Multi-spectral image intelligent processing method based on unmanned aerial vehicle

Abstract

The invention relates to the technical field of image processing, in particular to a multi-spectral image intelligent processing method based on an unmanned aerial vehicle, and the method comprises the steps: obtaining real-time data; determining a low illumination grid; determining a temporary grid; determining an extraction grid; determining an end member set; and adjusting and outputting the threshold value. According to the method, by comprehensively analyzing multi-dimensional data such as the real-time illumination intensity, the crop coverage rate, the ground height and the multi-spectral image of the farmland area, complex farmland environment changes such as factors such as uneven illumination and crop shielding can be effectively dealt with, and grids suitable for end member extraction are accurately screened out. By dynamically adjusting the light intensity threshold value and the synchronous threshold value, the end member extraction result can be optimized according to the actual situation, the accuracy and robustness of end member extraction are remarkably improved, noise interference is reduced, and the problems that spectral information is lost and processing is unstable due to the fact that graying processing and threshold value setting depend on manual adjustment are effectively solved.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a multispectral image intelligent processing method based on an unmanned aerial vehicle. Background Art

[0002] With the continuous advancement of UAV technology and remote sensing technology, multispectral images based on UAVs are increasingly used in the agricultural field, especially in precision agriculture monitoring. However, due to the complexity and variability of farmland environment, traditional image processing methods face challenges in the endmember extraction process and it is difficult to fully utilize the spatial and spectral information in multispectral images.

[0003] The patent document with publication number CN105631847A discloses a multispectral image processing method and device, the method comprising: calculating the grayscale image of the multispectral image and counting the histogram of the grayscale image; performing mean filtering on the histogram; constructing a differential symbol array and a support array filter, wherein the number of elements of the differential symbol array is equal to the dimension of the histogram, the initial value of the element of the differential symbol array is 0, and if the differential value obtained by performing differential calculation on the histogram after mean filtering is positive, the corresponding element in the differential symbol array is 1, if the difference value is negative, the corresponding element in the difference symbol array is -1; the number of elements of the support array filter is 2m, and the first m elements are 1, and the last m elements are -1, where m is a positive integer greater than or equal to 5; the support array filter is used as a mask to filter in the difference symbol array, and it is determined whether the obtained filter value is greater than the set threshold, and if so, the filter value is determined to be a peak of the histogram; and the number of calculated peaks is used as the number of clusters, and each pixel point of the multispectral image is clustered based on the k-means clustering algorithm.

[0004] It can be seen that the multispectral image processing method has the following problems: the method first converts the multispectral image into a grayscale image and processes it through histogram statistics, which will lead to the loss of some spectral information, especially when the spectral channel information in the multispectral image has high independence and diversity, the characteristics of each band cannot be fully retained after conversion to a grayscale image; the method relies on setting a threshold to determine whether the filter value is the peak of the histogram. The setting of this threshold needs to be adjusted according to the specific image, and lacks an automatic optimization mechanism, resulting in unstable performance in different environments or image types. Summary of the invention

[0005] To this end, the present invention provides a multispectral image intelligent processing method based on unmanned aerial vehicle, which is used to overcome the problems of spectral information loss and unstable processing in the prior art due to grayscale processing and threshold setting relying on manual adjustment through multi-scale feature fusion and dynamic threshold adjustment.

[0006] To achieve the above object, the present invention provides a multispectral image intelligent processing method based on unmanned aerial vehicle, comprising:

[0007] Obtain the real-time light intensity, real-time crop coverage, real-time average ground height and real-time multispectral image of each grid to be processed in the farmland drone collection area based on grid division;

[0008] Determine a number of low-irradiation grids according to the real-time light intensity and a preset light intensity threshold;

[0009] Determine a plurality of temporary grids according to the real-time crop coverage rate of each of the low-irradiation grids and the real-time multispectral image;

[0010] Determine a number of extraction grids according to the real-time multispectral image, the real-time average ground height and a preset synchronization threshold in each of the temporary grids;

[0011] determining a set of endmembers based on the real-time multispectral images of all the extraction grids;

[0012] Adjusting the preset light intensity threshold according to the end member set and the preset standard set to form an adjusted light intensity threshold;

[0013] Adjust the preset synchronization threshold according to the end member set formed based on the adjusted light intensity threshold within the preset adjustment time to form an adjusted synchronization threshold;

[0014] The end member set formed based on the adjusted synchronization threshold is output.

[0015] Further, determining a plurality of temporary grids according to the real-time crop coverage rate of each low-irradiation grid and the real-time multispectral image comprises:

[0016] Extracting the real-time spectral reflectance of the real-time multispectral image using a preset endmember extraction model;

[0017] Calculating an average value of the real-time spectral reflectance to form an average spectral reflectance;

[0018] Calculating a standard deviation of the real-time crop coverage rate within a preset first determined time period to form a coverage rate fluctuation value;

[0019] Calculating a standard deviation of the average spectral reflectance within the preset first determined time period to form a first reflectance fluctuation value;

[0020] A plurality of temporary grids are determined according to the coverage fluctuation value and the first reflectivity fluctuation value.

