CN119089318A - A multi-factor comprehensive lake island classification and identification method, system, product and medium - Google Patents

Abstract

A multi-factor comprehensive lake island classification and identification method, system, product and medium relate to the field of lake island classification. The method comprises: firstly obtaining multi-dimensional data of a target lake island, including image data, environmental data and influencing factors, and dividing the influencing factors into ecological factors and humanistic factors by text matching, thereby achieving a preliminary classification of lake island characteristics. Subsequently, the method conducts an in-depth analysis of the image and ecological data, extracts more detailed ecological and humanistic characteristics, and forms secondary factors. Finally, by matching the extracted characteristics with a preset classification list, the final classification type of the target lake island is obtained. This automated classification process not only improves the efficiency of classification, but also reduces human errors, making the lake island classification results more objective and accurate.

Description

Multi-factor comprehensive lake island classification and identification method, system, product and medium

Technical Field

The application relates to the field of classification of islands, in particular to a multi-factor comprehensive island classification and identification method, system, product and medium.

Background

With the increasing awareness of ecological environmental protection, research and management of the ecological system of the lake island is increasingly attracting attention. The classification and identification of the islands as unique ecological units is critical to the formulation of effective protection strategies and management measures. Accurate classification of the islands may help researchers and managers better understand the ecological characteristics, the humane features, and interactions between the islands.

In the related art, lake island classification and identification mainly depend on remote sensing image analysis and field investigation, researchers acquire image data of the lake island through satellites or aerial photography, and the lake island is classified by combining ecological and humane information collected by field investigation.

The classification method often needs manual data preprocessing and feature extraction, resulting in reduced efficiency of classification of the lake island.

Disclosure of Invention

The application provides a multi-factor comprehensive lake island classification and identification method, a system, a product and a medium, which are used for improving the efficiency of lake island classification.

In a first aspect, the application provides a multi-factor integrated lake island classification method, which is applied to a multi-factor integrated lake island classification system, and comprises the steps of acquiring image data of a target lake island in a target area, lake island environment data and influence factors, wherein the lake island environment data comprises lake island ecological data and lake island humane data, and the influence factors are factors influencing the classification of the target lake island; the method comprises the steps of carrying out text matching on the influence factors according to preset environment monitoring data and preset humanization data, determining the influence factors conforming to the preset environment monitoring data as primary ecological factors, determining the influence factors conforming to the preset humanization data as primary humanization factors, detecting the image data and the lake island ecological data, extracting ecological factor parameters of the target lake island to obtain primary ecological data, taking the primary ecological data as secondary ecological factors, extracting the humanization factor parameters of the target lake island according to the lake island humanization data to obtain primary humanization data, taking the primary humanization data as secondary humanization factors, determining the condition that the secondary ecological factors and the secondary humanization factors conform to a preset lake island classification list, obtaining a matching sequence, and determining the classification type of the target lake island according to the matching sequence.

By adopting the technical scheme, the multi-dimensional data of the target lake island is firstly obtained, the multi-dimensional data comprises image data, environment data and influence factors, the influence factors are subjected to text matching, the influence factors are divided into physiological factors and humane factors, the preliminary classification of the characteristics of the lake island is realized, then, the image and the ecological data are subjected to deep analysis by the method, more detailed ecological and humane characteristics are extracted, secondary factors are formed, finally, the extracted characteristics are matched with a preset classification list, the final classification type of the target lake island is obtained, the automatic classification process not only improves the classification efficiency, but also reduces human errors, and the classification result of the lake island is more objective and accurate.

In combination with some embodiments of the first aspect, in some embodiments, the step of performing text matching on the influence factor according to preset environmental monitoring data and preset humanization data, determining the influence factor conforming to the preset environmental monitoring data as a ecological factor, and determining the influence factor conforming to the preset humanization data as a humanization factor specifically includes performing semantic analysis on the influence factor, extracting an influence keyword of the influence factor, constructing a knowledge graph including the preset environmental monitoring data and the preset humanization data, calculating influence similarity of the influence factor and a node in the knowledge graph, determining the influence factor as a ecological factor if the influence similarity is greater than a preset influence threshold, and determining the influence factor as the humanization factor if the influence similarity is not greater than the preset influence threshold.

By adopting the technical scheme, the semantic analysis is carried out on the influence factors, the keywords are extracted, a foundation is laid for subsequent matching, a knowledge graph comprising preset environment monitoring data and human data is constructed, the influence similarity between the influence factors and the knowledge graph nodes is calculated as a reference standard for classification, each factor can be accurately judged to belong to ecological factors or human factors, the accuracy and the efficiency of classification are improved based on the classification method of the knowledge graph and the semantic analysis, the time and the manpower are saved in an automatic classification process, a large number of influence factors can be processed, and the comprehensive analysis of the characteristics of the lake island is realized.

In combination with some embodiments of the first aspect, in some embodiments, the step of detecting the image data and the lake island ecological data and extracting the ecological factor parameters of the target lake island to obtain first-level ecological data specifically includes the steps of segmenting the image data, extracting the geographic characteristic parameters of the target lake island, wherein the geographic characteristic parameters include a lake island area, a water body range and a vegetation coverage area, fusing the geographic characteristic parameters with the lake island ecological data to obtain ecological characteristic vectors of the target lake island, the lake island ecological data include vegetation types, cultivation conditions and construction conditions, calculating the statistical characteristic value of each parameter in the ecological characteristic vectors, and taking the statistical characteristic value of all parameters as first-level ecological data, wherein the first-level ecological data comprises the lake island area, the water body range, the vegetation coverage area, the vegetation types, cultivation conditions and construction conditions.

By adopting the technical scheme, the geographic characteristic parameters of the lake island, including the area, the water body range and the vegetation coverage area, are extracted by dividing the image data, so that basic data is provided for subsequent ecological characteristic analysis. And fusing the geographic characteristic parameters with the lake island ecological data to obtain ecological characteristic vectors. The method of data fusion enables the description of the ecological characteristics to be more comprehensive and accurate. The statistical characteristic value of each parameter in the ecological characteristic vector is calculated to obtain first-level ecological data, the multidimensional ecological data comprises physical characteristics of the lake island, and ecological information such as vegetation types, cultivation conditions, construction conditions and the like, so that comprehensive extraction and quantification of the ecological characteristics of the lake island are realized.

In combination with some embodiments of the first aspect, in some embodiments, the step of determining that the secondary ecological factor and the secondary humane factor conform to a preset lake island classification list to obtain a matching sequence, and determining the classification type of the target lake island according to the matching sequence specifically includes sequentially judging each classification condition in the secondary ecological factor and the secondary humane factor and the preset lake island classification list, if the secondary ecological factor and the secondary humane factor meet the classification condition, marking the corresponding position in the initial sequence as 1, if the secondary ecological factor and the secondary humane factor do not meet the classification condition, marking as 0, obtaining the matching sequence, and determining the classification type of the target lake island according to the matching sequence.

