Analysis method and system for population distribution of mangrove forest ecosystem

Abstract

The invention relates to the technical field of data processing, and discloses a mangrove forest ecosystem population distribution analysis method and system. The method comprises the following steps: acquiring a high-resolution remote sensing image and field data, and establishing a multi-dimensional database through an image enhancement algorithm; a random forest algorithm is used to identify community types, and plant position information is extracted through deep learning; calculating indexes such as relative multi-degree to obtain characteristic parameters; key patches are obtained through landscape index analysis; building a prediction model by combining the environmental parameters; and constructing an evaluation model based on analytic hierarchy process, and generating a management decision suggestion. According to the application, an analysis method for mangrove forest population distribution prediction and intelligent management decision is realized, so that the scientificity of mangrove forest ecosystem research and the effectiveness of management are improved.

Description

Technical Field

[0001] The present application relates to the field of data processing, and in particular to an analysis method and system for population distribution in mangrove ecosystems. Background Art

[0002] Mangroves are a type of woody plant community distributed in the intertidal zones of tropical and subtropical coasts. They play an important role in maintaining the stability of coastal ecosystems and protecting coastal biodiversity. Existing research on mangrove ecosystems mainly focuses on two aspects: remote sensing monitoring and sample plot surveys. Remote sensing monitoring obtains spatial distribution information of mangroves through satellite image interpretation, while sample plot surveys collect data such as species composition and growth status through field surveys. Currently, commonly used research methods include using remote sensing to monitor mangrove area and landscape changes, studying community structure and species diversity through sample plot surveys, and evaluating ecosystem health based on ecological indicators.

[0003] However, the existing research methods have the following shortcomings: First, remote sensing monitoring and sample plot surveys are often carried out separately, lacking an effective data integration mechanism, which makes it impossible to fully utilize the advantages of multi-source data; second, in the data analysis process, most of them use a single statistical method or ecological indicator, which makes it difficult to fully reflect the complexity of the mangrove ecosystem; third, the existing assessment methods mostly remain at the descriptive analysis level, lacking the ability to deeply explore and predict the distribution patterns of mangrove populations; finally, existing management decisions are mostly based on empirical judgments, lacking systematic and intelligent decision support tools. Summary of the invention

[0004] The present application provides an analysis method and system for the population distribution of a mangrove ecosystem, which is used to realize an analysis method for mangrove population distribution prediction and intelligent management decision-making, so as to improve the scientific nature of mangrove ecosystem research and the effectiveness of management.

[0005] In the first aspect, the present application provides an analysis method for the population distribution of a mangrove ecosystem, the analysis method for the population distribution of a mangrove ecosystem comprising: obtaining high-resolution remote sensing images through a remote sensing image collector, and collecting field survey data and environmental parameters in the study area, preprocessing the high-resolution remote sensing images through an image enhancement algorithm, and establishing a multidimensional database containing spatial information, species information and environmental parameters; using a random forest classification algorithm to identify community types of data in the multidimensional database, extracting location information of individual mangrove plants from the high-resolution remote sensing images through a deep learning target detection algorithm, and generating a community distribution vector data layer; based on the community distribution vector data layer, calculating the relative abundance, relative frequency and relative significance of mangrove plants in the woody layer to obtain important value indicators, and performing community distribution analysis on the relative abundance, relative frequency and relative significance of mangrove plants in the herbaceous layer. The relative cover and relative height of the objects are calculated to obtain the total dominance index, and the spatial distribution characteristic parameters of the population are calculated through a spatial statistical algorithm; according to the spatial distribution characteristic parameters of the population, the landscape pattern analysis is performed using a landscape index calculation model to obtain the maximum patch index, clumping index, similar adjacent percentage and patch aggregation index, and a connectivity analysis model is established based on a graph theory algorithm to obtain key patch identification results; the key patch identification results are correlated with environmental parameters, an environmental factor influence model is established through a gradient boosting algorithm, and uncertainty analysis is performed using a Bayesian network algorithm to obtain population distribution prediction results; based on the population distribution prediction results, combined with the important value index, total dominance index and landscape index, a health assessment model is constructed through a hierarchical analysis method, a knowledge graph reasoning system is established, and decision-making recommendations for mangrove ecosystem management are generated.

[0006] In a second aspect, the present application provides an analysis system for the population distribution of a mangrove ecosystem, the analysis system for the population distribution of a mangrove ecosystem comprising:

[0007] An acquisition module is used to acquire high-resolution remote sensing images through a remote sensing image collector, collect field survey data and environmental parameters in the study area, pre-process the high-resolution remote sensing images through an image enhancement algorithm, and establish a multidimensional database containing spatial information, species information and environmental parameters;

[0008] An identification module is used to identify the community type of the data in the multidimensional database using a random forest classification algorithm, extract the location information of a single mangrove plant from the high-resolution remote sensing image using a deep learning target detection algorithm, and generate a community distribution vector data layer;

[0009] A calculation module is used to calculate the relative abundance, relative frequency and relative significance of the mangrove plants in the woody layer based on the community distribution vector data layer to obtain the important value index, calculate the relative coverage and relative height of the mangrove plants in the herbaceous layer to obtain the total dominance index, and calculate the population spatial distribution characteristic parameters through a spatial statistical algorithm;

[0010] An analysis module is used to perform landscape pattern analysis using a landscape index calculation model based on the population spatial distribution characteristic parameters, obtain the maximum patch index, clumping index, similar adjacency percentage and patch aggregation index, establish a connectivity analysis model based on a graph theory algorithm, and obtain key patch identification results;

[0011] An association module is used to associate the key patch identification results with environmental parameters, establish an environmental factor impact model through a gradient boosting algorithm, and use a Bayesian network algorithm to perform uncertainty analysis to obtain population distribution prediction results;

[0012] A generation module is used to construct a health assessment model through a hierarchical analysis method based on the population distribution prediction results, combined with the important value indicators, total dominance indicators and landscape index, establish a knowledge graph reasoning system, and generate mangrove ecosystem management decision-making recommendations.