[0021] Further, determining a plurality of temporary grids according to the coverage fluctuation value and the first reflectivity fluctuation value includes:

[0022] Draw a coverage variation curve according to the coverage fluctuation value;

[0023] Draw a first reflectivity change curve according to the first reflectivity fluctuation value;

[0024] Calculating the cosine similarity of the coverage change curve and the first reflectivity change curve to form a change consistency;

[0025] When the change consistency is greater than a preset consistency threshold, the low-irradiation grid is determined to be a temporary grid.

[0026] Further, determining a plurality of extraction grids according to the real-time multispectral image in each of the temporary grids, the real-time average ground height, and a preset synchronization threshold comprises:

[0027] Calculating a standard deviation of the average spectral reflectance within a preset second determined time period to form a second reflectance fluctuation value;

[0028] Calculating the standard deviation of the real-time average ground height within the preset second determined time period to form a height fluctuation value;

[0029] Determine a plurality of abnormal grids according to the second reflectivity fluctuation value and the height fluctuation value;

[0030] The abnormal grids in all the grids to be processed are excluded to form a plurality of extracted grids.

[0031] Further, determining a number of abnormal grids according to the second reflectivity fluctuation value and the height fluctuation value includes:

[0032] Draw a second reflectivity change curve according to the second reflectivity fluctuation value;

[0033] Draw a height change curve according to the height fluctuation value;

[0034] Calculating the cosine similarity of the second reflectivity change curve and the height change curve to form a change synchronization degree;

[0035] When the change synchronization degree is less than the preset synchronization threshold, the temporary grid is determined to be an abnormal grid, and a plurality of abnormal grids are formed.

[0036] Further, determining an endmember set according to the real-time multispectral image of all the extraction grids comprises:

[0037] Preprocessing the real-time multispectral image to form a processed multispectral image;

[0038] The preset endmember extraction model is used to extract the real-time spectral reflectance of all the processed multispectral images to form an endmember set.

[0039] Further, adjusting the preset light intensity threshold according to the end member set and the preset standard set to form the adjusted light intensity threshold comprises:

[0040] Calculating the cosine similarity between the end member set and the preset standard set to form a set deviation;

[0041] When the collective deviation is greater than the preset deviation threshold, the preset light intensity threshold is adjusted according to the relative deviation between the collective deviation and the preset deviation threshold and a preset first adjustment coefficient to form an adjusted light intensity threshold.

[0042] Further, adjusting the preset synchronization threshold according to the end member set formed based on the adjusted light intensity threshold within the preset adjustment time period to form the adjusted synchronization threshold includes:

[0043] Calculating the standard deviation of the set quantity of the real-time endmember set within the preset adjustment time length to form a set quantity fluctuation value;

[0044] When the set quantity fluctuation value is greater than the preset quantity fluctuation threshold, the preset synchronization threshold is adjusted according to the relative deviation between the set quantity fluctuation value and the preset quantity fluctuation threshold and a preset second adjustment coefficient to form an adjusted synchronization threshold.

[0045] Further, determining a number of low illumination grids according to the real-time illumination intensity and a preset light intensity threshold comprises:

[0046] When the real-time light intensity is less than the preset light intensity threshold, the grid to be processed is determined to be a low-irradiation grid.

[0047] Further, preprocessing the real-time multispectral image to form a processed multispectral image includes:

[0048] A preset convex body optimization model is used to remove noise from the processed multispectral image to form a processed multispectral image.

[0049] Compared with the prior art, the beneficial effect of the present invention is that, by comprehensively analyzing multi-dimensional data such as real-time light intensity, crop coverage, ground height and multispectral images of farmland areas, it can effectively cope with complex farmland environmental changes, such as uneven lighting and crop occlusion, and accurately screen out grids suitable for extracting end members. By dynamically adjusting the light intensity threshold and synchronization threshold, the end member extraction results can be optimized according to actual conditions, the accuracy and robustness of end member extraction can be significantly improved, and noise interference can be reduced. This method not only improves the accuracy and reliability of multispectral image processing, but also provides accurate environmental monitoring and decision-making support for agricultural production, and has important application value. Through real-time feedback adjustment and optimization, it can better adapt to different farmland scenes, improve the intelligent management level in crop growth status assessment, pest and disease monitoring, fertilization and irrigation decision-making, etc., and effectively solve the problem of spectral information loss and unstable processing due to grayscale processing and threshold setting relying on manual adjustment.

[0050] Furthermore, by analyzing the fluctuations in crop coverage and average spectral reflectance, we can effectively screen out areas related to changes in the environment or crop growth status, help accurately identify temporary grids in farmland, improve the accuracy of data processing, and make subsequent abnormal grid identification and endmember set extraction more accurate, reducing misjudgments caused by environmental changes or image noise, thereby optimizing farmland monitoring and management effects.

