Key species population spatial distribution quantification method and system in complex ecosystem

Abstract

The invention provides a key species population spatial distribution quantification method and system in a complex ecological system, and the method comprises the steps: carrying out the feature extraction of biological and environmental data collected in a target ecological region, and carrying out the fusion of the extracted multi-modal features, and generating a comprehensive feature vector; based on the target large model, analyzing the comprehensive feature vector, identifying species in the target ecological region, and obtaining species identification information of each species; based on the species identification information and the comprehensive feature vector corresponding to the target ecological region, quantifying a species population space distribution condition, and generating species population quantification information at least comprising a population distribution diagram, a population density diagram and a population dynamic change diagram; and based on the species population quantitative information and the collected original data, generating and displaying a quantitative report. According to the method, the species can be intelligently, efficiently and accurately identified based on the large model, the distribution condition and dynamic change of the species can be visually displayed, a quantitative report is provided, and a powerful basis is provided for ecological protection and management.

Description

Technical Field

[0001] The present application relates to the field of ecological technology, and in particular to a method and system for quantifying the spatial distribution of key species populations in complex ecosystems. Background Art

[0002] With rapid socioeconomic development and a growing population, humanity's impact on ecosystems is becoming increasingly significant. The health and stability of ecosystems are closely linked to the development of human society. A healthy ecosystem not only ensures its normal function and biodiversity, but also provides humans with a wealth of ecosystem services, supporting economic, social, and cultural development.

[0003] Species diversity is a key factor in the health and stability of ecosystems. It plays a vital role in maintaining ecological balance, providing ecosystem services, and ensuring the sustainable development of human society. The spatial distribution of key species populations is crucial for ecological conservation, resource management, and environmental planning. Understanding species distribution through scientific surveys and monitoring can help develop more effective conservation and management measures.

[0004] Currently, traditional methods for surveying species distribution suffer from numerous limitations: low data collection efficiency, variable data quality, limited data coverage, poor data timeliness, difficulty integrating data, disturbances to the ecological environment during the collection process and susceptibility to environmental factors, cumbersome data processing, difficulty in accurately quantifying species distribution, and high costs. These drawbacks limit the depth and breadth of species distribution research and hinder the scientific nature and effectiveness of ecological conservation and management decisions.

[0005] Moreover, when using large models for species identification in existing technologies, the large models have obvious deficiencies in cross-modal data fusion, domain adaptability, and edge deployment. For example, multimodal pre-training lacks a hierarchical alignment mechanism, leading to a cross-modal semantic gap problem. Summary of the Invention

[0006] In view of the above problems, embodiments of the present application provide a method and system for quantifying the spatial distribution of key species populations in complex ecosystems that overcome the above problems or at least partially solve the above problems.

[0007] In a first aspect, the embodiments of the present application provide a method for quantifying the spatial distribution of key species populations in a complex ecosystem, including: Extracting features from biological data and environmental data collected in the target ecological area, and fusing the extracted multimodal features to generate a comprehensive feature vector, wherein the biological data includes at least species image data, species audio data, and species location data, and the environmental data includes at least meteorological data and natural resource data; Analyzing the comprehensive feature vector based on the target large model, identifying species in the target ecological area, and obtaining species identification information of each species in the target ecological area; quantifying the spatial distribution of species populations based on the species identification information corresponding to the target ecological region and the comprehensive feature vector, and generating species population quantitative information, wherein the species population quantitative information includes at least a population distribution map, a population density map, and a population dynamic change map; Based on the quantitative information of the species population and the collected original data, a quantitative report that comprehensively describes the distribution characteristics of the species population is generated and displayed.

[0008] In a second aspect, the embodiments of the present application provide a system for quantifying the spatial distribution of key species populations in a complex ecosystem, including: An extraction and generation module for extracting features from biological data and environmental data collected in the target ecological area, and fusing the extracted multimodal features to generate a comprehensive feature vector, wherein the biological data includes at least species image data, species audio data, and species location data, and the environmental data includes at least meteorological data and natural resource data; An analysis and identification module, configured to analyze the comprehensive feature vector based on the target large model, identify species within the target ecological area, and obtain species identification information of each species within the target ecological area; A quantification generation module is used to quantify the spatial distribution of species populations based on the species identification information corresponding to the target ecological region and the comprehensive feature vector, and generate species population quantitative information, wherein the species population quantitative information at least includes a population distribution map, a population density map, and a population dynamic change map; A generation and display module is used to generate and display a quantitative report that comprehensively describes the distribution characteristics of species populations based on the quantitative information of the species populations and the collected original data.

[0009] The technical solution of the embodiment of the present application extracts features from the biological data and environmental data collected in the target ecological area, fuses the extracted multimodal features to generate a comprehensive feature vector, analyzes the comprehensive feature vector based on the target large model, and obtains species identification information of each species in the target ecological area. While realizing intelligent and efficient species identification based on large model technology, it can use multimodal feature fusion to provide comprehensive feature representation, thereby improving the accuracy of species identification; by combining species identification information and comprehensive feature vectors for analysis to generate species population quantitative information, and combining species population quantitative information and original data to generate quantitative reports, it can intuitively display the distribution and dynamic changes of species, and provide comprehensive and accurate high-quality quantitative reports, providing a strong basis for ecological protection and management. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 A schematic diagram showing a method for quantifying the spatial distribution of key species populations in a complex ecosystem provided by an embodiment of the present application; Figure 2 A flowchart of feature fusion based on extracted features provided in an embodiment of the present application is shown; Figure 3 A flowchart showing an overall implementation of the method for quantifying the spatial distribution of key species populations in a complex ecosystem provided by an embodiment of the present application; Figure 4 A schematic diagram showing a system for quantifying the spatial distribution of key species populations in a complex ecosystem provided by an embodiment of the present application; Figure 5 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0011] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0012] It should be understood that references throughout this specification to "one embodiment" or "an embodiment" mean that a particular feature, structure, or characteristic associated with the embodiment is included in at least one embodiment of the present application. Therefore, the appearance of "in one embodiment" or "in an embodiment" throughout this specification does not necessarily refer to the same embodiment. Furthermore, these particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. The term "a plurality" in the embodiments of the present application may include two or more.

[0013] In the various embodiments of the present application, it should be understood that the size of the serial numbers of the following processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0014] The present invention provides a method for quantifying the spatial distribution of key species populations in a complex ecosystem. Figure 1 Shown, including: Step 101: Extract features from biological data and environmental data collected in the target ecological area, and fuse the extracted multimodal features to generate a comprehensive feature vector. The biological data includes at least species image data, species audio data, and species location data, and the environmental data includes at least meteorological data and natural resource data.

[0015] Collect biological and environmental data from the target ecological area using data collection equipment. For example, the data collection equipment can be set to collect data at a set time. For example, the data collection equipment can be unmanned equipment (drones, unmanned boats) equipped with various sensors. By using data collection equipment for data collection, data collection efficiency can be improved, labor costs can be reduced, and the coverage and timeliness of the collected data can be guaranteed.

