Ecological sensitive area monitoring method based on multispectral remote sensing

Abstract

The invention relates to the cross technical field of remote sensing technology and ecological environment monitoring, and discloses an ecological sensitive area monitoring method based on multispectral remote sensing, which comprises the following steps: multispectral data acquisition: performing multispectral data acquisition on a target ecological sensitive area by using an unmanned aerial vehicle carrying a multispectral camera, the multispectral camera at least comprises a plurality of wave bands such as visible light and near-infrared light; data preprocessing: preprocessing the acquired multispectral data, including radiation correction, image registration and orthoimage generation, and converting an original DN value into reflectivity data by radiation correction; and vegetation coverage extraction: extracting the vegetation coverage by using a normalized vegetation index based on the preprocessed multispectral data. The ecological sensitive area monitoring method based on multispectral remote sensing aims at solving the problems that a traditional ecological sensitive area monitoring method is limited in monitoring range, insufficient in data precision, poor in timeliness, bottleneck in data processing and analysis and the like.

Description

Technical Field

[0001] The present invention relates to the cross-technical field of remote sensing technology and ecological environment monitoring, and in particular to an ecologically sensitive area monitoring method based on multi-spectral remote sensing. Background Art

[0002] Ecologically sensitive areas, as indispensable natural treasures on the earth, refer to areas where the ecological environment is extremely fragile and easily susceptible to external interference. These areas often carry rich biodiversity and are habitats for many rare species. They play a vital role in maintaining the global ecological balance. They are like the "lungs" of nature. Through complex ecological processes, they regulate the climate, purify water quality, maintain soil health, and provide valuable resources for the reproduction of life on Earth.

[0003] However, traditional ecologically sensitive area monitoring methods often seem to be unable to cope with the demand for comprehensive and real-time monitoring. Most of these methods rely on manual field surveys or data collection from ground observation stations. Not only is the monitoring scope limited, making it difficult to cover the vast and unevenly distributed ecologically sensitive areas, but there are also obvious deficiencies in data accuracy and timeliness. Manual field surveys are easily restricted by various factors such as weather and terrain, resulting in long data collection cycles, high costs, and difficulty in capturing the dynamic changes in ecologically sensitive areas. Although ground observation stations can provide continuous data records, their coverage is limited and they are easily affected by the local environment, and cannot fully reflect the overall situation of ecologically sensitive areas. In addition, traditional monitoring methods also have bottlenecks in data processing and analysis, making it difficult to efficiently and accurately extract key ecological parameters and changing trends in ecologically sensitive areas. Summary of the invention

[0004] The purpose of the present invention is to solve the problems of limited monitoring range, insufficient data accuracy, poor timeliness and data processing and analysis bottlenecks existing in traditional ecological sensitive area monitoring methods, and to propose an ecological sensitive area monitoring method based on multispectral remote sensing.

[0005] The technical solution of the present invention to solve the above technical problems is as follows:

[0006] A method for monitoring ecologically sensitive areas based on multispectral remote sensing comprises the following steps:

[0007] S10, multispectral data acquisition: using a drone equipped with a multispectral camera to acquire multispectral data of the target ecologically sensitive area, wherein the multispectral camera includes at least visible light and near infrared bands;

[0008] S20, data preprocessing: preprocessing the acquired multispectral data, including radiation correction, image registration and orthophoto generation, wherein the radiation correction converts the original DN value into reflectance data;

[0009] S30, vegetation coverage extraction: extracting vegetation coverage using normalized vegetation index based on pre-processed multispectral data;

[0010] S40, Ecological vulnerability assessment: Combine vegetation coverage data with environmental factors such as topography and climate in ecologically sensitive areas to construct an ecological vulnerability assessment model for quantitative assessment;

[0011] S50, Dynamic monitoring and early warning: Based on the continuous acquisition and processing of multispectral data, dynamic monitoring of ecologically sensitive areas is achieved, and early warning thresholds are set to trigger the early warning mechanism.

[0012] Based on the above technical solution, the present invention can also be improved as follows.

[0013] Furthermore, in step S10, a Changguang Yuchen MS600V2 series multispectral camera is used for data collection, wherein the spectral camera is equipped with a high-precision positioning system, which can record the shooting location information in real time. According to the characteristics of the ecologically sensitive area and the monitoring objectives, a detailed shooting plan is formulated, including parameters such as shooting time, angle, and height, to ensure that the collected multispectral data is comprehensive and representative.

[0014] Furthermore, the radiation correction in step S20 adopts a linear correction model to improve the correction accuracy. Specifically, the radiation correction formula is as follows:

[0015]

[0016] Where R is the corrected reflectance, DN is the original DN value, and DN dark is the dark current value, DN white is the white balance value, k and c are correction coefficients, which are obtained by fitting known standard reflectance data.