[0013] In the technical solution provided in the present application, high-resolution remote sensing images are obtained through a remote sensing image collector, and combined with the collection of field survey data and environmental parameters, the effective fusion of multi-source data is realized, and the integrity and accuracy of the data are improved; the image enhancement algorithm is used to pre-process the remote sensing images, and a multidimensional database containing spatial information, species information and environmental parameters is established, which provides comprehensive data support for subsequent analysis; the random forest classification algorithm and the deep learning target detection algorithm are used to process the data, which not only realizes the accurate identification of community types, but also can extract the location information of individual mangrove plants, significantly improving the accuracy of spatial distribution analysis; by calculating various indicators of mangrove plants in the woody layer and the herbaceous layer, combined with spatial statistical algorithms, the quantitative expression of population distribution characteristics is realized; the landscape index calculation model is used to analyze the landscape pattern, and the connectivity analysis model is established through the graph theory algorithm, so that the identification of key patches is more accurate and objective; the gradient boosting algorithm and the Bayesian network algorithm are used to analyze the impact of environmental factors, which greatly improves the accuracy of population distribution prediction; finally, the hierarchical analysis method and the knowledge graph reasoning system are used to realize the intelligent and scientific management decision-making of mangrove ecosystems. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0015] Figure 1 A schematic diagram of an example of an analysis method for the distribution of mangrove ecosystem populations in an embodiment of the present application;

[0016] Figure 2 This is a schematic diagram of an embodiment of a system for analyzing the population distribution of a mangrove ecosystem in an embodiment of the present application. DETAILED DESCRIPTION

[0017] The present application embodiment provides a method and system for analyzing the distribution of populations in a mangrove ecosystem. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described here can be implemented in an order other than the content illustrated or described here. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0018] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 In the embodiments of the present application, an embodiment of the method for analyzing the population distribution of mangrove ecosystems includes:

[0019] Step S101, obtaining high-resolution remote sensing images through a remote sensing image collector, collecting field survey data and environmental parameters in the study area, preprocessing the high-resolution remote sensing images through an image enhancement algorithm, and establishing a multidimensional database containing spatial information, species information and environmental parameters;

[0020] Step S102: using a random forest classification algorithm to identify community types of data in a multidimensional database, extracting location information of individual mangrove plants from high-resolution remote sensing images through a deep learning target detection algorithm, and generating a community distribution vector data layer;

[0021] Step S103: Based on the community distribution vector data layer, the relative abundance, relative frequency and relative significance of the woody layer mangrove plants are calculated to obtain the important value index; the relative coverage and relative height of the herbaceous layer mangrove plants are calculated to obtain the total dominance index; and the population spatial distribution characteristic parameters are calculated by a spatial statistical algorithm;

[0022] Step S104: according to the population spatial distribution characteristic parameters, the landscape index calculation model is used to perform landscape pattern analysis, obtain the maximum patch index, clumping index, similarity adjacency percentage and patch aggregation index, establish a connectivity analysis model based on the graph theory algorithm, and obtain key patch identification results;

[0023] Step S105, performing correlation analysis on the key patch identification results and environmental parameters, establishing an environmental factor impact model through a gradient boosting algorithm, performing uncertainty analysis using a Bayesian network algorithm, and obtaining population distribution prediction results;

[0024] Step S106: Based on the population distribution prediction results, combined with the important value index, the total dominance index and the landscape index, a health assessment model is constructed through the hierarchical analysis method, a knowledge graph reasoning system is established, and decision-making recommendations for mangrove ecosystem management are generated.

[0025] It is understandable that the execution subject of the present application may be an analysis system for the distribution of mangrove ecosystem populations, or a terminal or a server, which is not specifically limited here. The present application embodiment is described by taking a server as the execution subject as an example.

[0026] Specifically, high-resolution remote sensing images were obtained through a remote sensing image collector. The collector mainly used the GeoEye-1 high-resolution satellite to perform multispectral imaging of the study area and obtain remote sensing data including visible light bands and near-infrared bands. At the same time, multiple sample strips were laid out in the study area, and sample plots were evenly set on each sample strip. The sample plots included two specifications: 10m×10m and 20m×5m. The tree height, breast diameter, crown width, species type, quantity, and coverage data of mangrove plants were recorded in the sample plots as field survey data, and environmental parameters including tide level, salinity, and soil pH were collected. The high-resolution remote sensing images obtained were preprocessed by geometric correction, radiation calibration, atmospheric correction, and image fusion to construct a multidimensional database containing spatial information, species information, and environmental parameters. Subsequently, the random forest classification algorithm was used to identify the community type of the data in the multidimensional database. The random forest algorithm was classified by constructing multiple decision trees. Each decision tree was trained using a randomly selected feature subset, and the classification result was finally determined by voting. In the specific processing process, spectral features and texture features are used as input variables, and a classification model is established through training samples to identify different types of mangrove communities. At the same time, a deep learning target detection algorithm is used to extract the location information of individual mangrove plants from high-resolution remote sensing images. The deep learning target detection algorithm uses a convolutional neural network structure to extract image features through multi-layer convolution and pooling operations, and finally locates the position of individual plants through bounding box regression. The community type recognition results are integrated with the location information of individual plants to generate a community distribution vector data layer.

[0027] Based on the generated community distribution vector data layer, the population dominance analysis of woody mangrove plants was conducted to calculate the relative abundance (the percentage of individuals of a certain plant to the total number of individuals of all species), relative frequency (the percentage of the frequency of a certain plant in each sample plot to the sum of the frequencies of all species) and relative significance (the percentage of the cross-sectional area at breast height of a certain plant to the sum of the cross-sectional areas of all species). The sum of the three was used to obtain the important value index. For the herbaceous mangrove plants, the relative cover (the ratio of the coverage area of ​​a certain plant to the total area) and relative height (the ratio of the height of a certain plant to the average height of all species) were calculated to obtain the total dominance index. The spatial distribution characteristic parameters of the population, including the variance / mean ratio, the negative binomial distribution parameter, the clustering index and the average crowding index, were calculated by spatial statistical algorithms to quantify the spatial distribution pattern of the population. Based on the calculated population spatial distribution characteristic parameters, the landscape index calculation model was used to analyze the landscape pattern. First, the maximum patch index is calculated, that is, the proportion of the largest patch area to the total area; secondly, the clumping index is calculated to reflect the degree of aggregation of patches of the same type; then the similarity adjacency percentage is calculated to indicate the degree of adjacency of patches of the same type; finally, the patch aggregation index is calculated to characterize the aggregation of landscape types. A connectivity analysis model is established based on graph theory algorithms, mangrove patches are regarded as nodes, and the connection relationship between patches is regarded as edges. Key patches are identified by calculating indicators such as node degree and centrality.

[0028] The identified key patches were analyzed for association with environmental parameters, and the environmental factor impact model was established through the gradient boosting algorithm. The gradient boosting algorithm constructed multiple weak learners in an iterative manner. Each iteration trained the residuals of the previous round of predictions, and finally combined the results of all weak learners to obtain a strong learner. In the modeling process, environmental parameters were used as feature variables, and mangrove distribution characteristics were used as target variables. The influence of environmental factors on mangrove distribution was obtained through training. The Bayesian network algorithm was used for uncertainty analysis. The Bayesian network represented the conditional dependency relationship between variables through a directed acyclic graph, calculated the conditional probability distribution of each node, and finally obtained the population distribution prediction results. Based on the prediction results, combined with the important value index, the total dominance index and the landscape index, a health assessment model was constructed through the hierarchical analysis method. The hierarchical analysis method hierarchically constructed a hierarchical structure for the evaluation indicators, constructed a judgment matrix through pairwise comparison, calculated the characteristic vector to obtain the weight of each indicator, and finally comprehensively evaluated the health status of the mangrove ecosystem. A knowledge graph reasoning system was established to construct the various elements and relationships of the mangrove ecosystem into a semantic network, and to perform knowledge reasoning through inference rules to provide decision-making recommendations for the management of the mangrove ecosystem.