[0016] Collected biological data includes at least species image data, species audio data, and species location data. Collected environmental data includes at least meteorological data and natural resource data. Natural resource data includes, for example, land resource data (land use type, soil quality), water resource data, and vegetation cover. Specifically, environmental data includes, but is not limited to, temperature, humidity, soil conditions, light intensity, water resource conditions, air pressure, wind speed, season, vegetation cover, and other data related to the species' living environment.

[0017] After data collection using the acquisition device, feature extraction is performed on the collected biological and environmental data. The extracted multimodal features are then fused to generate a comprehensive feature vector. Multimodal features include both biological and environmental features. Biological features include image features extracted from image data, audio features extracted from audio data, and positioning features extracted from positioning data. The comprehensive feature vector generated through feature fusion contains rich information and can comprehensively reflect the characteristics of species populations under specific temporal and spatial conditions.

[0018] Step 102: Analyze the comprehensive feature vector based on the target large model, identify the species in the target ecological area, and obtain species identification information of each species in the target ecological area.

[0019] After determining the comprehensive feature vector that can comprehensively reflect the characteristics of the species population based on feature fusion, the comprehensive feature vector is inferred and analyzed based on the target large model that can perform species identification, and the species identification information of each species in the target ecological area output by the target large model is obtained, so as to use the large model for intelligent species identification.

[0020] By identifying species within the target ecological area based on large-scale model technology, large-scale data can be processed automatically, human intervention can be reduced, and species identification can be completed in a shorter time, thereby achieving intelligent species identification while ensuring identification efficiency; and the fusion of multimodal features can improve the robustness of the model and reduce the impact of noise and uncertainty of single-modal data on identification results.

[0021] Step 103: Based on the species identification information and comprehensive feature vector corresponding to the target ecological region, the spatial distribution of the species population is quantified to generate species population quantitative information. The species population quantitative information at least includes a population distribution map, a population density map, and a population dynamic change map.

[0022] After obtaining species identification information of each species in the target ecological area through species identification based on the target large model, the species identification information and comprehensive feature vectors are comprehensively analyzed to quantify the spatial distribution of species populations and generate species population quantitative information.

[0023] The species identification information provided by the target large model directly reflects the species presence within the target ecological region. Combined with temporal and spatial information, it can also reveal species distribution. The comprehensive feature vector, which incorporates features from image, audio, location, and environmental data, provides richer context, such as the species' habitat and geographic location, helping to understand the relationship between species distribution and environmental factors. By combining species identification information with the comprehensive feature vector, a more comprehensive understanding of species distribution can be achieved, including species type, abundance, distribution range, and their relationship to environmental factors.

[0024] Combining species identification information and comprehensive feature vectors to quantify the spatial distribution of species populations, the generated quantitative information includes at least population distribution maps, population density maps, and population dynamics maps. Population distribution maps can visually display the geographic distribution of species, helping to identify species habitats and ranges; population density maps are used to show the density distribution of species in different regions, helping to assess species survival and ecological pressures; population dynamics maps are used to show the distribution changes of species at different points in time, helping to predict future trends of species and provide a scientific basis for ecological protection and management.

[0025] When generating a population distribution map, the location information of the identified species is visualized to show the distribution of the species in geographic space, such as using GIS (geographic information system) software to plot the latitude and longitude information of the species on a map; when generating a population density map, the density distribution of species in different areas is calculated and visualized; when generating a population dynamic change map, the distribution changes of species at different time points are analyzed and visualized.

[0026] Step 104: Based on the species population quantitative information and the collected original data, a quantitative report that comprehensively describes the species population distribution characteristics is generated and displayed.

[0027] After quantifying the spatial distribution of species populations and generating species population quantitative information, a comprehensive analysis is conducted by combining the species population quantitative information with the original data provided by the collection equipment, and a quantitative report is generated for display. The generated quantitative report can describe the species population distribution characteristics relatively comprehensively and accurately.

[0028] Species population quantitative information, including population distribution maps, population density maps, and population dynamics maps, can visually demonstrate the geographic distribution and dynamics of species, providing insights into species distribution patterns across regions and time. Raw data provided by acquisition equipment can provide even more detailed information. Raw data can be used to verify the accuracy of quantitative results. For example, examining raw image and audio data can confirm the correctness of species identification. The detailed context provided by raw data can also help understand the reasons behind species distribution. For example, analyzing environmental data (such as water quality, soil quality, and meteorological data) can reveal the relationship between species distribution and environmental factors. Combining species population quantitative information with raw data to generate quantitative reports provides more comprehensive, accurate, and in-depth analysis, enhancing the accuracy and reliability of the analysis and resulting in high-quality quantitative reports.

[0029] The embodiment of the present application provides a method for quantifying the spatial distribution of key species populations in complex ecosystems. After feature extraction of biological data and environmental data collected in the target ecological area, the extracted multimodal features are fused to generate a comprehensive feature vector. The comprehensive feature vector is analyzed based on the target large model to obtain species identification information of each species in the target ecological area. While realizing intelligent and efficient species identification based on large model technology, multimodal feature fusion is used to provide a comprehensive feature representation, thereby improving the accuracy of species identification. By combining species identification information and comprehensive feature vectors for analysis to generate species population quantitative information, and combining species population quantitative information and original data to generate a quantitative report, the distribution and dynamic changes of species can be intuitively displayed, and a comprehensive and accurate high-quality quantitative report can be provided, providing a strong basis for ecological protection and management.

[0030] The following describes a scheme for generating a comprehensive feature vector based on feature fusion. This involves extracting features from biological and environmental data collected in the target ecological area, and fusing the extracted multimodal features to generate a comprehensive feature vector. Perform data preprocessing on the collected biological data and environmental data; Perform feature extraction on the preprocessed data to obtain multimodal features, where the multimodal features include biometric features and environmental features. The biometric features include image features, audio features, and positioning features. Align multimodal features in time and space; The aligned image features, audio features, positioning features, and environmental features are combined with corresponding weights to perform feature concatenation and generate a comprehensive feature vector. Among them, the weight corresponding to each modal feature is dynamically adjusted based on the feature contribution.

[0031] Data preprocessing is required before feature extraction can be performed on collected data. Preprocessing biological data can include, for example, performing image denoising, image enhancement, and image normalization on species image data to obtain images to be analyzed; removing background noise from species audio data and further processing the audio using normalization and pre-emphasis to obtain audio to be analyzed; and converting species location data into planar coordinates and combining it with auxiliary data (such as data from an inertial measurement unit) to obtain location data to be analyzed. Preprocessing environmental data can also include standardizing environmental data to a uniform range to obtain environmental data to be analyzed.

[0032] After data preprocessing, image features, audio features, positioning features, and environmental features are extracted from the preprocessed data to obtain multimodal features. For example, convolutional neural networks are used to extract image features, extract mel-spectrogram features from audio, extract positioning features from positioning data, and extract environmental features from environmental data. Environmental features are obtained by converting environmental data into feature vectors that can reflect the state of the species' living environment. For example, environmental data such as temperature and humidity are standardized and used as environmental features. When extracting features, it is necessary to focus on extracting features related to species identification and spatial distribution. After feature extraction, multimodal features are aligned in time and space based on timestamps. This means that the spatial location information of biological data and environmental data needs to be integrated to ensure the consistency of the data in time and space dimensions, providing a basis for subsequent population distribution analysis.

