Crop fertilization decision-making method and system

Abstract

The invention relates to the field of agriculture, in particular to a crop fertilization decision-making method, which comprises the following steps of: acquiring a shot crop spectrogram, determining a weight according to a central wavelength of a wave band, if a pixel value in the spectrogram of the wave band is equal to a preset value, adjusting the pixel value to be 1, otherwise, adjusting the pixel value to be 0, multiplying the weight by the adjusted spectrogram according to bits to obtain a spectrogram after wave band adjustment, and accumulating the spectrograms of all wave bands to obtain a feature map corresponding to a preset value; correcting the spectrogram of each wave band based on the characteristic pattern corresponding to the preset value and the central wavelength of the wave band; obtaining a vegetation index by using the corrected spectrogram, and obtaining a fertilizer application amount map of crops according to the vegetation index; the fertilization amount is adjusted in real time according to the fertilization amount map in the fertilization process. By correcting the spectrum, the influence of atmospheric scattering and the like on the spectrum is reduced, and the calculation accuracy of the fertilizing amount is improved.

Description

Technical Field

[0001] The present invention relates to the field of agriculture, and in particular to a crop fertilization decision-making method and system. Background Art

[0002] Reasonable fertilization can increase crop yields and improve quality while reducing environmental pollution and resource waste. However, traditional fertilization methods rely on farmers' experience or general fertilizer application standards, and often fail to accurately respond to the specific needs of different plots, climate conditions, or crop growth stages. Hyperspectral imaging technology, with its high resolution, can capture the needs of crop leaves for nutrients such as nitrogen, phosphorus, and potassium, and is a key tool for assessing crop nutritional status. Spectral images taken by drones or satellites can monitor crop health in real time and dynamically, and then make reasonable fertilization decisions based on specific nutritional needs.

[0003] However, hyperspectral images face many problems in practical applications, among which atmospheric influence is an important source of error. Aerosols, clouds, water vapor and other factors in the atmosphere will absorb and scatter the spectral signals reflected from the surface, resulting in a difference between the spectrum received by the sensor and the actual reflectivity of the crop. Hyperspectral images that have not been corrected for the atmosphere will affect the accuracy of fertilization decisions and may lead to misjudgment of the nutritional status of crops, thus causing problems of over- or under-fertilization. Summary of the invention

[0004] In a first aspect, the present invention provides a method for crop fertilization decision-making, the method comprising the following steps: Get the crop spectrum image taken, determine the weight according to the central wavelength of the band, if the pixel value in the spectrum image of the band is equal to the preset value, adjust the pixel value to 1, otherwise adjust it to 0, multiply the weight and the adjusted spectrum image bit by bit to obtain the spectrum image after the band adjustment, and accumulate the spectrum images of all bands to obtain the feature image corresponding to the preset value; Correcting the spectrum of each band based on the characteristic graph corresponding to the preset value and the central wavelength of the band; The vegetation index is obtained by using the corrected spectrum map, and a fertilizer application amount map of crops is obtained according to the vegetation index; during the fertilization process, the fertilizer application amount is adjusted in real time according to the fertilizer application amount map.

[0005] In combination with the first aspect, in some implementations, the spectral graph of each band is corrected based on the characteristic graph corresponding to the preset value and the central wavelength of the band, specifically: Determine the division interval of the band based on the central wavelength of the band, determine the number of selections corresponding to the division interval, sort the preset values ​​in ascending order, determine the preset value of the preset number starting from the first one after sorting, and add the characteristic graphs corresponding to the preset values ​​to obtain the characteristic graph of the band; The characteristic map of the band is used to obtain the adjustment factor of the original spectrum of the band, and the adjustment factor is subtracted from each pixel value in the original spectrum of the band to obtain a corrected spectrum; if the pixel value minus the adjustment factor is a negative number, the pixel value is set to 0.

[0006] In combination with the first aspect, in some implementations, the adjustment factor of the original spectrum graph of the band is obtained by using the characteristic graph of the band, specifically: Determine the position of the maximum value in the characteristic map of the band, and use the pixel value of the position in the original spectrum map of the band as the adjustment factor; or, determine the position that is not 0 in the characteristic map of the band, and use the average value of the pixel values ​​at the position in the original spectrum map of the band as the adjustment factor; or, multiply the characteristic map of the band and the original spectrum map of the band bit by bit and then sum them up, sum the elements of the characteristic map of the band, and use the ratio of the previous summation result to the next summation result as the adjustment factor.

