Remote sensing monitoring and evaluation method and device for vegetation diseases and insect pests in loess hilly region

Abstract

The application provides a loess hilly vegetation disease and pest remote sensing monitoring and evaluation method and device, relates to the technical field of remote sensing monitoring and evaluation methods, and comprises the following steps: calculating the coverage of vegetation by using a double threshold plane method, taking the coverage of vegetation of each grid area as a correction term of ground surface temperature; constructing a temperature difference anomaly index of each grid area; based on the temperature difference anomaly index and transpiration efficiency index of each grid area, calculating the transpiration efficiency index of each grid area by using a vegetation temperature difference method, constructing a vegetation classification model by using a logistic regression method, and classifying each grid area according to a preset threshold value to distinguish a risk grid area and a normal grid area. A prediction model is constructed through the statistical relationship between the vegetation index and the ground surface temperature, the temperature difference anomaly index is quantified, early stress signals of vegetation that are difficult to find can be recognized, the transpiration efficiency is introduced to comprehensively evaluate the water utilization of vegetation, and through a dynamic threshold value and a statistical model, potential risk grids can be recognized before the outbreak of diseases and pests.

Description

Technical Field

[0001] This invention relates to the field of remote sensing monitoring and assessment methods, specifically to remote sensing monitoring and assessment methods and devices for vegetation diseases and pests in the Loess Hilly Area. Background Technology

[0002] Vegetation diseases and pests pose a serious challenge to ecological restoration and agricultural production in the Loess Plateau region. Traditional field monitoring methods are time-consuming, labor-intensive, and have limited coverage, making it difficult to achieve large-scale, dynamic early warning. Existing remote sensing monitoring technologies mostly rely on single vegetation indices or surface temperatures for diagnosis. However, relying solely on single indicators such as decreased vegetation greenness or increased temperature is easily confused with apparent changes caused by factors such as water stress, soil background, or seasonal variations, leading to high misjudgment rates and insufficient warning accuracy.

[0003] In the existing technology, the document with publication number CN118097459A proposes a pest and disease generation analysis model, which analyzes the causes of pests and diseases through cluster analysis. However, this method does not construct a temperature difference anomaly index, and cannot sensitively capture the vegetation transpiration dysfunction and energy balance abnormalities caused by pests and diseases. It does not establish a temperature baseline for healthy vegetation and calculate the transpiration efficiency index, so it cannot quantify temperature anomalies under unit vegetation cover and cannot directly reflect the decline of vegetation physiological functions. It does not accurately use vegetation cover information obtained based on red light and near-infrared bands to correct the temperature information retrieved from the thermal infrared band, and achieve deep fusion and physical consistency coordination of multi-source remote sensing data. It does not fundamentally improve the scientificity and practicality of the entire monitoring and evaluation system. Therefore, there is an urgent need for remote sensing monitoring and evaluation methods and devices for vegetation pests and diseases in the Loess Plateau.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a remote sensing monitoring and assessment method and device for vegetation pests and diseases in the Loess Hilly Area, so as to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] Remote sensing monitoring and assessment methods for vegetation diseases and pests in the Loess Hilly Areas, including the following specific steps:

[0008] S1: Acquire multi-band spectral images of the vegetation area to be monitored, and divide it into grids to obtain multiple grid areas. Extract the spectral sub-images corresponding to each grid area from the spectral images. Simultaneously acquire the surface temperature of each grid area during high-temperature and low-temperature periods, which are used as the highest and lowest surface temperatures of the corresponding grid areas.

[0009] S2: Based on the spectral sub-images corresponding to each grid region, the reflectance data of the preset feature bands are extracted. The vegetation parameters of each pixel in the spectral sub-image are constructed using the normalized difference method. The vegetation parameters of all pixels in the same grid region are averaged to obtain the vegetation index of the corresponding grid region. Based on the vegetation index, the vegetation coverage of the corresponding grid region is calculated using the double threshold averaging method. The vegetation coverage is used as a temperature correction factor to correct the surface temperature of the corresponding grid region.

[0010] S3: Obtain vegetation index and corresponding land surface temperature data for healthy vegetation samples, and perform parameter fitting to obtain the fitting relationship between vegetation index and land surface temperature. Substitute the vegetation index of each grid area into the fitting relationship to obtain the predicted land surface temperature of each grid area, and calculate the standard deviation of the predicted land surface temperature of all grid areas. Standardize the corrected highest and lowest land surface temperatures of each grid area using the standard deviation to obtain the temperature difference anomalous index of each grid area. Combined with the corresponding predicted land surface temperature, calculate the transpiration efficiency index of each grid area using the vegetation temperature difference method.

[0011] S4: Based on the temperature difference anomalous index and transpiration efficiency index of each grid area, a vegetation status classification model is constructed using logistic regression. Each grid area is then classified according to a preset threshold to distinguish between risky grid areas and normal grid areas.