[0033] After feature alignment, the aligned image features, audio features, positioning features, and environmental features are combined with corresponding weights for feature splicing to generate a comprehensive feature vector. When generating the comprehensive feature vector, the aligned image features, audio features, positioning features, and environmental features can first be feature spliced to generate a first feature vector. Then, based on the importance and relevance of the different modal data, initial weights are assigned to each modality, and the assigned initial weights are integrated into the first feature vector to generate a second feature vector. The initial weights of environmental features are determined based on the degree of impact of environmental data on species survival and identification. The initial weights of each modal feature in the biological feature are, for example, based on the estimated importance of each modal feature in species identification and quantification of species spatial distribution. When generating the second feature vector, the comprehensive feature vector is generated by dynamically adjusting the weights. The dynamic weight adjustment process is based on the contribution of biological and environmental features to species identification and quantification of species spatial distribution. When dynamically adjusting the weights, the weights of each modal feature can also be adjusted based on an attention mechanism. It should be noted that since the embodiments of the present application require quantification of the spatial distribution of species populations, more emphasis should be placed on features related to species location when assigning initial weights and when dynamically adjusting weights. Traditional methods often rely on single-modal data (such as images or audio alone), resulting in limited recognition accuracy. Traditional modal fusion also lacks an alignment mechanism for cross-modal feature fusion, resulting in a semantic gap. Furthermore, the weights assigned to each modality are often fixed, making them inadequate for varying data quality across scenarios. This solution involves cross-modal feature extraction of biological data (images, audio, and positioning) and environmental data (meteorology and natural resources). Data preprocessing improves data quality, providing a better data foundation and stronger data support for subsequent species identification. Temporal and spatial alignment of multimodal features ensures data consistency across both temporal and spatial dimensions. By considering the weights of each modal feature during feature fusion and prioritizing features related to species location, this approach can provide comprehensive feature vectors tailored to the spatial distribution quantification and species identification tasks.

[0034] The following is combined with Figure 2 The specific implementation scheme of extracting features from biological data and environmental data and fusing features based on the extracted features is introduced with examples.

[0035] A. Extract basic features through a pre-trained multimodal encoder. The underlying network parameters are shared across modalities, and the top layer generates standardized feature representations through modality-specific adapters.

[0036] B. Feature fusion using three-level contrastive learning constraints. Specifically, the comparison is performed using the following three methods: (1) instance-level comparison: ensuring that cross-modal samples of the same ecological event are adjacent in the embedding space; (2) feature-level comparison: minimizing the Frobenius distance between feature distributions of different modalities; and (3) semantic-level comparison: aligning high-level semantic concepts using the CLIP loss. Hierarchical alignment is achieved through three-level contrastive learning (instance-level, feature-level, and semantic-level).

[0037] The overall solution for feature extraction and feature fusion is as follows: 1. Instance-level comparison,constructing cross-modal positive and negative samples.

[0038] In the data preprocessing stage: for image data, use YOLOv5 to detect the target area, then perform center cropping, and finally adjust the image resolution to, for example, 224×224; for audio data, use STFT (short-time Fourier transform) to convert it into a 128-dimensional Mel-spectrogram; for environmental data, normalize it to, for example, a 10-dimensional feature vector.

[0039] During the sample pair generation phase: Positive sample pairs are generated by combining data from different modalities at the same spatiotemporal point, such as pairing images with audio or images with environmental data. Negative sample pairs are generated by randomly combining data from different modalities across spatiotemporal points, such as pairing images with audio at different time points or images with environmental data at different time points.

[0040] In the loss calculation stage: the contrast loss function is used to calculate the ratio of the similarity between the positive sample pairs and the similarity between the negative sample pairs to optimize the model so that the similarity of the positive sample pairs is higher and the similarity of the negative sample pairs is lower. Among them, the similarity metric is, for example, the dot product form of the feature vector s ( v , a )= middle, T represents the transpose operation of the matrix, W is a learnable projection matrix, and Belongs to the feature vector. For example, the initial temperature coefficient τ = 0.1. The temperature coefficient τ is a hyperparameter in the contrast loss function, which is used to control the "temperature" of the similarity, that is, the sensitivity of the similarity. Here τ is used to scale the similarity s ( v , a ), when τ is small, the similarity is highly exponential, which means that a small change in the similarity will lead to a large change in the exponential term.

[0041] Through instance-level comparison, we can learn the association between data of different modalities, dynamically adjust the strength of the discriminant boundaries of positive and negative sample pairs, and generate preliminary feature representations.

[0042] 2. Feature-level comparison and modal distribution alignment.

[0043] Shared Encoder Architecture: We build a shared encoder consisting of shared convolutional layers (using a pre-trained ResNet34) and adapters for different modalities. The adapters map features from different modalities to a shared 256-dimensional space. Specifically, the adapter for the image modality maps 512-dimensional features to 256-dimensional, the adapter for the audio modality maps 128-dimensional features to 256-dimensional, and the adapter for the ambient modality maps 10-dimensional features to 256-dimensional.

[0044] Distribution alignment loss: Calculate the mean and covariance of each modal feature, using metrics such as the Frobenius norm and Euclidean norm to measure the difference between the two modal feature distributions to achieve distribution alignment. Use a sliding average to update the statistic, with a momentum coefficient of β = 0.9.

[0045] Through feature-level comparison, the generated features are consistent in the shared space, which controls the strictness of cross-modal distribution alignment and reduces the distribution differences between different modalities.

[0046] 3. Compare semantic levels and conduct concept space mapping.

[0047] Label semantic expansion: Build an ecological knowledge graph that links species with habitats and environmental factors, such as species → habitat → environmental factors. Use a model like BERT to encode species description text and extract concept embeddings.

[0048] Loss function: Calculate the cosine similarity between visual features and concept embedding vectors, and calculate the cosine similarity between auditory features and concept embedding vectors, and then return the mean square error loss of these two similarities to achieve alignment of visual features and auditory features with semantic concepts.

[0049] By comparing the semantic level and balancing the clustering granularity of the concept space, the generated features are not only consistent in distribution but also consistent in semantic level.

[0050] During adjustment, for example, adaptive temperature control is performed in the following manner. In the three-level contrastive learning framework, the adaptive temperature control mechanism does not directly affect the feature extraction process, but indirectly improves the quality of feature extraction by adjusting the feature representation optimization in the contrastive learning stage.

[0051] 1. Confidence assessment: Confidence assessment is the starting point of the entire dynamic temperature coefficient adjustment process. Feature quality is assessed based on the feature norm. The norm of the feature vector is calculated and normalized to obtain the confidence level.

