CN118351431A - A rapid detection method for forestry seedling nutrient information based on specific band hyperspectral imaging equipment - Google Patents

Abstract

The invention provides a forestry seedling nutrient information rapid detection method based on a specific-band hyperspectral imaging device, and belongs to the technical field of spectral imaging. On the blade scale, the method comprises blade background rejection and sample extraction, blade weather information monitoring sensitive wave band characteristic analysis, blade nutrient content chemical value statistical analysis, blade spectrum data pretreatment, blade nutrient characteristic variable extraction, blade nutrient prediction model establishment based on characteristic variables and blade nutrient visual distribution. On the canopy scale, the method comprises canopy scale background rejection and sample extraction, canopy scale physical information monitoring sensitive wave band characteristic analysis, canopy nutrient content chemical value statistical analysis, canopy spectral data preprocessing, canopy nutrient characteristic variable extraction, canopy nutrient prediction model establishment based on characteristic variables and canopy nutrient visual distribution. By combining the research of two scales of the blade and the canopy, more comprehensive plant information can be provided, and the quick acquisition, real-time diagnosis and accurate management of the nutrient content of the forestry seedling are realized, so that the seedling management and production can be optimized, and the quality and yield of the forestry seedling can be improved.

Description

A rapid detection method for forestry seedling nutrient information based on a specific band hyperspectral imaging device

Technical Field

The present invention belongs to the technical field of spectral imaging, and in particular relates to a method for quickly detecting nutrient information of forestry seedlings based on a hyperspectral imaging device with a specific waveband.

Background technique

The nutrient content of seedlings is one of the important indicators for assessing their quality and health. Detecting the nutrient information of seedlings can help identify high-quality seedlings, ensure good growth potential and stress resistance, and thus improve the survival rate and growth rate of trees. The traditional nutrient detection method is to collect samples in the field and then measure them through laboratory chemical analysis methods. It has the advantages of high detection sensitivity and accurate results, but the detection cycle is long, the cost is high, the operation is complicated, and it is impossible to continuously monitor dynamically. It has obvious hysteresis, non-dynamicity and destructiveness, which is not conducive to large-scale promotion and use. Spectral detection technology realizes quantitative and qualitative analysis of crop characteristics based on crop spectral reflectance characteristics, but spectral detection technology cannot obtain images and spatial information containing crop morphology, color, texture and other characteristics, and cannot present the spatial distribution and spatiotemporal dynamic changes of crop nutrients in an intuitive way. It cannot fully characterize the characteristic information reflecting the nutritional status and growth of crops, thereby limiting the application of spectral technology in crop nutrient detection.

In recent years, hyperspectral imaging technology has been widely used for rapid detection of crop nutrients based on its high resolution and spectral-image integration. It can realize comprehensive analysis of the characteristic information of the research object from the perspective of spectrum, image and spectral information fusion, thereby obtaining richer characteristic information and more analytical perspectives. However, full-spectrum or wide-spectrum hyperspectral systems usually contain a large number of bands and generate a large amount of data, which requires higher costs for sensor design, manufacturing and maintenance, as well as more computing resources and time for processing and analysis. To this end, it is necessary to develop hyperspectral spectra with specific bands to improve cost-effectiveness, simplify data processing, and meet the specific needs of agricultural and forestry applications. Current research on the use of hyperspectral imaging technology in forestry mainly focuses on a single scale, usually studying the relationship between leaf spectral data and nutrient content. Since crop leaves can only reflect the nutrient status of crops in specific parts and cannot fully and effectively reflect the nutritional status of crops, there are certain limitations in practical applications. The crop spectrum at the canopy scale can comprehensively reflect the group growth status of crops. Estimating the nutrient content at the crop canopy scale is also of great significance for crop growth and development and growth monitoring.

Summary of the invention

1. Problems to be solved

In view of the above technical problems, the purpose of the present invention is to propose a method for rapid detection of nutrient information of forestry seedlings based on a hyperspectral imaging device with a specific wavelength band, so as to solve at least one of the above problems.