[0012] Furthermore, the preset feature bands include the thermal infrared band and the red light band, and the reflectance data are the average reflectance of the thermal infrared band and the average reflectance of the red light band. The steps for constructing the vegetation parameters of each pixel in the spectral sub-image are as follows:

[0013] The difference between the average reflectance of the thermal infrared band of a pixel and the average reflectance of the red band of the same pixel is defined as the first parameter.

[0014] The sum of the average reflectance of the thermal infrared band of the corresponding pixel and the average reflectance of the red band of the same pixel is used as the second parameter.

[0015] The ratio of the first parameter to the second parameter is defined as the vegetation index of that pixel.

[0016] Furthermore, the vegetation cover rate of the corresponding grid area is calculated using the double threshold averaging method. The specific steps are as follows:

[0017] Vegetation coverage is calculated using a double threshold flattening method based on vegetation indices:

[0018]

[0019] in, Indicates the first Vegetation coverage of each grid area; and The empirical thresholds for vegetation indices, representing bare soil areas and dense vegetation areas respectively, are determined by expert scoring. Indicates the first Vegetation index of each grid area;

[0020] The vegetation cover of each grid area will be used as a temperature correction factor:

[0021]

[0022] in,

[0023]

[0024] in, Indicates the surface emissivity; Indicates the corrected number Surface temperature of each grid area; Indicates the first Surface temperature of each grid area.

[0025] Furthermore, the fitting relationship between vegetation index and surface temperature was obtained, and the specific steps were as follows:

[0026] The vegetation index and corresponding land surface temperature data of healthy vegetation samples were obtained to form a set of sample data. The least squares method was used to perform linear regression analysis on the sample data to obtain the fitting relationship between the vegetation index and the land surface temperature.

[0027] Furthermore, the temperature difference anomaly index for each grid region is obtained, and the specific steps are as follows:

[0028] Substituting the vegetation index of each grid region into the fitting relationship, the predicted land surface temperature of each grid region is obtained, and the standard deviation of the predicted land surface temperature for all grid regions is calculated. The corrected highest and lowest land surface temperatures of each grid region are then standardized using the standard deviation to obtain the temperature anomaly index for each grid region.

[0029]

[0030] in, Indicates the first Temperature difference anomaly index for each grid region; Indicates the corrected number The highest surface temperature during the high-temperature period in each grid area; The revised version of the first The lowest surface temperature during the low-temperature period in each grid area; The standard deviation of predicted surface temperature for all grid regions.

[0031] Further, the transpiration efficiency index of each grid region is obtained, and the specific steps are as follows:

[0032] Based on the predicted surface temperature of each grid region, the transpiration efficiency index of each grid region is calculated using the vegetation temperature difference method:

[0033]

[0034] in, Indicates the first The evaporation efficiency index of each grid area.

[0035] Furthermore, the risk grid area and the normal grid area are distinguished, and the specific steps are as follows:

[0036] Based on the temperature difference anomaly index and transpiration efficiency index of each grid region, a vegetation status classification model is constructed using logistic regression:

[0037]

[0038] in, Indicates the first The probability of pest and disease occurrence in each grid area;

[0039] Each grid region is classified according to a preset threshold, distinguishing between risky and normal grid regions. Specifically: when A value greater than or equal to 0.7 indicates that the grid area belongs to a high-risk area. A value greater than or equal to 0.4 but less than 0.7 indicates that the grid area belongs to a potential risk area. A value less than 0.4 indicates that the area is in a healthy state.

[0040] The present invention also provides a remote sensing monitoring and assessment device for vegetation diseases and pests in the Loess Hilly Area, the assessment device being used to perform the above-described assessment method, including:

[0041] The data acquisition module is used to acquire multi-band spectral images of the vegetation area to be monitored, and to divide the area into grids to obtain multiple grid regions. The module extracts spectral sub-images corresponding to each grid region from the spectral images. Simultaneously, the module acquires the surface temperature of each grid region during high-temperature and low-temperature periods, which are used as the highest and lowest surface temperatures of the corresponding grid regions.

[0042] The surface temperature correction module is used to extract reflectance data of preset feature bands based on the spectral sub-images corresponding to each grid region, construct vegetation parameters of each pixel in the spectral sub-image using the normalized difference method, and perform regional averaging on the vegetation parameters of all pixels in the same grid region to obtain the vegetation index of the corresponding grid region. Based on the vegetation index, the vegetation coverage of the corresponding grid region is calculated using the double threshold averaging method, and the vegetation coverage is used as a temperature correction factor to correct the surface temperature of the corresponding grid region.

[0043] The parameter construction module is used to acquire vegetation index and corresponding land surface temperature data of healthy vegetation samples, and perform parameter fitting to obtain the fitting relationship between vegetation index and land surface temperature. The vegetation index of each grid area is substituted into the fitting relationship to obtain the predicted land surface temperature of each grid area, and the standard deviation of the predicted land surface temperature of all grid areas is calculated. The corrected highest and lowest land surface temperatures of each grid area are standardized using the standard deviation to obtain the temperature difference anomalous index of each grid area. Combined with the corresponding predicted land surface temperature, the transpiration efficiency index of each grid area is calculated using the vegetation temperature difference method.