By adopting the technical scheme, the secondary ecological factors and the secondary humane factors are judged with each classification condition in the preset lake island classification list to form a matching sequence, when the classification conditions are met, the corresponding positions in the initial sequence are marked as 1, and when the classification conditions are not met, the corresponding positions are marked as 0, the complex lake island characteristics are converted into concise digital representation by the generation process of the matching sequence, the subsequent classification judgment is convenient, the classification type of the target lake island is determined according to the matching sequence, the automation and standardization of the classification process are realized, the matching process of the lake island characteristics and the preset classification conditions is quantized, and the classification result is more objective and reliable.

In combination with some embodiments of the first aspect, in some embodiments, the computer determines that the secondary ecological factor and the secondary humane factor conform to a preset lake island classification list, obtains a matching sequence, and after determining a classification type of the target lake island according to the matching sequence, the method further includes obtaining historical classification data of the target lake island, where the historical classification data includes a type of the target lake island at a historical time point, constructing a time sequence model, inputting the historical classification data into the time sequence model, obtaining a classification type change trend of the target lake island, predicting a type of the target lake island at a preset future time point according to the classification type change trend, obtaining a prediction type, calculating a type similarity between the prediction type and the classification type, and if the type similarity is greater than a preset similarity threshold, generating lake island evolution early warning information and sending the lake island evolution early warning information to a client.

By adopting the technical scheme, the historical classification data of the target lake island is obtained, and a time sequence model is constructed to obtain the classification type change trend of the target lake island. And predicting the type of the target lake island at a preset future time point by utilizing the change trend to obtain a predicted type. Calculating the similarity between the prediction type and the current classification type, and generating lake island evolution early warning information and sending the information to a client when the similarity is larger than a preset threshold. The method not only realizes classification of the current state of the lake island, but also predicts the future change trend of the lake island, and the introduction of the time sequence model enables the classification of the lake island to be changed from static analysis to dynamic prediction, thereby being beneficial to taking protective measures in advance, preventing ecological deterioration of the lake island and realizing sustainable management of the lake island.

In combination with some embodiments of the first aspect, in some embodiments, the step of predicting the type of the target lake island at a preset future time point according to the classification type variation trend specifically includes calculating a type variation period and a type variation rate parameter of the target lake island according to the historical classification data, and inputting the type variation period and the type variation rate parameter into a type prediction model to obtain a prediction type.

By adopting the technical scheme, the type change period and the type change rate parameters of the target lake island are calculated according to the historical classification data, and the parameters are input into a type prediction model to obtain a prediction type. The introduction of the variation period and rate parameters enables the predictive model to capture the regularity characteristics of the variation of the type of the lake island. The use of the type prediction model converts the complex lake island evolution process into a quantifiable and predictable mathematical model, improves the prediction accuracy and can also provide a longer-range planning basis for lake island management.

With reference to some embodiments of the first aspect, in some embodiments, after the step of calculating the type similarity between the prediction type and the classification type, the method includes generating evolution prompt information according to the prediction type and sending the evolution prompt information to the client if the type similarity is not greater than a preset similarity threshold.

By adopting the technical scheme, when the similarity between the prediction type and the current classification type is not greater than the preset similarity threshold, the system generates evolution prompt information according to the prediction type and sends the evolution prompt information to the client. The setting of the preset similarity threshold ensures that the system can identify significant changes in the type of the islands of the lake. The generation process of the evolution prompt information considers the characteristics of the prediction type and provides specific and targeted information for the user. And the prompt information is sent to the client, so that timely feedback of the lake island classification result is realized.

In a second aspect, embodiments of the present application provide a multi-factor integrated lake island classification recognition system comprising one or more processors and memory coupled to the one or more processors, the memory for storing computer program code comprising computer instructions that the one or more processors invoke to cause the multi-factor integrated lake island classification recognition system to perform a method as described in the first aspect and any possible implementation of the first aspect.

In a third aspect, embodiments of the present application provide a computer program product comprising instructions that, when run on a multi-factor integrated lake island classification recognition system, cause the multi-factor integrated lake island classification recognition system to perform a method as described in the first aspect and any one of the possible implementations of the first aspect.

In a fourth aspect, embodiments of the present application provide a computer readable storage medium comprising instructions which, when run on a multi-factor integrated lake island classification recognition system, cause the multi-factor integrated lake island classification recognition system to perform a method as described in the first aspect and any one of the possible implementations of the first aspect.

It will be appreciated that the multi-factor integrated lake island classification recognition system provided in the second aspect, the computer program product provided in the third aspect and the computer storage medium provided in the fourth aspect are each configured to perform the method provided by the embodiment of the present application. Therefore, the advantages achieved by the method can be referred to as the advantages of the corresponding method, and will not be described herein.

One or more technical solutions provided in the embodiments of the present application at least have the following technical effects or advantages:

1. According to the method, the multi-dimensional data of the target lake island is firstly obtained, the multi-dimensional data comprises image data, environment data and influence factors, the influence factors are subjected to text matching, the influence factors are divided into physiological factors and humane factors, the primary classification of the characteristics of the lake island is realized, then, the image and the ecological data are subjected to deep analysis, more detailed ecological and humane characteristics are extracted, a secondary factor is formed, and finally, the extracted characteristics are matched with a preset classification list, so that the final classification type of the target lake island is obtained.

2. According to the method, the key words are extracted through semantic analysis on the influence factors, a foundation is laid for subsequent matching, the knowledge graph comprising preset environment monitoring data and human data is constructed, the influence similarity between the influence factors and the knowledge graph nodes is calculated as a reference standard for classification, whether each factor belongs to an ecological factor or a human factor can be accurately judged, the accuracy and the efficiency of classification are improved based on the classification method of the knowledge graph and the semantic analysis, time and manpower are saved in an automatic classification process, a large number of influence factors can be processed, and comprehensive analysis on characteristics of the lake island is realized.

3. According to the application, the secondary ecological factors and the secondary humane factors are judged with each classification condition in the preset lake island classification list to form the matching sequence, when the classification condition is met, the corresponding position in the initial sequence is marked as 1, and when the classification condition is not met, the corresponding position is marked as 0, the generation process of the matching sequence converts the complex lake island characteristics into a concise digital representation, the subsequent classification judgment is convenient, the classification type of the target lake island is determined according to the matching sequence, the automation and standardization of the classification process are realized, and the matching process of the lake island characteristics and the preset classification condition is quantized, so that the classification result is more objective and reliable.

