Automatic qualitative and quantitative analysis method for freshwater phytoplankton based on artificial intelligence

Abstract

The invention discloses an artificial intelligence-based automatic qualitative and quantitative analysis method for freshwater phytoplankton. The method comprises the following steps: data acquisition: capturing an algae image in a water sample by using high-resolution image acquisition equipment; image preprocessing: denoising, contrast enhancement and binarization preprocessing are carried out on the acquired image so as to improve the image quality and make preparations for subsequent analysis; image segmentation: using an image segmentation algorithm to distinguish algal cells and a background, and extracting an image area of the algal cells; feature extraction: extracting features of algae cells from the segmented image for subsequent identification and classification; the method has the beneficial effects that the artificial intelligence technology automatically recognizes algae species, and compared with a traditional artificial microscopic examination recognition method, the working efficiency can be greatly improved; and a user can create and store a personalized analysis template, so that reuse in similar projects is facilitated, and the working efficiency is improved.

Description

Technical Field

[0001] The invention belongs to the technical field of qualitative and quantitative analysis of phytoplankton, and in particular relates to an automatic qualitative and quantitative analysis method of freshwater phytoplankton based on artificial intelligence. Background Art

[0002] Phytoplankton refers to tiny plants that float freely in water bodies. They cannot swim actively, but move with water currents and wind direction. Phytoplankton mainly include single-celled or simple multi-celled plants. They produce organic matter through photosynthesis and are primary producers in aquatic ecosystems. They are the basis of the food chain, providing food for small zooplankton, and then supporting higher-level consumers. They play a fundamental role in energy flow and material circulation.

[0003] Phytoplankton is also an important component of aquatic ecosystems. The changes in its species composition and quantity can reflect the ecological health of the water body. Monitoring phytoplankton helps to evaluate the ecosystem service functions. The distribution and abundance of phytoplankton can be used as important indicators for water quality assessment, especially for the indication of nutritional status and pollution levels.

[0004] The excessive reproduction of certain phytoplankton can lead to algal blooms, which are harmful to human health and ecosystems. Qualitative and quantitative monitoring can help to warn and manage algal bloom events. At the same time, phytoplankton communities are sensitive to climate change, physical and chemical indicators, etc. Monitoring data can be used to study the impact of various factors on aquatic ecosystems.

[0005] Phytoplankton is almost entirely composed of algae, including blue-green algae, green algae, diatoms, etc.; algal bloom refers to the rapid reproduction of algae in a short period of time in the water body due to certain conditions, forming a visible aggregation phenomenon. This phenomenon is usually accompanied by changes in the color of the water, such as green, blue or red, which is due to the pigments in the algae cells; there are many possible conditions for the formation of algal blooms, nutrients, temperature, sunlight, water flow, pH value, etc., all of which may lead to changes in algal communities.

[0006] The harm caused by algae blooms is also very serious. After the algae die, their organic matter is decomposed by microorganisms, which consume a large amount of dissolved oxygen and cause hypoxia in the water. Some algae can produce harmful toxins, such as microcystins, which pose health risks to humans, animals and aquatic life. Algae blooms affect the transparency of water bodies, reduce sunlight penetration, affect the photosynthesis of underwater plants, and thus affect the entire aquatic ecosystem. Algae blooms can have a negative impact on fisheries, tourism, water sports and drinking water supply, leading to economic losses. Algae blooms change the composition and diversity of species in aquatic ecosystems and destroy the ecological balance.

[0007] The existing qualitative and quantitative analysis of freshwater phytoplankton often uses the following methods:

[0008] Microscopy: It is very time-consuming and requires professional skills. It is inefficient and depends on the subjectivity of the operator. In addition, errors may occur in various operations such as sample collection, fixation, storage and processing. There may be differences in operating techniques between different laboratories or operators, which may affect the consistency and repeatability of the results.

[0009] Flow cytometer: Samples require pre-treatment and preparation, and high-speed flow and laser irradiation may cause physical damage or pressure to cells, affecting the measurement; flow cytometers have certain requirements for the cell concentration in the sample. Too low or too high a concentration may affect the accuracy of the count. Flow cytometers may not be able to effectively distinguish very small cells or particles, and the interpretation of data may sometimes be subjective;

[0010] Molecular biology: requires professional operational knowledge, and the quality of sample processing is very high. Sample degradation or contamination affects experimental results. The amount of data is usually large, requiring professional bioinformatics tools and knowledge to analyze and interpret, and it is difficult to accurately measure low-abundance or rare molecules.