[0044] The threshold comparison module is used to construct a vegetation status classification model based on the temperature difference anomalous index and transpiration efficiency index of each grid area using logistic regression, and classify each grid area according to a preset threshold to distinguish between risk grid areas and normal grid areas.

[0045] Compared with the prior art, the beneficial effects of the present invention are:

[0046] By simultaneously acquiring and correcting diurnal surface temperatures in vegetated areas, a temperature anomaly index was constructed. This index can sensitively detect vegetation transpiration dysfunction and energy balance anomalies caused by pests and diseases. Simultaneously, by establishing a baseline for healthy vegetation and calculating the transpiration efficiency index, temperature anomalies per unit of vegetation cover were quantified, directly reflecting the decline in vegetation physiological function. Finally, these two stress indicators, extracted from different physical mechanisms, were used to make comprehensive decisions through a logistic regression model, effectively distinguishing between complex physiological anomalies caused by pests and diseases and disturbances caused by single environmental factors such as drought and soil background. This not only significantly improves the accuracy and reliability of monitoring, achieving true early warning, but also outputs a pest and disease risk probability map through gridding. Attached Figure Description

[0047] Figure 1 This is a schematic diagram of the overall method flow of the present invention;

[0048] Figure 2 This is a graph showing the relationship between the temperature difference anomaly index and the corresponding probability of pest and disease occurrence.

[0049] Figure 3This is a schematic diagram of the overall device of the present invention. Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0051] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0052] Example:

[0053] Please see Figures 1-2 The present invention provides a technical solution:

[0054] Remote sensing monitoring and assessment methods for vegetation diseases and pests in the Loess Hilly Areas, including the following specific steps:

[0055] S1: Acquire multi-band spectral images of the vegetation area to be monitored, and divide it into grids to obtain multiple grid areas. Extract the spectral sub-images corresponding to each grid area from the spectral images. Simultaneously acquire the surface temperature of each grid area during high-temperature and low-temperature periods, which are used as the highest and lowest surface temperatures of the corresponding grid areas.

[0056] In the above process, a large-scale vegetation area was discretized into uniform analysis units through "grid partitioning," which standardized the monitoring units and facilitated independent calculations and feature extraction for each grid, ultimately achieving precise spatial positioning of pest and disease risks. Simultaneously, by acquiring the surface temperature of the same grid area during high and low temperature periods, core data was prepared for subsequent calculations of the temperature difference anomaly index, which directly reflects abnormal vegetation physiological states.

[0057] The fundamental purpose of obtaining surface temperature data during high and low temperature periods is to capture and quantify anomalies in vegetation's energy balance and transpiration cooling function during the diurnal cycle. This is the core physical mechanism for diagnosing biological stresses such as pests and diseases. Under normal water supply and healthy physiological functions, vegetation effectively cools itself through vigorous transpiration during the day, preventing excessive increases in daytime surface temperature. At night, healthy vegetation canopy structure and moisture status also possess a certain capacity for heat preservation or heat dissipation regulation. Therefore, healthy vegetation typically exhibits a relatively stable and unique "diurnal temperature variation rhythm." When vegetation is infected by pests and diseases, its stomatal conductance decreases, and its vascular system is damaged, leading to a significant reduction in transpiration. During high-temperature periods, the decreased transpiration cooling capacity prevents the vegetation canopy from effectively dissipating heat, resulting in a significantly higher temperature than healthy vegetation under the same environmental conditions. During low-temperature periods, damaged physiological structures and moisture status may also alter the thermal inertia and heat dissipation rate of the vegetation canopy, causing its temperature to differ from that of healthy vegetation. Therefore, pest and disease stress can alter the original diurnal temperature variation pattern. Analyzing daytime or nighttime temperatures in isolation can be strongly influenced by environmental factors such as weather and soil background. However, analyzing diurnal temperature range or temperature range patterns as a whole can reveal thermal anomalies caused by the disorder of the vegetation's own physiological functions more sensitively and specifically.

[0058] The high-temperature period typically refers to around noon local time, approximately from 10:00 AM to 2:00 PM. During this period, solar radiation is strongest, and the earth's surface and vegetation receive the most energy. Transpiration is also most active during this time, and the temperature is the result of the combined effects of vegetation transpiration cooling capacity and environmental heating. Therefore, the temperature at this time is most sensitive to vegetation water stress and physiological dysfunction. The low-temperature period typically refers to around dawn local time, approximately from 3:00 AM to 6:00 AM. During this period, solar radiation is zero or extremely weak, and the earth's surface and vegetation mainly dissipate heat to outer space through long-wave radiation, with the temperature dropping to its daily minimum. The temperature at this time reflects more the thermal inertia of the vegetation and the release of previously stored heat, and can indirectly reflect the structure of the vegetation canopy, water content, and the moisture status of the underlying soil.

[0059] S2: Based on the spectral sub-images corresponding to each grid region, the reflectance data of the preset feature bands are extracted. The vegetation parameters of each pixel in the spectral sub-image are constructed using the normalized difference method. The vegetation parameters of all pixels in the same grid region are averaged to obtain the vegetation index of the corresponding grid region. Based on the vegetation index, the vegetation coverage of the corresponding grid region is calculated using the double threshold averaging method. The vegetation coverage is used as a temperature correction factor to correct the surface temperature of the corresponding grid region.

