calculate multiple vegetation assessment indices for each pixel. This indices are then used to generate a vegetation index rating, thereby forming a temporally ordered observable state vector for each pixel. Based on these observable and latent state vectors, a hidden Markov model is trained to obtain a parameter set corresponding to each pixel. This parameter set is then stored in a database for future use.

[0050] The parameter set includes: the initial health status probability distribution of crops in a preset area, the transition probability distribution between health states, and the probability distribution of health states expressed as remote sensing indicator states.

[0051] Furthermore, before training the hidden Markov model with the health status label data and the remote sensing indicator status data corresponding to each pixel, the hidden Markov model is initialized. This initialization is an average initialization, meaning that each state has an equal initial probability, and the probability of transitioning to other states is also equal. For example, assuming the initialization results in a probability of 1 / 11 for hidden state 0, then the probability of transitioning to hidden states 0-10 is also 1 / 11. The sum of the probabilities of transitioning to hidden state 0 is (1 / 11)*11=1, which meets the requirements of probability theory. Furthermore, the probability of hidden state 0 manifesting as visible states 0-10 is also 1 / 11, and the sum of all probabilities of hidden state 0 manifesting as visible states 0-10 is also 1.

[0052] Step 140: Determine the health status of crops corresponding to each pixel in the current preset area according to the parameter set and vegetation indicator state vector corresponding to each pixel.

[0053] The vegetation indicator state vector for each pixel is derived by combining historical vegetation indicator data for a pre-defined area with current satellite remote sensing imagery. Therefore, the parameter set and vegetation indicator state vector corresponding to each pixel can be used to determine the crop health status of each pixel within the pre-defined area. The health status of each pixel can then be used to determine whether the crop health status has reached the preset alarm threshold. If the threshold is reached, a corresponding alarm signal is generated to alert relevant personnel and provide notification services to relevant parties, thereby minimizing or eliminating risks and reducing potential losses.

[0054] Among them, the health status of crops includes the degree of impact of pests and diseases on crops; based on the parameter set and vegetation indicator state vector corresponding to each pixel, the degree of impact of pests and diseases corresponding to each pixel in the current preset area can be determined, so that relevant personnel can take corresponding measures in a timely and reasonable manner to eliminate or avoid risks and reduce possible losses based on the current degree of impact of pests and diseases on crops.

[0055] In the technical solution of this embodiment, the working principle of the farmland health status prediction method based on remote sensing imagery is as follows: by inviting experts to make health status annotations for crops in a preset area during a preset period or time, thereby obtaining time-sequential crop health status annotation data; obtaining satellite remote sensing image data for a preset area during the same preset period by downloading from the Internet, wherein the satellite remote sensing image data contains multiple pixels, and each pixel is processed through projection transformation, remapping, and calculation of multiple vegetation assessment indices to obtain remote sensing indicator status data corresponding to each pixel; obtaining historical vegetation indicator data for the preset area and combining it with current satellite remote sensing image data to determine the vegetation indicator state vector corresponding to each pixel; then, training the health status annotation data and the remote sensing indicator state data corresponding to each pixel through a hidden Markov model to obtain a parameter set corresponding to each pixel; finally, determining the health status of each pixel in the current preset area based on the parameter set corresponding to each pixel and the vegetation indicator state vector. This method can achieve: intelligent prediction of the health status of farmland crops in a certain area, reduce labor costs, shorten the prediction cycle, and improve the accuracy of the parameter set through model training, thereby improving the accuracy and efficiency of prediction. The system can also determine whether the health status of each pixel's crops has reached a preset threshold based on the corresponding crop health status. If the threshold is reached, a corresponding alarm signal is generated to alert relevant personnel, providing notification services to stakeholders to minimize or eliminate risks and reduce potential losses. Alternatively, it can provide stakeholders with timely, large-scale quantitative assessment results for reference, enabling timely resource allocation, knowledge dissemination, and risk assessment, thereby preventing delayed responses that could lead to irreparable losses.

