Ecological river channel revetment design method and system
By dividing the monitoring area into sub-areas and combining feature engineering and support vector machine models, the electromagnetic wave emission density is dynamically adjusted, which solves the overexposure problem of high-reflection areas in remote sensing monitoring, achieves high-precision and reliable environmental monitoring, and improves the feasibility and accuracy of ecological river bank protection design.
Patent Information
- Application Number
- CN202510749614.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-06
- Publication Date
- 2025-09-16
AI Technical Summary
Existing remote sensing monitoring technology is prone to overexposure in highly reflective areas, resulting in the inability to identify key surface changes and disaster risk signals, thereby reducing disaster warning capabilities.
The monitoring area is divided into multiple sub-areas, and reflectivity data is obtained in real time. Combined with feature engineering and support vector machine models, the electromagnetic wave emission density is dynamically adjusted, and a lower emission density is used for monitoring in highly reflective areas.
It effectively avoids overexposure, ensures the complete capture of ground details, improves the accuracy and reliability of monitoring data, enhances the adaptability and flexibility of the system, and is able to identify environmental changes and warn of potential disaster risks.
Smart Images

Figure CN120654300A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of ecological river bank protection design, and in particular to an ecological river bank protection design method and system. Background Art
[0002] The Ecological Riverbank Design System is a comprehensive design platform that integrates ecology, engineering, and information technology. It aims to achieve the organic integration of bank protection engineering and the ecological environment through intelligent analysis and multi-objective optimization methods. Based on river hydrological data, geological conditions, and ecological characteristics, the system utilizes remote sensing monitoring, 3D modeling, numerical simulation, and other technologies. Combined with ecological engineering measures such as plant root enhancement, ecological grids, and permeable materials, it designs bank protection solutions that prevent riverbank erosion and enhance river channel stability, while also protecting aquatic habitats and restoring the natural landscape. The system also features dynamic assessment and optimization capabilities, allowing it to adjust design solutions based on changes in the river environment to ensure long-term ecological benefits and project stability.
[0003] In the design of ecological river bank protection, the role of remote sensing monitoring is mainly reflected in the real-time dynamic monitoring of the river environment and the precise data collection, providing a reliable basis for the scientific design of bank protection. Through remote sensing technology, key information such as the hydrological characteristics of the river, shoreline changes, vegetation coverage, soil erosion level, etc. can be comprehensively obtained from a macro perspective, and high-risk areas (such as collapse, scour, leakage, etc.) can be effectively identified. At the same time, remote sensing monitoring can capture seasonal changes and long-term trends, such as water level fluctuations, flow rate changes, vegetation growth, etc., to provide dynamic reference data for bank protection design. In addition, during the implementation and later maintenance stages of the project, remote sensing monitoring can also be used to evaluate the effectiveness of bank protection and warn of potential risks, so as to achieve continuous optimization of the design and long-term protection of the ecological environment.
[0004] The existing technology has the following deficiencies:
[0005] When using remote sensing for monitoring, existing technologies typically use a constant electromagnetic wave emission density for surface monitoring. This aims to ensure that the acquired surface reflectance data has stable spectral feature contrast and comparability, thereby achieving consistent surface information monitoring across time, space, and different environmental conditions. However, when the surface reflectivity is high, continuing to use this monitoring method can lead to serious information loss and monitoring errors. In highly reflective areas (such as snow-covered areas, deserts, bare rock, and building roofs), the reflectivity of the surface objects is typically as high as 70% to 90%. The constant electromagnetic wave emission density causes the reflected signal received by the sensor to exceed the dynamic range, resulting in overexposure and completely losing the detailed features of the surface features. This situation can prevent the monitoring system from identifying key surface changes, such as changes in snow thickness, the spread of desertification, rock weathering, and other environmental processes. It can even make it difficult to capture disaster risk signals such as surface cracks and landslide signs, significantly reducing disaster early warning capabilities.
[0006] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not form the prior art that is already known to a person of ordinary skill in the art. Summary of the Invention
[0007] The purpose of the present invention is to provide an ecological river bank protection design method and system. By dividing the monitoring area into multiple sub-areas and obtaining reflectivity data in real time, combined with the prediction results of feature engineering and support vector machine models, accurate prediction of the reflectivity of each sub-area and dynamic adjustment of the electromagnetic wave emission density can be achieved. For surface areas with high reflectivity, a lower emission density is used for monitoring to avoid overexposure, ensure the complete capture of ground details, improve the accuracy and reliability of monitoring data, and solve the problems in the above-mentioned background technology.
[0008] In order to achieve the above-mentioned object, the present invention provides the following technical solution: a method for designing an ecological river bank protection, specifically comprising the following steps:
[0009] Divide the surface area to be monitored into multiple sub-areas. By dividing the sub-areas, the complexity of the entire monitoring area is reduced, making it easier to set up independent monitoring and analysis strategies for each sub-area.
[0010] In each sub-area, the surface reflectivity information is obtained in real time through remote sensing equipment;
[0011] Extract key features reflecting high surface reflectivity from real-time reflectivity data. Conduct in-depth analysis of the extracted key features through feature engineering to preliminarily identify high reflectivity patterns in each sub-region within a complex surface environment.