[0060] The preset feature bands include the thermal infrared band and the red light band. The reflectance data are the average reflectance of the thermal infrared band and the average reflectance of the red light band. The steps to construct the vegetation parameters of each pixel in the spectral sub-image are as follows:

[0061] The difference between the average reflectance of the thermal infrared band of a pixel and the average reflectance of the red band of the same pixel is defined as the first parameter.

[0062] The sum of the average reflectance of the thermal infrared band of the corresponding pixel and the average reflectance of the red band of the same pixel is used as the second parameter.

[0063] The ratio of the first parameter to the second parameter is defined as the vegetation index of that pixel.

[0064] In the above process, the reflectance of the thermal infrared band was creatively used to replace the near-infrared band in traditional vegetation indices, constructing an effective vegetation status indicator parameter. This allows for the identification and quantification of vegetation cover using only the thermal infrared and red light bands. The fundamental reason for this calculation is based on the systematic differences in the reflectance characteristics of vegetation in the thermal infrared and red light bands. Healthy, dense vegetation canopies, due to their complex structure and high water content, typically exhibit higher reflectance in the thermal infrared band than bare soil, similar to the near-infrared band. In the red light band, vegetation has extremely low reflectance due to chlorophyll absorption. By calculating the difference between thermal infrared and red light reflectance and comparing it with their sum, the core idea of ​​the classic vegetation index is essentially simulated. This ratio effectively offsets environmental interferences such as light and topography, amplifies the spectral contrast between vegetation and the background, and thus generates a vegetation index that, while not standard, is stable and usable.

[0065] The vegetation cover rate of the corresponding grid area is calculated using the double threshold averaging method. The specific steps are as follows:

[0066] Vegetation coverage is calculated using a double threshold flattening method based on vegetation indices:

[0067]

[0068] in, Indicates the first Vegetation coverage of each grid area; and The empirical thresholds for vegetation indices, representing bare soil areas and dense vegetation areas respectively, are determined by expert scoring. Indicates the first Vegetation index of each grid area;

[0069] The vegetation cover of each grid area will be used as a temperature correction factor:

[0070]

[0071] in,

[0072]

[0073] in, Indicates the surface emissivity; Indicates the corrected number Surface temperature of each grid area; Indicates the first Surface temperature of each grid area.

[0074] In the above process, by introducing dual thresholds for bare soil and dense vegetation, the continuously changing vegetation index is nonlinearly converted into a coverage rate that better reflects the actual spatial distribution characteristics of vegetation. The surface radiance of each grid is dynamically calculated using the coverage rate, thereby performing radiance correction on the original surface temperature based on physical mechanisms. Finally, a corrected temperature that more realistically reflects the thermal state of the vegetation canopy and removes the influence of the background soil is obtained, laying a reliable data foundation for the subsequent accurate diagnosis of vegetation physiological abnormalities.

[0075] and A comprehensive approach combining statistical analysis of typical samples with expert verification was adopted. Specifically, typical samples from both bare soil areas and healthy, densely vegetated areas were selected from remote sensing images, and the mean vegetation index of these samples was calculated as initial reference values. Then, the initial reference values ​​were verified and fine-tuned by combining field survey data, historical data, and expert knowledge to ultimately determine the final vegetation index. That is, the typical value of bare soil, which is usually close to 0 or slightly negative, and That is, the typical value for dense vegetation, usually between 0.7 and 0.9, which are two empirical thresholds;

[0076] The core reason for designing the correction formula in this way is to establish a quantitative physical relationship between vegetation cover and surface radiance, thereby achieving physical correction of surface temperature based on vegetation cover. The coefficients 0.004 and 0.986 are empirical coefficients obtained by linear regression analysis of measured radiance data or high-precision remote sensing products of typical land features, i.e., from completely bare soil to dense vegetation. Among them, 0.986 represents the typical radiance value of pure bare soil, i.e., when the vegetation cover is 0%, and 0.004 represents the slope of the linear increase of surface radiance as the vegetation cover increases. This combination of coefficients ensures that the calculated radiance is within a physically reasonable range, usually between 0.96 and 0.99, thereby significantly improving the accuracy of temperature inversion.

[0077] Dependent variable Indicates the corrected number The technique of analyzing the surface temperature of each grid area eliminates the influence of variations in surface emissivity caused by differences in vegetation cover on the original retrieved temperature, thus obtaining more accurate temperature data that better reflects the true thermal state of vegetation. The higher the elevation, the greater the surface emissivity, and surface radiation... The Stefan-Boltzmann law directly affects thermal infrared radiation energy, and thus influences the surface temperature inversion value based on radiance. Therefore, it is necessary to go through right Corrections were made, and the corrected surface temperature was... and It is directly proportional to the surface emissivity. The square root of the radiance is inversely proportional to the radiance, which is consistent with the physical law that, for the same radiance, a surface with higher emissivity corresponds to a lower physical temperature.