[0056] The technical solution of this embodiment provides a method for predicting the health status of farmland based on remote sensing images. The method comprises: obtaining crop health status annotation data and satellite remote sensing image data within a preset area during the same preset period; wherein the satellite remote sensing image data includes multiple pixels, and each pixel corresponds to a remote sensing indicator state data; obtaining historical vegetation indicator data for the preset area and determining a vegetation indicator state vector corresponding to each pixel in combination with current satellite remote sensing image data; training the health status annotation data and the remote sensing indicator state data corresponding to each pixel through a hidden Markov model to obtain a parameter set corresponding to each pixel; and determining the health status of each crop pixel within the current preset area based on the parameter set corresponding to each pixel and the vegetation indicator state vector. This method can achieve: intelligent prediction of the health status of farmland crops within a certain area, reduce labor costs, shorten the prediction cycle, and improve the accuracy of the parameter set through model training, thereby improving the accuracy and efficiency of the prediction.

[0057] On the basis of the above technical solution, optionally, the parameter set includes: the probability distribution of the initial health state of crops in a preset area, the probability distribution of transitions between health states, and the probability distribution of the health state as a remote sensing indicator state.

[0058] Among them, the transition state between health states represents the change of crop health status (or pest and disease status) over time; the health state is expressed as the remote sensing indicator state, that is, the hidden state is expressed as the visible state represents the observed vegetation indicator. These two probabilities are independent of each other and do not affect each other.

[0059] Optionally, the historical vegetation index data includes at least a normalized vegetation index and a ratio vegetation index.

[0060] The historical vegetation index data may be the Normalized Difference Vegetation Index and / or the Ratio Vegetation Index. Furthermore, the historical vegetation index data may also include the Difference Ambient Vegetation Index, the Soil Brightness Adjusted Vegetation Index, the Greenness Vegetation Index, and the Vertical Vegetation Index. The specific index to be used may be determined based on actual circumstances and is not specifically limited here.

[0061] Example 2

[0062] Figure 3 This is a flowchart of a method for predicting farmland health status based on remote sensing images provided in the second embodiment of the present invention. Figure 3 , the method specifically comprises the following steps:

[0063] Step 210: Acquire crop health status labeling data and satellite remote sensing image data within a preset area during the same preset period;

[0064] Step 220: Acquire historical vegetation index data of the preset area, and determine the vegetation index state vector corresponding to each pixel in combination with current satellite remote sensing image data;

[0065] Step 230: The health status label data and the remote sensing indicator status data corresponding to each pixel are trained through a hidden Markov model to obtain a parameter set corresponding to each pixel;

[0066] Step 240: Based on the parameter set and vegetation indicator state vector corresponding to each pixel, a maximum likelihood estimation method is used to determine the health status of the crops corresponding to each pixel.

[0067] Specifically, the maximum likelihood estimation method is used to estimate the current health status of crops at each pixel based on the probability distribution of the initial health status of crops in the preset area corresponding to each pixel, the transition probability distribution between health states, the probability distribution of the health status expressed as remote sensing indicator states, and the vegetation indicator state vector.

[0068] Step 250: Determine the risk level of the farmland in the current preset area based on the health status of the crops corresponding to each pixel.

[0069] Each pixel represents satellite remote sensing imagery data from a pre-set region, reflecting observable data for that region. Therefore, the current risk level of the farmland in that region can be determined based on the current crop health status corresponding to each pixel. After determining the current risk level of the farmland in that region, the credit risk of the farmland in that region can be assessed based on the current risk level. For example, after signing an agricultural credit product, changes in the crop vegetation index within the new growth cycle are regularly collected, and the credit risk of the pre-set region is assessed using the aforementioned crop health status prediction method and farmland risk level determination method. If the health status of the farmland in the pre-set region is assessed to be poor, reaching an alarm threshold, an alarm message is issued to the creditor, debtor, and other relevant parties of the credit product, prompting them to take measures to eliminate or avoid risks and minimize losses.