[0052] 2. Temperature update rule. The temperature update rule is the core of dynamic temperature coefficient adjustment. It dynamically adjusts the temperature coefficient according to the results of confidence evaluation. Specifically, it dynamically adjusts the temperature coefficient according to the confidence average value of the features within the batch, the basic temperature coefficient and the attenuation coefficient. For example, based on the basic temperature coefficient τ base =0.15, attenuation coefficient k=0.05, each batch (B=256) is dynamically adjusted; each batch (B=256) means that each batch contains 256 samples.

[0053] 3. Hardware acceleration. Hardware acceleration is a means of implementing dynamic temperature coefficient adjustment. When deployed using TensorRT, temperature coefficient adjustment is implemented as a custom plug-in layer to improve computing efficiency.

[0054] Dynamic temperature coefficient adjustment enables adaptive adjustment of the similarity scaling. This mechanism indirectly adjusts the optimization target of the feature encoder to automatically reduce the impact of low-quality data on feature extraction, achieve modal balance, and prevent a single modality from dominating the feature representation space. The above embodiment, through a three-level adaptive temperature adjustment mechanism, constructs a framework from feature extraction to cross-modal fusion, effectively improving the hierarchical alignment mechanism of multimodal pre-training and narrowing the cross-modal semantic gap.

[0055] During cross-modal feature fusion, the adaptive temperature control mechanism indirectly affects the modal weight allocation in the subsequent fusion stage by adjusting the feature representation optimization of contrastive learning. In the cross-modal fusion link, the features of different modalities are spliced, and the relationship between the modalities is modeled through the attention mechanism, and finally fused through a weight matrix. The attention mechanism adopts, for example, the dot product attention form, and the dimension of the weight matrix is determined according to the feature dimension. The embodiment of the present application adopts a dynamic weight adjustment mechanism to adaptively allocate the weights of each modal feature based on the feature contribution (such as confidence), adapting to the data quality differences in different scenarios.

[0056] In the above process, instance-level comparison is the basis. By constructing positive and negative sample pairs, the comprehensive feature vector generation model (as a comprehensive multimodal learning system for generating high-quality comprehensive feature vectors) is strengthened to learn the association between different modal data; feature-level comparison is based on instance-level comparison, and further optimizes the consistency of features by aligning feature distributions; semantic-level comparison is based on feature-level comparison, and further aligns the features of different modalities by introducing semantic information to ensure their consistency at the semantic level; dynamic temperature coefficient adjustment runs through the entire training process. By dynamically adjusting the temperature coefficient, the training process of the model (comprehensive feature vector generation model) is optimized to improve the model's performance and generalization ability; the connection with the comprehensive feature vector is the ultimate goal, ensuring that the dynamic temperature coefficient adjustment mechanism is closely integrated with the final generated comprehensive feature vector to generate high-quality comprehensive feature vectors for subsequent tasks.

[0057] As a specific application example, in a wetland bird monitoring scenario, the input data includes visual data (egret flight images), audio data (200-800Hz call spectrograms), and environmental data (temperature 28°C, humidity 70%). The processing steps include: 1. Instance-level comparison to strengthen the positive association between "egret image and specific call"; 2. Feature-level alignment to eliminate feature distribution shifts caused by illumination variations; and 3. Semantic-level mapping to associate the concepts of "wading birds" and "wetland environment." In the output comprehensive feature vector, for example, the first 256 dimensions are visual features: [0.12, -0.05, ..., 0.78]; the middle 256 dimensions are audio features: [0.33, 0.41, ..., -0.12]; and the last 256 dimensions are environmental features: [0.91, 0.02, ..., 0.15].

[0058] In an optional embodiment of the present application, when species identification is performed based on a large target model, the process includes: Analyze the comprehensive feature vector based on the target large model to obtain identification reference information of each species in the target ecological area, where the identification reference information includes at least one reference name and the confidence level corresponding to the reference name; For each species in the target ecological area, the species identification reference information is corrected based on the species prior information to obtain species identification information.

[0059] After obtaining the comprehensive feature vector, the target large model is optimized based on the comprehensive feature vector to enhance the target large model's ability to understand multimodal data. This optimization method, for example, involves fusing the comprehensive feature vector with the feature representation of the target large model to enhance the target large model's ability to understand multimodal data. After optimizing the target large model, the optimized target large model performs inference analysis on the comprehensive feature vector and provides identification reference information for each species within the target ecological area. The species identification reference information includes at least one reference name and a corresponding confidence level.

[0060] After obtaining the species identification reference information, the identification reference information is modified based on the species' prior information to obtain the species identification information. Species prior information refers to known or assumed information about the species' characteristics, behavior, distribution, and other characteristics. Species prior information helps to better understand and predict the species' behavior and distribution. Modifying the identification reference information based on species prior information actually involves modifying the confidence level corresponding to the reference name. After modifying the confidence level corresponding to the reference name using the species prior information, the species identification information (species identification information includes the species name and the species name confidence level) is determined based on the confidence level corresponding to at least one of the modified reference names. For example, the reference name with the highest confidence level is determined as the final species name.

[0061] The above implementation process optimizes the target large model's ability to understand multimodal data, thereby enhancing the model's recognition accuracy and improving cross-modal reasoning accuracy. By processing the integrated feature vectors based on the optimized target large model and outputting species identification reference information, and correcting this identification reference information based on species prior information, species identification can be performed relatively accurately and efficiently. This effectively addresses the problems of traditional models' poor adaptability to multimodal data and insufficient edge deployment capabilities, as well as the lack of prior knowledge correction, which makes low-confidence recognition results susceptible to noise.

[0062] The following describes the process of quantifying the spatial distribution of species populations. Species identification information includes species names and species name confidence scores. When quantifying the spatial distribution of species populations based on the species identification information and comprehensive feature vectors corresponding to the target ecological region, generating quantitative species population information includes: Species whose names have a confidence score below a preset threshold within the target ecoregion are filtered out, and a species set is determined based on the retained species, at least some of which are key species. Key species are species whose impact on the target ecoregion meets preset conditions. Based on the species identification information and comprehensive feature vectors corresponding to the key species in the species collection, a population distribution map indicating the distribution location of key species, a population density map indicating the density distribution of key species, and a population dynamic change map indicating the changes of key species at different time points are generated.

[0063] After obtaining species identification information for each species within the target ecological area, species filtering is performed based on the confidence level of the species names contained in the species identification information to remove species with a confidence level below a preset threshold. Because high-confidence identification results are generally more reliable, filtering out low-confidence identification results ensures that species with high identification accuracy are provided.

[0064] After filtering out species with a confidence score below a preset threshold, a species set is determined based on the remaining species. At least some of the species in this set are key species. Key species can be understood as species whose impact on the target ecological region meets preset conditions. For example, species that meet at least one of the following conditions are considered key species: 1. They play an important role in maintaining the structure and function of the ecosystem; 2. They have a significant impact on the survival of other species and the ecosystem; 3. They play a unique role in the ecosystem and are difficult to be replaced by other species.