[0012] The analyzed key features are input into the pre-trained support vector machine model to predict the surface reflectivity of each sub-region;
[0013] Based on the prediction results of the support vector machine model, the actual electromagnetic wave emission density of each sub-area is dynamically adjusted. For surfaces with high reflectivity, remote sensing monitoring is carried out with a lower electromagnetic wave emission density to avoid overexposure.
[0014] Preferably, the specific steps of dividing the surface area to be monitored into multiple sub-areas are as follows:
[0015] First, the spatial scope of the entire monitoring area should be clarified and the geographical boundaries should be determined;
[0016] According to the heterogeneity of surface characteristics, appropriate classification criteria are selected in combination with monitoring objectives and required analysis accuracy;
[0017] Within the identified monitoring area, the area is divided into multiple sub-areas based on the selected criteria;
[0018] Each sub-area after division is numbered or marked to facilitate subsequent monitoring, data processing and result analysis, ensuring that each sub-area has an independent identification mark to facilitate dynamic monitoring and data tracking.
[0019] Preferably, key features reflecting high surface reflectivity are extracted from the reflectivity data acquired in real time, wherein the extracted features include the reflectivity ratios of different bands and the slope changes of the spectral reflectivity of the ground objects within the specified band range. The extracted reflectivity ratios of different bands and the slope changes of the spectral reflectivity of the ground objects within the specified band range are deeply analyzed through feature engineering, and the reflectivity ratio quantification index and the surface spectral slope quantification index are generated respectively. The reflectivity ratio quantification index and the surface spectral slope quantification index are used to preliminarily identify the high reflectivity pattern of each sub-area in a complex surface environment.
[0020] Preferably, the reflectivity ratio quantification index and the surface spectral slope quantification index generated after analyzing the reflectivity ratio of different bands and the slope change of the spectral reflectance of the ground object within the specified band range are input into a pre-trained support vector machine model, and the reflection coefficient is generated by the support vector machine model, and the surface reflectivity of each sub-area is predicted by the reflection coefficient.
[0021] Preferably, the specific steps of generating a reflectivity ratio quantitative index by performing an in-depth analysis of the extracted reflectivity ratios of different bands through feature engineering are as follows:
[0022] The near-infrared band and the visible light band are selected to calculate the reflectance ratio between different bands. The calculation expression is:
[0023]
[0024] Where: R ratio Represents the reflectivity ratio, R NIR Represents the reflectivity in the near-infrared band, R VIS Indicates the reflectivity of visible light band;
[0025] The reflectivity ratio quantification index is generated by performing a nonlinear transformation on the reflectivity ratio to reflect the significance of high reflectivity in the sub-region. The reflectivity ratio quantification index is calculated by combining the exponential function and the band weighting coefficient. The calculation expression is:
[0026]
[0027] , where: R ratio_indexis the reflectivity ratio quantization index, α is the exponential coefficient of the reflectivity ratio, which is used to adjust the influence of the reflectivity ratio in the calculation and control the sensitivity of the ratio. β is a nonlinear weighting coefficient that controls the influence of the reflectivity ratio deviation, making the relationship between the reflectivity ratio and the high reflectivity characteristics of the surface more accurate. exp(·) is an exponential function used to further enhance the contrast between high-reflectivity and low-reflectivity areas.
[0028] Preferably, the specific steps of generating a surface spectral slope quantitative index by performing an in-depth analysis of the slope change of the ground feature spectral reflectance within the extracted specified band range through feature engineering are as follows:
[0029] The spectral slope is calculated based on the reflectance data of the ground objects within the specified band range extracted from the remote sensing data. Assume that the selected band range is λ1 to λ n , the spectrum slope is calculated by the following formula, and the calculation expression is:
[0030]
[0031] Where: S λ (i) is the slope of point i, R λ (i) is the reflectivity of the ground object measured at wavelength λ(i), R λ (i+1) is the reflectance of the ground object measured at wavelength λ(i+1), λ(i) is the wavelength value, and i is the discrete sampling point in the band;
[0032] The calculated slope data are further analyzed comprehensively. By considering the slope differences between bands and the overall spectral changes in the band range, a quantitative index of the surface spectral slope is constructed. The construction expression of the quantitative index of the surface spectral slope is:
[0033]
[0034] Among them: max(S λ (1), S λ (2),…,S λ (n-1)) is the maximum value of the slope within the band, min(S λ (1), S λ (2),…,S λ (n-1)) is the minimum slope value within the band, n is the number of selected bands, S SSI It is a quantitative index of the surface spectral slope.