Drawings

FIG. 1 is a schematic flow chart of a multi-factor integrated lake island classification and identification method in an embodiment of the application;

FIG. 2 is another flow chart of a multi-factor integrated lake island classification and identification method according to an embodiment of the application;

fig. 3 is a schematic structural diagram of an entity device of a multi-factor integrated lake island classification and identification system according to an embodiment of the application.

Detailed Description

The terminology used in the following embodiments of the application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used in the specification of the present application, the singular forms "a", "an", "the" and "the" are intended to include the plural forms as well, unless the context clearly indicates to the contrary. It should also be understood that the term "and/or" as used in this disclosure is intended to encompass any or all possible combinations of one or more of the listed items.

The terms "first," "second," and the like, are used below for descriptive purposes only and are not to be construed as implying or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include one or more such feature, and in the description of embodiments of the application, unless otherwise indicated, the meaning of "a plurality" is two or more.

For ease of understanding, the method provided in this embodiment is described in the following. Referring to fig. 1, a flow chart of a multi-factor integrated lake island classification and identification method according to an embodiment of the application is shown.

S101, acquiring image data of a target lake island in a target area, lake island environment data and influence factors, wherein the lake island environment data comprises lake island ecological data and lake island humane data, and the influence factors are factors influencing the classification of the target lake island.

Wherein the target area represents a specific geographical area where lake island classification identification is required. The target lake island refers to a specific lake island which needs to be classified and identified in the target area, and the image data is used for representing the visual information of the lake island, such as satellite images or aerial photos. The lake island environmental data refers to various data describing the ecology and the humanity of the lake island, the influence factors represent various factors which can influence the classification of the lake island, the lake island ecological data comprises natural environmental information such as vegetation coverage, water quality and the like, and the lake island humanity data comprises social environmental information such as population, economic activities and the like.

Specifically, before the lake island classification and identification process is started, related data of the target lake island needs to be acquired first. The system acquires high-resolution satellite images or aerial photographs of specific islands in the target area through remote sensing equipment or a database as image data. Meanwhile, the system collects ecological environment parameters and humane social indexes of the lake island from environmental monitoring sites, field investigation reports and other channels to form lake island environmental data.

S102, performing text matching on the influence factors according to preset environment monitoring data and preset humanization data, determining the influence factors conforming to the preset environment monitoring data as first-level ecological factors, and determining the influence factors conforming to the preset humanization data as first-level humanization factors.

Wherein the preset environmental monitoring data represents a predefined series of environmental indicators and parameters. The preset humanization data refers to a preset set of humanization social indexes. The text matching refers to comparing the text similarity of the influence factors and preset data. The primary ecological factor represents a directly related environmental impact element. The primary humane factors represent directly related social impact elements.

Specifically, after the impact factors are obtained, the system needs to perform a preliminary classification of these factors to determine whether they are ecological or humane. The system first invokes the preset environmental monitoring dataset and the humane dataset. Then, the system performs text analysis on each influence factor, and extracts keywords and semantic features of the influence factors. Then, the system matches the features with a preset data set to calculate the similarity. And classifying the environmental monitoring data as a first-level ecological factor if the matching degree of a certain influence factor and the environmental monitoring data is high, and classifying the environmental monitoring data as a first-level human factor if the matching degree of the influence factor and the human data is high.

In some embodiments, factor classification may be achieved in a number of ways, optionally, the system may employ text matching and factor classification by first word segmentation and part-of-speech tagging of influencing factors using natural language processing techniques. And then constructing a word vector model of preset environment monitoring data and human data. And then calculating cosine similarity between the influence factors and the preset data word vector. Then, a similarity threshold is set, and the similarity higher than the threshold is classified as a corresponding first-order factor. And finally, for factors which are difficult to judge, further classifying by adopting a machine learning classifier. Alternatively, the system may also factor the classification by first building a knowledge graph of environmental monitoring and human data. And extracting semantic features of the influence factors, and mapping the semantic features into a knowledge graph. Then, calculating the semantic distance between the influence factors and the knowledge graph nodes. The factor category is then determined based on the minimum value of the semantic distance.

S103, detecting the image data and the ecological data of the lake island, extracting ecological factor parameters of the target lake island to obtain primary ecological data, and taking the primary ecological data as a secondary ecological factor.

Wherein the image data represents visual information of the target lake island, such as satellite images or aerial photographs. The ecological data of the lake island refers to various data describing the natural environment characteristics of the lake island, including water quality, vegetation coverage and the like. Detection refers to the processing and analysis of images and data by computer vision and data analysis techniques. The ecological factor parameter is used to represent various elements that affect the lake shore ecosystem, such as water body area, vegetation index, etc. The first-level ecological data refers to quantitative information directly extracted from images and ecological data.

Specifically, after acquiring the image data and the lake island ecological data, the system needs to perform deep analysis on the raw data to extract more valuable ecological information. First, the system uses image processing techniques to pre-process satellite images or aerial photographs, including denoising, enhancing, and correcting. Then, the system performs feature extraction on the processed image by using a computer vision algorithm, and identifies key elements such as a water body boundary, a vegetation coverage area, an artificial building and the like. Meanwhile, the system performs statistical analysis on the lake island ecological data, and calculates statistics such as mean value, variance and the like of each ecological index. Then, the system combines the image analysis result with the ecological data statistical result to extract a series of quantitative parameters capable of comprehensively reflecting the ecological condition of the lake island, and the parameters form first-level ecological data.

In some embodiments, the extraction and processing of ecological data may be accomplished in a variety of ways:

alternatively, the system may employ the following steps for data processing:

Noise removal, contrast enhancement and geometric correction are carried out on the satellite image, image quality and geographic position accuracy are ensured, and the near infrared band reflectivity threshold is set to identify the water body region by utilizing the characteristics of the multispectral image. The water body boundary is refined through morphological operation to obtain an accurate lake island outline, then a normalized vegetation index (NDVI) is calculated, a vegetation coverage area is divided according to the NDVI value, the vegetation coverage degree is estimated, the water quality parameter distribution of the whole lake area is estimated by using a spatial interpolation method in combination with water quality data acquired in the field, and finally the extracted information of the water body area, vegetation coverage rate, water quality parameters and the like is integrated into structural data to form first-level ecological data.

Optionally, the system can process data by collecting seasonal satellite images for many years, analyzing seasonal changes and annual change trends of the water area, calculating other vegetation indexes such as Enhanced Vegetation Index (EVI) except NDVI, comprehensively evaluating the vegetation health condition, analyzing the relationship between human activity intensity and water quality parameters by utilizing night light data, comprehensively considering the water change, vegetation condition, water quality and human activity influence, constructing a simple model to evaluate the vulnerability of an ecological system, manufacturing the extracted ecological parameters into various charts, and intuitively displaying the ecological condition of a lake island.

S104, extracting the humane factor parameters of the target lake island according to the lake island humane data to obtain primary humane data, and taking the primary humane data as a secondary humane factor.