[0029] For example, remote sensing image data of the area was acquired, and four main community types were identified after preprocessing: Sonneratia apetala community, Kandelia candel community, Pteris solani community, and Acanthus sphaerocephala community. The population dominance of woody layer plants was calculated, and the importance value of Sonneratia apetala was the highest, reaching 105.39%, indicating that it is the dominant species in the region. Spatial distribution characteristics analysis showed that the aggregation of the Laguan wood population was the highest. Landscape pattern analysis found that the Sonneratia apetala community accounted for the largest area, reaching 82.01%. Environmental factor analysis showed that the species was most concentrated in the middle and low tidal flats. Based on these analysis results, the generated management decision recommendations include: strengthening the planting of native mangrove plants, moderately controlling the scale of introduction of Sonneratia apetala, and maintaining the species diversity and ecological balance of the mangrove ecosystem.

[0030] In the embodiment of the present application, high-resolution remote sensing images are obtained through a remote sensing image collector, and combined with the collection of field survey data and environmental parameters, effective fusion of multi-source data is achieved, and the integrity and accuracy of the data are improved; the remote sensing images are pre-processed by an image enhancement algorithm, and a multidimensional database containing spatial information, species information and environmental parameters is established, which provides comprehensive data support for subsequent analysis; the data is processed using a random forest classification algorithm and a deep learning target detection algorithm, which not only realizes the accurate identification of community types, but also can extract the location information of individual mangrove plants, significantly improving the accuracy of spatial distribution analysis; by calculating various indicators of mangrove plants in the woody layer and the herbaceous layer, combined with spatial statistical algorithms, the quantitative expression of population distribution characteristics is achieved; the landscape index calculation model is used to analyze the landscape pattern, and a connectivity analysis model is established through a graph theory algorithm, so that the identification of key patches is more accurate and objective; the gradient boosting algorithm and the Bayesian network algorithm are used to analyze the impact of environmental factors, which greatly improves the accuracy of population distribution prediction; finally, the hierarchical analysis method and the knowledge graph reasoning system are used to realize the intelligent and scientific management decision-making of the mangrove ecosystem.

[0031] In a specific embodiment, the process of executing step S101 may specifically include the following steps:

[0032] (1) Imaging the study area through a remote sensing image collector to obtain high-resolution remote sensing images;

[0033] (2) Using random sampling methods, sample strips were set up in the study area, and sample plots were arranged in each sample strip to collect field survey data and environmental parameters;

[0034] (3) Use the geometric correction algorithm to spatially register the high-resolution remote sensing image, use the radiometric calibration algorithm to perform spectral correction, and use the atmospheric correction algorithm to eliminate the atmospheric effect to obtain the corrected high-resolution remote sensing image;

[0035] (4) Performing image enhancement processing on the corrected high-resolution remote sensing image through a histogram equalization algorithm to obtain an enhanced high-resolution remote sensing image;

[0036] (5) Establish data standardization tables and determine data storage formats based on enhanced high-resolution remote sensing images, field survey data, and environmental parameters;

[0037] (6) Through data structuring processing, the data standardization table is converted into a multidimensional database containing spatial information, species information, and environmental parameters.

[0038] Specifically, the multispectral remote sensing images of the study area were obtained through the remote sensing image collector carried by the GeoEye-1 satellite. The remote sensing image collector contains multiple sensor bands, which collect the reflection information of the visible light and near-infrared bands respectively. The visible light band can reflect the chlorophyll content of vegetation, while the near-infrared band is more sensitive to the biomass and canopy structure of vegetation. High-resolution remote sensing images are obtained through multi-band combination. A random sampling method is used to set up multiple sample strips in the study area. The setting of the sample strips fully considers the changes in the tidal gradient. A fixed-size sample plot is arranged on each sample strip, including two specifications of 10m×10m and 20m×5m. The sample plots record various growth parameters of mangrove plants, such as tree height, breast diameter, and crown width data, and also record community parameters such as species type, quantity and coverage. While sampling the sample plots, environmental parameter data are collected, including ecological environmental factors such as tidal elevation, soil salinity, and pH value.

[0039] The acquired high-resolution remote sensing images are first subjected to geometric correction. The conversion relationship between the image coordinates and the geographic coordinates is established by selecting ground control points to eliminate the geometric deformation of the image. Then, the digital number value of the image is converted into the actual spectral reflectance using the radiometric calibration algorithm to correct the difference in spectral response. Then, the atmospheric correction algorithm is used to remove the influence of atmospheric scattering and absorption on the image quality, and the corrected image reflecting the true spectral characteristics of the ground object is obtained. The corrected high-resolution remote sensing image is subjected to image enhancement processing, and the grayscale distribution of the image is adjusted using the histogram equalization algorithm. Histogram equalization redistributes the pixel grayscale values ​​so that the entire grayscale range is fully utilized, the image contrast is enhanced, and the texture characteristics and boundary information of the mangrove canopy are highlighted, thereby obtaining an enhanced image with better visual effects.

[0040] The enhanced high-resolution remote sensing images are integrated with field survey data and environmental parameters to establish a unified data standardization table. Data standardization includes three aspects: format standardization, unit standardization, and scale standardization. Format standardization converts data from different sources into the same data format, unit standardization ensures that all measurement data use a unified unit of measurement, and scale standardization normalizes data of different dimensions. Through data structuring, the standardized table is converted into a multidimensional database. Data structuring establishes the relationship between spatial information, species information, and environmental parameters, forming a hierarchical data storage system. The multidimensional database is stored in a matrix form, and each data record contains information in multiple dimensions such as spatial coordinates, species attributes, and environmental variables.

[0041] For example, a multispectral image of the area is obtained through a remote sensing image collector, which contains data from four bands: blue, green, red, and near-infrared. Three sample strips perpendicular to the coastline are set up in the study area according to the tidal gradient. Sample plots are arranged at intervals of 25 meters on each sample strip to record the species composition and growth status within the sample plot. The coordinate transformation relationship is established through spatial control points to complete the geometric correction of the image. The corrected image is subjected to histogram equalization processing to make the spectral characteristics of different community types more obvious. Finally, all the data are integrated into a multidimensional database, and a complete data record including spatial location, community characteristics, and environmental factors is established to provide data support for subsequent analysis. This database not only records the specific location and species composition of each sample plot, but also contains the environmental parameter information of the location, realizing the effective integration of multi-source data.