[0007] In combination with the first aspect, in some implementations, obtaining an optimal fertilizer application amount map for crops according to the vegetation index is specifically: The correlation between different vegetation indices and different growth stages of crops and required nutrients is obtained, at least one vegetation index with the greatest correlation is selected to construct a fertilization amount model, and the vegetation index and crop growth stage are input into the fertilization amount model to obtain a fertilization amount map.

[0008] In conjunction with the first aspect, in some implementations, selecting at least one vegetation indicator with the greatest correlation to construct a fertilization amount model is specifically: A fitting function is constructed, wherein the input of the fitting function is the at least one vegetation index and the crop growth stage, the output of the fitting function is the fertilizer application amount of the nutrient component, and the fitting function is used as a fertilizer application amount model.

[0009] In conjunction with the first aspect, in some implementations, selecting at least one vegetation indicator with the greatest correlation to construct a fertilization amount model is specifically: Constructing training samples, wherein the characteristics of the training samples include at least the at least one indicator with the greatest correlation and the crop growth stage, and the training samples also include the fertilizer application amount corresponding to the nutrient components, using the training samples to train the integrated learning model, and using the trained integrated learning model as a fertilizer application amount model.

[0010] In a second aspect, the present invention provides a crop fertilization decision-making system, the system comprising the following modules: The spectrum acquisition module is used to obtain the captured crop spectrum, determine the weight according to the central wavelength of the band, and adjust the pixel value to 1 if the pixel value in the spectrum of the band is equal to the preset value, otherwise it is adjusted to 0. The weight and the adjusted spectrum are multiplied bit by bit to obtain the spectrum of the band adjustment, and the spectrum of all bands is accumulated to obtain the characteristic map corresponding to the preset value; A spectrum correction module, used to correct the spectrum graph of each band based on the characteristic graph corresponding to the preset value and the central wavelength of the band; The fertilization control module is used to obtain a vegetation index using the corrected spectrum map, and obtain a fertilizer amount map for crops according to the vegetation index; and adjust the fertilizer amount in real time according to the fertilizer amount map during the fertilization process.

[0011] In conjunction with the second aspect, in some implementations, the spectral graph of each band is corrected based on the characteristic graph corresponding to the preset value and the central wavelength of the band, specifically: Determine the division interval of the band based on the central wavelength of the band, determine the number of selections corresponding to the division interval, sort the preset values ​​in ascending order, determine the preset value of the preset number starting from the first one after sorting, and add the characteristic graphs corresponding to the preset values ​​to obtain the characteristic graph of the band; The characteristic map of the band is used to obtain the adjustment factor of the original spectrum of the band, and the adjustment factor is subtracted from each pixel value in the original spectrum of the band to obtain a corrected spectrum; if the pixel value minus the adjustment factor is a negative number, the pixel value is set to 0.

[0012] In conjunction with the second aspect, in some implementations, the adjustment factor of the original spectrum graph of the band is obtained by using the characteristic graph of the band, specifically: Determine the position of the maximum value in the characteristic map of the band, and use the pixel value of the position in the original spectrum map of the band as the adjustment factor; or, determine the position that is not 0 in the characteristic map of the band, and use the average value of the pixel values ​​at the position in the original spectrum map of the band as the adjustment factor; or, multiply the characteristic map of the band and the original spectrum map of the band bit by bit and then sum them up, sum the elements of the characteristic map of the band, and use the ratio of the previous summation result to the next summation result as the adjustment factor.

[0013] In conjunction with the second aspect, in some implementations, obtaining an optimal fertilizer application amount map for crops according to the vegetation index is specifically: The correlation between different vegetation indices and different growth stages of crops and required nutrients is obtained, at least one vegetation index with the greatest correlation is selected to construct a fertilization amount model, and the vegetation index and crop growth stage are input into the fertilization amount model to obtain a fertilization amount map.