[0035] Preferably, the actual electromagnetic wave emission density of each sub-region is dynamically adjusted according to the prediction results of the support vector machine model. The specific steps are as follows:
[0036] According to the prediction results of the support vector machine model, the reflection coefficient R of each sub-region is generated pred (u), where u represents the index of the subregion, based on the predicted reflection coefficient R pred (u), and further calculate the adjustment factor of the electromagnetic wave emission density. The adjustment factor determines the electromagnetic wave emission density required for each sub-area. The calculation expression is:
[0037]
[0038] Where: α(u) is the adjustment factor of the u-th sub-region, which represents the required electromagnetic wave emission density ratio; β' is the parameter that controls the adjustment sensitivity; θ is the reflection coefficient threshold, which is used to distinguish high-reflection and low-reflection areas. The sub-region with a reflection coefficient greater than the reflection coefficient threshold is a high-reflection area;
[0039] The actual electromagnetic wave emission density of each sub-area is dynamically adjusted using an adjustment factor. For highly reflective areas, the electromagnetic wave emission density is reduced to prevent signal overexposure; for low-reflective areas, the emission density is increased to enhance signal strength. The dynamic adjustment formula for the electromagnetic wave emission density is as follows:
[0040] P emission (u)=P base (1-α(u))
[0041] , where: P emission (u) is the electromagnetic wave emission density of the u-th sub-region, P base It is the benchmark electromagnetic wave emission density, that is, the default standard emission density.
[0042] An ecological river bank protection design system includes a region division module, a remote sensing data acquisition module, a feature extraction and analysis module, a prediction model module, and an emission density adjustment module;
[0043] The regional division module divides the surface area to be monitored into multiple sub-areas. By dividing the sub-areas, the complexity of the entire monitoring area is reduced, making it easier to set independent monitoring and analysis strategies for each sub-area.
[0044] Remote sensing data acquisition module, in each sub-area, obtains surface reflectivity information in real time through remote sensing equipment;
[0045] The feature extraction and analysis module extracts key features reflecting high surface reflectivity from real-time reflectivity data. Through feature engineering, the extracted key features are deeply analyzed to preliminarily identify high reflectivity patterns in each sub-region in a complex surface environment.
[0046] The prediction model module inputs the analyzed key features into the pre-trained support vector machine model to predict the surface reflectivity of each sub-area;
[0047] The emission density adjustment module dynamically adjusts the actual electromagnetic wave emission density of each sub-area based on the prediction results of the support vector machine model. For surfaces with high reflectivity, remote sensing monitoring is performed with a lower electromagnetic wave emission density to avoid overexposure.
[0048] In the above technical solution, the technical effects and advantages provided by the present invention are:
[0049] This method divides the monitoring area into multiple sub-areas and acquires reflectivity data in real time. Combining the prediction results of feature engineering and support vector machine models, this method accurately predicts the reflectivity of each sub-area and dynamically adjusts the electromagnetic wave emission density. For surface areas with high reflectivity, a lower emission density is used for monitoring to avoid overexposure, ensure complete capture of ground features, and improve the accuracy and reliability of monitoring data. Furthermore, dynamically adjusting the electromagnetic wave emission density not only optimizes the remote sensing monitoring process but also enhances the system's adaptability and flexibility, enabling effective identification and early warning of potential environmental changes or disaster risks in complex environments. Ultimately, this improves the feasibility and accuracy of ecological riverbank protection design. BRIEF DESCRIPTION OF THE DRAWINGS
[0050] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction to the drawings required for use in the embodiments will be given below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.
[0051] Figure 1 This is a flow chart of a method for designing an ecological river bank protection according to the present invention.
[0052] Figure 2 This is a module schematic diagram of an ecological river bank protection design system of the present invention. DETAILED DESCRIPTION
[0053] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of this disclosure will be thorough and complete and will fully convey the concepts of the example embodiments to those skilled in the art.
[0054] The present invention provides Figure 1 The ecological river bank protection design method shown in the figure specifically includes the following steps:
[0055] Divide the surface area to be monitored into multiple sub-areas. This is the first step in the monitoring process and a key step in ensuring monitoring accuracy and targeting. By dividing the sub-areas, the complexity of the entire monitoring area is reduced, making it easier to set up independent monitoring and analysis strategies for each sub-area.
[0056] The complexity and diversity of surface areas require customized monitoring tailored to each surface feature. Subdividing a large area into smaller, manageable subregions allows for personalized analysis of each subregion, minimizing the impact of local variations on global results. By breaking down a large area into smaller, manageable subregions, the monitoring system can more accurately capture surface variations in each region, particularly the differences between high- and low-reflectivity areas.
[0057] The specific steps for dividing the surface area to be monitored into multiple sub-areas are as follows:
[0058] Determine the boundaries of the monitoring area: First, clarify the spatial scope of the entire monitoring area. This can be done by determining the geographic boundaries through maps, remote sensing images or GPS data.
[0059] Selecting a classification criterion: Based on the heterogeneity of surface features, appropriate classification criteria should be chosen. Common criteria include topographic features (e.g., mountains, rivers, plains), ecological zoning (e.g., forests, grasslands, water bodies), or reflectivity distribution (e.g., the difference between high- and low-reflectivity areas). The choice of criterion should be based on the monitoring objective and the required analytical accuracy.
[0060] Perform spatial partitioning: Within the defined monitoring area, divide the area into multiple sub-areas based on selected criteria. This can be done using a uniform grid or adaptive partitioning based on features, terrain variations, and other factors. Each sub-area should be of a reasonable size to ensure sufficient detail and operability for subsequent analysis.
[0061] Mark and number sub-areas: Number or label each sub-area to facilitate subsequent monitoring, data processing, and analysis. Each sub-area should have an independent identification mark to facilitate dynamic monitoring and data tracking.