The lake island humane data refer to various data describing the social environment and human activity characteristics of the lake island, including demographics, economic indexes and the like, the humane factor parameters are used for representing various elements affecting the social culture characteristics of the lake island, such as population density, economic development level and the like, the primary humane data refer to quantitative information directly extracted from the original humane data, and the secondary humane factors refer to the humane influence elements subjected to primary processing and classification for subsequent lake island classification analysis.

Specifically, after the lake island humane data are acquired, the system needs to systematically process and analyze the data so as to extract key information capable of accurately reflecting the lake island humane characteristics. Firstly, the system performs data cleaning and standardization processing on various kinds of personal data, and ensures consistency and comparability of the data. Then, the system calculates descriptive statistics of various human indexes, such as mean, median, standard deviation and the like, by using a statistical analysis method. Then, the system performs correlation analysis, identifies the associated humane factors, and performs appropriate data dimension reduction or clustering. And then, the system combines expert knowledge and a data mining technology to extract key parameters which can best embody the humane features of the lake island from the processed data, and the parameters form first-stage humane data. Finally, the system further classifies and sorts the primary human data to form a structured secondary human factor.

For example, primary factors include residential population, regulatory requirements, ecological and construction conditions, human resources, and island scale, among others. The original lake island humane data acquired by the system comprises demographic data, administrative region division, travel facility distribution and the like, and the system firstly cleans and processes the original data to obtain a normalized data set. And then the system extracts relevant data representing the primary factors according to the provided examples, the system counts the number of resident population in the lake island for the living population condition, classifies the resident population number as a secondary factor of 'occupied person' or 'unoccupied person', carries out natural language processing on administrative division data according to the administrative division data processing to obtain that the island belongs to a single administrative division or a plurality of administrative divisions, obtains corresponding secondary factors, analyzes satellite images for ecology and construction conditions, extracts construction area distribution, combines topography and topography data, determines vegetation coverage and construction density, obtains secondary factors reflecting construction activity intensity, counts the number of historical sites and scenic spots on the island for human resources, judges whether the corresponding secondary factors are abundant or not, calculates island area for island scale, judges whether the island area is more than or less than 100 hectares, obtains the corresponding secondary factors, finally, selects factors representing human characteristics according to the extracted secondary factors, such as population conditions, construction conditions, human resources and the like, determines the secondary factors as the living characteristics, determines the ecological factors as the ecological characteristics, and the ecological factors representing the ecological characteristics.

In some embodiments, the extraction of primary human data may be accomplished in a variety of ways:

Optionally, the system may employ the steps of extracting values representing indexes such as living population, management and control requirements, and human resources from the raw data according to the primary factors in the examples, processing and calculating the index data to obtain primary human data expressing the primary factors, such as unmanned or living, and the like, and corresponding the primary human data to the secondary factors in the examples to form standardized secondary human factors, and selecting primary data representing human characteristics to form a result according to the definition of the secondary factors.

Optionally, the system may further construct an ontology model knowledge graph including an example primary factor, map text data to knowledge graph nodes, activate related branches, determine expressed primary human data according to the activation condition of the branches, and convert the primary data to obtain a standardized secondary human factor in the example, which may be understood that other manners may be used to implement extraction of the primary human data, which is not limited herein.

S105, determining that the secondary ecological factors and the secondary humane factors accord with a preset lake island classification list, obtaining a matching sequence, and determining the classification type of the target lake island according to the matching sequence.

The secondary ecological factors represent ecological data indexes after being processed and classified and are used for reflecting ecological environment characteristics of the lake island. The secondary humane factor refers to a processed and integrated humane data index used for representing the humane features of the lake island. The preset lake island classification list refers to a predefined lake island type and corresponding ecological and humane characteristic standards thereof. The matching sequence represents the matching degree of the secondary factor of the target lake island and a preset classification standard. The classification type refers to the category of the lake island determined according to the matching result.

The method is carried out after the extraction of the secondary ecological factors and the secondary humane factors is completed, and is used for determining the specific type of the target lake island. Specifically, the system firstly compares the extracted secondary ecological factors and secondary humane factors with various standards in a preset lake island classification list. In the comparison process, the system calculates the similarity or matching degree of each factor and various types of standards to form a matching sequence. This sequence contains the degree of matching of the target lake islands under various classification criteria. The system then analyzes the matching sequence, typically using a highest degree of matching principle or a comprehensive scoring method, to determine the final lake island classification type. If multiple types of matching are similar, the system may describe the characteristics of the lake island by adopting a suboptimal matching or mixed type mode.

In some embodiments, the matching and determining process of the lake island classification may be accomplished in a variety of ways:

Alternatively, the system may employ the step of normalizing the secondary ecological factors and the secondary humane factors to make the indices of different dimensions comparable. For example, a Z-score normalization or Min-Max scaling method may be used to calculate euclidean distance or cosine similarity between the normalized factors and various types of standards in the preset classification list, to obtain a similarity matrix, for each preset type, calculate the comprehensive matching degree between the normalized factors and the target lake island on all factors, and rank the matching degrees from high to low, set a matching degree threshold, screen out candidate types exceeding the threshold, if only one type exceeds the threshold, directly determine the type, and if there are multiple types, select the type with the highest matching degree as the final classification result. Optionally, the system can also classify by assigning importance weights to different secondary factors according to expert experience or a data-driven method (such as principal component analysis), constructing a fuzzy relation matrix, converting the secondary factor values into membership degrees, performing fuzzy comprehensive operation to obtain membership degrees of each preset type, performing fuzzy evaluation on ecological factors and humane factors respectively, performing fuzzy comprehensive on two evaluation results again to obtain a final classification result, calculating a confidence score for each possible classification result, reflecting the reliability of the classification result, and marking a plurality of types of labels for the lake island if the matching degrees of the plurality of types are similar and higher, so that the composite characteristics of the lake island are reflected.

In combination with the above scenario, a further more specific flow of the method provided in this embodiment will be described below. Referring to fig. 2, another flow chart of the multi-factor integrated lake island classification and identification method according to the embodiment of the application is shown.

S201, acquiring image data of a target lake island in a target area, lake island environment data and influence factors, wherein the lake island environment data comprises lake island ecological data and lake island humane data, and the influence factors are factors influencing the classification of the target lake island.

It is understood that this step is similar to step S101, and will not be described here.

S202, carrying out semantic analysis on the influence factors, and extracting influence keywords of the influence factors.

The semantic analysis is a process of obtaining semantic information and keywords expressed by a text by deep analysis of the text by using a natural language processing technology. The influencing factors refer to various factors that may influence the classification of the islands. Extracting the influence keywords refers to identifying core words capable of expressing influence properties of the influence factors from text descriptions of the influence factors.

