An inversion method for the density distribution of coral polyps and xanthophytes

Abstract

This invention relates to the field of marine ecosystem monitoring technology, and particularly to an inversion method for the density distribution of coral polyps and xanthophytes, comprising the following steps: spectral data preparation, acquiring the reflectance and fluorescence spectra of corals; collecting three-dimensional point cloud information of corals to obtain a multidimensional dataset; correcting the reflectance and fluorescence spectra and combining them with the multidimensional dataset to create a comprehensive dataset; reconstructing the three-dimensional point cloud information of corals to generate a three-dimensional mesh model and a solution model for the spatial distribution of coral optical property parameters; mapping the optical properties of corals to their health status to visualize coral health; and obtaining the inversion framework for the density distribution of coral polyps and xanthophytes. This invention provides an inversion method for the density distribution of coral polyps and xanthophytes, which, by acquiring three-dimensional information of underwater corals based on hyperspectral remote sensing technology and combining it with an inversion algorithm, achieves non-destructive, large-scale, and high-precision monitoring of coral reefs, contributing to a better understanding and prediction of changes in coral reef ecosystems.

Description

Technical Field

[0001] This invention relates to the field of marine ecosystem monitoring technology, and in particular to an inversion method for the density distribution of coral polyps and xanthophytes. Background Technology

[0002] Zooxanthellae, a type of dinoflagellate, are golden-yellow intercellular symbiotic bacteria found in various marine animals and protozoa. They are among the most important primary producers in marine ecosystems, exhibiting rich morphological diversity. Coral ecosystems almost entirely depend on dinoflagellates, represented by zooxanthellae, which are symbiotic with reef-building coral polyps. The primary productivity contributed by these zooxanthellae through photosynthetic carbon fixation forms the basis for coral growth and reef formation. Therefore, detecting the density distribution of zooxanthellae can reflect the health status of coral reefs.

[0003] Current methods for assessing coral reef health include in-situ sampling and analysis, underwater photogrammetry, remote sensing image analysis, and simple spectral reflectance models. Existing technologies have limitations in accuracy, non-destructiveness, coverage, and resolution. These limitations are primarily due to the complexity and diversity of coral reef ecosystems, and the impact of the complex three-dimensional structure of coral colonies on light transmission. A major challenge with existing technologies is accurately simulating the light transmission process within complex coral structures and retrieving the spatial distribution of zooxanthellae from spectral information without damaging the integrity of the corals.

[0004] Therefore, there is a need for an inversion method that achieves high accuracy without damaging the coral reef structure. Summary of the Invention

[0005] The main objective of this invention is to provide a method for inverting the density distribution of coral polyps and xanthophytes, aiming to solve the problems of existing coral reef detection methods being unable to meet the requirements of non-destructiveness and accuracy.

[0006] To achieve the above objectives, this invention proposes an inversion method for the density distribution of coral polyps and xanthophytes, which includes the following steps:

[0007] Spectral data preparation, obtaining the reflectance and fluorescence spectra of corals;

[0008] Three-dimensional point cloud information acquisition: Collect three-dimensional point cloud information of coral, and integrate the three-dimensional point cloud information of coral with reflectance spectrum and fluorescence spectrum to obtain multidimensional dataset;

[0009] Data processing involves correcting reflectance and fluorescence spectra and combining them with a multidimensional dataset to create a comprehensive dataset.

[0010] A model is established based on a comprehensive dataset and other prior data to reconstruct the three-dimensional point cloud information of corals, generate a three-dimensional mesh model and a solution model for the spatial distribution of coral optical property parameters;

[0011] Record data, map the optical properties of each area of ​​the coral to the health status of the coral, and assign corresponding color levels to the three-dimensional mesh model according to the health level to visualize the health status of the coral.

[0012] Model validation and optimization: After iterating through the dataset using a stratified sampling strategy, an inversion framework for the density distribution of coral polyps and yellow algae was obtained.