[0051] Furthermore, by analyzing the consistency of changes in coverage and reflectivity fluctuations, grid areas that reflect consistent crop growth trends can be effectively identified, avoiding misjudgments caused by environmental noise and local interference, thereby improving the accuracy of grid screening.

[0052] Furthermore, by combining the fluctuation values ​​of the average spectral reflectance and the ground height, abnormal grids can be effectively identified, thereby filtering out unreliable grid data and ensuring that reliable and stable data are selected in the subsequent processing process, thereby improving the accuracy of data analysis and the overall processing effect.

[0053] Furthermore, by analyzing the changing trends of reflectivity and altitude and their synchronization, abnormal grids can be effectively identified, avoiding interference from erroneous data caused by environmental changes or equipment errors, thereby improving the accuracy and reliability of data processing and ensuring the effectiveness of subsequent analysis and decision-making.

[0054] Furthermore, by preprocessing multispectral images and accurately extracting spectral reflectance, noise and irrelevant factors can be effectively eliminated, improving the quality and reliability of image data. By forming endmember sets, the characteristics of different regions can be more accurately characterized, providing strong support for subsequent crop monitoring, land management and precision agriculture.

[0055] Furthermore, by calculating the cosine similarity of the end member set and the standard set and dynamically adjusting the light intensity threshold based on the set deviation, it can be ensured that the light intensity threshold more accurately reflects the actual image features, thereby improving the system's processing accuracy in low-illuminated areas. It can be optimized according to different image features, making the light intensity threshold more adaptable in different environments, thereby improving the overall acquisition effect and processing quality.

[0056] Furthermore, by adjusting the synchronization threshold, the response to changes in the light intensity threshold can be optimized and the recognition accuracy can be improved. Using the set number fluctuation value to determine whether to make adjustments can effectively avoid unnecessary adjustments, while ensuring that efficient working performance is maintained under large fluctuations, and enhancing the ability to adapt to changes in complex environments.

[0057] Furthermore, by introducing a preset light intensity threshold, areas with insufficient light can be effectively screened out, helping to accurately identify low-light areas that require special attention, optimize subsequent processing and analysis, and improve the ability to respond to changes in environmental conditions in crop monitoring, agricultural management or related fields.

[0058] Furthermore, by applying the preset convex body optimization model, the image quality can be effectively improved and the impact of noise on subsequent analysis can be reduced, thereby improving the accuracy and reliability of image processing and providing more accurate data support for subsequent feature extraction, analysis and decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 This is a flow chart of the multispectral image intelligent processing method based on drone in this embodiment;

[0060] Figure 2 This is a decision logic diagram for determining a temporary grid in this embodiment;

[0061] Figure 3 This is a logic diagram for determining abnormal grids in this embodiment;

[0062] Figure 4 This is a decision logic diagram for determining a low-irradiation grid in this embodiment. DETAILED DESCRIPTION

[0063] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0064] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0065] See also Figure 1As shown, it is a flow chart of the multispectral image intelligent processing method based on unmanned aerial vehicle in this embodiment;

[0066] This embodiment provides a multispectral image intelligent processing method based on a drone, including:

[0067] Obtain the real-time light intensity, real-time crop coverage, real-time average ground height and real-time multispectral image of each grid to be processed in the farmland drone collection area based on grid division;

[0068] Determine a number of low-irradiation grids according to the real-time light intensity and a preset light intensity threshold;

[0069] Determine a plurality of temporary grids according to the real-time crop coverage rate of each of the low-irradiation grids and the real-time multispectral image;

[0070] Determine a number of extraction grids according to the real-time multispectral image, the real-time average ground height and a preset synchronization threshold in each of the temporary grids;

[0071] determining a set of endmembers based on the real-time multispectral images of all the extraction grids;

[0072] Adjusting the preset light intensity threshold according to the end member set and the preset standard set to form an adjusted light intensity threshold;

[0073] Adjust the preset synchronization threshold according to the end member set formed based on the adjusted light intensity threshold within the preset adjustment time to form an adjusted synchronization threshold;

[0074] The end member set formed based on the adjusted synchronization threshold is output.

[0075] The real-time light intensity, real-time crop coverage, real-time average ground height and real-time multispectral images of each grid to be processed in the farmland drone collection area based on grid division mainly rely on the various sensors and imaging equipment carried by the drone. Specifically, the drone collects the light intensity data of the farmland area in real time through the integrated light sensor, and obtains image data of different bands through the high-resolution multispectral camera, which is used to calculate the crop coverage and spectral reflectivity. At the same time, the ground height information of the farmland is obtained by using laser radar or other ground height measurement technology. According to the spatial distribution of the farmland area, the collected data is gridded so that each grid unit can be analyzed one by one, and the real-time light, crop coverage, ground height and multispectral image data of each grid are obtained, which provides a basis for subsequent processing and analysis.

[0076] The preset light intensity threshold is a standard value used to distinguish between low-light areas and high-light areas. It depends on the lighting conditions in the farmland area and the response capability of the sensor. It is usually set between 100l ux and 500l ux. In this embodiment, it is set to 300 lux, which can ensure that low-light areas where the lighting conditions may not be suitable for precise extraction can be effectively identified, preparing for subsequent further analysis.