Specifically, the multi-factor comprehensive lake island classification recognition system firstly utilizes natural language processing technologies such as word segmentation, part of speech tagging and the like to carry out semantic analysis on text description of influencing factors, recognizes important word classes such as nouns, verbs, adjectives and the like in the text, determines keywords such as words such as pollution, industry, population and the like which can accurately summarize influence properties of influencing factors by counting indexes such as word frequencies, association relations and the like of the words, extracts the keywords as semantic features of influencing factors, and provides basis for subsequent factor classification and matching. For example, if words such as "industrial wastewater", "chemical fertilizer pesticide", etc. appear in the descriptive text of an impact factor, the system may determine that the impact factor is related to water pollution and extract "pollution" as an impact key.

S203, constructing a knowledge graph containing the preset environment monitoring data and the preset humane data.

Knowledge graph is a structured knowledge base that expresses concepts and relationships between concepts. The preset environmental monitoring data refers to predefined environmental monitoring indexes and threshold values. The preset personal data refers to predetermined statistical data related to personal activities. The construction of the knowledge graph is to organize environment monitoring data and human data by using knowledge representation technology and adopting a network model mode to express the relationship between concepts so as to form a structured knowledge base.

Specifically, the multi-factor comprehensive lake island classification and identification system can construct a knowledge graph by adopting the following modes that firstly, the system sorts environment monitoring data, monitors indexes such as COD, ammonia nitrogen, PM2.5 and the like which represent aspects of water quality, atmosphere and the like are identified, and each index and a standard threshold value thereof are expressed as a conceptual node. Secondly, the system collates the humane data, identifies social statistical indexes such as population, GDP and the like, and also represents each index as a concept node. Then, the system analyzes the internal relation between indexes and establishes semantic association between nodes by means of attributes, relations and the like. Finally, the system displays the finally constructed knowledge graph in a certain visual mode. For example, a "COD" node may have a "reflecting" relationship with a "water quality" node, and a "GDP" node may have a "positive correlation" relationship with an "economic activity strength".

S204, calculating the influence similarity between the influence factors and the nodes in the knowledge graph.

Calculating the influence similarity refers to quantitatively determining the semantic relation degree of the influence factors and each node in the knowledge graph by using a similarity algorithm. The higher the impact similarity, the more semantically the impact factor is represented to the node.

Specifically, the multi-factor integrated lake island classification recognition system can calculate influence similarity by extracting key words of influence factors to be expressed as vectors, accessing a knowledge graph, traversing each node, equally expressing names and related attributes of the nodes as vectors, calling a word vector similarity algorithm, such as word co-occurrence similarity, word vector cosine similarity and the like, to calculate the similarity between the influence factor vector and each node vector, and finally returning the highest similarity to be used as the influence similarity between the influence factors and the nodes. For example, if the keyword of the influence factor is "pollution", and the "COD" node is present in the knowledge graph, the system may calculate cosine similarity of the two word vectors, and since the two semantic terms are closer, the similarity is higher, so as to determine that the influence factor has a stronger semantic relevance with the "COD" node.

S205, if the influence similarity is larger than a preset influence threshold, determining the influence factor as a ecological factor, and if the influence similarity is not larger than the preset influence threshold, determining the influence factor as a human factor.

The influence similarity refers to the semantic matching degree of the influence factors and the knowledge graph nodes. The preset influence threshold is a threshold preset for influence similarity. Ecological factors refer to factors related to environmental monitoring data. A humane factor refers to a factor related to a humane activity.

Specifically, after the influence similarity between the influence factors and each node is calculated, the multi-factor comprehensive lake island classification recognition system needs to judge the category of the influence factors according to a threshold value, wherein the system firstly presets a threshold value of the influence similarity, and the threshold value can be determined through training samples. And then, the system judges whether the maximum influence similarity between each influence factor and the knowledge graph node is larger than a preset threshold value. If the influence factor is larger than the environmental monitoring data, the influence factor is judged to be more relevant to the environmental monitoring data, and the influence factor is classified as a physiological factor. If the data is not more than the human data, the data is judged to be more relevant to the human data and classified as the human factor. For example, if the maximum similarity of "pollution" is 0.8, exceeding a preset 0.7 threshold, it is determined as an ecological factor related to water quality. Conversely, if the maximum similarity of the "industrial structure" is 0.5, it is determined as a humane factor related to the economic activity.

S206, segmenting the image data, and extracting geographic characteristic parameters of the target lake island, wherein the geographic characteristic parameters comprise the area of the lake island, the water body range and the vegetation coverage area.

Image segmentation is the division of an image into regions with different features. The geographic characteristic parameter is a quantization index describing the geographic information of the lake island. Including lake island area, water body range, vegetation coverage, and the like. The geographic characteristic parameters are extracted by image segmentation, and quantitative information of the lake island region is obtained by a segmentation algorithm.

Specifically, the multi-factor comprehensive lake island classification and identification system can adopt the following technical means to carry out image segmentation, namely firstly, carrying out preprocessing on satellite or aerial images to realize interference removal and image enhancement. Then, an image segmentation algorithm based on boundaries or regions, such as Canny edge detection, is applied to segment out the contour region of the lake island. Then, different threshold conditions are set, and the water range and different types of vegetation coverage areas are continuously segmented. And finally, calculating the area of each divided area to obtain the geographic element parameters such as the area of the lake island, the area of the water body, the vegetation coverage area and the like. For example, the system may first partition the boundary of the lake island using a Sobel edge detection algorithm, and then partition the water body and the land coverage type using a maximum inter-class variance method, so as to obtain geographical features such as area parameters.

S207, fusing the geographic characteristic parameters with the ecological data of the lake island to obtain ecological characteristic vectors of the target lake island, wherein the ecological data of the lake island comprise vegetation types, cultivation conditions and construction conditions.

The geographic characteristic parameter is a quantitative indicator obtained by image segmentation. The lake island ecological data is qualitative information describing the ecological condition of the lake island. And fusing the two to obtain an ecological feature vector, namely synthesizing the two types of data to construct a quantitative vector for expressing the ecological features of the lake island.

Specifically, the multi-factor integrated lake island classification and identification system can perform data fusion by firstly, arranging geographic characteristic parameters obtained by image segmentation by the system. Then, the system processes the lake island ecological data, encodes qualitative indicators, such as different vegetation types into different values. And then, the system splices the coded ecological data with the geographic characteristic parameters to construct a high-dimensional vector for describing the ecological condition of the lake island as an ecological characteristic vector. Each dimension in the vector represents a geographic or ecological parameter. Finally, the system can normalize the ecological feature vector to make the features of each dimension comparable, and then input the feature vector into the classification model. For example, the system may construct an ecological feature vector comprising 10 dimensions, where the first 3 dimensions represent lake island area, water body range, vegetation rate, and the second 7 dimensions represent coverage of different vegetation types.