[0078] S3: Obtain vegetation index and corresponding land surface temperature data for healthy vegetation samples, and perform parameter fitting to obtain the fitting relationship between vegetation index and land surface temperature. Substitute the vegetation index of each grid area into the fitting relationship to obtain the predicted land surface temperature of each grid area, and calculate the standard deviation of the predicted land surface temperature of all grid areas. Standardize the corrected highest and lowest land surface temperatures of each grid area using the standard deviation to obtain the temperature difference anomalous index of each grid area. Combined with the corresponding predicted land surface temperature, calculate the transpiration efficiency index of each grid area using the vegetation temperature difference method.

[0079] The specific steps to obtain the fitting relationship between vegetation index and land surface temperature are as follows:

[0080] The vegetation index and corresponding land surface temperature data of healthy vegetation samples were obtained to form a set of sample data. The least squares method was used to perform linear regression analysis on the sample data to obtain the fitting relationship between the vegetation index and the land surface temperature.

[0081] The formula used in the above process is:

[0082]

[0083] in, This represents the slope of the fitted curve; Represents the intercept of the fitted curve; Indicates the first Vegetation index of a healthy sample; The formula for fitting the curve represents the first... Predicted surface temperature of a healthy vegetation sample;

[0084] Solving for parameters using the least squares method and Its goal is to find the sum of squared residuals between the observed temperature and the model prediction for all healthy vegetation samples. Minimum, that is:

[0085]

[0086] in, Represents the sum of squared residuals; This represents the total number of healthy vegetation samples.

[0087] Through the Regarding the parameters respectively and By taking the partial derivatives and setting them equal to zero, we can obtain the normal equations and then solve for the optimal parameters. and ;

[0088]

[0089]

[0090] This yields a linear fitting equation that describes the average expected change in surface temperature for every unit change in vegetation index under healthy conditions, as well as the baseline temperature when the vegetation index is zero. This fitted curve will serve as the benchmark for subsequent calculations of temperature deviation, i.e., the transpiration efficiency index.

[0091] The fitted equation is a linear regression equation because, at the regional scale, especially in the Loess Hilly Vegetation Area where surface cover and climate conditions are relatively uniform, numerous remote sensing studies have shown that there is usually a significant linear negative correlation between the healthy vegetation index and surface temperature; that is, the more lush the vegetation, the lower the surface temperature. This is mainly driven by the dominant physical process of enhanced transpiration cooling effect due to increased vegetation cover. Linear models have fewer parameters and a clear physical meaning, namely the slope. The intercept reflects the temperature change caused by a change in the vegetation index. It reflects the background temperature of bare soil, and the least squares method can robustly and efficiently fit a baseline representing the overall trend from a limited sample. Although this linear approximation may have deviations under extreme cover conditions, it is sufficient to provide a stable and reliable health status reference for subsequent calculation of the evapotranspiration efficiency index.

[0092] The specific steps to obtain the temperature difference anomaly index for each grid region are as follows:

[0093] Substituting the vegetation index of each grid region into the fitting relationship, the predicted land surface temperature of each grid region is obtained, and the standard deviation of the predicted land surface temperature for all grid regions is calculated. The corrected highest and lowest land surface temperatures of each grid region are then standardized using the standard deviation to obtain the temperature anomaly index for each grid region.

[0094]

[0095] in, Indicates the first Temperature difference anomaly index for each grid region; Indicates the corrected number The highest surface temperature during the high-temperature period in each grid area; The revised version of the first The lowest surface temperature during the low-temperature period in each grid area; The standard deviation of predicted surface temperature for all grid regions.

[0096] In the above process, the diurnal surface temperature difference corrected for each grid is divided by the overall standard deviation of the predicted temperature of all grids, thereby eliminating the influence of regional background temperature level differences and dimensions. This allows the degree of temperature difference abnormality between different grids to be objectively compared and evaluated on the same standardized scale, thus accurately identifying those abnormal grids that significantly deviate from the normal regional pattern in diurnal temperature change rhythm, providing key thermal feature inputs for subsequent disease and pest risk diagnosis.

[0097] Dependent variable Indicates the first The temperature difference anomaly index for each grid region quantifies the degree to which the diurnal temperature range of that grid deviates from the normal fluctuation range of the region as a whole, thereby revealing anomalies in the vegetation's energy balance function in a standardized manner. The diurnal temperature range of the grid itself There is a positive correlation; the greater the temperature difference, the larger the index value, which is related to the standard deviation of regional temperature fluctuations. They exhibit an inverse correlation; the greater the overall fluctuation in the region, the smaller the relative index value.

[0098] The specific steps to obtain the transpiration efficiency index for each grid region are as follows:

[0099] Based on the predicted surface temperature of each grid region, the transpiration efficiency index of each grid region is calculated using the vegetation temperature difference method:

[0100]

[0101] in, Indicates the first The evaporation efficiency index of each grid area.