[0062] In each sub-area, the surface reflectivity information is obtained in real time through remote sensing equipment;
[0063] Within each sub-region, surface reflectance information can be acquired through a variety of remote sensing equipment, including satellite remote sensors, sensors carried by unmanned aerial vehicles (UAVs), aerial sensors, and ground-based remote sensing equipment. Satellite remote sensors (such as Landsat and Sentinel-2) provide wide-area coverage and are suitable for macro-scale monitoring; sensors carried by UAVs (such as hyperspectral imagers and infrared sensors) can provide higher spatial resolution within smaller areas and are suitable for more detailed and precise monitoring; aerial sensors, carried by aircraft, can provide high-resolution data over larger areas and are often used for large-scale monitoring missions; and ground-based remote sensing equipment (such as laser scanners and portable spectrometers) is used for high-precision, small-scale surface reflectance measurements. These devices enable real-time acquisition of surface reflectance information and precise analysis of surface characteristics in different regions.
[0064] Reflectivity, the ratio of electromagnetic waves reflected by the Earth's surface, is typically considered a core characteristic of land features in remote sensing monitoring. Obtaining real-time reflectivity information is a core data source for monitoring and forms the basis for subsequent analysis and prediction. The reflectivity of land features is affected by a variety of factors, such as surface material, humidity, and lighting conditions. By monitoring reflectivity in real time, the system can dynamically reflect changes in the surface. For example, reflectivity is higher in snow-covered areas and lower in bodies of water or wetlands. Real-time acquisition of this data enables monitoring systems to respond promptly to surface changes and provides a high-quality data source for subsequent feature extraction and input into machine learning models.
[0065] Extract key features reflecting high surface reflectivity from real-time reflectivity data. Conduct in-depth analysis of the extracted key features through feature engineering to preliminarily identify high reflectivity patterns in each sub-region within a complex surface environment.
[0066] Key features reflecting high surface reflectivity are extracted from the reflectivity data acquired in real time. The extracted features include the reflectivity ratios of different bands (such as the near-infrared and visible light bands) and the slope changes of the spectral reflectivity of the objects within the specified band range. Through feature engineering, the extracted reflectivity ratios of different bands and the slope changes of the spectral reflectivity of the objects within the specified band range are deeply analyzed to generate the reflectivity ratio quantification index and the surface spectral slope quantification index respectively. The reflectivity ratio quantification index and the surface spectral slope quantification index are used to preliminarily identify the high reflectivity pattern of each sub-region in a complex surface environment.
[0067] When the reflectivity ratio of different bands (such as the near-infrared and visible light bands) in a sub-region is high, it usually indicates that the surface reflectivity in the sub-region is high, especially in highly reflective areas such as bare rocks, snow, and deserts. In remote sensing monitoring, the reflectivity ratio reflects the spectral characteristics of the ground object. The reflectivity in the near-infrared band is usually high, especially for plants and some highly reflective materials, which can effectively reflect light in the near-infrared band, while the reflectivity in the visible light band is lower. When the reflectivity ratio of the near-infrared and visible light bands in a sub-region is high, it usually means that the region has strong reflectivity. For example, the reflectivity of snow is usually very high, and it has obvious high reflectivity characteristics in both the visible and near-infrared bands. This increase in the reflectivity ratio can help identify highly reflective areas and ensure that the remote sensing system can adjust its monitoring strategy in a timely manner to avoid the loss of key details due to overexposure.
[0068] The specific steps for generating a quantitative reflectance ratio index by performing an in-depth analysis of the extracted reflectance ratios of different bands through feature engineering are as follows:
[0069] The near-infrared band and the visible light band are selected to calculate the reflectance ratio between different bands. The calculation expression is:
[0070]
[0071] Where: R ratio Represents the reflectivity ratio, R NIR Indicates the reflectivity of the near-infrared band, which is usually used to reflect the reflective characteristics of plants, wetlands, etc. VIS It indicates the reflectivity of the visible light band, which mainly reflects the visible light reflection characteristics of the surface, such as soil, buildings, deserts, etc.
[0072] By calculating the reflectance ratio between the near-infrared and visible bands, we can distinguish between high- and low-reflection areas. In high-reflection areas, the ratio is typically larger, while in low-reflection areas, particularly those with dense water and vegetation, it is typically lower. This ratio provides the basis for the subsequent generation of a quantitative reflectance ratio index.
[0073] The reflectivity ratio quantification index is generated by performing a nonlinear transformation on the reflectivity ratio to reflect the significance of high reflectivity in the sub-region. The reflectivity ratio quantification index is calculated by combining the exponential function and the band weighting coefficient. The calculation expression is:
[0074]
[0075] Where: R ratio_indexis the reflectivity ratio quantification index, α is the exponential coefficient of the reflectivity ratio, which is used to adjust the influence of the reflectivity ratio in the calculation. It is usually set according to the actual surface type and monitoring target to control the sensitivity of the ratio. β is a nonlinear weighting coefficient that controls the influence of the reflectivity ratio deviation, making the relationship between the reflectivity ratio and the high reflectivity characteristics of the surface more accurate. exp(·) is an exponential function used to further enhance the contrast between high-reflectivity and low-reflectivity areas, especially when the deviation of the reflectivity ratio in the high-reflectivity area is more obvious.