[0070] Optionally, determining the risk level of farmland in the current preset area based on the health status of the crops corresponding to each pixel includes:

[0071] Based on the current health status of crops in each pixel, the statistical parameters of the health status of crops in each pixel are calculated, and the risk level of farmland in the current preset area is determined based on the statistical parameters.

[0072] The risk level of farmland in a preset area is equal to the average health status (or degree of pests and diseases) of all pixels in the preset area.

[0073] Optionally, the statistical parameters include at least: an expected value and a maximum likelihood estimate.

[0074] Specifically, the expected value and / or maximum likelihood estimation value of the health status of each pixel of crops can be calculated according to the current health status of each pixel of crops, so as to estimate the overall risk level of the farmland in the current preset area.

[0075] Example 3

[0076] Figure 4 This is a structural block diagram of a farmland health status prediction device based on remote sensing images provided in the third embodiment of the present invention. The embodiment of the present invention also provides a farmland health status prediction device based on remote sensing images, refer to Figure 4 , the apparatus 100 comprises:

[0077] The health status labeling data and satellite remote sensing image data acquisition module 10 is used to obtain the health status labeling data and satellite remote sensing image data of crops in a preset area during the same preset period; wherein the satellite remote sensing image data includes a plurality of pixels, and each pixel corresponds to a remote sensing indicator status data;

[0078] The historical vegetation index data acquisition module 20 is used to acquire the historical vegetation index data of a preset area;

[0079] The vegetation index state vector determination module 30 is used to determine the vegetation index state vector corresponding to each pixel based on the historical vegetation index data of the preset area and the current satellite remote sensing image data;

[0080] The parameter set acquisition module 40 is used to obtain the parameter set corresponding to each pixel by training the health status label data and the remote sensing indicator status data corresponding to each pixel through a hidden Markov model;

[0081] The health status determination module 50 is used to determine the health status of crops corresponding to each pixel in the current preset area according to the parameter set and vegetation indicator state vector corresponding to each pixel.

[0082] The technical solution of this embodiment provides a farmland health status prediction device based on remote sensing images, which includes: a health status annotation data and satellite remote sensing image data acquisition module, which is used to obtain health status annotation data and satellite remote sensing image data of crops in a preset area at the same preset period; wherein the satellite remote sensing image data includes multiple pixels, and each pixel corresponds to a remote sensing indicator status data; a historical vegetation indicator data acquisition module, which is used to obtain historical vegetation indicator data of the preset area; a vegetation indicator state vector determination module, which is used to determine the vegetation indicator state vector corresponding to each pixel based on the historical vegetation indicator data of the preset area and combined with the current satellite remote sensing image data; a parameter set acquisition module, which is used to train the health status annotation data and the remote sensing indicator status data corresponding to each pixel through a hidden Markov model to obtain a parameter set corresponding to each pixel; and a health status determination module, which is used to determine the health status corresponding to each pixel of the crop in the current preset area based on the parameter set and vegetation indicator state vector corresponding to each pixel. This device can achieve the following: intelligent prediction of the health status of farmland and crops in a certain area, reduce labor costs, shorten the prediction cycle, and improve the accuracy of parameter sets through model training, thereby improving the accuracy and efficiency of predictions.

[0083] Optionally, the parameter set includes: a probability distribution of the initial health state of crops in a preset area, a probability distribution of transitions between health states, and a probability distribution of the health state as a remote sensing indicator state.

[0084] Optionally, the health status determination module 50 includes:

[0085] The health status determination unit is used to determine the health status of the crops corresponding to each pixel using the maximum likelihood estimation method based on the parameter set and vegetation indicator state vector corresponding to each pixel.