(II) Technical solution

The present invention is achieved through the following technical scheme, and the rapid detection method of the nutrient content of forestry seedlings at leaf scale includes background removal and sample extraction of forestry seedling leaves, sensitive band characteristic analysis of leaf phenological information monitoring, statistical analysis of chemical values of leaf nutrient content, leaf spectral data preprocessing, extraction of leaf nutrient characteristic variables, establishment of a leaf nutrient prediction model based on characteristic variables, and visualization distribution of leaf nutrients.

The rapid detection method for the nutrient content of forestry seedlings at the canopy scale includes canopy scale background removal and sample extraction, canopy scale phenological information monitoring sensitive band characteristic analysis, canopy nutrient content chemical value statistical analysis, canopy spectral data preprocessing, canopy nutrient characteristic variable extraction, canopy nutrient prediction model establishment based on characteristic variables, and canopy nutrient visualization distribution.

In some embodiments, the forestry seedling leaf background removal and sample extraction mainly use image segmentation algorithms to effectively separate the background information and the forestry seedling leaf information.

In some embodiments, the sensitive band characteristics of leaf phenological information monitoring are analyzed, and spectral reflectance data within the visible light band (545-670nm), red edge band (680nm-750nm) and near infrared band (800nm-910nm) are selected.

In some embodiments, the statistical analysis of leaf nutrient chemical values is performed, the nitrogen content of forestry seedling leaves is determined by the indophenol blue colorimetric method; the phosphorus content of forestry seedling leaves is determined by the molybdenum antimony colorimetric method; and the potassium content of forestry seedling leaves is determined by the flame photometry.

In some embodiments, leaf spectral data is preprocessed by selecting a detrending method as a spectral preprocessing method for predicting the nitrogen content in leaves of forestry seedlings; selecting a second-order differential method as a spectral preprocessing method for predicting the phosphorus content in leaves of forestry seedlings; and selecting a multivariate scattering correction method as a spectral preprocessing method for predicting the potassium content in leaves of forestry seedlings.

In some embodiments, leaf nutrient characteristic variables are extracted, independent component analysis is selected for leaf nitrogen content characteristic variable selection, genetic algorithm is selected for leaf phosphorus content characteristic variable selection, and uninformation variable elimination algorithm is selected for leaf potassium content characteristic variable selection.

In some embodiments, the nutrient distribution of leaves is visualized, and the nutrient content corresponding to each pixel on the sample image of forest seedling leaves is calculated based on the nutrient content prediction model and the acquired hyperspectral imaging data. Combined with pseudo-color image encoding technology, a visual distribution map of nutrients in forest seedling leaves is drawn.

In some embodiments, canopy-scale background removal and sample extraction are performed, and grayscale images in the 750nm band are selected to separate forest seedling canopy leaves and background, which can distinguish forest seedling leaves from soil and hyperspectral equipment tripod background.

In some embodiments, the sensitive band characteristics of canopy-scale phenological information monitoring are analyzed, and spectral data of the visible light band (545-670nm), red edge band (680nm-750nm) and near infrared band (800nm-910nm) are selected for analysis and research.

In some embodiments, canopy spectral data is preprocessed, and a direct orthogonal signal correction method is selected as a spectral preprocessing method for predicting the nitrogen, potassium, and phosphorus content of canopy leaves.

In some embodiments, canopy nutrient characteristic variables are extracted by selecting the uninformation variable elimination method to extract canopy nitrogen content characteristic variables, and the continuous projection algorithm is selected to extract leaf phosphorus and potassium content characteristic variables.