[0017] Furthermore, in step S30, the normalized vegetation is calculated using an index formula, and the vegetation coverage level is divided in combination with the actual situation of the ecologically sensitive area. Specifically, the index calculation formula is as follows:

[0018]

[0019] Among them, ρ NIR is the reflectivity in the near-infrared band, ρ RED is the reflectivity of the red band, ∈ is a very small positive number used to avoid the denominator being zero.

[0020] Furthermore, in step S40, the ecological vulnerability assessment model adopts a random forest algorithm and combines it with a geographic information system for spatial analysis and visualization. The accuracy and generalization ability of the model are improved by constructing multiple decision trees and integrating their output results. Historical monitoring data and the actual ecological conditions of ecologically sensitive areas are used as training samples. The parameters of the random forest model are optimized and adjusted through cross-validation and grid search technology. The assessment results are displayed in the form of a map, and the ecological vulnerability levels are represented on the map in different colors or patterns through spatial interpolation and visualization technology.

[0021] Furthermore, in the training process of the machine learning algorithm, in addition to using historical monitoring data and actual ecological conditions of ecologically sensitive areas as training samples, external data sources are also introduced as auxiliary information. These external data sources include but are not limited to meteorological data, terrain data, and soil type data. These external data sources are integrated with multispectral data through feature selection and feature fusion methods to construct a more comprehensive and rich feature space. At the same time, data enhancement techniques such as rotation, scaling, and translation are used to pre-process the training samples to improve the robustness and generalization ability of the model.

[0022] Furthermore, the dynamic monitoring in step S50 not only includes regularly acquiring, processing and analyzing multispectral data, but also includes real-time monitoring of changes in ecologically sensitive areas. A detailed monitoring plan is formulated based on the characteristics of the ecologically sensitive areas and monitoring needs, including monitoring frequency, monitoring range, and parameters of monitoring indicators. By constructing a real-time monitoring system and a data processing platform, real-time collection, transmission and processing of multispectral data are achieved. At the same time, time series analysis and trend prediction methods are used to dynamically analyze and predict real-time monitoring data to promptly discover and warn of abnormal changes in ecologically sensitive areas.

[0023] Furthermore, the warning threshold set in the early warning mechanism is dynamically adjusted according to the actual situation of the ecologically sensitive areas and the ecological protection goals. The ecological conditions of the ecologically sensitive areas are reflected by constructing key indicators of the ecological vulnerability index and the vegetation coverage index. Reasonable warning thresholds are set according to the historical data and changing trends of these indicators. When the monitoring data reaches or exceeds the warning threshold, the system automatically triggers the warning signal and reminds the monitoring personnel by text message, email or system notification. At the same time, the early warning mechanism also has the function of automatic adjustment and optimization, and can dynamically adjust and optimize the warning threshold according to real-time monitoring data and the actual situation of the ecologically sensitive areas.

[0024] Furthermore, the setting of the warning threshold not only takes into account the actual conditions of the ecologically sensitive areas and the qualitative factors of the ecological protection objectives, but also introduces quantitative assessment methods and indicators. By constructing an ecological vulnerability assessment model and a vegetation coverage assessment model, the ecological conditions of the ecologically sensitive areas are quantitatively assessed, and reasonable warning thresholds are set based on the assessment results and ecological protection objectives. At the same time, the setting of the warning threshold also takes into account the characteristics and differences of different ecologically sensitive areas, as well as the severity and urgency of different ecological problems, and ensures the scientificity and rationality of the warning threshold through comprehensive analysis and evaluation.

[0025] Compared with the prior art, the technical solution of this application has the following beneficial technical effects:

[0026] The present invention uses an unmanned aerial vehicle equipped with a multispectral camera to acquire multispectral data of the target ecologically sensitive area. The unmanned aerial vehicle can cover a wider area. At the same time, the multispectral camera contains at least multiple bands such as visible light and near-infrared, providing richer and more accurate spectral information, laying a solid foundation for subsequent monitoring and analysis. The acquired multispectral data are preprocessed, including radiation correction, image registration and orthophoto generation. The radiation correction converts the original DN value into reflectance data. This processing process ensures the accuracy and comparability of the data and provides a reliable basis for subsequent analysis. Image registration and orthophoto generation further improve the spatial resolution and geometric accuracy of the data, which helps to more accurately identify and analyze the characteristics of ecologically sensitive areas. The vegetation coverage is extracted using the normalized vegetation index. The normalized vegetation index is a widely used vegetation monitoring indicator that can sensitively reflect the growth status and coverage of vegetation. The first feature makes this method highly accurate and real-time in vegetation coverage extraction, which provides an important basis for ecological vulnerability assessment. By combining vegetation coverage data with environmental factors such as topography and climate in ecologically sensitive areas, an ecological vulnerability assessment model is constructed for quantitative assessment. This feature comprehensively considers a variety of ecological factors, making the assessment results more comprehensive and accurate. By constructing an ecological vulnerability assessment model, the vulnerability of ecologically sensitive areas can be quantitatively analyzed, providing a scientific basis for the formulation of targeted protection measures. Finally, based on the continuous acquisition and processing of multispectral data, dynamic monitoring of ecologically sensitive areas is achieved, and warning thresholds are set to trigger warning mechanisms. This feature ensures the timeliness and accuracy of monitoring, and can promptly detect abnormal situations in ecologically sensitive areas and take corresponding countermeasures. By setting warning thresholds, real-time monitoring and early warning of ecologically fragile areas can be achieved, which helps prevent and reduce the occurrence of ecological disasters. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 The present invention is a method flow chart of an ecologically sensitive area monitoring method based on multispectral remote sensing. DETAILED DESCRIPTION

[0028] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0029] Combination Figure 1 As shown, a method for monitoring ecologically sensitive areas based on multispectral remote sensing of the present invention comprises the following steps:

[0030] S10. Multispectral data acquisition: Use a drone equipped with a multispectral camera to acquire multispectral data of the target ecologically sensitive area. The multispectral camera contains at least visible light and near-infrared bands;

[0031] S20, data preprocessing: preprocessing the acquired multispectral data, including radiation correction, image registration and orthophoto generation, wherein the radiation correction converts the original DN value into reflectance data;

[0032] S30, vegetation coverage extraction: extracting vegetation coverage using normalized vegetation index based on pre-processed multispectral data;

[0033] S40, Ecological vulnerability assessment: Combine vegetation coverage data with environmental factors such as topography and climate in ecologically sensitive areas to construct an ecological vulnerability assessment model for quantitative assessment;

[0034] S50, Dynamic monitoring and early warning: Based on the continuous acquisition and processing of multispectral data, dynamic monitoring of ecologically sensitive areas is achieved, and early warning thresholds are set to trigger the early warning mechanism.

[0035] In a preferred embodiment, the present invention can be further configured as follows: in step S10, a Changguang Yuchen MS600V2 series multispectral camera is used for data collection, wherein a high-precision positioning system is provided in the spectral camera, which can record the shooting location information in real time, and a detailed shooting plan is formulated according to the characteristics of the ecologically sensitive area and the monitoring target, including the parameters of shooting time, angle, and height, to ensure that the collected multispectral data is comprehensive and representative, and the Changguang Yuchen MS600V2 series multispectral camera is used for data collection to achieve multiple significant beneficial effects. First of all, this configuration greatly improves the accuracy and comprehensiveness of data collection. As a professional-grade spectral acquisition device, the Changguang Yuchen MS600V2 series multispectral camera has a built-in high-precision positioning system that can record the shooting location information in real time, ensuring that the collected data has accurate geographic coordinates, which is crucial for subsequent data analysis and spatial position matching. At the same time, the high spectral resolution of the camera can capture the subtle differences of surface objects in different spectral bands, making it possible for detailed monitoring of ecologically sensitive areas. During implementation, a detailed shooting plan is formulated according to the characteristics of the ecologically sensitive area and monitoring targets, such as vegetation type, soil moisture, water pollution, etc., including parameters such as shooting time, angle, and height. Such a shooting plan ensures that the collected multispectral data not only fully covers the entire ecologically sensitive area, but also can capture the spectral characteristics of key ecological elements under different time and space conditions. For example, in vegetation monitoring, it is possible to choose to shoot in seasons with vigorous plant growth and clear weather conditions to obtain accurate information on key parameters such as vegetation chlorophyll content and growth status. This data collection method not only improves the efficiency and accuracy of monitoring, but also provides strong data support for the formulation and implementation of ecological protection measures.

[0036] In a preferred embodiment, the present invention can be further configured as follows: the radiation correction in step S20 adopts a linear correction model to improve the correction accuracy. Specifically, the radiation correction formula is as follows:

[0037]

[0038] Where R is the corrected reflectance, DN is the original DN value, and DN dark is the dark current value, DN whiteis the white balance value, k and c are the correction coefficients, which are obtained by fitting known standard reflectance data. This correction method not only takes into account the nonlinear characteristics of the instrument itself, but also can adapt to the radiation correction needs under different lighting conditions and different surface cover types. For example, in the monitoring of ecologically sensitive areas, there are various types of surface cover, including forests, grasslands, water bodies, etc., and the reflectance characteristics of each type are different. By fitting the standard reflectance data of a specific ecologically sensitive area, the correction coefficient applicable to the area can be obtained, thereby ensuring the accuracy and pertinence of the radiation correction. This flexible and precise correction method provides a more reliable data basis for monitoring ecologically sensitive areas, and helps to improve the accuracy and scientificity of the monitoring results.