[0013] Furthermore, the steps for preparing the spectral data and obtaining the reflectance and fluorescence spectra of the coral include:

[0014] Collect coral health information;

[0015] Use a hyperspectral sensor to acquire reflectance spectral data of the coral surface;

[0016] Fluorescent proteins in coral tissue were excited using an excitation light source, and fluorescence signals were recorded and fluorescence spectra were collected.

[0017] The collected fluorescence spectra are smoothed to remove noise interference and obtain effective fluorescence spectra;

[0018] A dataset was established, and the reflectance spectrum was correlated with the effective fluorescence spectrum and coral health information to obtain the correspondence between the spectrum and the optical properties of coral.

[0019] Furthermore, the steps of acquiring the three-dimensional point cloud information, collecting the three-dimensional point cloud information of the coral, and integrating the three-dimensional point cloud information of the coral with the reflectance spectrum and fluorescence spectrum to obtain a multidimensional dataset include:

[0020] A structured light scanner was used to perform a 3D scan of the coral, obtaining 3D point cloud information of the coral;

[0021] By integrating the three-dimensional point cloud information of corals with reflectance and fluorescence spectra, and combining spatial registration technology, the spatial information in the three-dimensional point cloud information of corals is combined with the optical property information in the spectral data to obtain a multidimensional dataset.

[0022] Furthermore, the data processing steps, including correcting the reflectance and fluorescence spectra and combining them with a multidimensional dataset to create a comprehensive dataset, include:

[0023] Denoising and filtering are performed on the 3D point cloud information of corals to remove outliers and duplicate data, resulting in optimized 3D point cloud information of corals.

[0024] The optimized coral 3D point cloud information was smoothed, and the reflectance and fluorescence spectral data were corrected.

[0025] The corrected reflectance and fluorescence spectral data are remapped into a multidimensional dataset to create a comprehensive dataset.

[0026] Furthermore, the steps of establishing the model, based on a comprehensive dataset and other prior data, reconstructing the three-dimensional point cloud information of the coral, generating a three-dimensional mesh model, and solving the spatial distribution model of the coral's optical property parameters include:

[0027] Using binocular vision technology, a 3D mesh model of the coral is acquired, and the 3D point cloud information of the coral is reconstructed.

[0028] The reconstructed 3D point cloud information of the coral is processed into a mesh to generate an accurate 3D mesh model.

[0029] A likelihood function is constructed based on the forward model and observation error, and a posterior distribution function is calculated based on the prior probability distribution function to obtain the simulated coral model data.

[0030] The collected data were inverted and tested using simulated coral model data, and a spatial distribution solution model for coral optical property parameters was constructed.

[0031] The optical property parameters of the coral obtained by inversion are mapped point by point to a three-dimensional mesh model to obtain the spatial distribution of zooxanthellae density.

[0032] Furthermore, the recorded data maps the optical properties of each region of the coral to its health status, and assigns corresponding color levels to the three-dimensional mesh model according to the health level, thus visualizing the coral's health status. The steps include:

[0033] Calibrate the optical property parameters of corals and verify the relationship between the optical property parameters and the health status of corals;

[0034] The three-dimensional mesh model is reproduced as a solid three-dimensional model. The health status of the coral is combined with the three-dimensional solid model and mapped point by point. Color levels are set according to the health status of the coral to visualize the health status of the coral.

[0035] Furthermore, the model validation and optimization, which uses a stratified sampling strategy to iterate through the dataset to obtain the inversion framework for the density distribution of coral polyps and yellow algae, includes the following steps:

[0036] The comprehensive dataset is divided into training, validation and test sets using a stratified sampling strategy. Multiple iterations of training and testing are performed to obtain a stable dataset.

[0037] By regularizing stable datasets and constraining model complexity, an inversion framework for zooxanthellae density in underwater coral ecosystems is constructed.

[0038] This invention provides an inversion method for the density distribution of coral polyps and yellow algae. By collecting three-dimensional information of underwater corals based on hyperspectral remote sensing technology and combining it with an inversion algorithm, non-destructive, large-scale, and high-precision monitoring of coral reefs is achieved, which helps to better understand and predict changes in coral reef ecosystems. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the processes shown in these drawings without creative effort.