(III) Beneficial effects

It can be seen from the above technical scheme that the method for rapid detection of nutrient information of forestry seedlings based on a specific band hyperspectral imaging device provided by the present invention has the following beneficial effects:

(1) Through the study of both leaf and canopy scales, leaf-scale research can provide a deep understanding of the physiological characteristics, nutrient content, chlorophyll concentration, etc. of plants, while canopy-scale research can reveal the growth status, spatial distribution, canopy structure, etc. of plants. Combining the two scales of research can provide more comprehensive plant information and achieve rapid collection, real-time diagnosis and precise management of the nutrient content of forestry seedlings.

(2) Developing hyperspectral technology with specific wavelengths can centrally monitor and analyze key agricultural and forestry indicators and parameters, thereby reducing the cost of data collection and processing. Compared with full-spectrum or hyperspectral technology, hyperspectral technology with specific wavelengths can more specifically meet the specific needs of agriculture and forestry, avoiding unnecessary data collection and processing.

(3) Nutrient content feature extraction can be achieved through independent component analysis, genetic algorithm, uninformative variable elimination algorithm and other methods. The most representative and discriminative features can be extracted from hyperspectral data, reducing data redundancy and dimensionality, and retaining the most important information in the data.

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

FIG1 is a schematic diagram of a rapid detection method for nutrient content of forestry seedlings at leaf scale;

FIG2 is a schematic diagram of a rapid detection method for nutrient content of forestry seedlings at the canopy scale;

Implementation

The present invention proposes a method for rapid detection of nutrient information of forestry seedlings based on a hyperspectral imaging device with a specific band, and belongs to the field of spectral imaging technology. At the leaf scale, the method includes leaf background removal and sample extraction, sensitive band characteristic analysis of leaf phenological information monitoring, statistical analysis of chemical values of leaf nutrient content, leaf spectral data preprocessing, leaf nutrient characteristic variable extraction, establishment of a leaf nutrient prediction model based on characteristic variables, and leaf nutrient visualization distribution. At the canopy scale, the method includes canopy scale background removal and sample extraction, sensitive band characteristic analysis of canopy scale phenological information monitoring, statistical analysis of chemical values of canopy nutrient content, canopy spectral data preprocessing, canopy nutrient characteristic variable extraction, establishment of a canopy nutrient prediction model based on characteristic variables, and canopy nutrient visualization distribution. By combining the research of the two scales of leaves and canopies, more comprehensive plant information can be provided, and rapid collection, real-time diagnosis and precise management of nutrient content of forestry seedlings can be realized, which helps to optimize seedling management and production, and improve the quality and yield of forestry seedlings.

In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

To facilitate understanding of this embodiment, a method for rapid detection of nutrient information of forestry seedlings based on a hyperspectral imaging device with a specific wavelength band provided by an embodiment of the present invention is introduced in detail. The method for rapid detection of nutrient content of forestry seedlings at the leaf scale is shown in FIG1 , and includes:

Background removal and sample extraction of forest seedling leaves11, analysis of sensitive band characteristics for leaf phenological information monitoring12, statistical analysis of chemical values of leaf nutrient content13, preprocessing of leaf spectral data14, extraction of leaf nutrient characteristic variables15, establishment of leaf nutrient prediction model based on characteristic variables16, and visualization of leaf nutrient distribution17.

At the canopy scale, the rapid detection method for the nutrient content of forestry seedlings is shown in Figure 2, including:

Canopy scale background removal and sample extraction21, canopy scale phenological information monitoring sensitive band characteristics analysis22, canopy nutrient content chemical value statistical analysis23, canopy spectral data preprocessing24, canopy nutrient characteristic variable extraction25, canopy nutrient prediction model establishment based on characteristic variables26, and canopy nutrient visualization distribution27.

The forestry seedling leaf background elimination and sample extraction 11 mainly uses an image segmentation algorithm to effectively separate background information from forestry seedling leaf information. The image segmentation algorithm includes threshold segmentation, edge detection, region growth, etc.

The leaf phenology information monitoring sensitive band characteristic analysis 12 selects spectral reflectance data within the visible light band (545-670nm), red edge band (680nm-750nm) and near infrared band (800nm-910nm).