[0042] In a specific embodiment, the process of executing step S102 may specifically include the following steps:

[0043] (1) Using feature extraction algorithms to filter spectral features and texture features in a multidimensional database to obtain community feature data;

[0044] (2) Select training samples based on community characteristic data, substitute the training samples into the random forest classification algorithm, and obtain community type identification parameters;

[0045] (3) Filtering the community distribution information in the multidimensional database according to the community type identification parameters to form community type distribution data;

[0046] (4) Input high-resolution remote sensing images into the deep learning target detection algorithm, and generate target detection feature maps after processing by the convolution layer;

[0047] (5) Calculate the boundary range of mangrove plants from the target detection feature map, extract the location information of individual mangrove plants, and establish the coordinate data of individual plants;

[0048] (6) Combine community type distribution data and individual plant coordinate data for geospatial processing to generate a community distribution vector data layer.

[0049] Specifically, effective features are extracted from the multidimensional database through feature extraction algorithms. Feature extraction includes two aspects: spectral features and texture features. Spectral features refer to the reflectance characteristics of different mangrove plants in various bands, including the reflectance values ​​of visible light bands and near-infrared bands, while texture features describe the spatial structural characteristics of the mangrove canopy, such as mean, variance, entropy and other statistics. These features are screened by feature extraction algorithms to obtain feature combinations that can effectively distinguish different community types and form community feature data. Based on the obtained community feature data, representative samples are selected as training data. The selection of training samples needs to cover various community types in the study area, including Sonneratia apetala community, Kandelia candel community, Acanthus elegans community, etc. These training samples are input into the random forest classification algorithm. The random forest algorithm constructs multiple decision trees, each tree is trained using a randomly selected feature subset, and finally the classification results are determined by voting to obtain community type recognition parameters.

[0050] The community type identification parameters obtained are used to screen the community distribution information in the multidimensional database. During the screening process, the characteristic value of each spatial location is matched with the trained classifier to determine the community type of the location, thereby forming complete community type distribution data. These data contain the community type identification and spatial coordinate information of each location point.

[0051] The high-resolution remote sensing image is input into the deep learning target detection algorithm for processing. The algorithm uses a convolutional neural network structure to extract the hierarchical features of the image through multi-layer convolution operations. In the convolution layer processing process, convolution kernels of different scales are first used to extract local features, and then feature dimensionality reduction is performed through the pooling layer, and finally a feature map containing target location and category information is generated.

[0052] According to the target detection feature map, the boundary range of the mangrove plants is calculated. The boundary calculation uses edge detection and contour extraction algorithms to accurately locate the boundary position of each mangrove plant. The spatial position of a single mangrove plant is extracted through boundary information, the coordinates of its center point are determined, and a complete single plant coordinate database is established. These coordinate data accurately record the spatial distribution position of each mangrove plant in the study area. The community type distribution data and the single plant coordinate data are processed geospatially, and the community type to which each plant belongs is determined through spatial overlay analysis, and a hierarchical spatial data structure is established, and finally a community distribution vector data layer is generated. This layer contains complete mangrove spatial distribution information, including both the distribution pattern at the community scale and the precise location of a single plant.

[0053] For example, first, feature extraction is performed on the acquired multispectral remote sensing images. In the near-infrared band, the Sonneratia apetala community shows a higher reflectance value, while in the visible red band it shows a lower reflectance value, which is significantly different from other community types. At the same time, by calculating the grayscale co-occurrence matrix of the image and extracting texture features, it is found that the Sonneratia apetala community has a higher homogeneity and a lower entropy value. These features are combined to form a training sample and input into the random forest classifier for training. After the training is completed, the classifier is used to identify the community type of the entire study area and generate community distribution data. At the same time, a deep learning target detection algorithm is used to process high-resolution images. The algorithm can accurately identify the location of a single mangrove plant, especially for areas with high growth density, and can also accurately distinguish the boundaries of adjacent plants. The spatial range of each plant is determined by the boundary extraction algorithm, and a coordinate database is established. Finally, the community distribution data is combined with the individual plant location data to generate a complete spatial distribution layer, which clearly shows the spatial distribution pattern of different community types and the distribution characteristics of individual plants within each community.

[0054] In a specific embodiment, the process of executing step S103 may specifically include the following steps:

[0055] (1) Extract the number of mangrove plant individuals in the woody layer from the community distribution vector data layer, calculate the species occurrence frequency for each sample plot, and form the relative abundance;

[0056] (2) Count the number of occurrences of woody layer mangrove plants in the sample plots, divide the number of occurrences by the total number of sample plots, and generate the relative frequency;

[0057] (3) Read the DBH values ​​of the woody mangrove plants, calculate the sum of the cross-sectional areas, perform normalization, and output the relative significance;

[0058] (4) Combine relative abundance, relative frequency and relative significance to perform weighted summation to obtain the importance value index;

[0059] (5) Extract the coverage area and height information of mangrove plants in the herbaceous layer from the community distribution vector data layer, calculate the ratio and then find the average value to obtain the total dominance index;

[0060] (6) Select species distribution points in the community distribution vector data layer, perform spatial statistical operations using the variance-mean ratio method, negative binomial distribution parameter method, and clustering index method, and calculate the population spatial distribution characteristic parameters.

[0061] Specifically, the data of mangrove plants in the woody layer are extracted from the community distribution vector data layer. This data records the specific attributes of each mangrove plant in the woody layer, including species type, location coordinates, and growth parameters. Statistics are performed on individual plants in each sample plot, and the number of individuals of each species in the sample plot is recorded to obtain the frequency of occurrence of the species. The calculation process of relative abundance is to divide the number of individuals of a certain species by the total number of individuals of all species, reflecting the numerical advantage of the species in the community. After obtaining the relative abundance data, the distribution of mangrove plants in the woody layer in all sample plots is further counted. By recording the number of times each species appears in different sample plots and dividing this number of occurrences by the total number of sample plots surveyed, the relative frequency data is obtained. The relative frequency reflects the extent of the spatial distribution of a species. The larger the value, the more widely distributed the species is.

[0062] For mangrove plants in the woody layer, it is also necessary to read the value of their diameter at breast height for analysis. The diameter at breast height refers to the diameter of a tree measured 1.3 meters above the ground. The cross-sectional area of ​​the tree can be calculated based on the diameter at breast height. The cross-sectional area of ​​all individuals of each species is calculated, and the sum of the cross-sectional areas of a certain species is divided by the sum of the cross-sectional areas of all species, and the relative significance is obtained after normalization. The relative significance reflects the dominant position of a species in the community. The weighted summation operation of the three indicators of relative abundance, relative frequency and relative significance is performed to obtain the important value index. The important value index comprehensively reflects the status of the species in the community and is an important parameter for evaluating the importance of species. For different ecosystems, the weights of these three parameters can be adjusted according to actual conditions.