S208, calculating the statistical characteristic value of each parameter in the ecological characteristic vector, taking the statistical characteristic value of all the parameters as primary ecological data, wherein the primary ecological data comprises lake island area, water body range, vegetation coverage area, vegetation type, cultivation condition and construction condition, and taking the primary ecological data as a secondary ecological factor.

The first-level ecological data is a statistical data set which is obtained by summarizing and calculating each parameter in the ecological feature vector through a statistical analysis method and can comprehensively reflect the ecological state of the target lake island, and the second-level ecological factors are core parameters which are extracted from the first-level ecological data and can most represent the ecological features of the lake island.

Specifically, the multi-factor integrated lake island classification and identification system can adopt the following technical means to obtain first-level ecological data, wherein whether people live or not is taken as an example, the first-level ecological factor can be expressed as a binary variable, 1 represents that people live or not, and 0 represents that people live or not. The system can detect the characteristics related to human living in the ecological feature vector, such as night light data, if the ecological feature vector is larger than a certain threshold value, the ecological feature vector is judged to be 1, otherwise, the ecological feature vector is 0, and taking the building area as an example, the first-level ecological factor can be expressed as a specific numerical value, such as the building coverage area calculated according to square meters. The system can calculate and obtain the pixel number of the building area through an image segmentation algorithm, then the pixel number is converted into the actual area size, specifically, the multi-factor comprehensive lake island classification recognition system can obtain first-stage ecological data through the steps that firstly, the system obtains constructed ecological feature vectors which contain coded ecological data indexes, then, the system sets a judging rule or a calculating method for each ecological data index to obtain first-stage ecological data representing the factors, then, the system integrates all first-stage ecological factors to form a first-stage ecological data set, the ecological state of a target lake island is comprehensively reflected, and finally, the system can extract main second-stage ecological factors according to the first-stage ecological data set, and a key basis is provided for the classification of the subsequent lake island.

S209, extracting the humane factor parameters of the target lake island according to the lake island humane data to obtain primary humane data, and taking the primary humane data as a secondary humane factor.

It is understood that this step is similar to step S104, and will not be described here.

S210, judging the secondary ecological factors, the secondary humane factors and each classification condition in the preset lake island classification list in sequence.

The secondary ecological factors are key ecological parameters extracted from the primary ecological data. The secondary personality factors are key personality parameters extracted from the primary personality data. The preset lake island classification list is a classification condition of different types of lake islands defined in advance. The judgment is to compare the factor parameters with the classification conditions one by one.

Specifically, the lake island classification and identification system with multi-factor synthesis needs to be judged according to the following steps:

Firstly, a system acquires a preset lake island classification list which comprises characteristic conditions of different types of lake islands, such as area size, vegetation coverage rate and the like, then the system takes out a first classification condition, such as area greater than 1000 square kilometers, and carries out matching judgment with a secondary ecological factor of the lake island area, the system sequentially takes out each condition of the classification list, judges whether the secondary ecological factor and the secondary humane factor meet the condition, and the system completes one-by-one judgment of the secondary factor and the classification condition to form a judgment result sequence, thereby providing basis for subsequent final classification.

S211, if the secondary ecological factors and the secondary humane factors meet the classification conditions, the corresponding positions in the initial sequence are marked as 1, and if the secondary ecological factors and the secondary humane factors do not meet the classification conditions, the corresponding positions are marked as 0, so that the matching sequence is obtained.

After the judgment, the judgment result needs to be encoded into a sequence. The satisfaction condition is denoted as 1, and the non-satisfaction is denoted as 0. The matching sequence reflects the matching condition of the target lake island and each classification standard.

Specifically, the multi-factor integrated lake island classification recognition system can adopt the following technical means to generate a matching sequence, wherein firstly, the system initializes a null sequence, the length of the null sequence is equal to the number of conditions in a classification list, then, if one condition is judged, 1 is written in a corresponding sequence position if the factor is met, otherwise 0 is written in, and finally, after the system finishes all the judgment, the matching sequence formed by 0-1 is obtained, for example, if the classification list has 4 conditions, the judgment result is 'met, unsatisfied, met and unsatisfied', the matching sequence is {1,0,1,0}.

S212, determining the classification type of the target lake island according to the matching sequence.

For example, it is assumed that a lake island classification system of class 1-ecological conservation class, condition 1 of no living, condition 2 of single management area, condition 3 of natural vegetation coverage >90%, condition 4 of no water profit, condition 5 of class 2-cultivation conservation class, condition 1 of no living residents, condition 2 of single management area, condition 3 of natural vegetation coverage 60% -90%, condition 4 of small amount of farming activities, condition 5 of no permanent building, class 3-protection coordination class condition 1 of scattered residents distribution, condition 2 of single or double management area, condition 3 of natural vegetation coverage 30% -60%, condition 4 of presence of natural vegetation coverage >90%, condition 5 of area 1-10 square kilometers, class 4-protection utilization class, condition 1 of relative residential points, condition 2 of composite management area, condition 3 of natural vegetation coverage <30%, condition 4 of presence of town, condition 5 of area >10 square kilometers, and obtaining a matching sequence of the target lake island as a number of {1,1,1,1,1,0,0,1,0,1,0,1,0,0,0}, the number of matching classes 1 of the matching classes is a maximum number of the matching system of classes 1, and the matching condition 1 is determined as a comparison condition number of matching classes 1, and the matching system is most frequently included in the matching condition 1.

S213, acquiring historical classification data of the target lake island, wherein the historical classification data comprises the type of the target lake island at a historical time point.

The historical classification data refers to classification type results of the target lake island at various time points in the past. The historical time points may be the last years of the last years or may be more fine grained monthly or quarterly time periods. The type refers to the type of the lake island determined according to the classification system, such as ecological conservation and the like. Acquisition refers to extracting the lake island classification results of the historical time period from the database or record.

Specifically, the multi-factor integrated lake island classification recognition system can acquire historical classification data of target lake islands by connecting a lake island classification database, wherein the database stores classification results of various lake islands for many years, including information such as lake island IDs, time, classification labels and the like. The system extracts classification results of the target lake island in the past 5 years or 10 years and the like according to the ID of the lake island in time sequence. These time-stamped classification results constitute historical classification data for the lake island. For example, the last year classification category of the target lake island in the last 10 years can be obtained, reflecting the evolution of the lake island category.

S214, constructing a time sequence model, and inputting the historical classification data into the time sequence model to obtain the classification type change trend of the target lake island.

The time series model is a statistical analysis model reflecting the characteristics of the time series data. After the historical classification data is input into the time sequence model, the change trend and mode of the data can be detected. The classification type change trend refers to the overall direction of the change of the classification class of the target lake island with time.