[0077] The preset synchronization threshold is a parameter used to determine whether the spectral reflectance fluctuation and the ground height fluctuation change synchronously. It depends on the fluctuation characteristics of the crop growth cycle, environmental changes and farmland terrain. It is usually set between 0.2-0.5. In this embodiment, it is set to 0.3, which can effectively identify and exclude those grids whose spectral reflectance and ground height changes are not synchronized, thereby eliminating errors caused by environmental anomalies or data noise, and improving the accuracy and reliability of data analysis.

[0078] Through the farmland area divided based on the grid, the real-time light intensity, crop coverage, ground height and multispectral images and other data are obtained, and multiple grids of the farmland are intelligently processed. First, the low-irradiation grids are screened out according to the real-time light intensity and the preset light intensity threshold. Then, the temporary grids are further screened out in combination with the crop coverage and multispectral images. Next, the extraction grid is determined based on the multispectral image, ground height and synchronization threshold. Through the preset endmember extraction model, the spectral reflectance is extracted and the endmember set is formed. According to the deviation between the endmember set and the preset standard set, the light intensity threshold is dynamically adjusted, and the synchronization threshold is optimized within the preset adjustment time, and finally the optimized endmember set is output.

[0079] By comprehensively analyzing multi-dimensional data such as real-time light intensity, crop coverage, ground height and multispectral images in farmland areas, it is possible to effectively cope with complex farmland environmental changes, such as uneven lighting and crop occlusion, and accurately screen out grids suitable for extracting endmembers. By dynamically adjusting the light intensity threshold and synchronization threshold, the endmember extraction results can be optimized according to actual conditions, significantly improving the accuracy and robustness of endmember extraction and reducing noise interference. This method not only improves the accuracy and reliability of multispectral image processing, but also provides accurate environmental monitoring and decision-making support for agricultural production, and has important application value. Through real-time feedback adjustment and optimization, it can better adapt to different farmland scenes, improve the intelligent management level of crop growth status assessment, pest and disease monitoring, fertilization and irrigation decision-making, and effectively solve the problem of spectral information loss and unstable processing due to grayscale processing and threshold setting relying on manual adjustment.

[0080] Specifically, determining a plurality of temporary grids according to the real-time crop coverage rate of each low-irradiation grid and the real-time multispectral image comprises:

[0081] Extracting the real-time spectral reflectance of the real-time multispectral image using a preset endmember extraction model;

[0082] Calculating an average value of the real-time spectral reflectance to form an average spectral reflectance;

[0083] Calculating a standard deviation of the real-time crop coverage rate within a preset first determined time period to form a coverage rate fluctuation value;

[0084] Calculating a standard deviation of the average spectral reflectance within the preset first determined time period to form a first reflectance fluctuation value;

[0085] Determine a plurality of temporary grids according to the coverage fluctuation value and the first reflectivity fluctuation value;

[0086] Among them, the preset endmember extraction model is an algorithm model designed for multispectral image data, usually based on machine learning or deep learning technology, and is used to extract representative spectral features from images. During the extraction process, the model first preprocesses the input real-time multispectral image to remove noise and irrelevant background information, and then extracts specific spectral reflectance values ​​from the spectral data of each pixel. These reflectance values ​​reflect information such as crops and their growth status, ground cover, etc. The endmember extraction model usually optimizes its parameters by learning a large amount of training data to identify and extract features that are important for crop identification or land cover classification. These spectral reflectances are used in subsequent analysis to help determine key indicators such as crop coverage and growth status.

[0087] The preset first determined time length is the length of the time window used to calculate the crop coverage and spectral reflectance fluctuation values, which depends on the crop growth cycle, environmental factors and the sampling frequency of the sensor. It is usually set between 1 hour and 1 day. In this embodiment, it is set to 6 hours, which can balance short-term and long-term changes and avoid misjudgment due to short-term fluctuations. At the same time, it can capture relatively stable trend changes and improve the reliability and effectiveness of data processing.

[0088] According to the real-time crop coverage and real-time multispectral image of the low-irradiation grid, the real-time spectral reflectance of each grid is first extracted using a preset end member extraction model. Then, the standard deviation of the real-time crop coverage within the preset first determined time length is calculated to obtain the coverage fluctuation value; at the same time, the standard deviation of the real-time spectral reflectance within the same time length is calculated to form a first reflectance fluctuation value. According to the changes in these two fluctuation values, several temporary grids are determined, and by comparing the fluctuations of crop coverage and spectral reflectance, areas with potential feature changes are identified.

[0089] The fluctuations of crop coverage and average spectral reflectance directly reflect the changes in crop growth status and environmental conditions in the farmland area. The fluctuations of crop coverage can reveal the changes in the density and spatial distribution of crop growth, while the fluctuations of spectral reflectance can reflect the differences in plant health and soil types. By analyzing the fluctuations of these two parameters, abnormal changes or uneven areas in the farmland can be effectively identified, which helps to more accurately assess potential problems such as crop growth and the occurrence of pests and diseases.