[0014] In conjunction with the second aspect, in some implementations, selecting at least one vegetation indicator with the greatest correlation to construct a fertilizer application amount model is specifically: A fitting function is constructed, wherein the input of the fitting function is the at least one vegetation index and the crop growth stage, the output of the fitting function is the fertilizer application amount of the nutrient component, and the fitting function is used as a fertilizer application amount model.

[0015] In conjunction with the second aspect, in some implementations, selecting at least one vegetation indicator with the greatest correlation to construct a fertilizer application amount model is specifically: Constructing training samples, wherein the characteristics of the training samples include at least the at least one indicator with the greatest correlation and the crop growth stage, and the training samples also include the fertilizer application amount corresponding to the nutrient components, using the training samples to train the integrated learning model, and using the trained integrated learning model as a fertilizer application amount model.

[0016] In a third aspect, the present application provides a computer-readable storage medium, characterized in that instructions are stored in the computer-readable storage medium, and when the instructions are run on an electronic device, the method described in the first aspect is executed.

[0017] In a fourth aspect, the present application provides a computer program product, characterized in that the computer program product includes computer instructions, and when the computer program product is executed by an electronic device, the electronic device executes the method described in the first aspect.

[0018] In order to solve the problem that uncorrected spectral data may be affected by the atmosphere, resulting in inaccurate calculation of vegetation index and further affecting the scientific nature of fertilization decision-making, the present invention determines the positions of low pixel values ​​in the spectral graph from multiple bands according to the different effects of atmospheric scattering on different bands, and then further corrects the spectral graph of each band in combination with the central wavelength of the band, thereby avoiding inaccurate selection of dark pixels and using different adjustment factors for different bands in a targeted manner, thereby achieving more accurate correction and ensuring the accuracy of fertilizer calculation. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a flow chart of Embodiment 1; Figure 2 A schematic diagram for adjusting the spectrum according to preset values; Figure 3 A schematic diagram of weighting the adjusted spectrum; Figure 4 A schematic diagram of obtaining a characteristic graph corresponding to a preset value. DETAILED DESCRIPTION

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for describing the embodiments.

[0021] The terms "including" and "having" and any variations thereof mentioned in the description of this application are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device comprising a series of steps or units is not limited to the listed steps or units, but may optionally include other steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or devices.

[0022] It should be noted that, in the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0023] In the description of this application, unless otherwise specified, the meaning of "plurality" refers to two or more. "And / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone.

[0024] Applying fertilizers at the right time, in the right amount and in the right place can help improve the growth rate and health of crops, thereby directly improving the yield and quality of crops. Traditional fertilization methods may rely on experience and often cannot accurately control the amount and location of fertilizer application. Too little fertilization may cause a decrease in crop yields, while too much fertilization may lead to fertilizer waste and may also cause environmental problems such as groundwater pollution, eutrophication of rivers and lakes, etc.

[0025] Different substances have different absorption and reflection characteristics for light, so by analyzing the spectrum, we can obtain information about the physical and chemical characteristics of the substance. When light hits the surface of the plant, part of the light is reflected back to the sensor, and crops at different growth stages and different developmental conditions reflect light differently.

[0026] In a first embodiment of the present invention, a crop fertilization decision method is provided, comprising the following steps: Step 1: Obtain the captured crop spectrum, determine the weight according to the central wavelength of the band, if the pixel value in the spectrum of the band is equal to the preset value, adjust the pixel value to 1, otherwise adjust it to 0, multiply the weight and the adjusted spectrum bit by bit to obtain the spectrum of the band adjustment, and accumulate the spectrum of all bands to obtain the feature map corresponding to the preset value; The spectral image of the crops is captured by a multispectral sensor or a hyperspectral sensor carried by an unmanned aerial vehicle. In another embodiment, the spectral image of the crops is acquired by using a remote sensing satellite. Preferably, the spectral image is a hyperspectral image.

[0027] There are multiple channels in the spectrogram, each channel corresponds to a band, for example, the near-infrared band corresponds to a channel in the spectrogram. In remote sensing images, the light signal received by the sensor will first be converted into an analog electrical signal. Due to the influence of atmospheric scattering, the data collected by the spectrogram cannot be completely the reflection of the plant. Moreover, different wavelengths have different sensitivities to atmospheric scattering. The long-wave band is less affected by atmospheric scattering, which means that most of the light signals received by the sensor are signals reflected by crops, not scattered by the atmosphere; and the short-wave band is affected by atmospheric scattering much more than the long-wave band.