Specifically, the multi-factor integrated lake island classification recognition system can obtain classification variation trend by firstly, the system calls a time sequence analysis algorithm, such as an ARIMA model, to construct a time sequence model suitable for describing classification data. The system then enters the acquired historical classification data into the model in chronological order. Then, the system is simulated and trained to obtain model parameters. Finally, the system predicts the trend of the input classification sequence in the future time, and outputs the change trend of the classification type, such as gradually turning into ecological type, etc.

S215, calculating the type change period and the type change rate parameters of the target lake island according to the historical classification data.

Specifically, the lake island classification recognition system with multi-factor synthesis can calculate the parameters by the following technical means that the system firstly loads a time sequence of historical classification data, then adopts signal analysis algorithms such as Fourier transformation and the like to detect main frequency domain components of the sequence, and determines the period length of type change. The system can also calculate the variation of classification category near two time points and determine the variation rate of different time periods. Finally, the system performs statistical analysis on the change period and the change rate to obtain quantization parameters capable of expressing the classified dynamic change characteristics.

S216, inputting the type transformation period and the type transformation rate parameters into a type prediction model to obtain a prediction type.

The type change period reflects the characteristic that the classification type of the target lake island changes circularly according to a certain period. The type change rate reflects how fast the classification type changes. The two parameters can reasonably and quantitatively describe the dynamic characteristics in the classification evolution process of the lake island.

The type prediction model may be built by a machine learning algorithm for analyzing historical data and predicting future trends. The two parameters are input into the prediction model, so that the model can be helped to learn the periodic rule and the rate characteristic of the classification change of the lake island, and more accurate classification prediction is obtained.

Specifically, the following technical means may be adopted to build the type prediction model:

An LSTM recurrent neural network is constructed, a memory unit is arranged to store historical states and is used for model time series data, a network input layer access type change period and change rate are used as model feature parameters, hidden layers containing circulating structures in the middle of the network are arranged, time correlation knowledge is learned, an output layer is provided with a plurality of nodes to represent different classification types, prediction probabilities are generated through softmax, a marked historical classification data set is provided, a network model is trained, the change period and change rate parameters of test data are input, the operation model predicts, the output probability is analyzed, the classification type with the highest probability is determined to serve as a prediction result, the type prediction model can learn and simulate a lake island classification evolution process through inputting key time series feature parameters, and finally future prediction classification types are output.

S217, calculating the type similarity of the prediction type and the classification type.

The prediction type is a result of prediction by a time series model. The classification type is the classification result at the current time. And calculating the similarity of the two, and judging the deviation degree of the prediction result and the actual situation.

Specifically, a multi-factor integrated lake island classification recognition system may calculate type similarity by first representing both the predicted type and the classified type as vector forms of type codes. Then, the system calls a text similarity algorithm, such as hamming distance, editing distance, cosine similarity, and the like, and calculates the similarity degree of the two vectors. Finally, the system outputs a similarity value, if approaching 1 indicates that the prediction is accurate, if lower, the prediction generates larger deviation.

S218, if the type similarity is larger than a preset similarity threshold, generating lake island evolution early warning information and sending the information to the client.

The preset similarity threshold is a threshold predefined by the system for judging the prediction bias. Too low a degree of similarity in type indicates that a large deviation is predicted, and there may be a case where an abnormal change in the type of the lake island occurs. At the moment, the user needs to be reminded, and early warning information is generated and sent to the client.

Specifically, the multi-factor integrated lake island classification recognition system can generate and send early warning information by first presetting a threshold value of type similarity, such as 0.7. The system then determines whether the calculated similarity is below a threshold. If the function text or the audio alarm information is lower than the preset threshold, the system calls the early warning information generating module to generate the function text or the audio alarm information. And finally, the system sends early warning information to the client of the user terminal through an API interface so as to remind the user of paying attention to the abnormal evolution of the state of the lake island.

S219, if the type similarity is not greater than a preset similarity threshold, generating evolution prompt information according to the prediction type and sending the evolution prompt information to the client.

The prediction type is a future classification result predicted by the time series model. When the prediction deviates, prompt information needs to be generated according to the prediction type result to inform the user of the predicted classification and the possible evolution direction.

Specifically, the multi-factor integrated lake island classification recognition system can generate and send prompt information through the following technical means that the system firstly calculates the type similarity of prediction classification and actual classification, judges whether the similarity is lower than a preset threshold value of 0.8, and enters a prompt generation flow if the similarity is lower than the preset threshold value. And the system calls a natural language generation algorithm and generates text prompt information according to the prediction classification result. Such as "the system predicts that the lake island will evolve into protection coordination in the future". Speech synthesis may also be invoked to generate voice prompts.

The following describes a multi-factor integrated lake island classification and identification system in the embodiment of the present application from the perspective of hardware processing, please refer to fig. 3, which is a schematic diagram of a physical device structure of the multi-factor integrated lake island classification and identification system in the embodiment of the present application.

It should be noted that, the structure of the multi-factor integrated lake island classification recognition system shown in fig. 3 is only an example, and should not impose any limitation on the functions and application scope of the embodiments of the present invention.

As shown in fig. 3, the multi-factor integrated lake island classification recognition system includes a central processing unit (Central Processing Unit, CPU) 301 that can perform various appropriate actions and processes, such as performing the methods described in the above embodiments, according to a program stored in a Read-Only Memory (ROM) 302 or a program loaded from a storage section 308 into a random access Memory (Random Access Memory, RAM) 303. In the RAM 303, various programs and data required for the system operation are also stored. The CPU 301, ROM 302, and RAM 303 are connected to each other through a bus 304. An Input/Output (I/O) interface 305 is also connected to bus 304.

Connected to the I/O interface 305 are an input section 306 including an audio input device, a push button switch, and the like, an output section 307 including a Liquid crystal display (Liquid CRYSTAL DISPLAY, LCD) and an audio output device, an indicator lamp, and the like, a storage section 308 including a hard disk, and the like, and a communication section 309 including a network interface card such as a LAN (Local Area Network) card, a modem, and the like. The communication section 309 performs communication processing via a network such as the internet. The drive 310 is also connected to the I/O interface 305 as needed. A removable medium 311 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is installed on the drive 310 as needed, so that a computer program read therefrom is installed into the storage section 308 as needed.

In particular, according to embodiments of the present invention, the processes described above with reference to flowcharts may be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising a computer program for performing the method shown in the flowchart. In such an embodiment, the computer program may be downloaded and installed from a network via the communication portion 309, and/or installed from the removable medium 311. When the computer program is executed by a Central Processing Unit (CPU) 301, various functions defined in the present invention are performed.

Specific examples of a computer-readable storage medium include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-Only Memory (ROM), an erasable programmable read-Only Memory (Erasable Programmable Read Only Memory, EPROM), a flash Memory, an optical fiber, a portable compact disc read-Only Memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.