[0011] Spectrophotometry: The preparation and extraction process is cumbersome, and the extraction of chlorophyll is incomplete; other absorbing substances, these substances will absorb at the same wavelength, resulting in deviations in the measurement results. For low concentration substances, spectrophotometry may not be sensitive enough and difficult to detect or accurately quantify;

[0012] Fluorescence method: There is no standard; algae species composition, algae cluster size, and environmental conditions may all have an impact. Suspended particles, dirt, or other pollutants or chemicals may adhere to the sensor, affecting the measurement of the fluorescence signal, resulting in inaccurate readings, and requiring regular calibration and maintenance; other non-detectable fluorescent substances are also calculated, and the algae density is semi-quantitative, only the cyanobacteria density is monitored, and there is no total algae density;

[0013] Satellite remote sensing method: The spatial resolution may not be sufficient to detect small-scale or localized algal blooms, and clouds may block the view of satellite sensors, resulting in ineffective monitoring during cloudy or rainy seasons; suspended matter, sediments and other optically active substances in the water body may affect the interpretation of remote sensing signals, making it more difficult to distinguish algal blooms from other factors; different types of algae and other water body components may have similar spectral characteristics, making it difficult to distinguish and identify specific algae based on spectral data alone; there may be a time interval between satellite transit and data collection, which may limit the ability to monitor rapidly changing events in real time; satellite sensors may drift over time and require regular calibration to maintain data accuracy and consistency. Extracting useful information from remote sensing data requires professional data processing technology and experience, which may increase the complexity of monitoring. Summary of the invention

[0014] The purpose of the present invention is to provide an automatic qualitative and quantitative analysis method for freshwater phytoplankton based on artificial intelligence, so as to realize rapid, accurate and automated analysis of phytoplankton.

[0015] To achieve the above object, the present invention provides the following technical solution: an automatic qualitative and quantitative analysis method of freshwater phytoplankton based on artificial intelligence, comprising the following steps:

[0016] Data acquisition: Use high-resolution image acquisition equipment to capture images of algae in water samples;

[0017] Image preprocessing: denoising, contrast enhancement, and binarization are performed on the collected images to improve image quality and prepare for subsequent analysis;

[0018] Image segmentation: Apply image segmentation algorithm to distinguish algae cells from background and extract the image area of ​​algae cells;

[0019] Feature extraction: extract the features of algae cells from the segmented images for subsequent recognition and classification;

[0020] Deep learning model training and classification: Train a deep neural network to learn and identify different types of algae, and then use the trained model to automatically classify and identify algae cells;

[0021] Quantitative analysis: Based on the number of algae cells identified in the image and the sampling volume, the algae density and possible biomass of the water sample are calculated;

[0022] Data management: Develop a database system to store and manage algae images, analysis results, and related metadata for subsequent data retrieval and analysis;

[0023] User interface and interaction: Design an intuitive and easy-to-use user interface that allows users to easily upload images, start analysis, view results, and manage data;

[0024] Result visualization: Provides a visual display of the results to make the analysis results more intuitive and easy to understand.

[0025] As a preferred technical solution of the present invention, the image segmentation algorithm is one or a combination of at least two of threshold segmentation, edge detection, and deep learning-based segmentation.

[0026] As a preferred technical solution of the present invention, the features of algae cells are extracted from the segmented image, including shape, size, texture and color.

[0027] As a preferred technical solution of the present invention, a visual display of the results is provided, including an algae distribution map, a density curve map and a biomass estimation map.

[0028] As a preferred technical solution of the present invention, a deep neural network is trained to learn and identify different types of algae, and then the trained model is used to automatically classify and identify algae cells. The specific implementation method is as follows:

[0029] Collect data: First, you need to collect a large amount of algae image data. These data contain a variety of algae species, and each algae should have enough samples to ensure the generalization ability of the model.