[0028] In an optional embodiment, the weight is determined according to the central wavelength of the band. Specifically, the shorter the central wavelength of the band, the greater the weight. In this way, if the pixel value of a short-wave band is 0 or close to 0, it indicates that the influence of atmospheric scattering at this pixel is indeed very low, and the credibility is higher than that of the long-wave band. In a more specific embodiment, the weight is determined by a negative exponential function, the input of which is the central wavelength of the band, and the output is the weight.

[0029] In the atmospheric correction of the spectrum by dark pixels, the selection of dark pixels is directly related to the correction effect. In many cases, the darkest pixel in the spectrum is directly selected. However, the spectral characteristics of different bands are quite different. Some pixels may appear as the darkest pixels in one band, but may not be the darkest in other bands. Moreover, if there is only one dark pixel point in a spectrum, this point may be caused by noise, etc. Directly selecting the darkest pixel in a certain band as the dark pixel of the whole band may lead to incorrect correction. In the present invention, multiple preset values ​​are pre-set. In one embodiment, the preset values ​​are all values ​​less than a certain value. Specifically, if a certain value is 5, the preset values ​​are 0, 1, 2, 3, 4; in another embodiment, the multiple preset values ​​are a number of consecutive integers starting from 0. In order to be able to perform atmospheric correction on the spectrum, the number of preset values ​​is within a certain number, such as 5 or 6.

[0030] There are multiple preset values, and a feature map is obtained for each preset value. Specifically, for the spectrum map of each band, each pixel value (pixel point) of the spectrum map is compared with a specific preset value. If the pixel value is equal to the preset value, the pixel value is replaced by 1; if it is not equal to the preset value, it is replaced by 0, such as Figure 2For each band of the spectral image, the pixel value is first adjusted to 1 or 0, and then multiplied by the corresponding weight of the band, and then the contribution value of each pixel is adjusted according to the weight, so as to generate a band-adjusted spectral image, as shown in Figure 3 The adjusted spectral graphs of each band are accumulated, and the pixels are added one by one, and the information of multiple bands is fused together to finally generate a feature graph, which is the feature graph corresponding to the preset value, as shown in Figure 4 shown.

[0031] For example, if the preset value is 0, the pixel values ​​in the spectrum of band 1 that are equal to the preset value, i.e., 0, are adjusted to 1, and the pixel values ​​in the spectrum of band 1 that are not equal to the preset value, i.e., 0, are adjusted to 0, and then the weight of band 1 and the adjusted spectrum of band 1 are multiplied, and the same operation is performed for other bands such as bands 2 and 3. Finally, the feature map obtained by adding the spectrum of all bands after the above operation is the feature map of the preset value 0. The larger the value on the feature map, the greater the possibility that the pixel is 0 in all bands, that is, the greater the possibility that it is a dark pixel. Similarly, feature maps of preset values ​​1, 2, etc. can be obtained.

[0032] Step 2, correcting the spectrum of each band based on the characteristic graph corresponding to the preset value and the central wavelength of the band; Due to the influence of sensor noise, etc., some pixels are very low but not really places with low reflectivity. In addition, different bands are affected differently by atmospheric scattering due to different wavelengths. For example, the short-wave band is more seriously affected by atmospheric scattering, and the pixel values ​​of its corresponding spectral graph are generally larger than the true values. Different bands should have different correction methods.

[0033] In an optional embodiment, the spectral graph of each band is corrected based on the characteristic graph corresponding to the preset value and the central wavelength of the band, specifically: Determine the division interval of the band based on the central wavelength of the band, determine the number of selections corresponding to the division interval, sort the preset values ​​in ascending order, determine the preset value of the preset number starting from the first one after sorting, and add the characteristic graphs corresponding to the preset values ​​to obtain the characteristic graph of the band; The characteristic map of the band is used to obtain the adjustment factor of the original spectrum of the band, and the adjustment factor is subtracted from each pixel value in the original spectrum of the band to obtain a corrected spectrum; if the pixel value minus the adjustment factor is a negative number, the pixel value is set to 0.