[0063] For the mangrove plants in the herbaceous layer, the coverage area and height information are extracted from the community distribution vector data layer. The coverage area refers to the projected area of ​​the herbaceous plant on the ground, which is obtained through field measurement or remote sensing image interpretation. The height information is the plant height data recorded through field measurement. The coverage area and height information are calculated by ratio, and the average value of multiple samples is obtained to obtain the total dominance index. The total dominance index reflects the dominance of the herbaceous layer plants in the community. Finally, the species distribution point data in the community distribution vector data layer are selected, and the spatial distribution characteristics are analyzed by three different statistical methods. The variance-mean ratio method judges the distribution pattern by calculating the ratio of the variance of the number of species individuals to the mean; the negative binomial distribution parameter method evaluates the spatial distribution pattern based on the frequency distribution characteristics of individual data; the clustering index law quantifies the distribution characteristics by calculating the spatial aggregation degree of species individuals. The comprehensive application of these three methods can comprehensively reflect the spatial distribution characteristics of the population.

[0064] For example: First, the distribution data of Sonneratia apetala was extracted from the community distribution vector data layer, and its individual number was counted in 20 plots. Through statistics, it was found that Sonneratia apetala accounted for the largest proportion of the total number of individuals in all plots, and a higher relative abundance was calculated. Further analysis found that Sonneratia apetala was distributed in 15 plots, and its relative frequency was calculated. The diameter at breast height of each Sonneratia apetala was measured, and it was found that its average diameter at breast height was large, and the total cross-sectional area accounted for a significant proportion, resulting in a higher relative significance. The weighted sum of these three indicators confirmed that Sonneratia apetala had the highest important value index in the woody layer. For herbaceous layer plants such as Halophyta, the total dominance index was calculated by measuring their coverage area and plant height. Finally, the spatial distribution of Sonneratia apetala was analyzed. The calculation results of three spatial statistical methods showed that Sonneratia apetala showed significant clustered distribution characteristics in the midtidal zone, and this distribution characteristic was closely related to environmental factors such as tide level and salinity.

[0065] In a specific embodiment, the process of executing step S104 may specifically include the following steps:

[0066] (1) Patch area data were extracted from the population spatial distribution characteristic parameters, and the maximum patch area was divided by the total area of ​​the study area to obtain the maximum patch index;

[0067] (2) Extract the adjacency information of each patch in the population spatial distribution characteristic parameters, and generate the clumping index by calculating the spatial adjacency probability;

[0068] (3) Statistics are collected for the patch adjacency information in the population spatial distribution characteristic parameters, the similar adjacency ratio is calculated based on the total boundary length, and the similar adjacency percentage is output;

[0069] (4) Filter the patch pair data from the population spatial distribution characteristic parameters, calculate the ratio of the number of adjacent patches to the maximum possible number of adjacent patches, and obtain the patch aggregation index;

[0070] (5) Based on the maximum patch index, clumping index, similarity adjacency percentage, and patch aggregation index, an adjacency matrix was constructed, and the node connectivity was calculated using a graph theory algorithm;

[0071] (6) Network analysis is performed on the node connectivity and patch spatial distribution data to obtain key patch identification results.

[0072] Specifically, the patch area data were extracted from the population spatial distribution characteristic parameters. The patch refers to a continuous spatial unit with the same landscape type. The largest continuous patch area was identified through spatial analysis tools, and the area was divided by the total area of ​​the entire study area to calculate the maximum patch index. The maximum patch index reflects the relative size of the dominant patch in the landscape. The larger the value, the more dominant the patch type in the landscape.

[0073] For the adjacency relationship analysis between patches, it is necessary to extract the adjacency information of each patch in the population spatial distribution characteristic parameters. The adjacency information includes the spatial relationship data of each patch and its adjacent patches. The clumping index is generated by calculating the spatial adjacency probability, and its calculation formula is:

[0074]

[0075] Among them, α ij represents the adjacent length between patch i and patch j, β ij represents the similarity coefficient of patch types, γ represents the total boundary length, δ represents the number of landscape types, λ represents the normalization coefficient, n represents the total number of patches, and m represents the number of patches adjacent to patch i. The clumping index reflects the degree of spatial aggregation of patches of the same type.

[0076] Continue to perform statistical analysis on the patch adjacency information in the population spatial distribution characteristic parameters, calculate the ratio of the boundary length within each landscape type to the total boundary length of the type, and obtain the similar adjacency percentage. In the data processing process, first identify the boundary lines of all patches, count the common boundary lengths of each patch and the adjacent patches, then add up the boundary lengths between patches of the same type, and finally divide by the total boundary length of patches of this type. When screening patch pair data from the population spatial distribution characteristic parameters, it is necessary to identify the adjacency relationship of each patch. A patch pair refers to two patch units that are adjacent to each other in space. Calculate the ratio of the current number of adjacent patches to the theoretically maximum number of adjacent patches to obtain the patch aggregation index. The patch aggregation index reflects the degree of spatial aggregation of landscape types.

[0077] Based on the maximum patch index, clumping index, similar adjacency percentage and patch aggregation index calculated above, an adjacency matrix is ​​constructed. The adjacency matrix is ​​a two-dimensional array used to represent the connection relationship between patches. The element values ​​in the matrix reflect the connection strength between patches. The node connectivity is calculated by graph theory algorithm. Each patch is regarded as a node in the network, and the connection relationship between patches is regarded as an edge. The network characteristic parameters such as degree and centrality of each node are calculated. Finally, the node connectivity and patch spatial distribution data are used for network analysis. By evaluating the importance of nodes, key patches are identified. Key patches refer to patch units that have an important connection role in the entire ecological network. These patches are of great significance to maintaining the connectivity and stability of the ecosystem.

[0078] For example, the mangrove communities in the study area were first divided into patches, and different types of patches such as the Sonneratia apetala community, the Kandelia candel community, and the Pteris solani community were identified through remote sensing image interpretation. After area statistics, it was found that the Sonneratia apetala community constituted the largest continuous patch. By analyzing the spatial relationship between patches, it was found that the Sonneratia apetala community patches had more boundary contacts with other types of patches, showing a higher clumping index. When calculating the similar adjacency percentage, it was found that the patches of the same type of community tended to be adjacent to each other, especially in the mid-tidal zone, where the patches of the Sonneratia apetala community showed obvious spatial aggregation characteristics. The analysis of patch aggregation showed that patches of different community types showed different aggregation degrees in spatial distribution, among which the Sonneratia apetala community had the highest aggregation degree. By constructing an adjacency matrix and conducting network analysis, the patches that play a key role in the connectivity of the entire mangrove ecosystem were identified. These key patches are mainly distributed in the middle area of ​​the intertidal zone and play an important role in maintaining the spatial continuity of the mangrove ecosystem.

[0079] In a specific embodiment, the process of executing step S105 may specifically include the following steps:

[0080] (1) Align the key patch identification results with environmental parameters to generate patch-environment association data;

[0081] (2) Input the patch environment association data into the gradient boosting algorithm and distinguish the training data from the validation data by randomly splitting the data set;

[0082] (3) The gradient boosting algorithm is used to calculate the feature weights of the training data and establish an environmental factor impact model;

[0083] (4) Perform verification calculations on the environmental factor impact model based on the verification data and output environmental factor impact weight data;

[0084] (5) Construct a probability distribution network of environmental factors affecting weight data through the Bayesian network algorithm and calculate the conditional probability of each node;

[0085] (6) Convert the conditional probability into population distribution probability and combine it with the environmental factor influence model to obtain the population distribution prediction result.