[0030] Annotated data: Manually annotate the collected images;

[0031] Data augmentation: To improve the robustness and generalization ability of the model, a series of transformations can be performed on the original image to generate more training samples;

[0032] Building a neural network model

[0033] Select model architecture: Select the appropriate convolutional neural network architecture according to task requirements;

[0034] Configure model parameters: set parameters of convolutional layer, pooling layer, activation function, fully connected layer components, as well as learning rate, optimization algorithm, and loss function training parameters;

[0035] Training the model

[0036] Data division: divide the labeled data set into training set, validation set and test set;

[0037] Training process: Use the training set data to train the model. During the training process, the model will try to learn the mapping relationship between the input image and the annotation result, and adjust the model parameters through the back propagation algorithm to minimize the loss function;

[0038] Validation and Adjustment: Regularly use the validation set to evaluate model performance and adjust model parameters, training strategies, or model architecture as needed to improve the model's performance on the validation set.

[0039] Model evaluation and optimization

[0040] Evaluate the model: Use the test set to evaluate the performance of the final model, including accuracy, recall, and F1 score indicators;

[0041] Optimize the model: further optimize the model structure or training strategy based on test results and actual needs.

[0042] To improve model performance;

[0043] Apply models for automatic classification and recognition

[0044] Load the trained model: load the trained model into the application;

[0045] Input the image to be identified: Input the algae image to be identified into the model;

[0046] Automatic classification and recognition: The model automatically classifies and recognizes algae cells in the input image and outputs the classification results.

[0047] As a preferred technical solution of the present invention, a series of transformations include rotation, scaling, cropping, flipping, and color adjustment; and convolutional neural network architectures include AlexNet, VGG, ResNet, and Inception.

[0048] As a preferred technical solution of the present invention, it also includes subsequent processing: subsequent processing of the classification results output by the model, including calculation of algae density and biomass.

[0049] As a preferred technical solution of the present invention, a database system is developed to store and manage algae images, analysis results and related metadata. The specific implementation method is as follows:

[0050] Clarify requirements: First, clarify what types of data the system needs to store and the relationships between these data;

[0051] Planning system architecture: Plan the overall architecture of the database system according to requirements;

[0052] Define entities and relationships: Use entity-relationship diagram tools to define algae images, analysis results, samples, user entities, and the relationships between them;

[0053] Design table structure: Design the database table structure according to the data model;

[0054] Evaluate requirements: Evaluate the requirements for the database management system based on data volume, query performance requirements, and concurrent access factors;

[0055] Choose a database: Choose a suitable database management system;

[0056] Create database and table: Create database and table in the selected DBMS, and design corresponding indexes and views according to the table structure;

[0057] Storing image data: For algae images, store them in the file system and store the path or reference to the image in the database;

[0058] Store metadata: Store the metadata of algae samples and analysis results in corresponding tables;

[0059] Develop the data access layer: write code to access the database and perform add, delete, modify and query operations;

[0060] Optimize queries: optimize commonly used queries to improve query efficiency;

[0061] Implement security measures: set user permissions, encrypt sensitive data, and regularly audit database activities to ensure data security;

[0062] Develop backup strategies: Develop full and incremental backup strategies, and test the backup and recovery processes to ensure their effectiveness.

[0063] As a preferred technical solution of the present invention, the overall architecture of the database system is planned according to the requirements, including the front-end interface, back-end logic, and database services; the database table structure is designed according to the data model, including the table name, field name, data type, and constraints.

[0064] Compared with the prior art, the present invention has the following beneficial effects:

[0065] Fully meet HJ1216 standards, with multiple analysis methods including visual field method, diagonal method, row grid method and full film method;

[0066] Artificial intelligence technology automatically identifies algae species, which can greatly improve work efficiency compared with traditional manual microscopic identification methods;

[0067] Users can create and save personalized analysis templates to facilitate reuse in similar projects and improve work efficiency;

[0068] Allows users to perform secondary average analysis on data from two projects to assess consistency and reliability between the same samples;

[0069] Based on AI recognition, manual verification and editing functions are provided to ensure the accuracy of analysis results;

[0070] It can not only identify and count algae, but also calculate the density and biomass of algae, and provide comprehensive statistical analysis;

[0071] Display species distribution in the form of bar charts and pie charts to make the results more intuitive and easy to understand;