[0040] Figure 1 This is a schematic flowchart of one embodiment of the method for inverting the density distribution of coral polyps and yellow algae. Detailed Implementation

[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0042] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indication will also change accordingly.

[0043] Furthermore, the use of terms such as "first" and "second" in this invention is for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of that feature. Additionally, the technical solutions of the various embodiments can be combined with each other, but only on the basis of being achievable by those skilled in the art. When the combination of technical solutions is contradictory or impossible to implement, such a combination of technical solutions should be considered non-existent and not within the scope of protection claimed by this invention.

[0044] While current technologies for assessing coral reef health, particularly for the precise measurement of zooxanthellae density, are diversified, each still faces significant limitations. These technologies include: in-situ sampling and analysis, underwater photogrammetry, remote sensing image analysis, and simple spectral reflectance models. In-situ sampling and analysis offers high accuracy, precisely determining zooxanthellae density through microscopic observation and biochemical analysis, but it damages the coral reef structure, limiting its application in large-scale, long-term monitoring. Underwater photogrammetry estimates coral coverage and health through image processing; while non-destructive, it is susceptible to interference from complex underwater lighting conditions. Remote sensing image analysis utilizes multispectral and hyperspectral data captured by satellites or aerial platforms, enabling monitoring of the spatial distribution and health of coral reefs over large areas; however, its spatial resolution limitations hinder precise monitoring of the health status of individual coral colonies. Simple spectral reflectance models achieve non-destructive measurement, but they cannot fully account for the complex influence of the coral's three-dimensional structure on light transmission, resulting in limited accuracy in zooxanthellae density estimation.

[0045] Based on this, embodiments of this application provide an inversion method for the density distribution of coral polyps and xanthophytes, referring to... Figure 1 , Figure 1 This is a schematic flowchart of one embodiment of the method for inverting the density distribution of coral polyps and yellow algae.

[0046] This invention provides a method for inverting the density distribution of coral polyps and xanthophytes, specifically including the following steps:

[0047] S10, Spectral data preparation: Collect coral health information, use a hyperspectral sensor to acquire reflectance spectral data of the coral surface, use an excitation light source to excite fluorescent proteins in coral tissue and record fluorescence signals, collect fluorescence spectra; smooth the collected fluorescence spectra to remove noise interference and obtain effective fluorescence spectra; correlate the reflectance spectra with the effective fluorescence spectra and coral health information to obtain the correspondence between the spectra and the optical properties of the coral.

[0048] In detail, this invention requires first collecting a dataset of coral optical properties, including visible light (VIS) / near infrared (NIS) spectral information and corresponding coral health information. Specifically, a hyperspectral sensor is used to acquire reflectance spectral data of the coral surface, ensuring that the measurement covers different areas of the coral surface to fully consider the influence of coral skeletal structure and tissue layers. Simultaneously, utilizing the properties of fluorescent proteins in corals, a specific wavelength excitation light source (typically in the blue light range, approximately 450-490 nm) is used to excite the fluorescent proteins in the coral tissue, and fluorescence signals are recorded within different emission wavelength ranges (typically in the green to red light range, approximately 500-650 nm), collecting fluorescence spectra. Understandably, when dealing with different species of corals with different fluorescence characteristics, multiple light sources can be used for excitation, including ultraviolet light (350-400 nm), blue light (450-490 nm), and cyan light (490-520 nm), to capture the fluorescence characteristics of various corals. During the measurement process, environmental conditions that affect coral fluorescence performance, such as water temperature, salinity, and light intensity, are recorded simultaneously.

[0049] Understandably, noise interference is unavoidable when collecting the spectral characteristics of corals, requiring technicians to perform spectral smoothing on VIR / NIS data to reduce noise interference and improve the signal-to-noise ratio. In this invention, convolutional smoothing (Savitzky-Golay, SG) is used for noise removal. Convolutional smoothing is a filtering method based on local polynomial least squares fitting. It smooths the data by fitting a polynomial within a local window, thus preserving important signal characteristics such as shape, width, and trend while smoothing the data and reducing noise.