[0086] Specifically, the spatial location information of each key patch is spatially superimposed with the environmental parameter data of the corresponding location, including tidal level, salinity, soil pH and other data. The environmental characteristic description of each patch is established through the spatial correspondence to form patch environmental association data. The generated patch environmental association data is input into the gradient boosting algorithm for processing. The gradient boosting algorithm is an iterative decision tree algorithm that improves the prediction accuracy by constructing multiple decision trees. First, the data set is randomly split, 70% of the data is used as training data, and 30% of the data is used as verification data to ensure the generalization ability of the model.

[0087] For the training data, the gradient boosting algorithm is used to calculate the feature weights, and the calculation formula is:

[0088]

[0089] Among them, W ij represents the influence weight of the i-th environmental factor on the j-th patch, ω k represents the weight coefficient of the kth positive impact factor, represents the intensity value of the kth positive factor affecting the plaque, θ m represents the weight coefficient of the mth negative impact factor, ψ ijm represents the intensity value of the mth negative factor affecting the plaque, ξ ij represents the normalized adjustment coefficient, p represents the number of positive factors, and q represents the number of negative factors. The environmental factor impact model is established through this calculation.

[0090] The environmental factor impact model was verified and calculated using the verification data. During the verification process, the difference between the model prediction value and the actual observation value was compared. The model parameters were optimized through repeated iterations, and the environmental factor impact weight data were finally output. These weight data reflect the degree of influence of different environmental factors on the distribution of mangroves. Based on the environmental factor impact weight data, a probability distribution network was constructed using the Bayesian network algorithm. The Bayesian network is a probabilistic graph model that represents the conditional dependency relationship between variables through a directed acyclic graph. In the network, each node represents an environmental factor, and the connection between nodes represents the mutual influence relationship between factors. The conditional probability of each node is obtained by calculation.

[0091] Finally, the conditional probability is converted into population distribution probability, and a comprehensive analysis is performed with the environmental factor impact model to obtain the prediction results of mangrove population distribution. The prediction results include the possibility of mangrove distribution in different regions and the main influencing factors.

[0092] For example, in an analysis of a mangrove reserve, the identified key patches are first spatially matched with environmental monitoring data. In the intertidal zone, each patch corresponds to a set of environmental parameter data, including the tidal elevation, soil salinity, pH value, etc. at that location. Through data matching, it was found that the community of Sonneratia apetala is mainly distributed in the mid-tidal zone, which has a specific combination of environmental characteristics. These matching data were input into the gradient boosting algorithm, and after training, it was found that the tidal elevation had the greatest impact on the community distribution, followed by soil salinity. In the verification stage, the community distribution predicted by the model was highly consistent with the actual observation results. Bayesian network analysis showed that the distribution probability of the Sonneratia apetala community is highest when the tidal elevation is within a certain range and the soil salinity is moderate.

[0093] In a specific embodiment, the process of executing step S106 may specifically include the following steps:

[0094] (1) Integrate the population distribution prediction results, important value index, total dominance index and landscape index into a set of evaluation indicators, and determine the indicator weights through principal component analysis;

[0095] (2) Establish a discriminant matrix for the evaluation index set, calculate the eigenvector through the hierarchical analysis method, and output the relative importance of the index;

[0096] (3) Substitute the relative importance of the indicators into the hierarchical analysis method to calculate the consistency ratio and generate the indicator weight correction value;

[0097] (4) Construct a health assessment model based on the indicator weight correction value and calculate the mangrove ecosystem health score;

[0098] (5) Input the mangrove ecosystem health score into the knowledge graph reasoning system and extract management rules through semantic analysis;

[0099] (6) Combine management rules with the diagnostic results of the health assessment model to generate decision-making recommendations for mangrove ecosystem management.

[0100] Specifically, multidimensional evaluation indicators such as population distribution prediction results, important value indicators, total dominance indicators and landscape index are integrated. These indicators reflect the characteristics of different aspects of mangrove ecosystems. The indicators are reduced in dimension through principal component analysis to determine the weight of each indicator. Principal component analysis calculates the correlation between indicators and combines the correlated indicators into new comprehensive indicators to avoid information overlap between indicators. A discriminant matrix is ​​established for the integrated set of evaluation indicators, and the hierarchical analysis method is used to calculate the indicator weights. The elements in the discriminant matrix represent the relative importance of different indicators. The relative importance of each indicator is obtained by calculating the eigenvector of the discriminant matrix. The relative importance reflects the degree of influence of each indicator in the assessment of ecosystem health.

[0101] Substitute the relative importance of the indicators into the hierarchical analysis method to calculate the consistency ratio, and the calculation formula is:

[0102]

[0103] Among them, ρ ij represents the relative importance value of the i-th indicator and the j-th indicator, η ij represents the weight adjustment factor, σ ij represents the correlation coefficient between indicators, ∈ represents the maximum characteristic root, τ represents the matrix dimension, υ represents the random consistency index, ζ represents the correction coefficient, and n represents the total number of indicators. By calculating the consistency ratio, the correction value of the indicator weight is generated to ensure the scientific nature of the evaluation results.

[0104] Based on the corrected indicator weights, a mangrove ecosystem health assessment model was constructed. The model comprehensively considers multiple aspects such as population distribution, community structure, and landscape pattern, and obtains the health score of the mangrove ecosystem through weighted calculation. The health score reflects the overall status of the mangrove ecosystem.

[0105] The calculated ecosystem health score is input into the knowledge graph reasoning system, and management rules are extracted through semantic analysis technology. The knowledge graph contains the associations between the components of the mangrove ecosystem and the management experience under different health conditions. Through semantic analysis, the health assessment results are matched with the existing management knowledge to extract applicable management rules. Finally, the extracted management rules are integrated with the diagnostic results of the health assessment model to generate targeted mangrove ecosystem management decision-making recommendations. These recommendations include specific measures such as species protection, habitat restoration, and community optimization.

[0106] For example, first collect the data of various evaluation indicators in the area. From the results of population distribution prediction, there are multiple species communities such as Sonneratia apetala, Kandelia candel, and Acanthus acanthus in the area. Through the analysis of important value indicators, it is found that Sonneratia apetala is dominant in the community. The total dominance index shows that the herbaceous layer is dominated by Halophyta, and the landscape index analysis shows that the mangroves in this area show obvious patchy distribution characteristics. After determining the weights of these indicators through principal component analysis, a discriminant matrix is ​​established for hierarchical analysis. After consistency test and weight correction, a more accurate indicator weight system is obtained. According to the calculation results of the health assessment model, the ecosystem in this area is in a sub-healthy state. The main problem is that the species diversity is low and the proportion of dominant species is too large. The management rules extracted through knowledge graph analysis show that it is necessary to strengthen the planting of native mangrove plants and moderately control the expansion of Sonneratia apetala. Combined with the specific diagnostic results, management decision recommendations including increasing species diversity, optimizing community structure, and protecting key habitats were finally formed.