[0065] After filtering the species within the identified target ecological region to determine the species set, population distribution maps, population density maps, and population dynamics maps are generated based on the species identification information and comprehensive feature vectors corresponding to the key species in the species set. The generated population distribution map is used to indicate the geographical location distribution of key species, the generated population density map is used to indicate the density distribution of key species in different regions, and the generated population dynamics map is used to indicate the changes in key species at different points in time.

[0066] It should be noted that the embodiment of the present application does not necessarily require a large model to process when comprehensively analyzing species identification information and comprehensive feature vectors, quantifying the spatial distribution of species populations, and generating quantitative information on species populations. It relies more on GIS tools, statistical analysis methods, and visualization technologies. However, a large model can be used to provide additional help in this process. For example, the large model can predict the future distribution and dynamic changes of species based on historical data and environmental factors; the large model can process complex multimodal data and provide more in-depth analysis and insights; the large model can automatically process large-scale data and improve analysis efficiency. The large model at this stage processes time series and geospatial data, which can be used to predict the future distribution and dynamic changes of species, as well as to provide in-depth analysis results; that is, the large model at this stage is different from the target large model used for species identification mentioned above.

[0067] Among them, when generating a population distribution map indicating the distribution location of key species based on the species identification information and comprehensive feature vectors corresponding to the key species in the species set, it includes: Extract the location information of key species at a specified time from the comprehensive feature vector; Based on the location information of key species at a specified time and the species identification information of key species, a population distribution map indicating the distribution of key species at a specified time is generated.

[0068] The population distribution map of this embodiment represents the distribution of key species at a specified time (e.g., a specific point in time or a specific time period). Before generating the population distribution map, key species within the species set must be identified. Generating the population distribution map primarily involves the following operations: 1. Geographic coordinate positioning; 2. Species information association; and 3. Map creation. During geographic coordinate positioning, location information related to the key species' observed location, such as longitude and latitude coordinates, is extracted from the comprehensive feature vector. This coordinate information is derived from the positioning data provided by the acquisition device. During species information association, the names of key species are associated with their corresponding geographic coordinates. For example, if a key species is identified at a specific coordinate point, the species name of the key species is marked at that coordinate point. During map creation, the associated species names and coordinate points are plotted on a map using the geographic region as a background. This can be accomplished using GIS software or related mapping tools. Each species can be represented by a different color or symbol, thereby visually displaying the spatial distribution of different key species populations and forming a population distribution map.

[0069] Among them, when generating a population density map indicating the density distribution of key species based on the species identification information and comprehensive feature vectors corresponding to the key species in the species set, it includes: Divide the target ecological area into multiple grids; Based on the location information of key species at a specified time extracted from the comprehensive feature vector, key species are assigned to corresponding grids; For each grid, the grid weighted density is determined based on the species weights corresponding to each key species contained in the grid. The species weights are determined based on the confidence level of the species names contained in the species identification information corresponding to the key species. The weighted density distribution of the target ecological region is determined based on the grid weighted density corresponding to each grid, and the weighted density distribution of the target ecological region is visualized to generate a population density map.

[0070] The population density map in this embodiment represents the density distribution of key species at a specified time. Before generating the population density map, key species within the species set must be identified. Generating a population density map primarily involves the following stages: 1. Regional demarcation; 2. Species count; 3. Density calculation; and 4. Density map creation. In the first stage, the target ecological area is divided into multiple grids. The grid size can be determined based on the research accuracy requirements and the density of the data. For example, smaller grids can be used in areas with denser species distribution. In the second stage, key species are assigned to corresponding grids based on the location information of key species at a specified time extracted from the comprehensive feature vector. In the third stage, the population density within each grid is calculated. The calculation of population density within a grid also requires consideration of the species weights corresponding to each key species within the grid. Species weights can be determined based on the confidence level of the species name, for example, using the species name confidence level or the square of the species name confidence level as the weight. After determining the species weights corresponding to each key species within the grid, the weights of each species are accumulated to obtain the grid-weighted density corresponding to the grid. The number of key species contained in the grid also needs to be considered during the calculation. If there are multiple key species of the same kind in the grid, then when calculating the grid weighted density, the species weights of each key species of the same kind need to be added together. In the fourth stage, the weighted density distribution of the target ecological area is determined based on the grid weighted density corresponding to each grid in the target ecological area, and the weighted density distribution is visualized to generate a population density map. For example, after calculating the grid weighted density corresponding to each grid, a population density map is drawn on the map using color gradients or contour lines. The darker the color or the denser the contour lines, the higher the population density, and vice versa, thus intuitively showing the density of key species populations in different areas.

[0071] Since confidence can be regarded as the probability of the actual existence of a species, generating a weighted population density map based on a weighted operation of confidence can make the population density map reflect not only the number of species, but also the probability of species existence, while making the density value in the map closer to the possibility of the actual existence of the species, thereby more realistically reflecting the distribution of species in geographical space; and the weighted population density map can more clearly highlight areas with higher species density, and this highlighting effect helps to quickly identify key ecological areas.

[0072] The population density map generated above reflects the overall distribution of all key species within the grid, but cannot directly reflect the density of a single species. If the density of a single species needs to be analyzed, the weighted density of a single key species in each grid can be calculated to obtain the weighted density distribution of a single key species in the entire target ecological area.

[0073] In particular, when generating a population dynamics change diagram indicating the changes of key species at different time points based on the species identification information and comprehensive feature vectors corresponding to the key species in the species set, it includes: Record multiple time nodes of the target large model analyzing the comprehensive feature vector within the target time interval to generate a species identification time series. The comprehensive feature vector is dynamically updated within the target time interval. At each time point in the species identification time series, based on the comprehensive feature vector corresponding to the current time point and the species identification information corresponding to the key species in the species set, the number of species corresponding to each key species is counted, and / or the species distribution corresponding to each key species is counted to generate the key species population information corresponding to the current time point; Compare the key species population information corresponding to each time node of the species identification time series to generate a population dynamic change diagram.

[0074] The target time interval in this embodiment is, for example, a historical time interval, a comprehensive time interval including a historical time interval and a future time interval, and may also be a future time interval in special scenarios. Based on the analysis of multiple time nodes of the comprehensive feature vector within the target time interval, the species identification time series is determined. The comprehensive feature vector is dynamically updated within the target time interval. For the case where the target time interval is associated with the future time interval, the comprehensive feature vector can be obtained through model prediction, such as generating a comprehensive feature vector at a certain time point in the future based on historical data and a prediction model. Correspondingly, since the species identification information is determined by analyzing the comprehensive feature vector by the target large model, if the comprehensive feature vector is obtained through prediction, the matching species identification information can be obtained based on the comprehensive feature vector predicted by the target large model analysis. Under normal circumstances, the species identification information obtained based on the analysis of different comprehensive feature vectors by the target large model is not much different, and the key species in the species set remain unchanged.