[0076] By calculating the reflectivity ratio quantification index, surface reflectivity characteristics can be more precisely quantified, enhancing the ability to identify highly reflective areas. This reflectivity ratio quantification index effectively amplifies the impact of areas with large reflectivity ratios, thereby highlighting highly reflective areas such as snow and desert surfaces in remote sensing monitoring. In environments with large variations in surface reflectivity, the introduction of an exponential function helps overcome the limitation of traditional linear reflectivity ratio calculations, which fail to fully reflect surface features.
[0077] The reflectance ratio quantification index, generated through feature engineering and in-depth analysis of reflectance ratios extracted from different bands, shows that larger values generally indicate a higher surface reflectivity within the subregion. Because high-reflectivity areas, particularly those with snow, desert, and bare rock, have a large reflectance ratio between the near-infrared and visible bands, the weighted transformation of the exponential function significantly increases the reflectance ratio index in these areas. Conversely, lower values indicate lower surface reflectivity within the subregion, typically found in low-reflectivity areas such as water bodies, dense vegetation, or wetlands. By nonlinearly enhancing the reflectance ratio, the reflectance ratio quantification index enhances the contrast between high-reflectivity and low-reflectivity areas, effectively reflecting the reflectance characteristics of different surface types.
[0078] When the slope of the spectral reflectance of an object within a specified band in a sub-region varies significantly, it generally indicates that the sub-region has a high high reflectivity. This is because the reflectance slope of an object is closely related to its surface characteristics and material. In certain bands, especially the red, near-infrared, and short-wave infrared bands, changes in the reflectance of an object generally show different trends as the surface characteristics change. For highly reflective areas (such as snow, deserts, bare rock, etc.), the slope changes in the spectral reflectance are more significant due to the greater surface reflection intensity. Larger changes in the slope indicate that the reflectance of the surface varies strongly within this band, usually because materials with higher reflectivity (such as white snow or light-colored rock) have higher reflectivity in this band. Therefore, larger changes in the spectral slope can effectively indicate the high reflectivity characteristics of the sub-region, helping to identify high-reflectivity areas and conduct further monitoring and analysis.
[0079] The specific steps for generating a quantitative index of the surface spectral slope by performing an in-depth analysis of the slope changes of the ground feature spectral reflectance within the specified band range through feature engineering are as follows:
[0080] The spectral slope is calculated based on the reflectance data of the ground objects within the specified band range extracted from the remote sensing data. Assume that the selected band range is λ1 to λ n , the spectrum slope is calculated by the following formula, and the calculation expression is:
[0081]
[0082] Where: S λ (i) is the slope of point i, R λ (i) is the reflectivity of the ground object measured at wavelength λ(i), R λ (i+1) is the reflectance of the ground object measured at wavelength λ(i+1), λ(i) is the wavelength value, and i is the discrete sampling point in the band;
[0083] This step calculates the reflectivity variation between adjacent bands and obtains the slope corresponding to each band, thereby capturing the rate at which reflectivity changes with wavelength. In areas of high surface reflectivity, the spectral slope is significantly larger. Therefore, this slope variation serves as the basis for subsequent analysis and can effectively reveal sensitive patterns of reflectivity variation.
[0084] The calculated slope data are further analyzed comprehensively. By considering the slope differences between bands and the overall spectral changes in the band range, a quantitative index of the surface spectral slope is constructed. The construction expression of the quantitative index of the surface spectral slope is:
[0085]
[0086] Among them: max(S λ (1), S λ (2),…,S λ (n-1)) is the maximum value of the slope within the band, min(S λ (1), S λ (2),…,S λ (n-1)) is the minimum slope value within the band, n is the number of selected bands, S SSI It is a quantitative index of the surface spectral slope;
[0087] This step comprehensively reflects the sensitivity of surface reflectance changes by calculating the standardized difference between each slope value and the maximum and minimum slopes. Highly reflective areas of the surface typically display larger values in the Surface Spectral Slope Quantification Index because their reflectance varies significantly across bands, resulting in more significant spectral slope differences. In this way, the Surface Spectral Slope Quantification Index effectively quantifies high surface reflectance characteristics and provides a clear indicator of high reflectance areas in remote sensing data.
[0088] The Surface Spectral Slope Quantification Index (SSI), generated through feature engineering and in-depth analysis of the slope variations of the surface object's spectral reflectance within a specified wavelength range, indicates that a higher SSI value indicates a higher high reflectivity in that subregion. Because the SSI is derived through a comprehensive analysis of reflectivity variations within a specified wavelength range, larger slope values typically indicate a more dramatic wavelength-dependent variation in surface reflectivity. This typically occurs in highly reflective areas, such as snow, bare rock, and deserts. These areas have high reflectivity, especially within a specified wavelength range, where the slope of the spectral reflectance varies significantly, resulting in higher SSI values. Conversely, low-reflectivity areas (such as water bodies, wetlands, or dense vegetation) exhibit more gradual reflectivity variations within the same wavelength range, with smaller slope variations and lower SSI values. Therefore, the SSI, as a quantitative indicator of high reflectivity, can effectively distinguish areas of varying surface reflectivity.