The statistical analysis of leaf nutrient chemical values 13 uses the indophenol blue colorimetric method to determine the nitrogen content in the leaves of forestry seedlings; the indophenol blue reagent will form a colorimetric reaction with nitrogen, and the color produced after the reaction is positively correlated with the nitrogen content. The absorbance of the reaction solution is measured using a colorimeter or a spectrophotometer to obtain an optical signal related to the nitrogen content. The molybdenum antimony colorimetric method is used to determine the phosphorus content in the leaves of forestry seedlings. The molybdenum antimony reagent will form a colorimetric reaction with phosphorus. The flame photometry method is used to determine the potassium content in the leaves of forestry seedlings. The leaf sample is ground into powder, and the potassium element in the leaf is dissolved into a measurable form using an appropriate solvent (such as hydrochloric acid or nitric acid). After the solution is appropriately diluted, it is measured using a flame photometer. In the flame, potassium produces a characteristic emission spectrum signal.

The leaf spectral data preprocessing 14, selecting the detrending method as the spectral preprocessing method for predicting the nitrogen content of forest seedling leaves, can help eliminate the trend and background influence in the spectral data, and improve the accuracy of the prediction model; selecting the second-order differential method as the spectral preprocessing method for predicting the phosphorus content of forest seedling leaves, this method is based on the quadratic derivative information of the spectral curve, can highlight the subtle changes in the spectrum, and improve the feature discrimination of the data; selecting the multivariate scattering correction method as the spectral preprocessing method for predicting the potassium content of forest seedling leaves, this method is mainly used to remove the interference signal caused by multiple scattering in the reflectance spectrum, thereby improving the accuracy of the prediction model.

The leaf nutrient characteristic variable extraction 15 selects the independent component analysis method for leaf nitrogen content characteristic variable selection, selects the genetic algorithm for leaf phosphorus content characteristic variable selection, and selects the uninformed variable elimination algorithm for leaf potassium content characteristic variable selection.

The leaf nutrient visualization distribution 17 is based on the nutrient content prediction model and the acquired hyperspectral imaging data, and the nutrient content corresponding to each pixel on the forest seedling leaf sample image is calculated, and the nutrient content of each pixel is mapped to the corresponding color according to the numerical range of the nutrient content using pseudo-color image encoding technology. A gradient color spectrum is used, such as a color transition from low to high, to express the difference in nutrient content. The pseudo-color-coded nutrient content is superimposed on the original image to generate a visualization distribution map of the nutrients in the forest seedling leaves. In this figure, different colors represent different nutrient content levels, thereby presenting the spatial changes in leaf nutrients.

The canopy scale background elimination and sample extraction 21 selects the grayscale image of the 750nm band to separate the canopy leaves of forestry seedlings from the background, and can distinguish the leaves of forestry seedlings from the soil and the background of the tripod of the hyperspectral equipment.

The sensitive band characteristics analysis of canopy-scale phenological information monitoring22 selected spectral data of the visible light band (545-670nm), red edge band (680nm-750nm) and near infrared band (800nm-910nm) for analysis and research.

The canopy spectral data preprocessing 24 selects the direct orthogonal signal correction method as the spectral preprocessing method for predicting the nitrogen, potassium and phosphorus contents of canopy leaves, which can remove non-correlated components and systematic changes and improve the correlation and prediction accuracy of spectral data.

The canopy nutrient characteristic variable extraction 25 selects the information-free variable elimination method to extract the canopy nitrogen content characteristic variable, and selects the continuous projection algorithm to extract the leaf phosphorus and potassium content characteristic variables.

The preferred embodiments of the present invention are described in detail above in conjunction with the accompanying drawings. However, the present invention is not limited to the specific details in the above embodiments. Within the technical concept of the present invention, a variety of simple modifications can be made to the technical solution of the present invention, and these simple modifications all belong to the protection scope of the present invention.