[0107] The above describes the analysis method for the population distribution of mangrove ecosystems in the embodiment of the present application. The following describes the analysis system for the population distribution of mangrove ecosystems in the embodiment of the present application. Figure 2 In one embodiment of the present application, an analysis system for the population distribution of a mangrove ecosystem includes:

[0108] An acquisition module is used to acquire high-resolution remote sensing images through a remote sensing image collector, collect field survey data and environmental parameters in the study area, pre-process the high-resolution remote sensing images through an image enhancement algorithm, and establish a multidimensional database containing spatial information, species information and environmental parameters;

[0109] An identification module is used to identify the community type of the data in the multidimensional database using a random forest classification algorithm, extract the location information of a single mangrove plant from the high-resolution remote sensing image using a deep learning target detection algorithm, and generate a community distribution vector data layer;

[0110] A calculation module is used to calculate the relative abundance, relative frequency and relative significance of the mangrove plants in the woody layer based on the community distribution vector data layer to obtain the important value index, calculate the relative coverage and relative height of the mangrove plants in the herbaceous layer to obtain the total dominance index, and calculate the population spatial distribution characteristic parameters through a spatial statistical algorithm;

[0111] An analysis module is used to perform landscape pattern analysis using a landscape index calculation model based on the population spatial distribution characteristic parameters, obtain the maximum patch index, clumping index, similar adjacency percentage and patch aggregation index, establish a connectivity analysis model based on a graph theory algorithm, and obtain key patch identification results;

[0112] An association module is used to associate the key patch identification results with environmental parameters, establish an environmental factor impact model through a gradient boosting algorithm, and use a Bayesian network algorithm to perform uncertainty analysis to obtain population distribution prediction results;

[0113] A generation module is used to construct a health assessment model through a hierarchical analysis method based on the population distribution prediction results, combined with the important value indicators, total dominance indicators and landscape index, establish a knowledge graph reasoning system, and generate mangrove ecosystem management decision-making recommendations.

[0114] Through the coordinated cooperation of the above components, high-resolution remote sensing images are obtained through remote sensing image collectors, and combined with the collection of field survey data and environmental parameters, the effective fusion of multi-source data is realized, and the integrity and accuracy of the data are improved; the image enhancement algorithm is used to pre-process the remote sensing images, and a multidimensional database containing spatial information, species information and environmental parameters is established, which provides comprehensive data support for subsequent analysis; the random forest classification algorithm and deep learning target detection algorithm are used to process the data, which not only realizes the accurate identification of community types, but also can extract the location information of individual mangrove plants, significantly improving the accuracy of spatial distribution analysis; by calculating various indicators of mangrove plants in the woody layer and herbaceous layer, combined with spatial statistical algorithms, the quantitative expression of population distribution characteristics is realized; the landscape index calculation model is used to analyze the landscape pattern, and the connectivity analysis model is established through the graph theory algorithm, making the identification of key patches more accurate and objective; the gradient boosting algorithm and Bayesian network algorithm are used to analyze the impact of environmental factors, which greatly improves the accuracy of population distribution prediction; finally, the hierarchical analysis method and knowledge graph reasoning system are used to realize the intelligent and scientific management decision-making of mangrove ecosystems.

[0115] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for analyzing the population distribution of mangrove ecosystems, characterized in that: The analysis method for the population distribution of mangrove ecosystems includes: Acquire high-resolution remote sensing images through a remote sensing image collector, collect field survey data and environmental parameters in the study area, pre-process the high-resolution remote sensing images through an image enhancement algorithm, and establish a multidimensional database containing spatial information, species information and environmental parameters; Using a random forest classification algorithm to identify community types of data in the multidimensional database, extracting location information of individual mangrove plants from the high-resolution remote sensing images through a deep learning target detection algorithm, and generating a community distribution vector data layer; Based on the community distribution vector data layer, the relative abundance, relative frequency and relative significance of the woody layer mangrove plants are calculated to obtain the important value index, the relative coverage and relative height of the herbaceous layer mangrove plants are calculated to obtain the total dominance index, and the population spatial distribution characteristic parameters are calculated by spatial statistical algorithm; According to the population spatial distribution characteristic parameters, the landscape index calculation model is used to analyze the landscape pattern, obtain the maximum patch index, clumping index, similarity adjacency percentage and patch aggregation index, and establish a connectivity analysis model based on the graph theory algorithm to obtain the key patch identification results; The key patch identification results are correlated with environmental parameters, an environmental factor impact model is established using a gradient boosting algorithm, and uncertainty analysis is performed using a Bayesian network algorithm to obtain population distribution prediction results; Based on the population distribution prediction results, combined with the important value indicators, total dominance indicators and landscape indexes, a health assessment model is constructed through the hierarchical analysis method, a knowledge graph reasoning system is established, and decision-making recommendations for mangrove ecosystem management are generated.

2. The method for analyzing the population distribution of mangrove ecosystem according to claim 1, characterized in that: The high-resolution remote sensing images are obtained by the remote sensing image collector, and field survey data and environmental parameters are collected in the study area. The high-resolution remote sensing images are preprocessed by an image enhancement algorithm to establish a multidimensional database containing spatial information, species information and environmental parameters, including: The study area is imaged by a remote sensing image collector to obtain high-resolution remote sensing images; Through random sampling method, transects were set up in the study area, and sample plots were laid out in each transect to collect field survey data and environmental parameters; The high-resolution remote sensing image is spatially registered by a geometric correction algorithm, spectral correction is performed by a radiometric calibration algorithm, and atmospheric correction algorithm is used to eliminate atmospheric influence, thereby obtaining a corrected high-resolution remote sensing image; Performing image enhancement processing on the corrected high-resolution remote sensing image by using a histogram equalization algorithm to obtain an enhanced high-resolution remote sensing image; Establishing a data standardization table and determining a data storage format based on the enhanced high-resolution remote sensing images, field survey data and environmental parameters; The data standardization table is converted into a multidimensional database containing spatial information, species information and environmental parameters through data structuring processing.

3. The method for analyzing the population distribution of mangrove ecosystem according to claim 1, characterized in that: The method uses a random forest classification algorithm to identify the community type of the data in the multidimensional database, extracts the location information of a single mangrove plant from the high-resolution remote sensing image through a deep learning target detection algorithm, and generates a community distribution vector data layer, including: The spectral features and texture features in the multidimensional database are screened by a feature extraction algorithm to obtain community feature data; Selecting training samples according to the community characteristic data, substituting the training samples into a random forest classification algorithm, and obtaining community type recognition parameters; Filtering the community distribution information in the multidimensional database according to the community type identification parameters to form community type distribution data; Input the high-resolution remote sensing image into a deep learning target detection algorithm, and generate a target detection feature map after processing by a convolutional layer; Calculate the boundary range of mangrove plants from the target detection feature map, extract the location information of individual mangrove plants, and establish the coordinate data of individual plants; The community type distribution data and the individual plant coordinate data are combined to perform geospatial processing to generate a community distribution vector data layer.