[0090] By analyzing the fluctuations in crop coverage and average spectral reflectance, we can effectively screen out areas related to changes in the environment or crop growth status, help accurately identify temporary grids in farmland, improve the accuracy of data processing, and make subsequent abnormal grid identification and endmember set extraction more accurate, reducing misjudgments caused by environmental changes or image noise, thereby optimizing farmland monitoring and management effects.

[0091] Please continue reading Figure 2 As shown, it is a decision logic diagram for determining a temporary grid in this embodiment;

[0092] Specifically, determining a number of temporary grids according to the coverage fluctuation value and the first reflectivity fluctuation value includes:

[0093] Draw a coverage variation curve according to the coverage fluctuation value;

[0094] Draw a first reflectivity change curve according to the first reflectivity fluctuation value;

[0095] Calculating the cosine similarity of the coverage change curve and the first reflectivity change curve to form a change consistency;

[0096] When the change consistency is greater than a preset consistency threshold, the low-irradiation grid is determined to be a temporary grid.

[0097] The preset consistency threshold is used to measure the similarity between the coverage change curve and the reflectivity change curve. It is usually set according to the characteristics of agricultural data and application requirements. Generally, it is set between 0 and 1. In this embodiment, it is set to 0.8, which can effectively filter out some interference and noise while ensuring recognition accuracy, ensuring that the extracted temporary grid reflects the actual changes in crop growth, reducing errors, and improving the accuracy of subsequent processing. The purpose of setting this value is to ensure the accuracy of grid screening, eliminate misjudgments caused by small-scale fluctuations, and thus improve the stability and credibility of the analysis.

[0098] According to the coverage fluctuation value and the first reflectance fluctuation value, the coverage change curve and the reflectance change curve are first drawn, and then the cosine similarity of the two curves is calculated to form the change consistency. When the change consistency is greater than the preset consistency threshold, the low-irradiation grid is judged as a temporary grid, and the comprehensive crop coverage and spectral reflectance change characteristics are used to help accurately identify stable grids with similar growth conditions.

[0099] By analyzing the consistency of changes in coverage and reflectivity fluctuations, grid areas that reflect consistent crop growth trends can be effectively identified, avoiding misjudgments caused by environmental noise and local interference, thereby improving the accuracy of grid screening.

[0100] Specifically, determining a number of extraction grids according to the real-time multispectral image in each temporary grid, the real-time average ground height and a preset synchronization threshold comprises:

[0101] Calculating a standard deviation of the average spectral reflectance within a preset second determined time period to form a second reflectance fluctuation value;

[0102] Calculating the standard deviation of the real-time average ground height within the preset second determined time period to form a height fluctuation value;

[0103] Determine a plurality of abnormal grids according to the second reflectivity fluctuation value and the height fluctuation value;

[0104] The abnormal grids in all the grids to be processed are excluded to form a plurality of extracted grids.

[0105] The preset second determination time length refers to the time interval used to calculate the standard deviation of the real-time spectral reflectance and the ground height, which is determined according to the crop growth cycle, the frequency of environmental fluctuations, and the analysis accuracy requirements. It is usually between a few minutes and tens of minutes. In this embodiment, it is set to 10 minutes, which can effectively capture short-term fluctuations while balancing data stability and response speed, thereby improving the recognition accuracy of abnormal grids and ensuring the accuracy of processing results.

[0106] First, the standard deviation of the average spectral reflectance within the preset second determined time period is calculated to obtain the second reflectance fluctuation value, and the standard deviation of the real-time average ground height is calculated to obtain the height fluctuation value. Subsequently, based on these two fluctuation values, the abnormal conditions of each temporary grid are analyzed, the grids with abnormal fluctuations are identified, these abnormal grids are excluded, and finally a group of extraction grids are determined.

[0107] The second reflectivity fluctuation value and height fluctuation value can determine abnormal grids because the crop coverage and ground height should maintain a certain stability under normal conditions, and their fluctuation values ​​reflect abnormal changes caused by environmental factors or acquisition errors. Larger reflectivity fluctuation values ​​and height fluctuation values ​​usually indicate that there are obvious interference or abnormal phenomena in the area, such as equipment failure, weather changes, etc. These changes will cause abnormalities in crops or ground features, thereby helping to identify these grids as abnormal grids. Therefore, by calculating and analyzing these two fluctuation values, abnormal grids can be effectively screened out, noise interference can be eliminated, and the accuracy of subsequent processing can be ensured.

[0108] By combining the fluctuation values ​​of average spectral reflectance and ground height, abnormal grids can be effectively identified, thereby filtering out unreliable grid data and ensuring that reliable and stable data are selected in the subsequent processing process, thereby improving the accuracy of data analysis and the overall processing effect.