[0050] In the feature extraction stage of this invention, to ensure that the feature bands extracted from underwater spectral data can characterize the health status of corals, the Pearson Correlation Coefficient Method is used to calculate the correlation coefficient between the spectral information and the target variables (such as the optical properties of corals, zooxanthellae density, etc.) to select feature bands. The specific formula is as follows:

[0051] in, This represents the reflectance value in the spectral band. The value of the target variable. This is the average value across the spectral band. The average value of the target variable.

[0052] Feature bands are selected based on the absolute value of the correlation coefficient. In this invention, the correlation coefficient threshold is set to an absolute value of 0.8. Bands with an absolute correlation coefficient greater than the set threshold are selected as feature bands to ensure that the selected bands have a strong correlation with the target variable.

[0053] S20, 3D point cloud information acquisition: Acquire 3D point cloud information of coral and integrate it with reflectance and fluorescence spectra to obtain a multidimensional dataset; Use a structured light scanner to perform 3D scanning of coral to obtain 3D point cloud information of coral; Integrate the 3D point cloud information of coral with reflectance and fluorescence spectra, and combine spatial registration technology to combine the spatial information in the 3D point cloud information of coral with the optical property information in the spectral data to obtain a multidimensional dataset.

[0054] In detail, this step uses a structured light scanner to perform a 3D scan of the coral community, including its branching structure, surface texture, and overall morphology. Each scanning viewpoint... Point cloud data can be represented as:

[0055] in, Indicates the first The three-dimensional coordinates of the points It's a perspective The total number of points captured.

[0056] Simultaneously, these 3D point cloud data are integrated with previously collected reflectance and fluorescence spectral data. Spatial registration technology is used during the integration process to ensure that each 3D point in the point cloud corresponds to the relevant reflectance and fluorescence spectral data. The spatial registration process can be represented as:

[0057] in, These are the coordinates in the spectral image. These are the corresponding coordinates in the 3D model.

[0058] Furthermore, by combining the spatial information from point cloud data with the optical properties from spectral data, a multidimensional dataset was created. This multidimensional dataset includes the geometric morphology information of the coral, the reflectance spectrum and fluorescence characteristics at each spatial point, and can be represented as:

[0059] in, These are the coordinates in the spectral image. yes The reflectance spectral values ​​of each band, yes Fluorescence spectral values ​​in each band.

[0060] S30, Data Processing: Correct the reflectance and fluorescence spectra and combine them with a multidimensional dataset to create a comprehensive dataset; perform noise reduction and filtering on the coral 3D point cloud information to remove outliers and duplicate data, obtaining optimized coral 3D point cloud information; perform smoothing processing on the optimized coral 3D point cloud information and correct the reflectance and fluorescence spectra data; remap the corrected reflectance and fluorescence spectra data into the multidimensional dataset to create a comprehensive dataset.

[0061] Specifically, this step involves denoising and filtering the collected point cloud data to remove outliers and duplicate data. In this process, the Statistical Outlier Removal (SOR) algorithm is used to calculate the average distance from each point to its neighbors and identify points that exceed the standard deviation threshold. Then, the moving least squares (MLS) method is used to smooth the point cloud to reduce the small noise generated during the scanning process while preserving the fine structure of the coral surface to obtain a three-dimensional mesh model.

[0062] After obtaining a three-dimensional model of the coral and correcting the spectral data, the bidirectional reflectance distribution function (BRDF) model is applied to describe the light reflection characteristics of the coral surface at different incident and observation angles. The BRDF model can be expressed as:

[0063] in, It is the incident light. It is the direction of observation. It is the diffuse reflectance coefficient. It is the specular reflection coefficient. It is the Lambert diffuse reflection term. It is the Cook-Torrance specular reflection term.

[0064] Based on this BRDF model, the spectral data of each 3D point cloud needs to be corrected. This correction process considers the incident light direction (usually the known location of the light source), the viewing direction (determined by the scanner's position), and the local surface normal vector (extracted from the 3D mesh model). The corrected reflectivity... It can be represented as: in, It is the original measured reflectance. It is the angle between the incident light and the surface normal.

[0065] Finally, the corrected spectral data is remapped onto a 3D mesh model to create a comprehensive dataset, in which each grid point contains not only location information but also corrected reflectance and fluorescence spectral data.