[0034] Since shortwave is more affected by atmospheric scattering, its value is higher than the true value. For example, if the true value is 1, the pixel value of the shortwave spectrum may be 2 or 3, which requires observing the situation of larger pixels. In one embodiment, the shorter the central wavelength of the band, the larger the number of selections corresponding to the division interval. For example, the central wavelength of band 1 is divided into the first division interval, and the corresponding number of selections is 2. If the central wavelength of band 2 is shorter than the central wavelength of band 1, it is divided into the second division interval, and the corresponding number of selections is 3. However, the maximum value of the number of selections does not exceed the number of preset values.

[0035] Arrange all given preset values ​​in ascending order. Assuming the preset values ​​are 0, 1, 2, 3, the order is 0, 1, 2, 3. Assuming that the number of selected band 1 is 2, the feature maps with preset values ​​of 0 and 1 will be selected, and the two feature maps will be added bit by bit to obtain the feature map of band 1. The feature map of band 2, the feature map of band 3, etc. can be determined in the same way.

[0036] Each band has a feature map, and the value of the pixel point on the feature map indicates the degree to which the pixel value at the position of this pixel point in all bands is relatively low. For example, the value of position (2, 3) on the feature map of band 1 is 3, and the value of position (3, 5) is 0, which means that the pixel values ​​in the spectrum maps of other bands at position (2, 3) are also very low, so this position is better as a dark pixel than (3, 5). In one embodiment, the adjustment factor of the original spectrum map of the band obtained by using the feature map of the band is specifically: Determine the position of the maximum value in the characteristic map of the band, and use the pixel value of the position in the original spectrum map of the band as the adjustment factor; or, determine the position that is not 0 in the characteristic map of the band, and use the average value of the pixel values ​​at the position in the original spectrum map of the band as the adjustment factor; or, multiply the characteristic map of the band and the original spectrum map of the band bit by bit and then sum them up, sum the elements of the characteristic map of the band, and use the ratio of the previous summation result to the next summation result as the adjustment factor.

[0037] There are several ways to determine the adjustment factor. One optional embodiment is to find the position with the largest pixel value in the characteristic map of the band, then check the pixel value of the original spectrum map of the band at the same position, and use this value as the adjustment factor. For example, the value of position (2, 3) on the characteristic map of band 1 is the largest, so the value of position (2, 3) in the original spectrum map of band 1 is used as the adjustment factor.

[0038] Another optional embodiment is to find all positions where the pixel values ​​are not 0 in the characteristic map of the band, then find the pixel values ​​of these positions in the original spectrum map of the band, and calculate the average value of these values ​​as the adjustment factor. For example, if the values ​​of positions (2, 3) and (9, 12) in the characteristic map of band 1 are not 0, then the average value of the values ​​of positions (2, 3) and (9, 12) in the original spectrum map of band 1 is used as the adjustment factor.

[0039] In another optional embodiment, the characteristic graph of the band and the original spectrum graph of the band are bitwise multiplied, that is, the pixel values ​​at the same position are multiplied, and then the product is element-wise summed. Next, the characteristic graph of the band itself is element-wise summed. Finally, the ratio of the two summation results is calculated, and the ratio is used as an adjustment factor. For example, the values ​​of positions (2, 3) and (9, 12) on the characteristic graph of band 1 are not 0, but 1 and 5 respectively, and the values ​​of positions (2, 3) and (9, 12) in the original spectrum graph of band 1 are 7 and 1. The characteristic graph of the band and the original spectrum graph of the band are bitwise multiplied, and then the product is element-wise summed, and the result is 12. The element-wise summation of the characteristic graph of the band itself is 6, and the ratio of the two is 2.

[0040] The above method can be used to calculate the adjustment coefficient of each band, so that each band will have an adjustment coefficient, and the adjustment factor is subtracted from each pixel value in the original spectrum of the band to obtain a corrected spectrum; if the pixel value minus the adjustment factor is a negative number, the pixel value is set to 0.