[0109] Please continue reading Figure 3 As shown, it is a determination logic diagram for determining abnormal grids in this embodiment;

[0110] Specifically, determining a number of abnormal grids according to the second reflectivity fluctuation value and the height fluctuation value includes:

[0111] Draw a second reflectivity change curve according to the second reflectivity fluctuation value;

[0112] Draw a height change curve according to the height fluctuation value;

[0113] Calculating the cosine similarity of the second reflectivity change curve and the height change curve to form a change synchronization degree;

[0114] When the change synchronization degree is less than the preset synchronization threshold, the temporary grid is determined to be an abnormal grid, and a plurality of abnormal grids are formed.

[0115] First, the reflectivity change curve and the height change curve are drawn according to the second reflectivity fluctuation value and the height fluctuation value. These curves reflect the changing trend of crop coverage and ground height in a specific period of time. Then, the change synchronization degree is obtained by calculating the cosine similarity of the two curves. When the change synchronization degree is less than the preset synchronization threshold, it means that the changes of reflectivity and height are not synchronized, which may be due to abnormal factors, so the grid is judged as an abnormal grid.

[0116] By analyzing the changing trends of reflectivity and height and their synchronization, abnormal grids can be effectively identified, avoiding interference from erroneous data caused by environmental changes or equipment errors, thereby improving the accuracy and reliability of data processing and ensuring the effectiveness of subsequent analysis and decision-making.

[0117] Specifically, determining an endmember set according to the real-time multispectral image of all the extraction grids includes:

[0118] Preprocessing the real-time multispectral image to form a processed multispectral image;

[0119] The preset endmember extraction model is used to extract the real-time spectral reflectance of all the processed multispectral images to form an endmember set.

[0120] First, the real-time multispectral images of all extracted grids are preprocessed to remove noise and irrelevant information to ensure image quality. Then, the preset endmember extraction model is used to extract real-time spectral reflectance from the processed multispectral images. These spectral reflectances can represent the characteristic information of the ground and crops. Finally, the spectral reflectances are extracted and these spectral reflectance data sets are combined to form an endmember set, which includes the spectral characteristics of different areas in the image for subsequent analysis and processing.

[0121] By preprocessing multispectral images and accurately extracting spectral reflectance, noise and irrelevant factors can be effectively removed, improving the quality and reliability of image data. By forming endmember sets, the characteristics of different regions can be more accurately characterized, providing strong support for subsequent crop monitoring, land management and precision agriculture.

[0122] Specifically, adjusting the preset light intensity threshold according to the end member set and the preset standard set to form the adjusted light intensity threshold includes:

[0123] Calculating the cosine similarity between the end member set and the preset standard set to form a set deviation;

[0124] When the collective deviation is greater than the preset deviation threshold, the preset light intensity threshold is adjusted according to the relative deviation between the collective deviation and the preset deviation threshold and the preset first adjustment coefficient to form an adjusted light intensity threshold, wherein the adjustment method is to increase, and the relative deviation between the collective deviation and the preset deviation threshold is positively correlated with the adjusted light intensity threshold.

[0125] The preset deviation threshold is a standard used to determine whether the deviation between the end member set and the preset standard set is significant. It depends on the data characteristics and actual application requirements. It is usually set between 0.1 and 0.3. In this embodiment, it is set to 0.2, which helps to make adjustments within a looser deviation range and avoid over-sensitivity or over-tolerance.

[0126] The preset first adjustment coefficient is a coefficient used to adjust the light intensity threshold, which depends on the sensitivity requirement of the adjustment and is usually set between 0.5 and 2. In this embodiment, it is set to 1.5, which can achieve a balance in the adjustment of the light intensity threshold after calculating the deviation, so that the adjustment process is not excessive or too conservative.

[0127] First, the cosine similarity is calculated based on the end member set and the preset standard set to obtain the set deviation. Then, when the set deviation is greater than the preset deviation threshold, the original light intensity threshold is adjusted by further calculating the relative deviation between the set deviation and the preset deviation threshold and combining the preset first adjustment coefficient. Finally, the adjusted light intensity threshold is obtained.

[0128] By calculating the cosine similarity of the end member set and the standard set and dynamically adjusting the light intensity threshold based on the set deviation, it can be ensured that the light intensity threshold reflects the actual image features more accurately, thereby improving the system's processing accuracy in low-illuminated areas. It can be optimized according to different image features, making the light intensity threshold more adaptable in different environments, thereby improving the overall acquisition effect and processing quality.

[0129] Specifically, adjusting the preset synchronization threshold according to the end member set formed based on the adjusted light intensity threshold within the preset adjustment time period to form the adjusted synchronization threshold includes:

[0130] Calculating the standard deviation of the set quantity of the real-time endmember set within the preset adjustment time length to form a set quantity fluctuation value;

[0131] When the set quantity fluctuation value is greater than the preset quantity fluctuation threshold, the preset synchronization threshold is adjusted according to the relative deviation between the set quantity fluctuation value and the preset quantity fluctuation threshold and the preset second adjustment coefficient to form an adjusted synchronization threshold, wherein the adjustment method is to increase, and the relative deviation between the set quantity fluctuation value and the preset quantity fluctuation threshold is positively correlated with the adjusted synchronization threshold.