4. The method for analyzing the population distribution of mangrove ecosystem according to claim 1, characterized in that: Based on the community distribution vector data layer, the relative abundance, relative frequency and relative significance of the woody layer mangrove plants are calculated to obtain the important value index, the relative coverage and relative height of the herbaceous layer mangrove plants are calculated to obtain the total dominance index, and the population spatial distribution characteristic parameters are calculated by the spatial statistical algorithm, including: Extracting the number of mangrove plant individuals in the woody layer from the community distribution vector data layer, calculating the species occurrence frequency for each sample plot, and forming a relative abundance; Count the number of occurrences of the woody layer mangrove plants in the sample plots, divide the number of occurrences by the total number of sample plots, and generate a relative frequency; Read the DBH values ​​of the woody mangrove plants, calculate the sum of the cross-sectional areas, perform normalization, and output the relative significance; Combining the relative abundance, relative frequency and relative significance to perform a weighted sum operation to obtain an important value index; Extracting the coverage area and height information of the mangrove plants in the herbaceous layer in the community distribution vector data layer, calculating the ratio and then averaging it to obtain the total dominance index; Species distribution points in the community distribution vector data layer are selected, and spatial statistical operations are performed using the variance-mean ratio method, the negative binomial distribution parameter method, and the clustering index method to calculate the population spatial distribution characteristic parameters.

5. The method for analyzing the population distribution of mangrove ecosystem according to claim 1, characterized in that: According to the population spatial distribution characteristic parameters, the landscape index calculation model is used to perform landscape pattern analysis, obtain the maximum patch index, clumping index, similar adjacency percentage and patch aggregation index, and establish a connectivity analysis model based on the graph theory algorithm to obtain key patch identification results, including: Extract patch area data from the population spatial distribution characteristic parameters, divide the maximum patch area by the total area of ​​the study area to obtain the maximum patch index; Extracting the adjacency information of each patch in the population spatial distribution characteristic parameters, and generating a clumping index by calculating the spatial adjacency probability; Statistics are collected for the patch adjacency information in the population spatial distribution characteristic parameters, similar adjacency ratios are calculated based on the total boundary length, and similar adjacency percentages are output; Selecting patch pair data from the population spatial distribution characteristic parameters, calculating the ratio of the number of adjacent patches to the maximum possible number of adjacent patches, and obtaining a patch aggregation index; An adjacency matrix is ​​constructed based on the maximum patch index, clumping index, similarity adjacency percentage, and patch aggregation index, and node connectivity is calculated using a graph theory algorithm; The node connectivity and patch spatial distribution data are subjected to network analysis to obtain key patch identification results.

6. The method for analyzing the population distribution of mangrove ecosystem according to claim 1, characterized in that: The key patch identification results are correlated with environmental parameters, an environmental factor impact model is established by a gradient boosting algorithm, and uncertainty analysis is performed using a Bayesian network algorithm to obtain population distribution prediction results, including: Matching and aligning the key plaque identification results with the environmental parameters to generate plaque environment association data; Inputting the patch environment associated data into the gradient boosting algorithm, and distinguishing the training data from the validation data by randomly splitting the data set; Using a gradient boosting algorithm to calculate feature weights on the training data, and establishing an environmental factor impact model; Perform verification calculation on the environmental factor impact model based on the verification data, and output environmental factor impact weight data; The probability distribution network of the environmental factor influence weight data is constructed by using the Bayesian network algorithm to calculate the conditional probability of each node; The conditional probability is converted into population distribution probability, and combined with the environmental factor impact model to obtain the population distribution prediction result.

7. The method for analyzing the population distribution of mangrove ecosystem according to claim 1, characterized in that: According to the population distribution prediction results, combined with the important value index, total dominance index and landscape index, a health assessment model is constructed through the hierarchical analysis method, a knowledge graph reasoning system is established, and decision-making recommendations for mangrove ecosystem management are generated, including: Integrate the population distribution prediction results, the important value index, the total dominance index and the landscape index into an evaluation index set, and determine the index weights through principal component analysis; Establish a discriminant matrix for the evaluation index set, calculate the characteristic vector through the hierarchical analysis method, and output the relative importance of the index; Substituting the relative importance of the indicators into the hierarchical analysis method to calculate the consistency ratio and generate the indicator weight correction value; Building a health assessment model based on the weight correction values ​​of the indicators and calculating the health score of the mangrove ecosystem; Input the mangrove ecosystem health score into the knowledge graph reasoning system and extract management rules through semantic analysis; The management rules are combined with the diagnostic results of the health assessment model to generate mangrove ecosystem management decision recommendations.

8. An analysis system for the population distribution of a mangrove ecosystem, used to implement the analysis method for the population distribution of a mangrove ecosystem as claimed in any one of claims 1 to 7, characterized in that: The analysis system for the population distribution of mangrove ecosystems comprises: An acquisition module is used to acquire high-resolution remote sensing images through a remote sensing image collector, collect field survey data and environmental parameters in the study area, pre-process the high-resolution remote sensing images through an image enhancement algorithm, and establish a multidimensional database containing spatial information, species information and environmental parameters; An identification module is used to identify the community type of the data in the multidimensional database using a random forest classification algorithm, extract the location information of a single mangrove plant from the high-resolution remote sensing image using a deep learning target detection algorithm, and generate a community distribution vector data layer; A calculation module is used to calculate the relative abundance, relative frequency and relative significance of the mangrove plants in the woody layer based on the community distribution vector data layer to obtain the important value index, calculate the relative coverage and relative height of the mangrove plants in the herbaceous layer to obtain the total dominance index, and calculate the population spatial distribution characteristic parameters through a spatial statistical algorithm; An analysis module is used to perform landscape pattern analysis using a landscape index calculation model based on the population spatial distribution characteristic parameters, obtain the maximum patch index, clumping index, similar adjacency percentage and patch aggregation index, establish a connectivity analysis model based on a graph theory algorithm, and obtain key patch identification results; An association module is used to associate the key patch identification results with environmental parameters, establish an environmental factor impact model through a gradient boosting algorithm, and use a Bayesian network algorithm to perform uncertainty analysis to obtain population distribution prediction results; A generation module is used to construct a health assessment model through a hierarchical analysis method based on the population distribution prediction results, combined with the important value indicators, total dominance indicators and landscape index, establish a knowledge graph reasoning system, and generate mangrove ecosystem management decision-making recommendations.

Images