[0086] Optionally, the farmland health status prediction device based on remote sensing images further includes: a farmland risk level determination module, configured to determine the risk level of the farmland in the preset area according to the health status of the crops corresponding to each pixel.

[0087] Optionally, the farmland risk level determination module includes:

[0088] a statistical parameter calculation unit, configured to calculate statistical parameters of the health status of crops corresponding to each pixel according to the health status of the crops corresponding to each pixel;

[0089] The risk level determination unit is used to determine the risk level of farmland in the current preset area based on statistical parameters.

[0090] Optionally, the statistical parameters include at least: an expected value and a maximum likelihood estimate.

[0091] Optionally, the historical vegetation index data includes at least a normalized vegetation index and a ratio vegetation index.

[0092] The farmland health status prediction device based on remote sensing images provided in an embodiment of the present invention can execute the farmland health status prediction method based on remote sensing images provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0093] Example 4

[0094] Figure 5 This is a structural diagram of a device provided in the fourth embodiment of the present invention, such as Figure 5 As shown, the device includes a processor 60, a memory 61, an input device 62, an output device 63 and a remote sensing satellite 64; the number of processors 60 in the device can be one or more. Figure 5 In the embodiment, a processor 60 is used as an example; the processor 60, the memory 61, the input device 62 and the output device 63 in the device can be connected by a bus or other means. Figure 5 The bus connection is taken as an example.

[0095] The memory 61, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules, such as the program instructions / modules corresponding to the remote sensing image-based farmland health status prediction method in the embodiments of the present invention (for example, the health status annotation data and satellite remote sensing image data acquisition module 10, the historical vegetation index data acquisition module 20, the vegetation index state vector determination module 30, the parameter set acquisition module 40, and the health status determination module 50 in the remote sensing image-based farmland health status prediction device). The processor 60 executes the software programs, instructions, and modules stored in the memory 61 to execute various functional applications and data processing of the device, thereby implementing the aforementioned remote sensing image-based farmland health status prediction method.

[0096] The memory 61 may primarily include a program storage area and a data storage area. The program storage area may store an operating system and at least one application required for a function; the data storage area may store data generated based on the use of the terminal. Furthermore, the memory 61 may include high-speed random access memory and non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state memory device. In some instances, the memory 61 may further include memory remotely located relative to the processor 60, and these remote memories may be connected to the device via a network. Examples of such networks include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0097] Input device 62 can be used to receive input digital or character information and generate key signal input related to user settings and function control of the device. Output device 63 can include a display device such as a display screen. Remote sensing satellite 64 is used to collect remote sensing image data of crops within a preset area.

[0098] Example 5

[0099] The fifth embodiment of the present invention further provides a storage medium containing computer-executable instructions. When the computer-executable instructions are executed by a computer processor, the method is used to perform a method for predicting farmland health status based on remote sensing images, including:

[0100] Acquire crop health status annotation data and satellite remote sensing image data within a preset area during the same preset period; wherein the satellite remote sensing image data includes a plurality of pixels, and each pixel corresponds to a remote sensing indicator status data;

[0101] Obtain historical vegetation index data for a preset area and determine the vegetation index state vector corresponding to each pixel in combination with current satellite remote sensing image data;

[0102] The health status label data and the remote sensing indicator status data corresponding to each pixel are trained through a hidden Markov model to obtain the parameter set corresponding to each pixel;

[0103] The health status of crops corresponding to each pixel in the current preset area is determined based on the parameter set and vegetation indicator state vector corresponding to each pixel.

[0104] Of course, the storage medium containing computer-executable instructions provided in an embodiment of the present invention is not limited to the method operations described above, and can also execute related operations in the farmland health status prediction method based on remote sensing images provided in any embodiment of the present invention.

[0105] Through the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented with the help of software and necessary general-purpose hardware, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment of the present invention.