[0039] In a preferred embodiment, the present invention can be further configured as follows: in step S30, the normalized vegetation index calculation formula is used, and the vegetation coverage level is divided in combination with the actual situation of the ecologically sensitive area. Specifically, the index calculation formula is as follows:

[0040]

[0041] Among them, ρ NIR is the reflectivity in the near-infrared band, ρ REd is the reflectivity of the red band, ∈ is a very small positive number, which is used to avoid the situation where the denominator is zero. In step S30, the calculation formula of the normalized difference vegetation index (NDVI) is adopted, and the vegetation coverage level is divided according to the actual situation of the ecologically sensitive area. This measure has brought many beneficial effects. First of all, NDVI, as a widely used vegetation index, can effectively reflect the key information such as the growth status, coverage and biomass of vegetation, and is an important tool for vegetation monitoring in ecologically sensitive areas. By calculating NDVI, a rapid and accurate assessment of vegetation coverage can be achieved, which is of great significance for monitoring the dynamic changes of vegetation in ecologically sensitive areas and assessing the health of the ecosystem.

[0042] The importance of dividing vegetation coverage levels in combination with the actual conditions of ecologically sensitive areas. This approach fully considers factors such as the regional characteristics, vegetation types and ecological functions of ecologically sensitive areas, making the assessment of vegetation coverage more in line with the actual situation, more scientific and targeted. For example, in some arid or semi-arid ecologically sensitive areas, the vegetation coverage is low, but the growth of vegetation is crucial to maintaining the stability of the ecosystem. Therefore, when dividing the vegetation coverage level, a relatively low threshold can be set to more accurately reflect the vegetation conditions in these areas. In humid ecologically sensitive areas, the vegetation coverage is high and the biomass is abundant. At this time, a higher threshold can be set to distinguish areas with different vegetation coverage. Through such a level division, the vegetation conditions in ecologically sensitive areas can be more accurately evaluated, providing a more scientific basis for ecological protection and management. At the same time, this division method combined with actual conditions also helps to improve the practicality and operability of monitoring results.

[0043] In a preferred embodiment, the present invention can be further configured as follows: in step S40, the ecological vulnerability assessment model adopts a random forest algorithm, and combines a geographic information system for spatial analysis and visualization display, and improves the accuracy and generalization ability of the model by constructing multiple decision trees and synthesizing their output results, taking historical monitoring data and the actual ecological conditions of ecologically sensitive areas as training samples, optimizing and adjusting the parameters of the random forest model through cross-validation and grid search technology, and displaying the assessment results in the form of a map, and representing the ecological vulnerability level on the map with different colors or patterns through spatial interpolation and visualization technology, introducing a random forest algorithm in step S40 for ecological vulnerability assessment, and combining a geographic information system (GIS) for spatial analysis and visualization display, this measure brings significant beneficial effects. First, as an integrated learning method, the random forest algorithm can significantly improve the accuracy and generalization ability of the model by constructing multiple decision trees and synthesizing their output results. In the ecological vulnerability assessment, this method can comprehensively consider the interaction of multiple ecological factors (such as vegetation coverage, soil erosion degree, water resources status, etc.), so as to more accurately reflect the vulnerability status of ecologically sensitive areas. This is of great significance for formulating targeted ecological protection measures and optimizing resource allocation.

[0044] Combining geographic information systems for spatial analysis and visualization makes the results of ecological vulnerability assessment more intuitive and easy to understand. By using historical monitoring data and the actual ecological conditions of ecologically sensitive areas as training samples, and optimizing and adjusting the parameters of the random forest model using cross-validation and grid search techniques, the accuracy and reliability of the model can be ensured. The assessment results are presented in the form of maps. Through spatial interpolation and visualization techniques, the ecological vulnerability levels are represented on the map in different colors or patterns, so that decision makers can intuitively see the spatial distribution and changing trends of ecological vulnerability. This visual expression method not only improves the efficiency of information transmission, but also helps decision makers better understand the spatial pattern and potential risks of ecological vulnerability, so as to make more scientific and reasonable ecological protection decisions. For example, in the planning of ecologically sensitive areas, high-risk areas can be identified based on ecological vulnerability maps, and protection measures can be taken as a priority to reduce the risk of damage to the ecosystem.