[0075] At each time point in the species identification time series, based on the comprehensive feature vector and species identification information corresponding to the current time point, at least one of the following items, namely, the number of species corresponding to each key species and the species distribution corresponding to each key species, is counted. Based on these statistics, key species population information corresponding to the current time point is generated. For any time point, its corresponding comprehensive feature vector can be determined based on the data collected during the period corresponding to that time point and the previous time point. If the time point is in the future, the corresponding comprehensive feature vector can be obtained through prediction.

[0076] When generating the key species population information corresponding to the current time node, it includes: after determining the key species in the species set, based on the species identification information of the key species, extracting at least one of the species quantity and location information of each key species in the comprehensive feature vector; generating the key species population information based on at least one of the species quantity and species distribution of each key species, and the species distribution of the key species is determined based on the location information.

[0077] After the key species are determined at the current time node, for each key species, the species number and / or location information of the key species is extracted from the comprehensive feature vector corresponding to the current time node. If the key species is an animal, the extracted location information is the species activity area information. After obtaining the location information of the key species, the species distribution of the key species can be determined. For any key species, after extracting the species number and / or location information, the statistical information corresponding to the key species can be determined, and the statistical information corresponding to each key species can be aggregated to obtain the key species population information at the current time node.

[0078] After counting the key species population information at each time node, the key species population information corresponding to each time node of the species identification time series is compared and visualized to generate a population dynamic change graph.

[0079] In another implementation scheme for generating a population dynamics map, the key species of the target ecological area are determined in advance, and data related to the observation time of the key species are extracted from multiple comprehensive feature vectors, and the observation records of the same key species at different time points are organized into time series data. For example, the number of observations and location information of a key species at different time intervals such as daily, weekly or monthly are recorded. For each key species, its number changes at different time points are analyzed, such as by calculating the difference in number or the rate of change between adjacent time points to quantify the dynamic changes in population size. The position changes of key species at different time points are analyzed to determine the changes in the migration path or activity range of the species. For example, by comparing the geographical coordinates at different times, it can be found that a key species migrated from one area to another over a period of time. For multiple key species, the information on population changes and location changes is combined and displayed dynamically on the map. For example, lines of different colors are used to represent the migration paths of species, and the thickness of the lines can indicate the scale of the migration; circles of different sizes are used to represent the number of species at different time points, and the position of the circles indicates the distribution location of the species. As time goes by, these graphic elements change on the map, thereby intuitively showing the dynamic changes in the populations of key species.

[0080] In the above implementation process, after determining the key species in the target ecological region, a population distribution map indicating the distribution location of key species, a population density map indicating the density distribution of key species, and a population dynamic change map indicating the changes of key species at different time points are generated to quantify the spatial distribution of key species populations. This plan introduces species weights to calculate grid weighted density, which more realistically reflects ecological pressure, can help researchers understand the ecosystem structure of the target ecological region, determine key areas for biodiversity conservation, and provide basic data for predicting ecosystem changes.

[0081] The following describes the solution for generating quantitative reports. Based on the quantitative information of species populations and the collected raw data, a quantitative report that comprehensively describes the distribution characteristics of species populations is generated and presented, including: Conduct a comprehensive analysis of population distribution maps, population density maps, population dynamic change maps, and collected raw data to generate statistical analysis results that include at least changes in population size, population distribution range, population density changes, hot spots, factors affecting population dynamic changes, and ecological pressure in target ecological areas; generate and display a quantitative report based on the statistical analysis results and population distribution maps, population density maps, and population dynamic change maps.

[0082] The raw data provided by the collection equipment can provide more detailed information to verify the accuracy of the quantitative results. The detailed context provided by the raw data can also help understand the reasons for species distribution. The generated population distribution map is used to show the geographical distribution of species, the generated population density map is used to show the density distribution of species in different areas, and the generated population dynamic change map is used to show the distribution and abundance changes of species at different points in time.

[0083] By combining population distribution maps, population density maps, population dynamic change maps and original data, the following analysis can be performed: Count the number of key species within each time period and analyze their changing trends to determine population size changes; analyze the geographic distribution of species to identify their activity areas and determine their population distribution ranges; calculate and visualize the density distribution and changes of species in different regions to determine population density changes; identify areas with high species density to identify species hotspots; analyze the impact of environmental data on population dynamics to identify factors influencing population dynamics; and assess ecological pressures in the target ecological region, including the combined impact of species abundance, distribution range, density changes, and environmental factors. It should be noted that when combining population distribution maps, population density maps, and population dynamics maps with raw data for analysis, multiple population distribution maps and population density maps can be combined to compare population distribution and density over time. Multiple population distribution maps and population density maps can be generated based on multiple adapted comprehensive feature vectors.

[0084] The statistical analysis results obtained based on the above analysis include: population size changes that show the changes in species numbers over time, population distribution ranges that show the distribution range of species in geographical space, population density changes that show the density distribution of species in different regions and its changes, hot spots with high species density, factors affecting population dynamics (environmental factors), and ecological pressures in target ecological regions.

[0085] After obtaining the statistical analysis results, the statistical analysis results and population distribution maps, population density maps, and population dynamic change maps are summarized, and a quantitative report containing statistical analysis (such as migration paths and environmental impacts) is automatically generated and displayed, providing researchers with high-quality quantitative reports and providing a scientific basis for ecological protection and management.

[0086] The following is an introduction to the method for quantifying the spatial distribution of key species populations in complex ecosystems provided by the embodiment of the present application through an overall implementation process. Figure 3 As shown, the following steps are included: Step 301: Preprocess the biological data and environmental data collected in the target ecological area.

[0087] Step 302: Extract features from the preprocessed data to obtain multimodal features, and align the multimodal features in time and space. The multimodal features include image features, audio features, positioning features, and environmental features.

[0088] Step 303: Combine the aligned image features, audio features, positioning features, and environmental features with corresponding weights to perform feature concatenation and generate a comprehensive feature vector.

[0089] Step 304: Analyze the comprehensive feature vector based on the target large model to obtain identification reference information of each species in the target ecological area. The identification reference information includes at least one reference name and a confidence level corresponding to the reference name.

[0090] Step 305: For each species in the target ecological area, the species identification reference information is modified based on the species prior information to obtain species identification information, where the species identification information includes the species name and the species name confidence level.

[0091] Step 306: Filter species in the target ecological area whose species name confidence is less than a preset threshold, and determine a species set based on the retained species, at least some of the species in the species set being key species.

[0092] Step 307: Based on the species identification information and comprehensive feature vectors corresponding to the key species in the species set, generate a population distribution map indicating the distribution location of the key species, a population density map indicating the density distribution of the key species, and a population dynamic change map indicating the changes of the key species at different time points.

[0093] Step 308: Comprehensively analyze the population distribution map, population density map, population dynamic change map, and the collected original data to generate statistical analysis results that include at least population size changes, population distribution range, population density changes, hot spots, factors affecting population dynamic changes, and ecological pressure in the target ecological area.

[0094] Step 309: Generate and display a quantitative report based on the statistical analysis results and the population distribution map, population density map, and population dynamic change map.