[0102] In the above process, the technical effect of this step is to construct a sensitive index that can directly quantify the degree of temperature anomaly under a unit of vegetation cover, thereby removing the influence of vegetation cover itself and specifically revealing the decline of vegetation transpiration cooling function. It standardizes the deviation between the actual surface temperature after correction for each grid and the healthy temperature predicted based on its vegetation index by dividing by the vegetation index of that grid, so that the temperature deviation can be fairly compared and evaluated regardless of whether the vegetation is dense or sparse, providing core evidence for diagnosing physiological function, i.e. transpiration disorder, caused by pests and diseases.

[0103] Dependent variable Indicates the first The transpiration efficiency index of each grid region is a technical effect that quantitatively characterizes the transpiration efficiency of a given vegetation cover level. The deviation of the actual vegetation temperature from the expected healthy temperature directly and sensitively indicates abnormalities in the efficiency of vegetation transpiration cooling. Representing the absolute deviation between the actual observed temperature and the expected healthy temperature, it reflects the degree of thermal anomaly in the vegetation. This index represents the vegetation cover level of the grid. By dividing the temperature deviation by the vegetation index, the absolute deviation is normalized to the basis of unit vegetation cover, thus eliminating the difference in the dimension of deviation caused by different vegetation cover levels. and There is a positive correlation; the higher the actual temperature exceeds the health expectation, the larger the index value, and the higher the vegetation index. They show an inverse correlation; under the same temperature deviation, the index value will be relatively diluted in areas with higher vegetation cover.

[0104] S4: Based on the temperature difference anomalous index and transpiration efficiency index of each grid area, a vegetation status classification model is constructed using logistic regression. Each grid area is then classified according to a preset threshold to distinguish between risky grid areas and normal grid areas.

[0105] The specific steps to distinguish between risky grid areas and normal grid areas are as follows:

[0106] Based on the temperature difference anomaly index and transpiration efficiency index of each grid region, a vegetation status classification model is constructed using logistic regression:

[0107]

[0108] in, Indicates the first The probability of pest and disease occurrence in each grid area;

[0109] In the above process, the technical effect of this step is to achieve quantitative, probabilistic, and automated diagnosis of pest and disease risks by constructing an integrated decision-making model based on multi-source stress indicators. It weights and fuses two independent indicators reflecting abnormal energy balance and physiological function decline—the temperature difference anomalous index and the transpiration efficiency index—and uses a logistic regression function to map them into an intuitive probability value of pest and disease occurrence. This method not only comprehensively considers the manifestation of pest and disease stress in different physical dimensions, enhancing the robustness and accuracy of diagnosis, but also, by setting clear probability thresholds, can automatically and objectively divide the monitoring area into high-risk, potential-risk, and healthy areas, ultimately generating a spatial risk classification map that can be directly used to guide prevention and control actions.

[0110] Dependent variable Indicates the first The probability of pest and disease occurrence in a grid area is achieved by integrating multiple remote sensing indicators reflecting vegetation thermal and physiological anomalies into a unified risk quantification value ranging from 0 to 1 through an interpretable mathematical model. This enables objective and automated risk classification of the monitored area. The independent variable... and The degree of vegetation stress was quantified from two independent but complementary dimensions: energy balance and physiological function. These dimensions were then combined linearly. A comprehensive stress score is generated. The higher the score, the more significant the diurnal temperature variation and the decrease in transpiration efficiency the vegetation exhibits, indicating stronger comprehensive evidence of pest and disease occurrence. The logistic function maps this score to probability. Compared with the overall stress score A positive correlation is observed, namely, the temperature difference anomalous index. and transpiration efficiency index The higher the value, the greater the probability of pests and diseases occurring. The higher the value, the better; parameters 0.8 and 1.2 are for the two input features. and The weights reflect their relative importance in comprehensively judging the risk of pests and diseases; the constant term -2.0 is the decision threshold bias of the model, used to adjust the overall baseline level of the probability output;

[0111] Each grid region is classified according to a preset threshold, distinguishing between risky and normal grid regions. Specifically: when A value greater than or equal to 0.7 indicates that the grid area belongs to a high-risk area. A value greater than or equal to 0.4 but less than 0.7 indicates that the grid area belongs to a potential risk area. A value less than 0.4 indicates that the area is in a healthy state.

[0112] In the above process, the preset thresholds of 0.4 and 0.7 are mainly based on the statistical analysis of historical sample data and the trade-off between actual prevention and control needs. Typically, by analyzing the model's performance on the validation set, receiver operating characteristic curves or precision-recall curves are plotted, and combined with the results of field validation, a critical point that can balance monitoring sensitivity (i.e., no missed reports) and specificity (i.e., no false alarms) is selected. 0.4 serves as the initial boundary between health and risk status, aiming to capture minor abnormalities at an early stage; 0.7 serves as a high-risk threshold, corresponding to a high-confidence judgment of severe stress by the model, ensuring the reliability of high-risk early warning, thereby providing a clear basis for hierarchical management.