[0089] The analyzed key features are input into the pre-trained support vector machine (SVM) model to predict the surface reflectance of each sub-region;
[0090] The reflectivity ratio quantification index and surface spectral slope quantification index generated by analyzing the reflectivity ratio of different bands and the slope change of the spectral reflectance of the ground object within the specified band range are input into the pre-trained support vector machine model. The reflection coefficient is generated by the support vector machine model, and the surface reflectivity of each sub-area is predicted by the reflection coefficient.
[0091] A pre-trained support vector machine model refers to a machine learning model that has been trained and optimized in advance on a dataset. It can generate a reflectance coefficient based on input feature data (such as a quantitative index of reflectance ratio and a quantitative index of surface spectral slope), thereby predicting the surface reflectance of each sub-region. In a support vector machine (SVM), the core goal of the model is to perform classification or regression analysis on the input data by finding a hyperplane (or a set of hyperplanes). In this way, the support vector machine can find the optimal boundary between different categories or regression tasks, thereby making accurate predictions when new data arrives. In remote sensing monitoring applications, this pre-trained model is particularly important for the prediction of surface reflectance because it can map complex spectral data into a high-dimensional feature space, thereby revealing the reflectance pattern and changing trend of the ground object.
[0092] The training process of the support vector machine model usually involves a large amount of historical data, which contains known surface reflectance values and their corresponding spectral reflectance characteristics (such as reflectance ratios and slope changes in specified bands). During the training phase, the SVM algorithm learns how to find the optimal decision boundary in the feature space by continuously adjusting the parameters in the model. This process not only considers the impact of different band characteristics on surface reflectance, but also uses kernel functions (such as linear kernels, radial basis functions, etc.) to map the data to a higher-dimensional space, allowing more complex nonlinear relationships to be effectively expressed. In this way, the support vector machine can construct an efficient regression model that can model the input reflectance ratios and spectral slope changes, and accurately predict the surface reflectance of the target area.
[0093] In actual applications, the trained support vector machine model is used to make predictions by inputting newly collected remote sensing data, and predicting the surface reflectivity of each sub-area based on the reflection coefficient generated by the model. Specifically, when new remote sensing data is input into the trained model, the model calculates the surface reflectivity value of each sub-area based on the previously learned pattern (i.e., the relationship between features and reflectivity). This prediction result is based on the surface reflectivity pattern and the association of band features learned by the model during the training process, so it can achieve accurate reflectivity prediction, thereby providing support for applications such as remote sensing monitoring, environmental assessment, and disaster warning. The advantage of the pre-trained support vector machine model is that it can effectively handle complex nonlinear relationships and high-dimensional data. Therefore, it can cope with the challenges of different surface types, environmental conditions, and band changes in remote sensing monitoring, ensuring the accuracy and reliability of the prediction results.
[0094] The support vector machine model is not specifically limited here, and can realize the quantization index R of reflectivity ratio ratio_index and the surface spectral slope quantification index S SSI Perform comprehensive analysis to generate reflection coefficient Reflectcoeff The support vector machine model can be any of the above. In order to implement the technical solution of the present invention, the present invention provides a specific implementation method; the reflection coefficient Reflect coeff The resulting calculation formula is: Where, ω x 、ω y Reflectance ratio quantitative index R ratio_index and the surface spectral slope quantification index S SSI The preset proportional coefficient, and ω x 、ω y Both are greater than 0.
[0095] It can be seen from the reflection coefficient that the larger the performance value of the reflectivity ratio quantification index generated by in-depth analysis of the reflectivity ratio of different bands extracted through feature engineering, the larger the performance value of the surface spectral slope quantification index generated by in-depth analysis of the slope change of the spectral reflectance of the ground object within the specified band range extracted through feature engineering, that is, the larger the performance value of the reflection coefficient generated when the surface reflectivity of each sub-area is predicted by the pre-trained support vector machine model, the higher the surface reflectivity of the sub-area, and vice versa.
[0096] Based on the prediction results of the support vector machine model, the actual electromagnetic wave emission density of each sub-region is dynamically adjusted. For highly reflective surfaces, remote sensing monitoring is performed at a lower electromagnetic wave emission density to avoid overexposure.
[0097] According to the prediction results of the support vector machine model, the actual electromagnetic wave emission density of each sub-area is dynamically adjusted. The specific steps are as follows:
[0098] According to the prediction results of the support vector machine model, the reflection coefficient R of each sub-region is generated pred (u), where u represents the index of the sub-region. The reflection coefficient is calculated by the support vector machine model based on the input features and is used to characterize the surface reflectivity of each sub-region. Based on the predicted reflection coefficient R pred (u), and further calculate the adjustment factor of the electromagnetic wave emission density. The adjustment factor determines the electromagnetic wave emission density required for each sub-area. The calculation expression is:
[0099]
[0100] Where: α(u) is the adjustment factor of the u-th sub-region, which represents the required electromagnetic wave emission density ratio; β' is a parameter that controls the adjustment sensitivity (which can be adjusted through experiments); θ is the reflection coefficient threshold, which is used to distinguish high-reflection and low-reflection areas. The sub-region with a reflection coefficient greater than the reflection coefficient threshold is a high-reflection area;
[0101] The purpose of this step is to calculate the adjustment factor of the electromagnetic wave emission density of each sub-area based on the reflection coefficient of the sub-area. The adjustment factor reflects the need to reduce the electromagnetic wave emission density in high reflectivity areas to prevent overexposure.