[0066] S40. Model building: Based on a comprehensive dataset and other prior data, reconstruct the 3D point cloud information of corals, generating a 3D mesh model and a solution model for the spatial distribution of coral optical property parameters. Use binocular vision technology to acquire the 3D mesh model of corals and reconstruct the 3D point cloud information. Mesh the reconstructed 3D point cloud information to generate an accurate 3D mesh model. Construct a likelihood function based on the forward model and observation errors, and simultaneously calculate the posterior distribution function based on the prior probability distribution function to obtain simulated coral model data. Use the simulated coral model data to perform inversion testing and verification on the collected data, constructing a solution model for the spatial distribution of coral optical property parameters. Map the inverted coral optical property parameters point-by-point to the 3D mesh model to obtain the spatial distribution of zooxanthellae density.

[0067] This step requires establishing a spatial distribution solution model for coral optical property parameters based on reflectance and fluorescence spectroscopy measurements. This model uses reflectance and fluorescence spectral information obtained from hyperspectral sensors as observed values. (Reflectance and fluorescence spectra obtained from hyperspectral sensors), combined with other prior knowledge as constraints. The forward simulation model was thus constructed.

[0068] In detail, a 3D mesh model of the coral is first obtained using binocular vision technology. This involves simultaneously capturing images of the coral from different angles using two cameras with a fixed distance between them. Then, the 3D structure of the coral is reconstructed using triangulation principles. The SIFT (Scale Invariant Feature Transform) stereo vision algorithm, based on feature point matching, is used to identify corresponding points in the image pairs. RANSAC (Random Sample Consensus) is then used to estimate the fundamental matrix, and finally, a 3D point cloud is reconstructed using triangulation. The obtained 3D point cloud data is then meshed to generate an accurate 3D mesh model, which can be represented as follows: in, These are observed values. X represents the optical properties of each grid on the coral surface to be solved, X represents the known input parameters (such as seawater scattering properties, light source and sensor parameters), and ε represents the observation error.

[0069] This three-dimensional mesh model utilizes known seawater scattering properties, light source and sensor parameters as prior knowledge, and employs a Bayesian parameter estimation algorithm to inversely determine the positional optical properties of each mesh on the coral surface. When constructing this 3D mesh model, a likelihood function is built based on the forward model and the observation error model. At the same time, establish the prior probability distribution Given the observed value Y, calculate the parameters. The posterior probability distribution:

[0070] Prior distribution Prior probability density including optical property parameters of coral This constraint is used to limit the range of values ​​for optical property parameters and zooxanthellae density, and to consider the correlation between the spatial distribution of zooxanthellae and environmental factors (such as light and water flow conditions). A spatial smoothing constraint in the form of Tikhonov regularization is also introduced, and the modified posterior probability distribution can be expressed as: in, It is a spatial smoothing operator. It is the regularization parameter.

[0071] The Markov chain Monte Carlo method (MCMC) is used to compute the posterior distribution. This allows us to obtain the maximum likelihood estimate of the parameters. And uncertainty. The core of MCMC is to construct a Markov chain, the stationary distribution of which is the posterior distribution of the target:

[0072] in, It is a proposed distribution. This indicates the number of iterations. During the model validation and optimization phase, simulated coral model data and experimentally collected data are used to test and validate the algorithm. Adjustments and optimizations are then made based on the results to improve inversion accuracy and reliability.

[0073] Furthermore, this step utilizes the two-way mapping relationship between coral optical properties and symbiotic algae density to map the inverted coral optical property parameters point by point, thereby obtaining an estimate and uncertainty of zooxanthellae density on the coral surface, and finally obtaining the spatial distribution of zooxanthellae density, which is expressed as:

[0074] in, Spatial location Zooxanthindella density at the location It is a mapping function. The performance of the inversion model is influenced by the accuracy of the model input parameters and the selection of measurement and algorithm parameters, thus constructing an optimal inversion model for the spatial distribution of zooxanthellae density within the coral canopy. Model performance can be evaluated using the root mean square error (RMSE). in, This is an estimated zooxanthellae density. This is the actual density of zooxanthellae. It refers to the number of samples.