[0132] The preset number fluctuation threshold is a standard value used to determine whether the fluctuation of the number of collections is significant, which depends on the system's tolerance to changes in the number of collections and is usually set between 0.05 and 0.2. In this embodiment, it is set to 0.1 to balance the stability and response speed of the system, ensuring that the synchronization threshold is adjusted in time when the number of collections fluctuates greatly, thereby improving the sensitivity and adaptability of the system.

[0133] The preset second adjustment coefficient is a proportional coefficient used to adjust the synchronization threshold, which depends on the system's response to fluctuations and is usually set between 0.1 and 0.5. In this embodiment, it is set to 0.3 to ensure that the adjustment range can effectively cope with fluctuations without over-adjusting, thereby maintaining the stability and accuracy of the system.

[0134] The preset second adjustment coefficient is a proportional coefficient for adjusting the synchronization threshold, which depends on the system's response to fluctuations and is usually set between 0.1 and 0.5. In this embodiment, it is set to 0.3 to ensure that the adjustment range can effectively cope with fluctuations without over-adjustment, thereby maintaining the stability and accuracy of the system.

[0135] Within the preset adjustment time, the standard deviation of the set quantity of the real-time end member set is first calculated to obtain the set quantity fluctuation value. When the set quantity fluctuation value exceeds the preset quantity fluctuation threshold, the system adjusts the preset synchronization threshold according to the relative deviation between the set quantity fluctuation value and the preset quantity fluctuation threshold and the preset second adjustment coefficient, and finally forms the adjusted synchronization threshold.

[0136] By adjusting the synchronization threshold, the response to changes in light intensity threshold can be optimized and recognition accuracy can be improved. Using the set number fluctuation value to determine whether to make adjustments can effectively avoid unnecessary adjustments, while ensuring efficient working performance in the case of large fluctuations and enhancing the ability to adapt to changes in complex environments.

[0137] Please continue reading Figure 4 As shown, it is a decision logic diagram for determining a low-irradiation grid in this embodiment;

[0138] Specifically, determining a number of low illumination grids according to the real-time illumination intensity and a preset light intensity threshold comprises:

[0139] When the real-time light intensity is less than the preset light intensity threshold, the grid to be processed is determined to be a low-irradiation grid.

[0140] By monitoring the light intensity in real time and comparing it with the preset light intensity threshold, when the real-time light intensity is lower than the threshold, the system automatically identifies and marks these grids as low-light grids.

[0141] By introducing a preset light intensity threshold, areas with insufficient light can be effectively screened out, helping to accurately identify low-light areas that require special attention, optimize subsequent processing and analysis, and improve the ability to respond to changes in environmental conditions in crop monitoring, agricultural management or related fields.

[0142] Specifically, preprocessing the real-time multispectral image to form a processed multispectral image includes:

[0143] A preset convex body optimization model is used to remove noise from the processed multispectral image to form a processed multispectral image.

[0144] The preset convex body optimization model is an image processing method based on convex optimization theory, which aims to remove noise and maintain effective signals by optimizing the spectral reflectance and other features in the image. The model usually adopts mathematical optimization technology, minimizes the influence of noise by setting an objective function, and makes the spectral characteristics of the image more in line with the preset standard. By introducing convex body constraints, the model ensures that the structure of the image remains convex during the optimization process, so that external interference can be effectively removed while retaining the effective information of the image. Common applications include denoising, enhancing the contrast of the image, etc. In this embodiment, the preset convex body optimization model is a Total Variation (TV) denoising model of the denoising method based on convex optimization theory, which removes noise by minimizing the total change of the image gradient and maintains the edges and details in the image. This method regards the noise in the image as a high-frequency component by setting an objective function, and removes these components through a convex optimization process, thereby retaining the effective spectral information of the image.

[0145] First, the real-time multispectral image is preprocessed, and the noise in the image is removed by applying a preset convex body optimization model to obtain a clear processed multispectral image. The optimization model eliminates unnecessary interference signals by optimizing the spectral characteristics of the image.

[0146] By applying the preset convex body optimization model, the image quality can be effectively improved and the impact of noise on subsequent analysis can be reduced, thereby improving the accuracy and reliability of image processing and providing more accurate data support for subsequent feature extraction, analysis and decision-making.

[0147] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0148] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A multispectral image intelligent processing method based on unmanned aerial vehicle, characterized in that: include: Obtain the real-time light intensity, real-time crop coverage, real-time average ground height and real-time multispectral image of each grid to be processed in the farmland drone collection area based on grid division; Determine a number of low-irradiation grids according to the real-time light intensity and a preset light intensity threshold; Determine a plurality of temporary grids according to the real-time crop coverage rate of each of the low-irradiation grids and the real-time multispectral image; Determine a number of extraction grids according to the real-time multispectral image, the real-time average ground height and a preset synchronization threshold in each of the temporary grids; determining a set of endmembers based on the real-time multispectral images of all the extraction grids; Adjusting the preset light intensity threshold according to the end member set and the preset standard set to form an adjusted light intensity threshold; Adjust the preset synchronization threshold according to the end member set formed based on the adjusted light intensity threshold within the preset adjustment time to form an adjusted synchronization threshold; The end member set formed based on the adjusted synchronization threshold is output.