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

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

BRIEF DESCRIPTION OF THE DRAWINGS

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

FIG1 is a schematic diagram of a rapid detection method for nutrient content of forestry seedlings at leaf scale;

FIG2 is a schematic diagram of a rapid detection method for nutrient content of forestry seedlings at the canopy scale;

Detailed ways

The present invention proposes a method for rapid detection of nutrient information of forestry seedlings based on a hyperspectral imaging device with a specific band, and belongs to the field of spectral imaging technology. At the leaf scale, the method includes leaf background removal and sample extraction, sensitive band characteristic analysis of leaf phenological information monitoring, statistical analysis of chemical values of leaf nutrient content, leaf spectral data preprocessing, leaf nutrient characteristic variable extraction, establishment of a leaf nutrient prediction model based on characteristic variables, and leaf nutrient visualization distribution. At the canopy scale, the method includes canopy scale background removal and sample extraction, sensitive band characteristic analysis of canopy scale phenological information monitoring, statistical analysis of chemical values of canopy nutrient content, canopy spectral data preprocessing, canopy nutrient characteristic variable extraction, establishment of a canopy nutrient prediction model based on characteristic variables, and canopy nutrient visualization distribution. By combining the research of the two scales of leaves and canopies, more comprehensive plant information can be provided, and rapid collection, real-time diagnosis and precise management of nutrient content of forestry seedlings can be realized, which helps to optimize seedling management and production, and improve the quality and yield of forestry seedlings.

In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

To facilitate understanding of this embodiment, a method for rapid detection of nutrient information of forestry seedlings based on a hyperspectral imaging device with a specific wavelength band provided by an embodiment of the present invention is introduced in detail. The method for rapid detection of nutrient content of forestry seedlings at the leaf scale is shown in FIG1 , and includes:

Background removal and sample extraction of forest seedling leaves11, analysis of sensitive band characteristics for leaf phenological information monitoring12, statistical analysis of chemical values of leaf nutrient content13, preprocessing of leaf spectral data14, extraction of leaf nutrient characteristic variables15, establishment of leaf nutrient prediction model based on characteristic variables16, and visualization of leaf nutrient distribution17.

At the canopy scale, the rapid detection method for the nutrient content of forestry seedlings is shown in Figure 2, including:

Canopy scale background removal and sample extraction21, canopy scale phenological information monitoring sensitive band characteristics analysis22, canopy nutrient content chemical value statistical analysis23, canopy spectral data preprocessing24, canopy nutrient characteristic variable extraction25, canopy nutrient prediction model establishment based on characteristic variables26, and canopy nutrient visualization distribution27.

The forestry seedling leaf background elimination and sample extraction 11 mainly uses an image segmentation algorithm to effectively separate background information from forestry seedling leaf information. The image segmentation algorithm includes threshold segmentation, edge detection, region growth, etc.

The leaf phenology information monitoring sensitive band characteristic analysis 12 selects spectral reflectance data within the visible light band (545-670nm), red edge band (680nm-750nm) and near infrared band (800nm-910nm).

The statistical analysis of leaf nutrient chemical values 13 uses the indophenol blue colorimetric method to determine the nitrogen content in the leaves of forestry seedlings; the indophenol blue reagent will form a colorimetric reaction with nitrogen, and the color produced after the reaction is positively correlated with the nitrogen content. The absorbance of the reaction solution is measured using a colorimeter or a spectrophotometer to obtain an optical signal related to the nitrogen content. The molybdenum antimony colorimetric method is used to determine the phosphorus content in the leaves of forestry seedlings. The molybdenum antimony reagent will form a colorimetric reaction with phosphorus. The flame photometry method is used to determine the potassium content in the leaves of forestry seedlings. The leaf sample is ground into powder, and the potassium element in the leaf is dissolved into a measurable form using an appropriate solvent (such as hydrochloric acid or nitric acid). After the solution is appropriately diluted, it is measured using a flame photometer. In the flame, potassium produces a characteristic emission spectrum signal.