[0113] The core purpose of this classification is to achieve tiered and refined management of pest and disease risks. Continuous probability outputs are divided into three levels: healthy, potential risk, and high risk. This not only aligns with the gradual process of vegetation stress from occurrence and development to outbreak, but also provides differentiated action guidelines for management departments: healthy areas can be routinely inspected; potential risk areas require enhanced monitoring and investigation of causes; and high-risk areas necessitate emergency investigations and precise interventions. This classification method transforms complex continuous probability values ​​into intuitive and actionable decision categories, significantly improving the practicality of monitoring results and the efficiency of control actions.

[0114] In the above embodiments, 20 sets of data on temperature difference anomaly index and corresponding pest and disease occurrence probabilities are given to reflect the change in pest and disease occurrence probabilities with temperature difference anomaly index, as shown in Table 1:

[0115] Table 1: Relationship between Temperature Difference Anomaly Index and Corresponding Probability of Pest and Disease Occurrence

[0116]

[0117] In Table 1 above, given In the case of change The value shows the temperature difference index. The larger the size, the higher the probability of pests and diseases. The higher.

[0118] Please see Figure 3 The present invention also provides a remote sensing monitoring and assessment device for vegetation diseases and pests in the Loess Hilly Area, the assessment device being used to perform the above-mentioned assessment method, including:

[0119] The data acquisition module is used to acquire multi-band spectral images of the vegetation area to be monitored, and to divide the area into grids to obtain multiple grid regions. The module extracts spectral sub-images corresponding to each grid region from the spectral images. Simultaneously, the module acquires the surface temperature of each grid region during high-temperature and low-temperature periods, which are used as the highest and lowest surface temperatures of the corresponding grid regions.

[0120] The surface temperature correction module is used to extract reflectance data of preset feature bands based on the spectral sub-images corresponding to each grid region, construct vegetation parameters of each pixel in the spectral sub-image using the normalized difference method, and perform regional averaging on the vegetation parameters of all pixels in the same grid region to obtain the vegetation index of the corresponding grid region. Based on the vegetation index, the vegetation coverage of the corresponding grid region is calculated using the double threshold averaging method, and the vegetation coverage is used as a temperature correction factor to correct the surface temperature of the corresponding grid region.

[0121] The parameter construction module is used to acquire vegetation index and corresponding land surface temperature data of healthy vegetation samples, and perform parameter fitting to obtain the fitting relationship between vegetation index and land surface temperature. The vegetation index of each grid area is substituted into the fitting relationship to obtain the predicted land surface temperature of each grid area, and the standard deviation of the predicted land surface temperature of all grid areas is calculated. The corrected highest and lowest land surface temperatures of each grid area are standardized using the standard deviation to obtain the temperature difference anomalous index of each grid area. Combined with the corresponding predicted land surface temperature, the transpiration efficiency index of each grid area is calculated using the vegetation temperature difference method.

[0122] The threshold comparison module is used to construct a vegetation status classification model based on the temperature difference anomalous index and transpiration efficiency index of each grid area using logistic regression, and classify each grid area according to a preset threshold to distinguish between risk grid areas and normal grid areas.

[0123] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0124] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0125] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0126] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A remote sensing monitoring and assessment method for vegetation diseases and pests in the Loess Hilly Area, characterized in that, include: S1: Acquire multi-band spectral images of the vegetation area to be monitored, and divide the area into grids to obtain multiple grid regions. Extract the spectral sub-images corresponding to each grid region from the spectral images. Simultaneously acquire the surface temperature of each grid area during high-temperature and low-temperature periods, and use it as the highest and lowest surface temperature of the corresponding grid area; S2: Based on the spectral sub-images corresponding to each grid region, extract the reflectance data of the preset feature bands, use the normalized difference method to construct the vegetation parameters of each pixel in the spectral sub-image, and perform regional averaging on the vegetation parameters of all pixels in the same grid region to obtain the vegetation index of the corresponding grid region. Based on the vegetation index, the vegetation coverage rate of the corresponding grid area is calculated using the double threshold flat method, and the vegetation coverage rate is used as a temperature correction factor to correct the surface temperature of the corresponding grid area. S3: Obtain vegetation index and corresponding land surface temperature data for healthy vegetation samples, and perform parameter fitting to obtain the fitting relationship between vegetation index and land surface temperature. Substitute the vegetation index of each grid area into the fitting relationship to obtain the predicted land surface temperature of each grid area, and calculate the standard deviation of the predicted land surface temperature of all grid areas. Standardize the corrected highest and lowest land surface temperatures of each grid area using the standard deviation to obtain the temperature difference anomalous index of each grid area. Combined with the corresponding predicted land surface temperature, calculate the transpiration efficiency index of each grid area using the vegetation temperature difference method. S4: Based on the temperature difference anomalous index and transpiration efficiency index of each grid area, a vegetation status classification model is constructed using logistic regression. Each grid area is then classified according to a preset threshold to distinguish between risk grid areas and normal grid areas. The vegetation cover rate of the corresponding grid area is calculated using the double threshold averaging method. The specific steps are as follows: Vegetation coverage is calculated using a double threshold flattening method based on vegetation indices: in, Indicates the first Vegetation coverage of each grid area; and The empirical thresholds for vegetation indices, representing bare soil areas and dense vegetation areas respectively, are determined by expert scoring. Indicates the first Vegetation index of each grid area; The vegetation cover of each grid region is used as a temperature correction factor: in, in, Indicates the surface emissivity; Indicates the corrected number Surface temperature of each grid area; Indicates the first Surface temperature of each grid area; The specific steps to obtain the temperature difference anomaly index for each grid region are as follows: Substituting the vegetation index of each grid region into the fitting relationship, the predicted land surface temperature of each grid region is obtained, and the standard deviation of the predicted land surface temperature for all grid regions is calculated. The corrected highest and lowest land surface temperatures of each grid region are then standardized using the standard deviation to obtain the temperature anomaly index for each grid region. in, Indicates the first Temperature difference anomaly index for each grid region; Indicates the corrected number The highest surface temperature during the high-temperature period in each grid area; Indicates the corrected number The lowest surface temperature during the low-temperature period in each grid area; Standard deviation of predicted surface temperature for all grid regions; The specific steps to obtain the transpiration efficiency index for each grid region are as follows: Based on the predicted surface temperature of each grid region, the transpiration efficiency index of each grid region is calculated using the vegetation temperature difference method: in, Indicates the first The transpiration efficiency index of each grid region; Indicates the expected temperature for health; The specific steps to distinguish between risky grid areas and normal grid areas are as follows: Based on the temperature difference anomaly index and transpiration efficiency index of each grid region, a vegetation status classification model is constructed using logistic regression: in, Indicates the first The probability of pest and disease occurrence in each grid area; Indicates the first Temperature difference anomaly index for each grid region; Indicates the first The transpiration efficiency index of each grid region; Each grid region is classified according to a preset threshold, distinguishing between risky and normal grid regions. Specifically: when A value greater than or equal to 0.7 indicates that the grid area belongs to a high-risk area. A value greater than or equal to 0.4 but less than 0.7 indicates that the grid area belongs to a potential risk area. A value less than 0.4 indicates that the area is in a healthy state.