2. The multispectral image intelligent processing method based on unmanned aerial vehicle according to claim 1 is characterized in that: Determining a plurality of temporary grids according to the real-time crop coverage rate of each low-irradiation grid and the real-time multispectral image comprises: Extracting the real-time spectral reflectance of the real-time multispectral image using a preset endmember extraction model; Calculating an average value of the real-time spectral reflectance to form an average spectral reflectance; Calculating a standard deviation of the real-time crop coverage rate within a preset first determined time period to form a coverage rate fluctuation value; Calculating a standard deviation of the average spectral reflectance within the preset first determined time period to form a first reflectance fluctuation value; A plurality of temporary grids are determined according to the coverage fluctuation value and the first reflectivity fluctuation value.

3. The multispectral image intelligent processing method based on unmanned aerial vehicle according to claim 2 is characterized in that: Determining a number of temporary grids according to the coverage fluctuation value and the first reflectivity fluctuation value comprises: Draw a coverage variation curve according to the coverage fluctuation value; Draw a first reflectivity change curve according to the first reflectivity fluctuation value; Calculating the cosine similarity of the coverage change curve and the first reflectivity change curve to form a change consistency; When the change consistency is greater than a preset consistency threshold, the low-irradiation grid is determined to be a temporary grid.

4. The multispectral image intelligent processing method based on unmanned aerial vehicle according to claim 3 is characterized in that: Determining a number of extraction grids according to the real-time multispectral image in each of the temporary grids, the real-time average ground height, and a preset synchronization threshold comprises: Calculating a standard deviation of the average spectral reflectance within a preset second determined time period to form a second reflectance fluctuation value; Calculating the standard deviation of the real-time average ground height within the preset second determined time period to form a height fluctuation value; Determine a plurality of abnormal grids according to the second reflectivity fluctuation value and the height fluctuation value; The abnormal grids in all the grids to be processed are excluded to form a plurality of extracted grids.

5. The multispectral image intelligent processing method based on unmanned aerial vehicle according to claim 4 is characterized in that: Determining a number of abnormal grids according to the second reflectivity fluctuation value and the height fluctuation value comprises: Draw a second reflectivity change curve according to the second reflectivity fluctuation value; Draw a height change curve according to the height fluctuation value; Calculating the cosine similarity of the second reflectivity change curve and the height change curve to form a change synchronization degree; When the change synchronization degree is less than the preset synchronization threshold, the temporary grid is determined to be an abnormal grid, and a plurality of abnormal grids are formed.

6. The multispectral image intelligent processing method based on unmanned aerial vehicle according to claim 5 is characterized in that: Determining a set of endmembers based on the real-time multispectral images of all the extraction grids comprises: Preprocessing the real-time multispectral image to form a processed multispectral image; The preset endmember extraction model is used to extract the real-time spectral reflectance of all the processed multispectral images to form an endmember set.

7. The multispectral image intelligent processing method based on unmanned aerial vehicle according to claim 6 is characterized in that: Adjusting the preset light intensity threshold according to the end member set and the preset standard set to form the adjusted light intensity threshold comprises: Calculating the cosine similarity between the end member set and the preset standard set to form a set deviation; When the collective deviation is greater than the preset deviation threshold, the preset light intensity threshold is adjusted according to the relative deviation between the collective deviation and the preset deviation threshold and a preset first adjustment coefficient to form an adjusted light intensity threshold.

8. The multispectral image intelligent processing method based on unmanned aerial vehicle according to claim 7 is characterized in that: Adjusting the preset synchronization threshold according to the end member set formed based on the adjusted light intensity threshold within the preset adjustment time, forming the adjusted synchronization threshold includes: Calculating the standard deviation of the set quantity of the real-time endmember set within the preset adjustment time length to form a set quantity fluctuation value; When the set quantity fluctuation value is greater than the preset quantity fluctuation threshold, the preset synchronization threshold is adjusted according to the relative deviation between the set quantity fluctuation value and the preset quantity fluctuation threshold and a preset second adjustment coefficient to form an adjusted synchronization threshold.

9. The multispectral image intelligent processing method based on unmanned aerial vehicle according to claim 8 is characterized in that: Determining a number of low-irradiation grids according to the real-time illumination intensity and the preset light intensity threshold comprises: When the real-time light intensity is less than the preset light intensity threshold, the grid to be processed is determined to be a low-irradiation grid.

10. The multispectral image intelligent processing method based on unmanned aerial vehicle according to claim 9 is characterized in that: Preprocessing the real-time multispectral image to form a processed multispectral image includes: A preset convex body optimization model is used to remove noise from the processed multispectral image to form a processed multispectral image.

Images