[0072] Users can view and review the analytical data of historical projects and summarize historical data to facilitate long-term trend analysis and data comparison;

[0073] Provides a one-click export function to facilitate users to quickly obtain all relevant data tables. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 The figure is a flow chart of the analysis method of the present invention. DETAILED DESCRIPTION

[0075] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0076] Example 1

[0077] See also Figure 1 , which is the first embodiment of the present invention, provides an automatic qualitative and quantitative analysis method of freshwater phytoplankton based on artificial intelligence, comprising the following steps:

[0078] Data collection: Use high-resolution image acquisition equipment such as microscopes and cameras to capture images of algae in water samples;

[0079] Image preprocessing: Perform preprocessing steps such as denoising, contrast enhancement, and binarization on the collected images to improve image quality and prepare for subsequent analysis;

[0080] Image segmentation: Apply image segmentation algorithms (such as threshold segmentation, edge detection, or deep learning-based segmentation techniques) to distinguish algae cells from the background and extract the image area of ​​the algae cells;

[0081] Feature extraction: Extract the features of algae cells from the segmented images, including shape, size, texture, and color, for subsequent recognition and classification;

[0082] Deep learning model training and classification: training deep neural networks (such as convolutional neural networks (CNNs)) to learn and identify different types of algae, and then using the trained models to automatically classify and identify algae cells;

[0083] Quantitative analysis: Based on the number of algae cells identified in the image and the sampling volume, the algae density and possible biomass of the water sample are calculated;

[0084] Data management: Develop a database system to store and manage algae images, analysis results, and related metadata for subsequent data retrieval and analysis;

[0085] User interface and interaction: Design an intuitive and easy-to-use user interface that allows users to easily upload images, start analysis, view results, and manage data;

[0086] Result visualization: Provides visual display of results, such as algae distribution map, density curve map and biomass estimation map, making the analysis results more intuitive and easy to understand;

[0087] Report generation: Automatically generate detailed reports containing analysis results and statistical data to facilitate users to conduct further analysis and decision-making;

[0088] Quality control: integrated quality control module to ensure the accuracy and repeatability of data and improve the reliability of analysis results;

[0089] Continuous updates and maintenance: Establish a mechanism to regularly update the software, including algorithm optimization, user feedback integration, and new feature development, to keep the software advanced and competitive.

[0090] Example 2

[0091] See also Figure 1 , which is the second embodiment of the present invention, is based on the previous embodiment, except that:

[0092] Train a deep neural network to learn and identify different types of algae, and then use the trained model to automatically classify and identify algae cells. The specific implementation method is as follows:

[0093] Data preparation

[0094] Collect data: First, you need to collect a large amount of algae image data. This data should include multiple algae species, and each algae should have enough samples to ensure the generalization ability of the model.

[0095] Labeling data: Manually labeling the collected images, that is, marking the type or boundary of each algae cell in the image, usually done manually by experts or professionals;

[0096] Data enhancement: In order to improve the robustness and generalization ability of the model, a series of transformations (such as rotation, scaling, cropping, flipping, color adjustment, etc.) can be performed on the original image to generate more training samples;

[0097] Building a neural network model

[0098] Select model architecture: Choose a suitable convolutional neural network architecture according to task requirements, such as AlexNet, VGG, ResNet, Inception, etc., or design a custom model architecture;

[0099] Configure model parameters: set parameters of components such as convolutional layers, pooling layers, activation functions, fully connected layers, as well as training parameters such as learning rate, optimization algorithm, and loss function;

[0100] Training the model

[0101] Data division: divide the labeled data set into training set, validation set and test set; the training set is used to train the model, the validation set is used to adjust the model parameters and select the best model, and the test set is used to evaluate the performance of the final model;

[0102] Training process: Use the training set data to train the model. During the training process, the model will try to learn the mapping relationship between the input image and the annotation result, and adjust the model parameters through the back propagation algorithm to minimize the loss function;

[0103] Validation and Adjustment: Regularly use the validation set to evaluate model performance and adjust model parameters, training strategies, or model architecture as needed to improve the model's performance on the validation set.