The leaf spectral data preprocessing 14, selecting the detrending method as the spectral preprocessing method for predicting the nitrogen content of forest seedling leaves, can help eliminate the trend and background influence in the spectral data, and improve the accuracy of the prediction model; selecting the second-order differential method as the spectral preprocessing method for predicting the phosphorus content of forest seedling leaves, this method is based on the quadratic derivative information of the spectral curve, can highlight the subtle changes in the spectrum, and improve the feature discrimination of the data; selecting the multivariate scattering correction method as the spectral preprocessing method for predicting the potassium content of forest seedling leaves, this method is mainly used to remove the interference signal caused by multiple scattering in the reflectance spectrum, thereby improving the accuracy of the prediction model.

The leaf nutrient characteristic variable extraction 15 selects the independent component analysis method for leaf nitrogen content characteristic variable selection, selects the genetic algorithm for leaf phosphorus content characteristic variable selection, and selects the uninformed variable elimination algorithm for leaf potassium content characteristic variable selection.

The leaf nutrient visualization distribution 17 is based on the nutrient content prediction model and the acquired hyperspectral imaging data, and the nutrient content corresponding to each pixel on the forest seedling leaf sample image is calculated, and the nutrient content of each pixel is mapped to the corresponding color according to the numerical range of the nutrient content using pseudo-color image encoding technology. A gradient color spectrum is used, such as a color transition from low to high, to express the difference in nutrient content. The pseudo-color-coded nutrient content is superimposed on the original image to generate a visualization distribution map of the nutrients in the forest seedling leaves. In this figure, different colors represent different nutrient content levels, thereby presenting the spatial changes in leaf nutrients.

The canopy scale background elimination and sample extraction 21 selects the grayscale image of the 750nm band to separate the canopy leaves of forestry seedlings from the background, and can distinguish the leaves of forestry seedlings from the soil and the background of the tripod of the hyperspectral equipment.

The sensitive band characteristics analysis of canopy-scale phenological information monitoring22 selected spectral data of the visible light band (545-670nm), red edge band (680nm-750nm) and near infrared band (800nm-910nm) for analysis and research.

The canopy spectral data preprocessing 24 selects the direct orthogonal signal correction method as the spectral preprocessing method for predicting the nitrogen, potassium and phosphorus contents of canopy leaves, which can remove non-correlated components and systematic changes and improve the correlation and prediction accuracy of spectral data.

The canopy nutrient characteristic variable extraction 25 selects the information-free variable elimination method to extract the canopy nitrogen content characteristic variable, and selects the continuous projection algorithm to extract the leaf phosphorus and potassium content characteristic variables.

The preferred embodiments of the present invention are described in detail above in conjunction with the accompanying drawings. However, the present invention is not limited to the specific details in the above embodiments. Within the technical concept of the present invention, a variety of simple modifications can be made to the technical solution of the present invention, and these simple modifications all belong to the protection scope of the present invention.

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

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

Claims (11)