The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present invention. Where each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures.

Specifically, the multi-factor integrated lake island classification and identification system of the embodiment includes a processor and a memory, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the multi-factor integrated lake island classification and identification method provided by the embodiment is realized.

In another aspect, the present invention also provides a computer-readable storage medium, which may be included in the multi-factor integrated lake island classification and identification system described in the above embodiments, or may exist alone without being assembled into the multi-factor integrated lake island classification and identification system. The storage medium carries one or more computer programs which, when executed by a processor of the multi-factor integrated lake island classification recognition system, cause the multi-factor integrated lake island classification recognition system to implement the multi-factor integrated lake island classification recognition method provided in the above embodiment.

While the application has been described in detail with reference to the foregoing embodiments, it will be understood by those skilled in the art that the foregoing embodiments may be modified or equivalents may be substituted for some of the features thereof, and that the modifications or substitutions do not depart from the spirit of the embodiments.

As used in the above embodiments, the term "when..is interpreted as meaning" if..or "after..or" in response to determining..or "in response to detecting..is" depending on the context. Similarly, the phrase "when determining..or" if (a stated condition or event) is detected "may be interpreted to mean" if determined.+ -. "or" in response to determining.+ -. "or" when (a stated condition or event) is detected "or" in response to (a stated condition or event) "depending on the context.

Those of ordinary skill in the art will appreciate that implementing all or part of the above-described method embodiments may be accomplished by a computer program to instruct related hardware, the program may be stored in a computer readable storage medium, and the program may include the above-described method embodiments when executed. The storage medium includes a ROM or a random access memory RAM, a magnetic disk or an optical disk, and other various media capable of storing program codes.

Claims (10)

1. A multi-factor integrated lake island classification and identification method, which is characterized by being applied to a multi-factor integrated lake island classification and identification system, comprising:

acquiring image data of a target lake island, lake island environment data and influence factors in a target area, wherein the lake island environment data comprises lake island ecological data and lake island humane data, and the influence factors are factors influencing the classification of the target lake island;

Performing text matching on the influence factors according to preset environment monitoring data and preset humane data, determining the influence factors conforming to the preset environment monitoring data as first-level ecological factors, and determining the influence factors conforming to the preset humane data as first-level humane factors;

Detecting the image data and the lake island ecological data, extracting ecological factor parameters of the target lake island to obtain primary ecological data, and taking the primary ecological data as a secondary ecological factor;

Extracting the humane factor parameters of the target lake island according to the lake island humane data to obtain primary humane data, and taking the primary humane data as a secondary humane factor;

Determining the condition that the secondary ecological factors and the secondary humane factors accord with a preset lake island classification list, obtaining a matching sequence, and determining the classification type of the target lake island according to the matching sequence.

2. The method according to claim 1, wherein the step of text matching the influence factor according to the preset environmental monitoring data and the preset humane data, determining the influence factor conforming to the preset environmental monitoring data as a ecological factor, and determining the influence factor conforming to the preset humane data as a humane factor specifically comprises:

carrying out semantic analysis on the influence factors, and extracting influence keywords of the influence factors;

Constructing a knowledge graph comprising the preset environmental monitoring data and the preset humane data;

Calculating the influence similarity between the influence factors and the nodes in the knowledge graph;

And if the influence similarity is larger than a preset influence threshold, determining the influence factor as a ecological factor, and if the influence similarity is not larger than the preset influence threshold, determining the influence factor as a human factor.

3. The method according to claim 1, wherein the step of detecting the image data and the lake island ecological data and extracting the ecological factor parameters of the target lake island to obtain first-level ecological data specifically comprises:

Dividing the image data, and extracting geographic characteristic parameters of the target lake island, wherein the geographic characteristic parameters comprise a lake island area, a water body range and a vegetation coverage area;

Fusing the geographic characteristic parameters with the lake island ecological data to obtain ecological characteristic vectors of the target lake island, wherein the lake island ecological data comprises vegetation types, cultivation conditions and construction conditions;

Calculating the statistical characteristic value of each parameter in the ecological characteristic vector, and taking the statistical characteristic values of all the parameters as primary ecological data, wherein the primary ecological data comprises a lake island area, a water body range, a vegetation coverage area, a vegetation type, a cultivation condition and a construction condition.

4. The method according to claim 1, wherein the step of determining that the secondary ecological factor and the secondary humane factor meet a preset lake island classification list, obtaining a matching sequence, and determining the classification type of the target lake island according to the matching sequence specifically includes:

Judging the secondary ecological factors, the secondary humane factors and each classification condition in the preset lake island classification list in sequence;

If the secondary ecological factors and the secondary humane factors meet the classification conditions, the corresponding positions in the initial sequence are marked as 1, and if the secondary ecological factors and the secondary humane factors do not meet the classification conditions, the corresponding positions are marked as 0, so that a matching sequence is obtained;

and determining the classification type of the target lake island according to the matching sequence.

5. The method according to claim 1, wherein after the step of determining that the secondary ecological factor and the secondary humane factor meet a preset classification list of the lake island, obtaining a matching sequence, and determining the classification type of the target lake island according to the matching sequence, the method further comprises:

Acquiring historical classification data of the target lake island, wherein the historical classification data comprises the type of the target lake island at a historical time point;

Constructing a time sequence model, and inputting the historical classification data into the time sequence model to obtain a classification type change trend of the target lake island;

predicting the type of the target lake island at a preset future time point according to the classification type change trend to obtain a prediction type;

Calculating the type similarity of the prediction type and the classification type;

if the type similarity is larger than a preset similarity threshold, generating lake island evolution early warning information and sending the information to a client.

6. The method according to claim 5, wherein the step of predicting the type of the target lake island at a preset future point in time according to the classification type variation trend specifically comprises:

calculating the type change period and the type change rate parameters of the target lake island according to the historical classification data;

and inputting the type change period and the type change rate parameter into a type prediction model to obtain a prediction type.

7. The method according to claim 6, wherein after the step of calculating the type similarity of the prediction type and the classification type, the method comprises:

And if the type similarity is not greater than a preset similarity threshold, generating evolution prompt information according to the prediction type and sending the evolution prompt information to the client.

8. A multi-factor integrated lake island classification recognition system comprising one or more processors and memory coupled to the one or more processors, the memory to store computer program code comprising computer instructions that the one or more processors invoke to cause the multi-factor integrated lake island classification recognition system to perform the method of any of claims 1-7.

9. A computer readable storage medium comprising instructions which, when run on a multi-factor integrated lake island classification recognition system, cause the multi-factor integrated lake island classification recognition system to perform the method of any of claims 1-7.

10. A computer program product, characterized in that the computer program product, when run on a multi-factor integrated lake island classification recognition system, causes the multi-factor integrated lake island classification recognition system to perform the method of any one of claims 1-7.