[0106] It is worth noting that in the embodiment of the above-mentioned search device, the various units and modules included are only divided according to functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0107] Note that the above are only preferred embodiments of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments, and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments and may include many other equivalent embodiments without departing from the concept of the present invention. The scope of the present invention is determined by the scope of the appended claims.

Claims

1. A method for predicting farmland health status based on remote sensing images, characterized in that: include: Acquire crop health status annotation data and satellite remote sensing image data within a preset area during the same preset period; wherein the satellite remote sensing image data includes a plurality of pixels, and each pixel corresponds to a remote sensing indicator status data; Acquire historical vegetation index data of the preset area, and determine the vegetation index state vector corresponding to each pixel in combination with current satellite remote sensing image data; The health status label data and the remote sensing indicator status data corresponding to each pixel are trained through a hidden Markov model to obtain a parameter set corresponding to each pixel; Determine the health status of crops corresponding to each pixel in the current preset area according to the parameter set and vegetation indicator state vector corresponding to each pixel; The determining of the health status of crops corresponding to each pixel in the current preset area according to the parameter set and vegetation indicator state vector corresponding to each pixel includes: The maximum likelihood estimation method is used to determine the health status of the crops corresponding to each pixel based on the parameter set and vegetation indicator state vector corresponding to each pixel.

2. The method for predicting farmland health status based on remote sensing images according to claim 1, characterized in that: The parameter set includes: the initial health state probability distribution of the crops in the preset area, the transition probability distribution between health states, and the probability distribution of the health state as a remote sensing indicator state.

3. The method for predicting farmland health status based on remote sensing images according to claim 1, characterized in that: After determining the health status of crops corresponding to each pixel in the current preset area according to the parameter set and the vegetation indicator state vector corresponding to each pixel, the method further includes: The risk level of the farmland in the preset area is determined according to the health status of the crops corresponding to each pixel.

4. The method for predicting farmland health status based on remote sensing images according to claim 3, characterized in that: The determining of the risk level of the farmland in the preset area according to the health status of the crops corresponding to each pixel at present includes: According to the current health status of the crops corresponding to each pixel, statistical parameters of the health status of the crops corresponding to each pixel are calculated, and the risk level of the farmland in the preset area is determined according to the statistical parameters.

5. The method for predicting farmland health status based on remote sensing images according to claim 4, characterized in that: The statistical parameters include at least expected values ​​and maximum likelihood estimates.

6. The method for predicting farmland health status based on remote sensing images according to claim 1, characterized in that: The historical vegetation index data at least includes the normalized vegetation index and the ratio vegetation index.

7. A device for predicting farmland health status based on remote sensing images, characterized in that: include: A module for acquiring health status annotation data and satellite remote sensing image data, configured to acquire health status annotation data and satellite remote sensing image data of crops within a preset area during a preset period; wherein the satellite remote sensing image data includes a plurality of pixels, and each pixel corresponds to a remote sensing indicator status data; A historical vegetation index data acquisition module, used to acquire historical vegetation index data of the preset area; A vegetation index state vector determination module is used to determine the vegetation index state vector corresponding to each pixel based on the historical vegetation index data of the preset area and in combination with the current satellite remote sensing image data; A parameter set acquisition module is used to obtain a parameter set corresponding to each pixel by training the health status annotation data and the remote sensing indicator status data corresponding to each pixel through a hidden Markov model; A health status determination module is used to determine the health status of crops corresponding to each pixel in the current preset area based on the parameter set and vegetation indicator state vector corresponding to each pixel; The health status determination module includes: The health status determination unit is used to determine the health status of the crops corresponding to each pixel using the maximum likelihood estimation method based on the parameter set and vegetation indicator state vector corresponding to each pixel.

8. A prediction device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: Also includes: A remote sensing satellite for collecting remote sensing image data of crops in a preset area; wherein, when the processor executes the program, it implements the farmland health status prediction method based on remote sensing images as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method for predicting farmland health status based on remote sensing images as described in any one of claims 1 to 6 is implemented.

Images