[0041] Step three, using the corrected spectrum to obtain a vegetation index, and obtaining a fertilizer application map for crops according to the vegetation index; and adjusting the fertilizer application amount in real time according to the fertilizer application map during the fertilization process.

[0042] The corrected spectrum can reduce the impact of atmospheric scattering on the spectrum, so that the vegetation index of crops can be obtained more accurately, and the calculation of the corresponding fertilizer amount will be more accurate. After obtaining the vegetation index, the vegetation index is used to obtain the optimal fertilizer amount. Since each pixel of the spectrum represents a certain range of the surface in the remote sensing image, the fertilizer amount within the range of the crop represented by the pixel can be obtained. During the fertilization process, the fertilizer amount is adjusted or controlled in real time according to the optimal fertilizer amount map. For example, when using a drone for foliar fertilization, the pixel where the drone is located is determined based on the position of the drone, and the fertilizer amount is further controlled according to the value of the pixel in the optimal fertilizer amount map. The vegetation index includes but is not limited to NDVI (Normalized Vegetation Index), EVI (Enhanced Vegetation Index), SAVI (Soil Adjusted Vegetation Index), etc.

[0043] Different crops at different growth stages have different correlations with the same vegetation index. In an optional embodiment, the optimal fertilizer application amount map for crops is obtained according to the vegetation index, specifically: The correlation between different vegetation indices and different growth stages of crops and required nutrients is obtained, at least one vegetation index with the greatest correlation is selected to construct a fertilization amount model, and the vegetation index and crop growth stage are input into the fertilization amount model to obtain a fertilization amount map.

[0044] By analyzing the relationship between each vegetation index and the nutrients required by the crop, such as nitrogen, phosphorus, and potassium, find out which vegetation index can best reflect the nutritional status of the crop. Preferably, a linear regression or integrated learning model is used to construct a relationship model between the vegetation index and the nutritional requirements of the crop. The selected vegetation index and the current crop growth stage are input into the fertilizer application model to calculate the fertilizer requirements of each pixel point.

[0045] In an optional embodiment, the step of selecting at least one vegetation indicator with the greatest correlation to construct a fertilizer application amount model is as follows: A fitting function is constructed, wherein the input of the fitting function is the at least one vegetation index and the crop growth stage, the output of the fitting function is the fertilizer application amount of the nutrient component, and the fitting function is used as a fertilizer application amount model.

[0046] Through mathematical or statistical methods, a fitting function is constructed to represent the relationship between input (vegetation index and crop growth stage) and output (nutrient fertilizer application amount). For example, at least one vegetation index with the greatest correlation with crop nutritional requirements, such as NDVI or NDRE, and crop growth stages, such as germination, vegetative growth, and grain filling, are input into the fitting function, and the output is the nutrient fertilizer application amount, such as the specific fertilizer application amount of nitrogen, phosphorus, and potassium.

[0047] In another optional embodiment, the step of selecting at least one vegetation indicator with the greatest correlation to construct a fertilizer application amount model is as follows: Constructing training samples, wherein the characteristics of the training samples include at least the at least one indicator with the greatest correlation and the crop growth stage, and the training samples also include the fertilizer application amount corresponding to the nutrient components, using the training samples to train the integrated learning model, and using the trained integrated learning model as a fertilizer application amount model.

[0048] In a second aspect, the present invention provides a crop fertilization decision-making system, the system comprising the following modules: The spectrum acquisition module is used to obtain the captured crop spectrum, determine the weight according to the central wavelength of the band, and adjust the pixel value to 1 if the pixel value in the spectrum of the band is equal to the preset value, otherwise it is adjusted to 0. The weight and the adjusted spectrum are multiplied bit by bit to obtain the spectrum of the band adjustment, and the spectrum of all bands is accumulated to obtain the characteristic map corresponding to the preset value; A spectrum correction module, used to correct the spectrum graph of each band based on the characteristic graph corresponding to the preset value and the central wavelength of the band; The fertilization control module is used to obtain a vegetation index using the corrected spectrum map, and obtain a fertilizer amount map for crops according to the vegetation index; and adjust the fertilizer amount in real time according to the fertilizer amount map during the fertilization process.