[0095] exist Figure 3 In the overall implementation process shown, large model technology is used to achieve intelligent and efficient species identification, and multimodal feature fusion is used to provide comprehensive feature representation, which can improve the accuracy of species identification; by combining species identification information and comprehensive feature vectors for analysis to generate species population quantitative information, and combining species population quantitative information and original data to generate quantitative reports, it can intuitively display the distribution and dynamic changes of species, and provide comprehensive and accurate high-quality quantitative reports, providing a strong basis for ecological protection and management.

[0096] The present application embodiment provides a system for quantifying the spatial distribution of key species populations in a complex ecosystem, such as Figure 4 As shown, including: Extraction and generation module 401 is used to extract features from biological data and environmental data collected in the target ecological area, and fuse the extracted multimodal features to generate a comprehensive feature vector. The biological data includes at least species image data, species audio data, and species location data. The environmental data includes at least meteorological data and natural resource data. An analysis and identification module 402 is configured to analyze the comprehensive feature vector based on the target macro model, identify species within the target ecological area, and obtain species identification information of each species within the target ecological area; A quantitative generation module 403 is configured to quantify the spatial distribution of species populations based on the species identification information corresponding to the target ecological region and the comprehensive feature vector, and generate species population quantitative information, wherein the species population quantitative information includes at least a population distribution map, a population density map, and a population dynamic change map; The generation and display module 404 is used to generate and display a quantitative report that comprehensively describes the distribution characteristics of the species population based on the species population quantitative information and the collected original data.

[0097] Optionally, the extraction and generation module includes: The preprocessing submodule is used to preprocess the collected biological data and environmental data; An extraction submodule, configured to extract features from the preprocessed data to obtain multimodal features, wherein the multimodal features include biological features and environmental features, and the biological features include image features, audio features, and positioning features; an alignment submodule, configured to align the multimodal features in time and space; A splicing generation submodule is used to combine the aligned image features, audio features, positioning features, and environmental features with corresponding weights to perform feature splicing to generate the comprehensive feature vector; Among them, the weight corresponding to each modal feature is dynamically adjusted based on the feature contribution.

[0098] Optionally, the analysis and identification module includes: an analysis and acquisition submodule, configured to analyze the comprehensive feature vector based on the target macromodel to obtain identification reference information of each species in the target ecological area, wherein the identification reference information includes at least one reference name and a confidence level corresponding to the reference name; The correction and acquisition submodule is used to correct the species identification reference information for each species in the target ecological area based on the species prior information to obtain the species identification information.

[0099] Optionally, the species identification information includes the species name and the species name confidence; and the quantification generation module includes: A filtering and determining submodule, configured to filter species within the target ecological area whose species name confidence is less than a preset threshold, and determine a species set based on the retained species; a generation submodule for generating, based on the species identification information corresponding to the key species in the species set and the comprehensive feature vector, a population distribution map indicating the distribution location of the key species, a population density map indicating the density distribution of the key species, and a population dynamic change map indicating the change of the key species at different time points; Among them, at least some of the species in the species set belong to the key species, and the key species are species whose influence on the target ecological area meets preset conditions.

[0100] Optionally, the generating submodule includes: An extraction unit, configured to extract the location information of the key species at a specified time from the comprehensive feature vector; A generating unit is configured to generate a population distribution map indicating the distribution of the key species at the specified time based on the location information of the key species at the specified time and the species identification information of the key species.

[0101] Optionally, the generating submodule includes: A division unit, used for dividing the target ecological area into a plurality of grids; an allocating unit, configured to allocate the key species to corresponding grids based on the location information of the key species at a specified time extracted from the comprehensive feature vector; a determining unit configured to determine, for each grid, a grid weighted density based on species weights corresponding to respective key species contained in the grid, wherein the species weights are determined based on confidence levels of species names contained in species identification information corresponding to the key species; A determination generating unit is configured to determine a weighted density distribution of the target ecological region based on the grid weighted densities corresponding to each grid, and to visualize the weighted density distribution of the target ecological region to generate the population density map.

[0102] Optionally, the generating submodule includes: a record generation unit, configured to record a plurality of time nodes at which the target large model analyzes the comprehensive feature vector within a target time interval, and generate a species identification time series, wherein the comprehensive feature vector is dynamically updated within the target time interval; a statistics generating unit configured to, at each time node of the species identification time series, count the number of species corresponding to each key species and / or count the species distribution corresponding to each key species based on the comprehensive feature vector corresponding to the current time node and the species identification information corresponding to the key species in the species set, and generate key species population information corresponding to the current time node; The comparison and generation unit is used to compare the key species population information corresponding to each time node of the species identification time series to generate the population dynamic change graph.

[0103] Optionally, the statistics generating unit is further configured to: After determining the key species in the species set, extracting at least one of species quantity and location information of each key species from the comprehensive feature vector based on the species identification information of the key species; The key species population information is generated based on at least one of the species quantity and species distribution of each key species, and the species distribution of the key species is determined based on the location information.

[0104] Optionally, the generating and displaying module includes: An analysis and generation submodule is used to conduct a comprehensive analysis of the population distribution map, the population density map, the population dynamic change map, and the collected raw data to generate statistical analysis results including at least population size changes, population distribution range, population density changes, hot spots, factors affecting population dynamic changes, and ecological pressure in the target ecological area; A display submodule is generated, which is used to generate and display the quantitative report based on the statistical analysis results and the population distribution map, the population density map, and the population dynamic change map.

[0105] As for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0106] An embodiment of the present application also provides an electronic device, including: a processor, a memory, and a computer program stored in the memory and runnable on the processor. When the computer program is executed by the processor, the various processes of the embodiment of the method for quantifying the spatial distribution of key species populations in complex ecosystems described above are implemented, and the same technical effects can be achieved. To avoid repetition, they will not be described here.

[0107] For example, Figure 5 FIG. 1 shows a schematic diagram of the physical structure of an electronic device. Figure 5As shown, the electronic device may include: a processor 510, a communications interface 520, a memory 530, and a communication bus 540. The processor 510, the communications interface 520, and the memory 530 communicate with each other via the communication bus 540. The processor 510 may call logic instructions stored in the memory 530. The processor 510 is used to execute each process of the method for quantifying the spatial distribution of key species populations in a complex ecosystem according to an embodiment of the present application, which will not be elaborated on here.

[0108] In addition, the logic instructions in the aforementioned memory 530 can be implemented in the form of a software functional unit and, when sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the portion that contributes to the prior art, or the portion of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application.

[0109] The present application also provides a computer-readable storage medium having a computer program stored thereon. When executed by a processor, the computer program implements the various processes of the aforementioned embodiment of the method for quantifying the spatial distribution of key species populations in a complex ecosystem, and can achieve the same technical effects. To avoid repetition, the description is omitted here. The computer-readable storage medium may be, for example, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0110] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0111] Through the description of the above embodiments, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases the former is a more preferred embodiment. Based on this understanding, the technical solution of this application, or the part that contributes to the existing technology, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a terminal (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of this application.

[0112] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of this application, ordinary technicians in this field can also make many forms without departing from the purpose of this application and the scope of protection of the claims, all of which are within the protection of this application.