[0045] In a preferred embodiment, the present invention can be further configured as follows: in the training process of the machine learning algorithm, in addition to using historical monitoring data and the actual ecological conditions of the ecologically sensitive areas as training samples, external data sources are also introduced as auxiliary information, and these external data sources include but are not limited to meteorological data, topographic data, and soil type data. These external data sources are integrated with multispectral data through feature selection and feature fusion methods to construct a more comprehensive and rich feature space. At the same time, data enhancement techniques such as rotation, scaling, and translation are used to pre-process the training samples to improve the robustness and generalization ability of the model. In the training process of the machine learning algorithm, by introducing external data sources as auxiliary information, combining feature selection and feature fusion methods, and using data enhancement techniques to pre-process the training samples, significant beneficial effects are brought. First, this measure greatly enriches the feature space of the training samples, so that the machine learning model can learn more relevant information about the vulnerability of the ecologically sensitive areas. In addition to historical monitoring data and the actual ecological conditions of the ecologically sensitive areas, external data sources such as meteorological data, topographic data, and soil type data provide additional dimensions, which are essential for a comprehensive understanding of the causes and mechanisms of ecological vulnerability. By integrating these data sources, the model can more accurately capture the spatial distribution and changing trends of ecological vulnerability, thereby improving the accuracy and reliability of the assessment.

[0046] The methods of feature selection and feature fusion ensure the effective integration of external data sources with multispectral data. Feature selection can screen out the most influential features for ecological vulnerability assessment and avoid information redundancy and noise interference; while feature fusion can effectively integrate relevant information from different data sources to construct a more comprehensive and rich feature space. In addition, the use of data enhancement technology to pre-process training samples, such as rotation, scaling, translation and other operations, can increase the diversity and complexity of training samples, thereby improving the robustness and generalization ability of the model. This means that the model can make more accurate predictions and judgments when faced with new and unseen data. For example, in the actual monitoring of ecologically sensitive areas, various complex environmental conditions and surface cover types may be encountered. The model trained by data enhancement technology can better adapt to these changes and provide more accurate and reliable ecological vulnerability assessment results.

[0047] In a preferred embodiment, the present invention can be further configured as follows: the dynamic monitoring in step S50 not only includes periodically acquiring multispectral data and processing and analyzing it, but also includes real-time monitoring of changes in ecologically sensitive areas. A detailed monitoring plan is formulated according to the characteristics of ecologically sensitive areas and monitoring requirements, including monitoring frequency, monitoring range, and parameters of monitoring indicators. The real-time acquisition, transmission, and processing of multispectral data are realized by constructing a real-time monitoring system and a data processing platform. At the same time, the real-time monitoring data is dynamically analyzed and predicted by using time series analysis and trend prediction methods, and abnormal changes in ecologically sensitive areas are discovered and warned in time. The dynamic monitoring scheme proposed in step S50 not only covers the traditional monitoring method of periodically acquiring multispectral data and processing and analyzing it, but more importantly, it introduces a mechanism for real-time monitoring of changes in ecologically sensitive areas. This improvement ensures the continuity and timeliness of monitoring work, so that any abnormal changes in ecologically sensitive areas can be discovered and warned in time. This is of great significance for protecting the ecological environment and preventing ecological disasters. In particular, in the face of rapidly changing environmental conditions and potential ecological risks, real-time monitoring can provide a critical time window so that countermeasures can be taken in time.

[0048] The program has formulated a detailed monitoring plan based on the characteristics and monitoring needs of ecologically sensitive areas, including monitoring frequency, monitoring range, parameters of monitoring indicators, etc. This customized monitoring plan ensures the pertinence and effectiveness of monitoring work and avoids waste of resources and monitoring blind spots. For example, in ecologically fragile areas or ecologically sensitive points, higher monitoring frequencies and finer monitoring ranges can be set to capture any subtle ecological changes. At the same time, the construction of a real-time monitoring system and data processing platform realizes the real-time collection, transmission and processing of multispectral data, greatly improving monitoring efficiency and data quality. In addition, the use of time series analysis and trend prediction methods to dynamically analyze and predict real-time monitoring data can reveal the trends and laws of changes in ecologically sensitive areas and provide a scientific basis for ecological protection and management. For example, through the time series analysis of vegetation coverage, the growth status and change trend of vegetation in the future can be predicted, so as to formulate protection measures in advance to prevent ecological degradation. This dynamic monitoring program not only improves the intelligent level of ecological monitoring, but also provides strong support for ecological protection and sustainable development.