[0049] In conjunction with the second aspect, in some implementations, the spectral graph of each band is corrected based on the characteristic graph corresponding to the preset value and the central wavelength of the band, specifically: Determine the division interval of the band based on the central wavelength of the band, determine the number of selections corresponding to the division interval, sort the preset values ​​in ascending order, determine the preset value of the preset number starting from the first one after sorting, and add the characteristic graphs corresponding to the preset values ​​to obtain the characteristic graph of the band; The characteristic map of the band is used to obtain the adjustment factor of the original spectrum of the band, and the adjustment factor is subtracted from each pixel value in the original spectrum of the band to obtain a corrected spectrum; if the pixel value minus the adjustment factor is a negative number, the pixel value is set to 0.

[0050] In conjunction with the second aspect, in some implementations, the adjustment factor of the original spectrum graph of the band is obtained by using the characteristic graph of the band, specifically: Determine the position of the maximum value in the characteristic map of the band, and use the pixel value of the position in the original spectrum map of the band as the adjustment factor; or, determine the position that is not 0 in the characteristic map of the band, and use the average value of the pixel values ​​at the position in the original spectrum map of the band as the adjustment factor; or, multiply the characteristic map of the band and the original spectrum map of the band bit by bit and then sum them up, sum the elements of the characteristic map of the band, and use the ratio of the previous summation result to the next summation result as the adjustment factor.

[0051] In conjunction with the second aspect, in some implementations, obtaining an optimal fertilizer application amount map for crops according to the vegetation index is specifically: The correlation between different vegetation indices and different growth stages of crops and required nutrients is obtained, at least one vegetation index with the greatest correlation is selected to construct a fertilization amount model, and the vegetation index and crop growth stage are input into the fertilization amount model to obtain a fertilization amount map.

[0052] In conjunction with the second aspect, in some implementations, selecting at least one vegetation indicator with the greatest correlation to construct a fertilizer application amount model is specifically: A fitting function is constructed, wherein the input of the fitting function is the at least one vegetation index and the crop growth stage, the output of the fitting function is the fertilizer application amount of the nutrient component, and the fitting function is used as a fertilizer application amount model.

[0053] In conjunction with the second aspect, in some implementations, selecting at least one vegetation indicator with the greatest correlation to construct a fertilizer application amount model is specifically: Constructing training samples, wherein the characteristics of the training samples include at least the at least one indicator with the greatest correlation and the crop growth stage, and the training samples also include the fertilizer application amount corresponding to the nutrient components, using the training samples to train the integrated learning model, and using the trained integrated learning model as a fertilizer application amount model.

[0054] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more available media sets. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a magnetic tape), an optical medium (e.g., a high-density digital video disc (DVD), or a semiconductor medium. The semiconductor medium may be an SSD.

[0055] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A crop fertilization decision-making method, characterized in that: The method comprises the following steps: Get the crop spectrum image taken, determine the weight according to the central wavelength of the band, if the pixel value in the spectrum image of the band is equal to the preset value, adjust the pixel value to 1, otherwise adjust it to 0, multiply the weight and the adjusted spectrum image bit by bit to obtain the spectrum image after the band adjustment, and accumulate the spectrum images of all bands to obtain the feature image corresponding to the preset value; Correcting the spectrum of each band based on the characteristic graph corresponding to the preset value and the central wavelength of the band; The vegetation index is obtained by using the corrected spectrum map, and a fertilizer application amount map of crops is obtained according to the vegetation index; during the fertilization process, the fertilizer application amount is adjusted in real time according to the fertilizer application amount map.

2. The method according to claim 1, characterized in that The spectral graph of each band is corrected based on the characteristic graph corresponding to the preset value and the central wavelength of the band, specifically: Determine the division interval of the band based on the central wavelength of the band, determine the number of selections corresponding to the division interval, sort the preset values ​​in ascending order, determine the preset value of the preset number starting from the first one after sorting, and add the characteristic graphs corresponding to the preset values ​​to obtain the characteristic graph of the band; The characteristic map of the band is used to obtain the adjustment factor of the original spectrum of the band, and the adjustment factor is subtracted from each pixel value in the original spectrum of the band to obtain a corrected spectrum; if the pixel value minus the adjustment factor is a negative number, the pixel value is set to 0.