1. A method for rapid detection of nutrient information of forestry seedlings based on a specific band hyperspectral imaging device, characterized in that, at the leaf scale, the method for rapid detection of nutrient content of forestry seedlings includes background removal and sample extraction of forestry seedling leaves, characteristic analysis of sensitive bands for monitoring leaf phenological information, statistical analysis of chemical values of leaf nutrient content, preprocessing of leaf spectral data, extraction of leaf nutrient characteristic variables, establishment of a leaf nutrient prediction model based on characteristic variables, and visualization of leaf nutrient distribution; at the canopy scale, the method for rapid detection of nutrient content of forestry seedlings includes background removal and sample extraction at the canopy scale, characteristic analysis of sensitive bands for monitoring leaf phenological information, statistical analysis of chemical values of canopy nutrient content, preprocessing of canopy spectral data, extraction of canopy nutrient characteristic variables, establishment of a canopy nutrient prediction model based on characteristic variables, and visualization of canopy nutrient distribution. 2. According to claim 1, a method for rapid detection of nutrient information of forestry seedlings based on a specific band hyperspectral imaging device is characterized in that background removal and sample extraction of forestry seedling leaves mainly use image segmentation algorithm to effectively separate background information and forestry seedling leaf information. 3. According to claim 1, a method for rapid detection of nutrient information of forestry seedlings based on a specific band hyperspectral imaging device is characterized in that the sensitive band characteristic analysis of leaf phenological information monitoring selects spectral reflectance data within the visible light band (545-670nm), red edge band (680nm-750nm) and near infrared band (800nm-910nm). 4. According to claim 1, a method for rapid detection of nutrient information of forestry seedlings based on a specific band hyperspectral imaging device is characterized in that the statistical analysis of leaf nutrient chemical values uses indophenol blue colorimetry to determine the nitrogen content of forestry seedling leaves; uses molybdenum antimony colorimetry to determine the phosphorus content of forestry seedling leaves; and uses flame photometry to determine the potassium content of forestry seedling leaves. 5. According to claim 1, a method for rapid detection of nutrient information of forestry seedlings based on a specific band hyperspectral imaging device is characterized in that the leaf spectral data preprocessing selects a detrending method as a spectral preprocessing method for predicting the nitrogen content of forestry seedling leaves; selects a second-order differential method as a spectral preprocessing method for predicting the phosphorus content of forestry seedling leaves; and selects a multivariate scattering correction method as a spectral preprocessing method for predicting the potassium content of forestry seedling leaves. 6. According to a method for rapid detection of nutrient information of forestry seedlings based on a specific band hyperspectral imaging device as described in claim 1, it is characterized in that the independent component analysis method is selected for leaf nutrient characteristic variable extraction to select leaf nitrogen content characteristic variables, the genetic algorithm is selected for leaf phosphorus content characteristic variables, and the uninformation variable elimination algorithm is selected for leaf potassium content characteristic variables. 7. According to claim 1, a method for rapid detection of nutrient information of forestry seedlings based on a specific band hyperspectral imaging device is characterized in that the leaf nutrient visualization distribution is based on the nutrient content prediction model and the acquired hyperspectral imaging data, and the nutrient content corresponding to each pixel point on the forestry seedling leaf sample image is calculated, and combined with pseudo-color image coding technology, a visualization distribution map of the nutrients of the forestry seedling leaves is drawn. 8. According to claim 1, a method for rapid detection of nutrient information of forestry seedlings based on a specific band hyperspectral imaging device is characterized in that canopy-scale background elimination and sample extraction select a grayscale image in the 750nm band to separate the canopy leaves and background of forestry seedlings, and can distinguish the leaves of forestry seedlings from the soil and the background of the hyperspectral device tripod. 9. According to claim 1, a method for rapid detection of nutrient information of forestry seedlings based on a specific band hyperspectral imaging device is characterized in that the spectral data of the visible light band (545-670nm), red edge band (680nm-750nm) and near infrared band (800nm-910nm) are selected for analysis and research for the sensitive band characteristic analysis of canopy-scale phenological information monitoring. 10. According to claim 1, a method for rapid detection of nutrient information of forestry seedlings based on a specific band hyperspectral imaging device is characterized in that the canopy spectral data preprocessing selects a direct orthogonal signal correction method as a spectral preprocessing method for predicting the nitrogen, potassium and phosphorus content of canopy leaves. 11. According to claim 1, a method for rapid detection of nutrient information of forestry seedlings based on a specific band hyperspectral imaging device is characterized in that the canopy nutrient characteristic variable extraction selects the information-free variable elimination method to extract the canopy nitrogen content characteristic variable, and selects the continuous projection algorithm to extract the leaf phosphorus and potassium content characteristic variables.