[0102] The actual electromagnetic wave emission density of each sub-area is dynamically adjusted using an adjustment factor. For highly reflective areas, the electromagnetic wave emission density is reduced to prevent signal overexposure; for low-reflective areas, the emission density can be appropriately increased to enhance signal strength. The dynamic adjustment formula for the electromagnetic wave emission density is as follows:
[0103] P emission (u)=P base (1-α(u))
[0104] , where: P emission (u) is the electromagnetic wave emission density of the u-th sub-region, P base It is the benchmark electromagnetic wave emission density, that is, the default standard emission density;
[0105] In this formula, when the adjustment factor α(u) is large, indicating a sub-region with high reflectivity, the system will reduce its electromagnetic wave emission density. Conversely, when the adjustment factor α(u) is small (i.e., with low reflectivity), the electromagnetic wave emission density remains high to ensure sufficient signal strength. By dynamically adjusting the electromagnetic wave emission density, remote sensing monitoring is optimized, ensuring that the system can effectively capture the characteristics of objects in areas with different reflectivity, while also avoiding distortion of monitoring data or loss of key information due to overexposure in high-reflectivity areas.
[0106] The above solution can effectively solve the problem of overexposure caused by constant electromagnetic wave emission density in remote sensing monitoring. By dividing the monitoring area into multiple sub-areas and acquiring reflectivity data in real time, combined with feature engineering and the prediction results of the support vector machine (SVM) model, accurate prediction of the reflectivity of each sub-area and dynamic adjustment of the electromagnetic wave emission density are achieved. In this way, for surface areas with high reflectivity, a lower emission density is used for monitoring, avoiding overexposure, ensuring the complete capture of ground details, and improving the accuracy and reliability of monitoring data. In addition, dynamic adjustment of electromagnetic wave emission density not only optimizes the remote sensing monitoring process, but also improves the adaptability and flexibility of the system, enabling effective identification and warning of potential environmental changes or disaster risks in complex environments, ultimately improving the feasibility and accuracy of ecological river bank protection design.
[0107] An ecological river bank protection design method provided in an embodiment of the present invention is implemented through the above-mentioned ecological river bank protection design system. The specific method and process of an ecological river bank protection design system are detailed in the embodiment of the above-mentioned ecological river bank protection design method, which will not be repeated here.
[0108] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formulas are set by technicians in this field according to actual conditions.
[0109] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.
[0110] The above description is merely illustrative of certain exemplary embodiments of the present invention. It goes without saying that those skilled in the art will be able to modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and description are illustrative in nature and should not be construed as limiting the scope of protection of the claims.
Claims
1. A method for designing ecological river bank protection, characterized in that: The specific steps include: Divide the surface area to be monitored into multiple sub-areas. By dividing the sub-areas, the complexity of the entire monitoring area is reduced, making it easier to set up independent monitoring and analysis strategies for each sub-area. In each sub-area, the surface reflectivity information is obtained in real time through remote sensing equipment; Extract key features reflecting high surface reflectivity from real-time reflectivity data. Conduct in-depth analysis of the extracted key features through feature engineering to preliminarily identify high reflectivity patterns in each sub-region within a complex surface environment. The analyzed key features are input into the pre-trained support vector machine model to predict the surface reflectivity of each sub-region; Based on the prediction results of the support vector machine model, the actual electromagnetic wave emission density of each sub-area is dynamically adjusted. For surfaces with high reflectivity, remote sensing monitoring is carried out with a lower electromagnetic wave emission density to avoid overexposure.
2. The ecological river bank protection design method according to claim 1 is characterized in that: The specific steps for dividing the surface area to be monitored into multiple sub-areas are as follows: First, the spatial scope of the entire monitoring area should be clarified and the geographical boundaries should be determined; According to the heterogeneity of surface characteristics, appropriate classification criteria are selected in combination with monitoring objectives and required analysis accuracy; Within the identified monitoring area, the area is divided into multiple sub-areas based on the selected criteria; Each sub-area after division is numbered or marked to facilitate subsequent monitoring, data processing and result analysis, ensuring that each sub-area has an independent identification mark to facilitate dynamic monitoring and data tracking.
3. The ecological river bank protection design method according to claim 1 is characterized in that: Key features reflecting high surface reflectivity are extracted from the reflectivity data acquired in real time. The extracted features include the reflectivity ratios of different bands and the slope changes of the spectral reflectivity of the objects within the specified band range. Through feature engineering, the extracted reflectivity ratios of different bands and the slope changes of the spectral reflectivity of the objects within the specified band range are deeply analyzed to generate the reflectivity ratio quantification index and the surface spectral slope quantification index respectively. The reflectivity ratio quantification index and the surface spectral slope quantification index are used to preliminarily identify the high reflectivity pattern of each sub-area in a complex surface environment.
4. The ecological river bank protection design method according to claim 3 is characterized in that: The reflectivity ratio quantification index and surface spectral slope quantification index generated by analyzing the reflectivity ratio of different bands and the slope change of the spectral reflectance of the ground object within the specified band range are input into the pre-trained support vector machine model. The reflection coefficient is generated by the support vector machine model, and the surface reflectivity of each sub-area is predicted by the reflection coefficient.