2. The remote sensing monitoring and assessment method for vegetation diseases and pests in the Loess Hilly Areas according to claim 1, characterized in that, The preset feature bands include the thermal infrared band and the red light band. The reflectance data are the average reflectance of the thermal infrared band and the average reflectance of the red light band. The steps to construct the vegetation parameters of each pixel in the spectral sub-image are as follows: The difference between the average reflectance of the thermal infrared band of a pixel and the average reflectance of the red band of the same pixel is defined as the first parameter. The sum of the average reflectance of the thermal infrared band of the corresponding pixel and the average reflectance of the red band of the same pixel is used as the second parameter. The ratio of the first parameter to the second parameter is defined as the vegetation index of that pixel.

3. The remote sensing monitoring and assessment method for vegetation diseases and pests in the Loess Hilly Areas according to claim 1, characterized in that, The specific steps to obtain the fitting relationship between vegetation index and land surface temperature are as follows: The vegetation index and corresponding land surface temperature data of healthy vegetation samples were obtained to form a set of sample data. The least squares method was used to perform linear regression analysis on the sample data to obtain the fitting relationship between the vegetation index and the land surface temperature.

4. A remote sensing monitoring and assessment device for vegetation diseases and pests in loess hilly areas, characterized in that: The evaluation apparatus is used to perform the evaluation method according to any one of claims 1-3, including: The data acquisition module is used to acquire multi-band spectral images of the vegetation area to be monitored, and to divide the area into grids to obtain multiple grid regions. The module extracts spectral sub-images corresponding to each grid region from the spectral images. Simultaneously, the module acquires the surface temperature of each grid region during high-temperature and low-temperature periods, which are used as the highest and lowest surface temperatures of the corresponding grid regions. The surface temperature correction module is used to extract reflectance data of preset feature bands based on the spectral sub-images corresponding to each grid region, construct vegetation parameters of each pixel in the spectral sub-image using the normalized difference method, and perform regional averaging on the vegetation parameters of all pixels in the same grid region to obtain the vegetation index of the corresponding grid region. Based on the vegetation index, the vegetation coverage of the corresponding grid region is calculated using the double threshold averaging method, and the vegetation coverage is used as a temperature correction factor to correct the surface temperature of the corresponding grid region. The parameter construction module is used to acquire vegetation index and corresponding land surface temperature data of healthy vegetation samples, and perform parameter fitting to obtain the fitting relationship between vegetation index and land surface temperature. The vegetation index of each grid area is substituted into the fitting relationship to obtain the predicted land surface temperature of each grid area, and the standard deviation of the predicted land surface temperature of all grid areas is calculated. The corrected highest and lowest land surface temperatures of each grid area are standardized using the standard deviation to obtain the temperature difference anomalous index of each grid area. Combined with the corresponding predicted land surface temperature, the transpiration efficiency index of each grid area is calculated using the vegetation temperature difference method. The threshold comparison module is used to construct a vegetation status classification model based on the temperature difference anomalous index and transpiration efficiency index of each grid area using logistic regression, and classify each grid area according to a preset threshold to distinguish between risk grid areas and normal grid areas.

Images