[0104] Model evaluation and optimization

[0105] Evaluate the model: Use the test set to evaluate the performance of the final model, including accuracy, recall, F1 score and other indicators;

[0106] Optimize the model: Based on the test results and actual needs, further optimize the model structure or training strategy to improve model performance;

[0107] Apply models for automatic classification and recognition

[0108] Load the trained model: load the trained model into the application;

[0109] Input the image to be identified: Input the algae image to be identified into the model;

[0110] Automatic classification and recognition: The model will automatically classify and recognize algae cells in the input image and output the classification results;

[0111] Subsequent processing

[0112] Result processing: perform subsequent processing on the classification results output by the model, such as calculating algae density, biomass, etc.;

[0113] User feedback: Collect user feedback on classification results to further improve the model.

[0114] Develop a database system to store and manage algae images, analysis results, and related metadata. The specific implementation methods are as follows:

[0115] Demand Analysis and Planning

[0116] Clarify requirements: First, clarify what types of data the system needs to store (such as algae images, analysis results, sample information, experimental conditions, etc.) and the relationship between these data;

[0117] Planning system architecture: Plan the overall architecture of the database system according to requirements, including front-end interface, back-end logic, database services and other components;

[0118] Data model design

[0119] Define entities and relationships: Use tools such as entity-relationship (ER) diagrams to define entities such as algae images, analysis results, samples, users, and the relationships between them;

[0120] Design table structure: Design the database table structure according to the data model, including table name, field name, data type, constraints, etc.;

[0121] Choosing a database management system

[0122] Evaluate requirements: Evaluate the requirements for the database management system based on factors such as data volume, query performance requirements, and concurrent access volume;

[0123] Choose a database: Choose a suitable database management system (DBMS), such as MySQL, PostgreSQL, MongoDB (for non-relational data), etc.

[0124] Database Implementation

[0125] Create database and table: Create database and table in the selected DBMS, and design corresponding indexes and views according to the table structure;

[0126] Storing image data: For algae images, you can store them in the file system and store the path or reference to the image in the database. Another approach is to store the image data directly using a binary large object (BLOB) field supported by the database.

[0127] Storing metadata: storing metadata of algae samples (such as collection time, location, species, etc.) and analysis results (such as classification results, biomass, etc.) in corresponding tables;

[0128] Data access and query

[0129] Develop the data access layer: write code to access the database and perform CRUD operations; this can be done by writing SQL statements directly or using an ORM (object-relational mapping) framework;

[0130] Optimize queries: Optimize common queries to improve query efficiency, including creating indexes, optimizing query statements, adjusting database configuration, etc.

[0131] Data security and backup

[0132] Implement security measures: set user permissions, encrypt sensitive data, regularly audit database activities, etc. to ensure data security;

[0133] Develop a backup strategy: Back up database data regularly to prevent data loss or corruption. Develop full backup and incremental backup strategies, and test the backup and recovery process to ensure its effectiveness.

[0134] Although the embodiments of the present invention have been shown and described, as detailed above, it will be understood by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An automatic qualitative and quantitative analysis method for freshwater phytoplankton based on artificial intelligence, characterized by: The steps include: Data acquisition: Use high-resolution image acquisition equipment to capture images of algae in water samples; Image preprocessing: denoising, contrast enhancement, and binarization are performed on the collected images to improve image quality and prepare for subsequent analysis; Image segmentation: Apply image segmentation algorithm to distinguish algae cells from background and extract the image area of ​​algae cells; Feature extraction: extract the features of algae cells from the segmented images for subsequent recognition and classification; Deep learning model training and classification: Train a deep neural network to learn and identify different types of algae, and then use the trained model to automatically classify and identify algae cells; Quantitative analysis: Based on the number of algae cells identified in the image and the sampling volume, the algae density and possible biomass of the water sample are calculated; Data management: Develop a database system to store and manage algae images, analysis results, and related metadata for subsequent data retrieval and analysis; User interface and interaction: Design an intuitive and easy-to-use user interface that allows users to easily upload images, start analysis, view results, and manage data; Result visualization: Provides a visual display of the results to make the analysis results more intuitive and easy to understand.

2. The method for automatic qualitative and quantitative analysis of freshwater phytoplankton based on artificial intelligence according to claim 1, characterized in that: The image segmentation algorithm is one of threshold segmentation, edge detection, and deep learning-based segmentation, or a combination of at least two of them.