[0075] S50 records data, maps the optical properties of each area of ​​the coral to its health status, and assigns corresponding color levels to the 3D mesh model according to the health level, thus visualizing the coral's health status; calibrates the coral's optical property parameters, and verifies the relationship between the coral's optical property parameters and its health status; reproduces the 3D mesh model as a solid 3D model, combines the coral's health status with the 3D solid model and maps it point by point, sets different color levels from cool to warm or from dull to vibrant according to the different health levels of the coral from high to low, and then combines the color levels into the 3D solid model to visualize the coral's health status.

[0076] S60, Model Validation and Optimization: A stratified sampling strategy is used to iterate through the dataset to obtain an inversion framework for the density distribution of zooxanthellae in corals; the comprehensive dataset is divided into training, validation, and test sets using a stratified sampling strategy, and multiple iterations of training and testing are performed to obtain a stable dataset; the stable dataset is regularized and the model complexity is constrained to construct an inversion framework for the density of zooxanthellae in underwater coral ecosystems.

[0077] In detail, a stratified sampling strategy is used to divide the dataset into training, validation, and test sets. Ten-fold cross-validation (i.e., randomly splitting the dataset into ten equal parts, each called a "fold," using nine folds as the training set and the remaining fold as the test set, performing ten training and testing iterations, each using a different fold as the test set, and finally averaging the results of these ten tests to evaluate the overall model performance) reduces the impact of random errors and ensures consistency in data distribution across subsets. In model performance evaluation, the root mean square error (RMSE) is used to quantify prediction accuracy, and the coefficient of determination (COP) is used to determine the model's performance. The mean absolute error (MAE) measures the explanatory power of a model on real data, assesses the magnitude of prediction bias, and the relative RMSE (rRMSE) compares model performance across different scales. Understandably, model optimization can also employ a combination of grid search and random search to explore the hyperparameter space and find the optimal configuration. Simultaneously, to prevent overfitting, L1 and L2 regularization techniques can be introduced (i.e., adding regularization terms L1 (the sum of the absolute values ​​of the parameters) and L2 (the sum of the squares of the weight vectors) to the loss function to penalize it, thus making the model more stable and having stronger generalization ability), to constrain model complexity and enhance its generalization ability.

[0078] Furthermore, this invention also enables sensitivity analysis. It accurately identifies spectral features in zooxanthellae density retrieval through feature importance analysis and / or evaluates the sensitivity of each input parameter in the model using the Morris screening method. Together, these constitute a high-precision retrieval framework for zooxanthellae density in underwater coral ecosystems.

[0079] It is understood that this invention can also be applied to generate density distribution maps of zooxanthellae in underwater coral ecosystems, and combined with geographic information systems (GIS) to transform them into spatial distribution maps, showing regional differences in zooxanthellae density.

[0080] In combination with all the above embodiments, the present invention provides an inversion method for the density distribution of coral polyps and yellow algae. By acquiring three-dimensional information of underwater corals based on hyperspectral remote sensing technology, and using an improved inversion algorithm, as well as introducing Bayesian parameter estimation and MCMC algorithm (Markov chain Monte Carlo algorithm) to quantitatively assess the uncertainty of the inversion results, a reliable inversion estimation of the density distribution of underwater coral polyps and yellow algae is achieved. This enables non-destructive, large-scale, and high-precision monitoring of coral reefs, and helps to better understand and predict changes in coral reef ecosystems.

[0081] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural transformations made using the contents of the present invention's specification and drawings under the inventive concept of the present invention, or direct / indirect applications in other related technical fields, are included within the patent protection scope of the present invention.