3. The method according to claim 2, characterized in that The adjustment factor of the original spectrum graph of the band is obtained by using the characteristic graph of the band, specifically: Determine the position of the maximum value in the characteristic map of the band, and use the pixel value of the position in the original spectrum map of the band as the adjustment factor; or determine the position that is not 0 in the characteristic map of the band, and use the average value of the pixel value of the position in the original spectrum map of the band as the adjustment factor; Alternatively, the characteristic graph of the band and the original spectrum graph of the band are bitwise multiplied and then element-wise summed, the characteristic graph of the band is element-wise summed, and the ratio of the previous summation result to the next summation result is used as the adjustment factor.

4. The method according to claim 1, characterized in that The optimal fertilizer application amount map for crops is obtained according to the vegetation index, specifically: The correlation between different vegetation indices and different growth stages of crops and required nutrients is obtained, at least one vegetation index with the greatest correlation is selected to construct a fertilization amount model, and the vegetation index and crop growth stage are input into the fertilization amount model to obtain a fertilization amount map.

5. The method according to claim 1, characterized in that The step of selecting at least one vegetation indicator with the greatest correlation to construct a fertilizer application amount model is specifically as follows: A fitting function is constructed, wherein the input of the fitting function is the at least one vegetation index and the crop growth stage, the output of the fitting function is the fertilizer application amount of the nutrient component, and the fitting function is used as a fertilizer application amount model.

6. The method according to claim 1, characterized in that The step of selecting at least one vegetation indicator with the greatest correlation to construct a fertilizer application amount model is specifically as follows: Constructing training samples, wherein the characteristics of the training samples include at least the at least one indicator with the greatest correlation and the crop growth stage, and the training samples also include the fertilizer application amount corresponding to the nutrient components, using the training samples to train the integrated learning model, and using the trained integrated learning model as a fertilizer application amount model.

7. A crop fertilization decision system, characterized in that: The system includes the following modules: The spectrum acquisition module is used to obtain the captured crop spectrum, determine the weight according to the central wavelength of the band, and adjust the pixel value to 1 if the pixel value in the spectrum of the band is equal to the preset value, otherwise it is adjusted to 0. The weight and the adjusted spectrum are multiplied bit by bit to obtain the spectrum of the band adjustment, and the spectrum of all bands is accumulated to obtain the characteristic map corresponding to the preset value; A spectrum correction module, used to correct the spectrum graph of each band based on the characteristic graph corresponding to the preset value and the central wavelength of the band; The fertilization control module is used to obtain a vegetation index using the corrected spectrum map, and obtain a fertilizer amount map for crops according to the vegetation index; and adjust the fertilizer amount in real time according to the fertilizer amount map during the fertilization process.

8. The system according to claim 7, characterized in that The spectral graph of each band is corrected based on the characteristic graph corresponding to the preset value and the central wavelength of the band, specifically: Determine the division interval of the band based on the central wavelength of the band, determine the number of selections corresponding to the division interval, sort the preset values ​​in ascending order, determine the preset value of the preset number starting from the first one after sorting, and add the characteristic graphs corresponding to the preset values ​​to obtain the characteristic graph of the band; The characteristic map of the band is used to obtain the adjustment factor of the original spectrum of the band, and the adjustment factor is subtracted from each pixel value in the original spectrum of the band to obtain a corrected spectrum; if the pixel value minus the adjustment factor is a negative number, the pixel value is set to 0.

9. The system according to claim 8, characterized in that The adjustment factor of the original spectrum graph of the band is obtained by using the characteristic graph of the band, specifically: Determine the position of the maximum value in the characteristic map of the band, and use the pixel value of the position in the original spectrum map of the band as the adjustment factor; or determine the position that is not 0 in the characteristic map of the band, and use the average value of the pixel value of the position in the original spectrum map of the band as the adjustment factor; Alternatively, the characteristic graph of the band and the original spectrum graph of the band are bitwise multiplied and then element-wise summed, the characteristic graph of the band is element-wise summed, and the ratio of the previous summation result to the next summation result is used as the adjustment factor.

10. A computer storage device, wherein a computer program is stored on the storage device, characterized in that: When the computer program is executed by a processor, the computer program implements the method according to any one of claims 1 to 6.

Images