3. The method for automatic qualitative and quantitative analysis of freshwater phytoplankton based on artificial intelligence according to claim 1, characterized in that: The features of algae cells, including shape, size, texture, and color, are extracted from the segmented images.

4. The method for automatic qualitative and quantitative analysis of freshwater phytoplankton based on artificial intelligence according to claim 1, characterized in that: Provides visual presentation of results, including algal distribution maps, density curves, and biomass estimation maps.

5. The method for automatic qualitative and quantitative analysis of freshwater phytoplankton based on artificial intelligence according to claim 1, characterized in that: Train a deep neural network to learn and identify different types of algae, and then use the trained model to automatically classify and identify algae cells. The specific implementation method is as follows: Collect data: First, you need to collect a large amount of algae image data. These data contain a variety of algae species, and each algae should have enough samples to ensure the generalization ability of the model. Annotated data: Manually annotate the collected images; Data augmentation: To improve the robustness and generalization ability of the model, a series of transformations can be performed on the original image to generate more training samples; Building a neural network model Select model architecture: Select the appropriate convolutional neural network architecture according to task requirements; Configure model parameters: set parameters of convolutional layer, pooling layer, activation function, fully connected layer components, as well as learning rate, optimization algorithm, and loss function training parameters; Training the model Data division: divide the labeled data set into training set, validation set and test set; Training process: Use the training set data to train the model. During the training process, the model will try to learn the mapping relationship between the input image and the annotation result, and adjust the model parameters through the back propagation algorithm to minimize the loss function; Validation and Adjustment: Regularly use the validation set to evaluate model performance and adjust model parameters, training strategies, or model architecture as needed to improve the model's performance on the validation set. Model evaluation and optimization Evaluate the model: Use the test set to evaluate the performance of the final model, including accuracy, recall, and F1 score indicators; Optimize the model: Based on the test results and actual needs, further optimize the model structure or training strategy to improve model performance; Apply models for automatic classification and recognition Load the trained model: load the trained model into the application; Input the image to be identified: Input the algae image to be identified into the model; Automatic classification and recognition: The model automatically classifies and recognizes algae cells in the input image and outputs the classification results.

6. The method for automatic qualitative and quantitative analysis of freshwater phytoplankton based on artificial intelligence according to claim 5, characterized in that: A series of transformations include rotation, scaling, cropping, flipping, and color adjustment; convolutional neural network architectures include AlexNet, VGG, ResNet, and Inception.

7. The method for automatic qualitative and quantitative analysis of freshwater phytoplankton based on artificial intelligence according to claim 5, characterized in that: It also includes subsequent processing: subsequent processing of the classification results output by the model, including calculation of algae density and biomass.

8. The method for automatic qualitative and quantitative analysis of freshwater phytoplankton based on artificial intelligence according to claim 1, characterized in that: Develop a database system to store and manage algae images, analysis results, and related metadata. The specific implementation methods are as follows: Clarify requirements: First, clarify what types of data the system needs to store and the relationships between these data; Planning system architecture: Plan the overall architecture of the database system according to requirements; Define entities and relationships: Use entity-relationship diagram tools to define algae images, analysis results, samples, user entities, and the relationships between them; Design table structure: Design the database table structure according to the data model; Evaluate requirements: Evaluate the requirements for the database management system based on data volume, query performance requirements, and concurrent access factors; Choose a database: Choose a suitable database management system; Create database and table: Create database and table in the selected DBMS, and design corresponding indexes and views according to the table structure; Storing image data: For algae images, store them in the file system and store the path or reference to the image in the database; Store metadata: Store the metadata of algae samples and analysis results in corresponding tables; Develop the data access layer: write code to access the database and perform add, delete, modify and query operations; Optimize queries: optimize commonly used queries to improve query efficiency; Implement security measures: set user permissions, encrypt sensitive data, and regularly audit database activities to ensure data security; Develop backup strategies: Develop full and incremental backup strategies, and test the backup and recovery processes to ensure their effectiveness.

9. The method for automatic qualitative and quantitative analysis of freshwater phytoplankton based on artificial intelligence according to claim 8, characterized in that: Plan the overall architecture of the database system according to requirements, including the front-end interface, back-end logic, and database services; design the database table structure according to the data model, including table name, field name, data type, and constraints.

Images