[0049] In a preferred embodiment, the present invention can be further configured as follows: the warning threshold set in the early warning mechanism is dynamically adjusted according to the actual situation of the ecological sensitive area and the ecological protection target, and the ecological status of the ecological sensitive area is reflected by constructing the key indicators of the ecological vulnerability index and the vegetation coverage index. A reasonable early warning threshold is set according to the historical data and change trend of these indicators. When the monitoring data reaches or exceeds the early warning threshold, the system automatically triggers the early warning signal and reminds the monitoring personnel through SMS, email or system notification. At the same time, the early warning mechanism also has the function of automatic adjustment and optimization, and can dynamically adjust and optimize the early warning threshold according to the real-time monitoring data and the actual situation of the ecological sensitive area. By constructing key indicators such as the ecological vulnerability index and the vegetation coverage index, the ecological status of the ecological sensitive area can be fully and accurately reflected, providing a scientific basis for the setting of the early warning mechanism. The historical data and change trend of these indicators are used to set a reasonable early warning threshold, ensuring the accuracy and sensitivity of the early warning mechanism. When the monitoring data reaches or exceeds the early warning threshold, the system automatically triggers the early warning signal and quickly reminds the monitoring personnel through a variety of communication methods (such as SMS, email or system notification), thereby achieving instant response and rapid processing of abnormal changes in the ecological sensitive area.

[0050] In practical applications, the ecological conditions in ecologically sensitive areas may change due to a variety of factors, such as climate change, human interference, etc. Therefore, the warning threshold needs to be dynamically adjusted and optimized according to the real-time monitoring data and the actual situation of the ecologically sensitive areas to ensure the continued effectiveness and adaptability of the early warning mechanism. For example, in some ecologically fragile areas, due to climate change, the vegetation coverage has decreased. At this time, the early warning mechanism should automatically adjust the warning threshold of the vegetation coverage index to more accurately reflect the ecological conditions of the area and trigger the early warning signal in time. This automatic adjustment and optimization function not only improves the flexibility and accuracy of the early warning mechanism, but also greatly reduces the workload of monitoring personnel, providing a strong guarantee for the protection and sustainable development of ecologically sensitive areas.

[0051] In a preferred embodiment, the present invention can be further configured as follows: the setting of the early warning threshold not only considers the actual situation of the ecologically sensitive area and the qualitative factors of the ecological protection target, but also introduces quantitative evaluation methods and indicators, and quantitatively evaluates the ecological status of the ecologically sensitive area by constructing an ecological vulnerability assessment model and a vegetation coverage assessment model, and sets a reasonable early warning threshold according to the assessment results and the ecological protection target. At the same time, the setting of the early warning threshold also considers the characteristics and differences of different ecologically sensitive areas, as well as the severity and urgency of different ecological problems. Through comprehensive analysis and evaluation, the scientificity and rationality of the early warning threshold are ensured, and an innovative method of comprehensively considering qualitative and quantitative factors is adopted in the setting of the early warning threshold, which brings significant beneficial effects. First, by introducing quantitative evaluation methods and indicators, such as constructing an ecological vulnerability assessment model and a vegetation coverage assessment model, the ecological status of the ecologically sensitive area can be accurately and objectively measured, which provides a solid scientific basis for the setting of the early warning threshold. This quantitative evaluation not only improves the accuracy of the early warning mechanism, but also enhances its applicability and flexibility under different ecological conditions. The setting of the early warning threshold is no longer based on experience or subjective judgment alone, but is based on objective analysis of scientific models and data, thereby greatly improving the effectiveness and reliability of the early warning.

[0052] The characteristics and differences of different ecologically sensitive areas, as well as the severity and urgency of different ecological problems, are also fully considered. In practical applications, the ecological conditions of ecologically sensitive areas vary greatly, and the problems they face are also different. Therefore, the setting of early warning thresholds must be adapted to local conditions and fully consider the regional characteristics and the particularity of ecological problems. For example, in some arid and semi-arid areas, water shortage is the main ecological problem, and the setting of early warning thresholds should focus on changes in soil moisture and vegetation coverage; while in humid areas, more attention may be paid to risks such as floods and soil erosion. In addition, for some areas where ecological problems are serious and urgent, the early warning threshold should be set more sensitively to ensure timely response and effective treatment. By comprehensively analyzing and evaluating these factors, the scientificity and rationality of the early warning threshold are ensured, providing strong support for the protection and sustainable development of ecologically sensitive areas. For example, in a certain ecologically fragile area, by constructing an ecological vulnerability assessment model, it was found that the vegetation coverage in the area decreased significantly and soil erosion intensified. Based on this assessment result, the early warning mechanism automatically adjusted the early warning thresholds of vegetation coverage and soil erosion, ensuring that the early warning can be triggered in time when the ecological conditions in the area deteriorate, providing a valuable time window for ecological protection and management.