Claims

1. A method for quantifying the spatial distribution of key species populations in complex ecosystems, characterized by: include: Extracting features from biological data and environmental data collected in the target ecological area, and fusing the extracted multimodal features to generate a comprehensive feature vector, wherein the biological data includes at least species image data, species audio data, and species location data, and the environmental data includes at least meteorological data and natural resource data; Analyzing the comprehensive feature vector based on the target large model, identifying species in the target ecological area, and obtaining species identification information of each species in the target ecological area; quantifying the spatial distribution of species populations based on the species identification information corresponding to the target ecological region and the comprehensive feature vector, and generating species population quantitative information, wherein the species population quantitative information includes at least a population distribution map, a population density map, and a population dynamic change map; Based on the quantitative information of the species population and the collected original data, a quantitative report that comprehensively describes the distribution characteristics of the species population is generated and displayed.

2. The method for quantifying the spatial distribution of key species populations in complex ecosystems according to claim 1, characterized in that: The feature extraction of biological data and environmental data collected in the target ecological area and the fusion of the extracted multimodal features to generate a comprehensive feature vector include: Perform data preprocessing on the collected biological data and environmental data; Extracting features from the preprocessed data to obtain multimodal features, wherein the multimodal features include biological features and environmental features, and the biological features include image features, audio features, and positioning features; aligning the multimodal features in time and space; The aligned image features, audio features, positioning features, and environmental features are combined with corresponding weights to perform feature concatenation to generate the comprehensive feature vector; Among them, the weight corresponding to each modal feature is dynamically adjusted based on the feature contribution.

3. The method for quantifying the spatial distribution of key species populations in complex ecosystems according to claim 2, characterized in that: The step of analyzing the comprehensive feature vector based on the target large model, identifying species within the target ecological area, and obtaining species identification information of each species within the target ecological area includes: Analyzing the comprehensive feature vector based on the target macromodel to obtain identification reference information of each species in the target ecological area, wherein the identification reference information includes at least one reference name and a confidence level corresponding to the reference name; For each species in the target ecological area, the species identification reference information is corrected based on the species prior information to obtain the species identification information.

4. The method for quantifying the spatial distribution of key species populations in complex ecosystems according to any one of claims 1 to 3, characterized in that: The species identification information includes the species name and the confidence level of the species name; The step of quantifying the spatial distribution of species populations based on the species identification information corresponding to the target ecological region and the comprehensive feature vector to generate species population quantitative information includes: Filtering species within the target ecological area whose species name confidence is less than a preset threshold, and determining a species set based on the retained species; Based on the species identification information corresponding to the key species in the species set and the comprehensive feature vector, generating a population distribution map indicating the distribution location of the key species, a population density map indicating the density distribution of the key species, and a population dynamic change map indicating the changes of the key species at different time points; Among them, at least some of the species in the species set belong to the key species, and the key species are species whose influence on the target ecological area meets preset conditions.

5. The method for quantifying the spatial distribution of key species populations in complex ecosystems according to claim 4, characterized in that: Based on the species identification information corresponding to the key species in the species set and the comprehensive feature vector, a population distribution map indicating the distribution location of the key species is generated, including: extracting the location information of the key species at a specified time from the comprehensive feature vector; Based on the location information of the key species at the specified time and the species identification information of the key species, a population distribution map indicating the distribution of the key species at the specified time is generated.

6. The method for quantifying the spatial distribution of key species populations in complex ecosystems according to claim 4, characterized in that: Based on the species identification information corresponding to the key species in the species set and the comprehensive feature vector, a population density map indicating the density distribution of the key species is generated, including: Dividing the target ecological area into a plurality of grids; Based on the location information of the key species at a specified time extracted from the comprehensive feature vector, the key species are allocated to corresponding grids; For each grid, the grid weighted density is determined based on the species weights corresponding to the key species contained in the grid, wherein the species weights are determined based on the confidence level of the species name contained in the species identification information corresponding to the key species; The weighted density distribution of the target ecological region is determined based on the grid weighted densities corresponding to each grid, and the weighted density distribution of the target ecological region is visualized to generate the population density map.

7. The method for quantifying the spatial distribution of key species populations in complex ecosystems according to claim 4, characterized in that: Based on the species identification information corresponding to the key species in the species set and the comprehensive feature vector, a population dynamic change diagram indicating the change of the key species at different time points is generated, including: Recording multiple time nodes of the target large model analyzing the comprehensive feature vector within a target time interval to generate a species identification time series, wherein the comprehensive feature vector is dynamically updated within the target time interval; At each time node of the species identification time series, based on the comprehensive feature vector corresponding to the current time node and the species identification information corresponding to the key species in the species set, the number of species corresponding to each key species is counted, and / or the species distribution corresponding to each key species is counted to generate the key species population information corresponding to the current time node; The key species population information corresponding to each time node of the species identification time series is compared to generate the population dynamic change diagram.

8. The method for quantifying the spatial distribution of key species populations in complex ecosystems according to claim 7, characterized in that: The method of generating key species population information corresponding to the current time node by counting the number of species corresponding to each key species and / or counting the species distribution corresponding to each key species based on the comprehensive feature vector corresponding to the current time node and the species identification information corresponding to the key species in the species set includes: After determining the key species in the species set, extracting at least one of species quantity and location information of each key species from the comprehensive feature vector based on the species identification information of the key species; The key species population information is generated based on at least one of the species quantity and species distribution of each key species, and the species distribution of the key species is determined based on the location information.

9. The method for quantifying the spatial distribution of key species populations in complex ecosystems according to claim 1, characterized in that: Based on the quantitative information of the species population and the collected original data, a quantitative report that comprehensively describes the distribution characteristics of the species population is generated and displayed, including: Conduct a comprehensive analysis of the population distribution map, the population density map, the population dynamics map, and the collected raw data to generate statistical analysis results that include at least population size changes, population distribution ranges, population density changes, hotspot areas, factors affecting population dynamics, and ecological pressures in the target ecological region; Based on the statistical analysis results and the population distribution map, the population density map, and the population dynamic change map, the quantitative report is generated and displayed.

10. A system for quantifying the spatial distribution of key species populations in complex ecosystems, characterized by: include: An extraction and generation module for extracting features from biological data and environmental data collected in the target ecological area, and fusing the extracted multimodal features to generate a comprehensive feature vector, wherein the biological data includes at least species image data, species audio data, and species location data, and the environmental data includes at least meteorological data and natural resource data; An analysis and identification module, configured to analyze the comprehensive feature vector based on the target large model, identify species within the target ecological area, and obtain species identification information of each species within the target ecological area; A quantification generation module is used to quantify the spatial distribution of species populations based on the species identification information corresponding to the target ecological region and the comprehensive feature vector, and generate species population quantitative information, wherein the species population quantitative information at least includes a population distribution map, a population density map, and a population dynamic change map; A generation and display module is used to generate and display a quantitative report that comprehensively describes the distribution characteristics of species populations based on the quantitative information of the species populations and the collected original data.

Images