Claims

1. A method for inverting the density distribution of coral polyps and xanthophytes, characterized in that, Includes the following steps: Spectral data preparation, obtaining the reflectance and fluorescence spectra of corals; Three-dimensional point cloud information acquisition: Collect three-dimensional point cloud information of coral, and integrate the three-dimensional point cloud information of coral with reflectance spectrum and fluorescence spectrum to obtain multidimensional dataset; Data processing involves correcting reflectance and fluorescence spectra and combining them with a multidimensional dataset to create a comprehensive dataset. A model was established, and a 3D mesh model of coral was acquired using binocular vision technology. The 3D point cloud information of coral was reconstructed, and the reconstructed 3D point cloud information was meshed to generate an accurate 3D mesh model. A likelihood function was constructed based on the forward model and observation error. At the same time, based on the prior probability distribution function, the posterior distribution function was calculated to obtain the simulated coral model data. The simulated coral model data was used to invert and test the collected data, and a solution model for the spatial distribution of coral optical property parameters was constructed. The inverted coral optical property parameters were mapped point by point to the 3D mesh model to obtain the spatial distribution of zooxanthellae density. Record data, map the optical properties of each area of ​​the coral to the health status of the coral, and assign corresponding color levels to the three-dimensional mesh model according to the health level to visualize the health status of the coral. Model validation and optimization: After iterating through the dataset using a stratified sampling strategy, an inversion framework for the density distribution of coral polyps and yellow algae was obtained.

2. The method for inverting the density distribution of coral polyps and xanthophytes as described in claim 1, characterized in that, The steps for preparing the spectral data and obtaining the reflectance and fluorescence spectra of corals include: Collect coral health information; Use a hyperspectral sensor to acquire reflectance spectral data of the coral surface; Fluorescent proteins in coral tissue were excited using an excitation light source, and fluorescence signals were recorded and fluorescence spectra were collected. The collected fluorescence spectra are smoothed to remove noise interference and obtain effective fluorescence spectra; A dataset was established, and the reflectance spectrum was correlated with the effective fluorescence spectrum and coral health information to obtain the correspondence between the spectrum and the optical properties of coral.

3. The method for inverting the density distribution of coral polyps and xanthophytes as described in claim 1, characterized in that, The steps of acquiring 3D point cloud information, including collecting 3D point cloud information of coral, and integrating the 3D point cloud information of coral with reflectance and fluorescence spectra to obtain a multidimensional dataset, include: A structured light scanner was used to perform a 3D scan of the coral, obtaining 3D point cloud information of the coral; By integrating the three-dimensional point cloud information of corals with reflectance and fluorescence spectra, and combining spatial registration technology, the spatial information in the three-dimensional point cloud information of corals is combined with the optical property information in the spectral data to obtain a multidimensional dataset.

4. The method for inverting the density distribution of coral polyps and xanthophytes as described in claim 1, characterized in that, The data processing steps, including correcting reflectance and fluorescence spectra and combining them with a multidimensional dataset to create a comprehensive dataset, include: Denoising and filtering are performed on the 3D point cloud information of corals to remove outliers and duplicate data, resulting in optimized 3D point cloud information of corals. The optimized coral 3D point cloud information was smoothed, and the reflectance and fluorescence spectral data were corrected. The corrected reflectance and fluorescence spectral data are remapped into a multidimensional dataset to create a comprehensive dataset.

5. The method for inverting the density distribution of coral polyps and xanthophytes as described in claim 1, characterized in that, The recorded data maps the optical properties of each region of the coral to its health status, and assigns corresponding color levels to the three-dimensional mesh model according to the health level, thus visualizing the coral's health status. The steps include: Calibrate the optical property parameters of corals and verify the relationship between the optical property parameters and the health status of corals; The three-dimensional mesh model is reproduced as a solid three-dimensional model. The health status of the coral is combined with the three-dimensional solid model and mapped point by point. Color levels are set according to the health status of the coral to visualize the health status of the coral.

6. The method for inverting the density distribution of coral polyps and xanthophytes as described in claim 1, characterized in that, The model validation and optimization process, which uses a stratified sampling strategy to iterate through the dataset to obtain the inversion framework for the density distribution of coral polyps and xanthophytes, includes the following steps: The comprehensive dataset is divided into training, validation and test sets using a stratified sampling strategy. Multiple iterations of training and testing are performed to obtain a stable dataset. By regularizing stable datasets and constraining model complexity, an inversion framework for zooxanthellae density in underwater coral ecosystems is constructed.

Images