[0053] In step S10, a drone equipped with a multispectral camera is used to acquire multispectral data of the target ecologically sensitive area. The Changguang Yuchen MS600V2 series multispectral camera is specially used here. Its built-in high-precision positioning system can record the shooting location information in real time to ensure that the collected multispectral data is comprehensive and representative. The shooting plan is formulated according to the characteristics of the ecologically sensitive area and the monitoring objectives, including shooting time, angle, height and other parameters to maximize the accuracy and practicality of the data;

[0054] Next, in step S20, the acquired multispectral data is preprocessed. The preprocessing process includes radiation correction, image registration, and orthophoto generation. Radiation correction uses a linear correction model. The correction coefficient is obtained by fitting known standard reflectance data, and the original DN value is converted into reflectance data to improve the accuracy and availability of the data. Image registration and orthophoto generation further ensure the accuracy and consistency of the data.

[0055] In step S30, based on the pre-processed multispectral data, the vegetation coverage is extracted using the normalized vegetation index. The calculation formula of the normalized vegetation index combines the reflectance of the near-infrared and red bands. The vegetation coverage index is calculated and the vegetation coverage level is divided according to the actual situation of the ecologically sensitive area, providing key information for the subsequent ecological vulnerability assessment.

[0056] Entering step S40, combining vegetation coverage data and environmental factors such as topography and climate in ecologically sensitive areas, an ecological vulnerability assessment model is constructed for quantitative assessment. Here, the random forest algorithm is used, and combined with the geographic information system for spatial analysis and visualization. By constructing multiple decision trees and integrating their output results, the accuracy and generalization ability of the model are improved. At the same time, historical monitoring data and the actual ecological conditions of ecologically sensitive areas are used as training samples. The model parameters are optimized and adjusted through cross-validation and grid search technology to ensure the accuracy and reliability of the assessment results. The assessment results are displayed in the form of a map. The ecological vulnerability levels are represented on the map with different colors or patterns through spatial interpolation and visualization technology, which is convenient for intuitive understanding and analysis.

[0057] In step S50, dynamic monitoring and early warning are realized. Dynamic monitoring includes not only regularly acquiring multispectral data and processing and analyzing them, but also real-time monitoring of changes in ecologically sensitive areas. According to the characteristics of ecologically sensitive areas and monitoring needs, a detailed monitoring plan is formulated, including parameters such as monitoring frequency, monitoring range, and monitoring indicators. By building a real-time monitoring system and a data processing platform, real-time collection, transmission, and processing of multispectral data are realized. At the same time, time series analysis and trend prediction methods are used to dynamically analyze and predict real-time monitoring data, so as to timely discover and warn abnormal changes in ecologically sensitive areas.

[0058] The warning threshold set in the early warning mechanism is dynamically adjusted according to the actual situation of ecologically sensitive areas and ecological protection goals. It reflects the ecological status of ecologically sensitive areas by constructing key indicators such as the ecological vulnerability index and vegetation coverage index, and sets reasonable warning thresholds based on the historical data and changing trends of these indicators. When the monitoring data reaches or exceeds the warning threshold, the system automatically triggers the warning signal and reminds the monitoring personnel through text messages, emails or system notifications. In addition, the early warning mechanism also has the function of automatic adjustment and optimization, which can dynamically adjust and optimize the warning threshold according to real-time monitoring data and the actual situation of ecologically sensitive areas, to ensure the accuracy and effectiveness of the early warning mechanism.

[0059] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0060] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for monitoring ecologically sensitive areas based on multispectral remote sensing, characterized in that: The following steps are involved: S10, multispectral data acquisition: using a drone equipped with a multispectral camera to acquire multispectral data of the target ecologically sensitive area, wherein the multispectral camera includes at least visible light and near infrared bands; S20, data preprocessing: preprocessing the acquired multispectral data, including radiation correction, image registration and orthophoto generation, wherein the radiation correction converts the original DN value into reflectance data; S30, vegetation coverage extraction: extracting vegetation coverage using normalized vegetation index based on pre-processed multispectral data; S40, Ecological vulnerability assessment: Combine vegetation coverage data with environmental factors such as topography and climate in ecologically sensitive areas to construct an ecological vulnerability assessment model for quantitative assessment; S50, Dynamic monitoring and early warning: Based on the continuous acquisition and processing of multispectral data, dynamic monitoring of ecologically sensitive areas is achieved, and early warning thresholds are set to trigger the early warning mechanism.