5. The ecological river bank protection design method according to claim 3 is characterized in that: The specific steps for generating a quantitative reflectance ratio index by performing an in-depth analysis of the extracted reflectance ratios of different bands through feature engineering are as follows: The near-infrared band and the visible light band are selected to calculate the reflectance ratio between different bands. The calculation expression is: Where: R ratio Represents the reflectivity ratio, R NIR Represents the reflectivity in the near-infrared band, R VIS Indicates the reflectivity of visible light band; The reflectivity ratio quantification index is generated by performing a nonlinear transformation on the reflectivity ratio to reflect the significance of high reflectivity in the sub-region. The reflectivity ratio quantification index is calculated by combining the exponential function and the band weighting coefficient. The calculation expression is: Where: R ratio_index is the reflectivity ratio quantization index, α is the exponential coefficient of the reflectivity ratio, which is used to adjust the influence of the reflectivity ratio in the calculation and control the sensitivity of the ratio. β is a nonlinear weighting coefficient that controls the influence of the reflectivity ratio deviation, making the relationship between the reflectivity ratio and the high reflectivity characteristics of the surface more accurate. exp(·) is an exponential function used to further enhance the contrast between high-reflectivity and low-reflectivity areas.
6. The ecological river bank protection design method according to claim 3 is characterized in that: The specific steps for generating a quantitative index of the surface spectral slope by performing an in-depth analysis of the slope changes of the ground feature spectral reflectance within the specified band range through feature engineering are as follows: The spectral slope is calculated based on the reflectance data of the ground objects within the specified band range extracted from the remote sensing data. Assume that the selected band range is λ1 to λ n , the spectrum slope is calculated by the following formula, and the calculation expression is: Where: S λ (i) is the slope of point i, R λ (i) is the reflectivity of the ground object measured at wavelength λ(i), R λ (i+1) is the reflectance of the ground object measured at wavelength λ(i+1), λ(i) is the wavelength value, and i is the discrete sampling point in the band; The calculated slope data are further analyzed comprehensively. By considering the slope differences between bands and the overall spectral changes in the band range, a quantitative index of the surface spectral slope is constructed. The construction expression of the quantitative index of the surface spectral slope is: Among them: max(S λ (1), S λ (2),…,S λ (n-1)) is the maximum value of the slope within the band, min(S λ (1), S λ (2),…,S λ (n-1)) is the minimum slope value within the band, n is the number of selected bands, S SSI It is a quantitative index of the surface spectral slope.
7. The ecological river bank protection design method according to claim 4 is characterized in that: According to the prediction results of the support vector machine model, the actual electromagnetic wave emission density of each sub-area is dynamically adjusted. The specific steps are as follows: According to the prediction results of the support vector machine model, the reflection coefficient R of each sub-region is generated pred (u), where u represents the index of the subregion, based on the predicted reflection coefficient R pred (u), and further calculate the adjustment factor of the electromagnetic wave emission density. The adjustment factor determines the electromagnetic wave emission density required for each sub-area. The calculation expression is: Where: α(u) is the adjustment factor of the u-th sub-region, which represents the required electromagnetic wave emission density ratio, β ' is a parameter that controls the adjustment sensitivity, θ is the reflection coefficient threshold, which is used to distinguish high-reflection and low-reflection areas. The sub-area with a reflection coefficient greater than the reflection coefficient threshold is a high-reflection area; The actual electromagnetic wave emission density of each sub-area is dynamically adjusted using an adjustment factor. For highly reflective areas, the electromagnetic wave emission density is reduced to prevent signal overexposure; for low-reflective areas, the emission density is increased to enhance signal strength. The dynamic adjustment formula for the electromagnetic wave emission density is as follows: P emission (in)=P base ·(1-α(u)) Where: P emission (u) is the electromagnetic wave emission density of the u-th sub-region, P base It is the benchmark electromagnetic wave emission density, that is, the default standard emission density.
8. An ecological river bank protection design system, used to implement the ecological river bank protection design method according to any one of claims 1 to 7, characterized in that: It includes region division module, remote sensing data acquisition module, feature extraction and analysis module, prediction model module and emission density adjustment module; The regional division module divides the surface area to be monitored into multiple sub-areas. By dividing the sub-areas, the complexity of the entire monitoring area is reduced, making it easier to set independent monitoring and analysis strategies for each sub-area. Remote sensing data acquisition module, in each sub-area, obtains surface reflectivity information in real time through remote sensing equipment; The feature extraction and analysis module extracts key features reflecting high surface reflectivity from real-time reflectivity data. Through feature engineering, the extracted key features are deeply analyzed to preliminarily identify high reflectivity patterns in each sub-region in a complex surface environment. The prediction model module inputs the analyzed key features into the pre-trained support vector machine model to predict the surface reflectivity of each sub-area; The emission density adjustment module dynamically adjusts the actual electromagnetic wave emission density of each sub-area based on the prediction results of the support vector machine model. For surfaces with high reflectivity, remote sensing monitoring is performed with a lower electromagnetic wave emission density to avoid overexposure.