2. The method for monitoring ecologically sensitive areas based on multispectral remote sensing according to claim 1, characterized in that: In step S10, a Changguang Yuchen MS600V2 series multispectral camera is used for data collection. The spectral camera is equipped with a high-precision positioning system, which can record the shooting location information in real time. According to the characteristics of the ecologically sensitive area and the monitoring objectives, a detailed shooting plan is formulated, including parameters such as shooting time, angle, and height, to ensure that the collected multispectral data is comprehensive and representative.

3. The method for monitoring ecologically sensitive areas based on multispectral remote sensing according to claim 1, characterized in that: The radiation correction in step S20 adopts a linear correction model to improve the correction accuracy. Specifically, the radiation correction formula is as follows: Where R is the corrected reflectance, DN is the original DN value, and DN dark is the dark current value, DN white is the white balance value, k and c are correction coefficients, which are obtained by fitting known standard reflectance data.

4. The method for monitoring ecologically sensitive areas based on multispectral remote sensing according to claim 1, characterized in that: In step S30, the normalized vegetation is calculated using an index formula, and the vegetation coverage level is divided in combination with the actual situation of the ecologically sensitive area. Specifically, the index calculation formula is as follows: Among them, ρ NIR is the reflectivity in the near-infrared band, ρ RED is the reflectivity of the red band, ∈ is a very small positive number used to avoid the denominator being zero.

5. The method for monitoring ecologically sensitive areas based on multispectral remote sensing according to claim 1, characterized in that: In step S40, the ecological vulnerability assessment model adopts a random forest algorithm and combines it with a geographic information system for spatial analysis and visualization. The accuracy and generalization ability of the model are improved by constructing multiple decision trees and integrating their output results. Historical monitoring data and the actual ecological conditions of ecologically sensitive areas are used as training samples. The parameters of the random forest model are optimized and adjusted through cross-validation and grid search technology. The assessment results are displayed in the form of a map, and the ecological vulnerability levels are represented on the map in different colors or patterns through spatial interpolation and visualization technology.

6. The method for monitoring ecologically sensitive areas based on multispectral remote sensing according to claim 1, characterized in that: In the training process of the machine learning algorithm, in addition to using historical monitoring data and the actual ecological conditions of ecologically sensitive areas as training samples, external data sources are also introduced as auxiliary information. These external data sources include but are not limited to meteorological data, terrain data, and soil type data. These external data sources are integrated with multispectral data through feature selection and feature fusion methods to construct a more comprehensive and rich feature space. At the same time, data enhancement techniques such as rotation, scaling, and translation are used to pre-process the training samples to improve the robustness and generalization ability of the model.

7. The method for monitoring ecologically sensitive areas based on multispectral remote sensing according to claim 1, characterized in that: The dynamic monitoring in step S50 includes not only regularly acquiring, processing and analyzing multispectral data, but also real-time monitoring of changes in ecologically sensitive areas. A detailed monitoring plan is formulated based on the characteristics of the ecologically sensitive areas and monitoring needs, including monitoring frequency, monitoring range, and parameters of monitoring indicators. Real-time collection, transmission and processing of multispectral data are achieved by building a real-time monitoring system and a data processing platform. At the same time, time series analysis and trend prediction methods are used to dynamically analyze and predict real-time monitoring data to promptly discover and warn of abnormal changes in ecologically sensitive areas.

8. The method for monitoring ecologically sensitive areas based on multispectral remote sensing according to claim 7, characterized in that: The warning threshold set in the early warning mechanism is dynamically adjusted according to the actual situation of the ecologically sensitive area and the ecological protection goals. The ecological status of the ecologically sensitive area is reflected by constructing key indicators of the ecological vulnerability index and the vegetation coverage index. Reasonable warning thresholds are set according to the historical data and changing trends of these indicators. When the monitoring data reaches or exceeds the warning threshold, the system automatically triggers the warning signal and reminds the monitoring personnel by SMS, email or system notification. At the same time, the early warning mechanism also has the function of automatic adjustment and optimization, and can dynamically adjust and optimize the warning threshold according to the real-time monitoring data and the actual situation of the ecologically sensitive area.

9. The method for monitoring ecologically sensitive areas based on multispectral remote sensing according to claim 8, characterized in that: The setting of the early warning threshold not only takes into account the actual conditions of the ecologically sensitive areas and the qualitative factors of the ecological protection objectives, but also introduces quantitative assessment methods and indicators. By constructing an ecological vulnerability assessment model and a vegetation coverage assessment model, the ecological conditions of the ecologically sensitive areas are quantitatively assessed, and reasonable early warning thresholds are set according to the assessment results and the ecological protection objectives. At the same time, the setting of the early warning threshold also takes into account the characteristics and differences of different ecologically sensitive areas, as well as the severity and urgency of different ecological problems. The scientificity and rationality of the early warning threshold are ensured through comprehensive analysis and evaluation.

Images