Green tide algae biomass estimation method and system based on aerial remote sensing interpretation

Abstract

The invention discloses a green tide algae biomass estimation method and system based on aerial remote sensing interpretation. The method comprises the following steps: S1, collecting a target coastal aerial remote sensing image and green tide algae data; s2, interpreting a beach type by using a convolutional neural network; s3, constructing a biomass estimation model; and S4, carrying out dynamic optimization and verification on the biomass estimation model. According to the method, the beach type is interpreted through aerial remote sensing for the first time, the quantitative relation between the beach type and the green tide algae biomass is established, and the regionalized biomass estimation precision is improved. Through multi-source data fusion, in combination with an aerial remote sensing image, historical green tide algae distribution data and an on-site sampling result, a dynamic weight model is constructed, and biomass prediction of different seasons and tide levels is optimized.

Description

Technical Field

[0001] The present invention relates to the field of marine environment monitoring, and in particular to a method and system for estimating the biomass of green tide algae based on aerial remote sensing interpretation. Background Art

[0002] Outbreaks of green tide algae (such as Enteromorpha) pose a serious threat to offshore ecology, fisheries, and tourism. Traditional biomass monitoring relies on manual field sampling, which has problems such as high cost, poor timeliness, and difficulty covering large areas of sea. Existing remote sensing technologies mostly focus on the inversion of chlorophyll concentration in water bodies, but do not effectively incorporate the impact of differences in beach types on the distribution of green tide algae. The bottom sediments, tidal conditions, and human activities of different beach types (such as grass beaches, stone embankments, and aquaculture dams) significantly affect the attachment and growth of green tide algae, but existing technologies lack high-precision biomass estimation models based on beach types. In addition, the application of aerial remote sensing data in shoreline classification has not yet been deeply integrated with the dynamic monitoring of green tide algae, resulting in insufficient prediction accuracy. Summary of the Invention

[0003] The purpose of the present invention is to provide a method and system for estimating the biomass of green tide algae based on aerial remote sensing interpretation, so as to solve the problems raised in the above background technology.

[0004] To achieve the above-mentioned object, one aspect of the present invention provides a method for estimating the biomass of green tide algae based on aerial remote sensing interpretation, comprising the following steps:

[0005] Step S1, collecting aerial remote sensing images of the target coast and green tide algae data;

[0006] Step S2, using a convolutional neural network to interpret the beach type;

[0007] Step S3, constructing a biomass estimation model;

[0008] Step S4: Dynamically optimize and verify the biomass estimation model.

[0009] Furthermore, the green tide algae data includes green tide algae species, biomass, distribution location and shore type labels.

[0010] Furthermore, in step S2, a convolutional neural network is used to automatically classify the remote sensing image, output a shore distribution map, and combine the classification results with manual visual interpretation to generate high-precision shore type vector data.

[0011] Furthermore, the beach types include grass beach, reed salt wormwood, aquaculture dam, coastal stone embankment, estuary beach and others.

[0012] Furthermore, step S3 includes the following steps:

[0013] Step S301, calculating the historical biomass averages of different beach types;

[0014] Step S302: Establish a multiple regression model, which is expressed as follows:

[0015]

[0016] Where B is the total biomass, k i is the biomass coefficient of the i-th type of beach, A i is the area of this type of beach, C is the environmental correction factor, and n represents the number of beach types.

[0017] Furthermore, the dynamic optimization in step S4 introduces time series analysis to adjust the biomass coefficient according to seasonal changes.

[0018] Furthermore, in step S4, a cross-validation method is used to verify the model, and the average error of the model is calculated based on the field sampling data.

[0019] Another aspect of the present invention provides a green tide algae biomass estimation system based on aerial remote sensing interpretation, comprising a collection module, an interpretation module, a modeling module, and an optimization and verification module, wherein:

[0020] The acquisition module is used to collect aerial remote sensing images of the target coast and green tide algae data;

[0021] The interpretation module is used to interpret the shore type using a convolutional neural network;

[0022] The modeling module is used to build biomass estimation models;

[0023] The optimization and verification module is used to dynamically optimize and verify the biomass estimation model.

[0024] Compared with the prior art, the present system and method have the following advantages:

[0025] 1. This study, for the first time, uses aerial remote sensing to interpret beach types (grass beach, stone embankment, aquaculture dam, etc.), establishing a quantitative relationship between beach type and green tide algae biomass, thereby improving the accuracy of regional biomass estimation.

[0026] 2. This invention constructs a dynamic weight model by fusing multi-source data, combining aerial remote sensing images, historical green tide algae distribution data and field sampling results, and optimizes biomass prediction in different seasons and tide levels.

[0027] 3. The present invention develops an automatic shore type identification algorithm based on deep learning to reduce manual visual interpretation errors and improve data processing efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1Flowchart of the green tide algae biomass estimation method based on aerial remote sensing interpretation.

[0029] Figure 2 This is a schematic diagram of the beach type classification results. DETAILED DESCRIPTION

[0030] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0031] like Figure 1 The flowchart of the method of the present invention is shown. The present invention provides a method for estimating the biomass of green tide algae based on aerial remote sensing interpretation. The specific steps are as follows:

[0032] Step S1: collecting aerial remote sensing images of the target coast and green tide algae data.

[0033] Acquire aerial remote sensing images of the target coastline (resolution ≤ 1m), covering visible light and near-infrared bands; simultaneously collect field sampling data (green tide algae species, biomass, distribution location and shore type labels).

[0034] Step S2: using a convolutional neural network to interpret the shore type.

[0035] The convolutional neural network (CNN) is used to automatically classify remote sensing images and output distribution maps of six types of beaches, including grassy beaches, reed and salt wormwood, aquaculture dams, coastal stone embankments, estuary beaches, and others. The classification results are corrected by manual visual interpretation to generate high-precision beach type vector data. Figure 2 Shown is a schematic diagram of the generated beach type classification results.

[0036] Step S3: constructing a biomass estimation model.

[0037] The following steps are involved:

[0038] Step S301: Count the historical average biomass of different beach types (e.g., the average biomass of grass beach is 12.8 g / m², and the average biomass of stone embankment is 25.6 g / m²);

[0039] Step S302: Establish a multiple regression model, which is expressed as follows:

[0040]

[0041] Where B is the total biomass, k i is the biomass coefficient of the i-th type of beach, A iis the area of this type of beach, C is the environmental correction factor, and n represents the number of beach types.

[0042] Step S4: Dynamic Optimization and Validation. Time series analysis was introduced to adjust the biomass coefficient based on seasonal variations (e.g., a 15% increase in the biomass coefficient of the stone bank in spring). Cross-validation was used, using field sampling data as a benchmark, to calculate the average model error (target ≤ 10%).

[0043] The technical solution of the present invention is described below with reference to two embodiments.

[0044] Example 1: Estimation of beach biomass at Haozhi Harbor in Nantong.

[0045] Input aerial remote sensing images from April 2023 (resolution 0.5m); automatically classify and identify the grassy beach and stone embankment areas as 45% and 22%, respectively; use historical coefficients (grassy beach k1=12.8k1=12.8, stone embankment k2=25.6k2=25.6) to calculate the theoretical biomass: B=12.8×0.45+25.6×0.22=15.2 g / m 2 ; Adding the water temperature correction factor (+8%), the final biomass prediction output is 16.4 g / m², which has an error of 4.1% from the measured value of 17.1 g / m².

[0046] Example 2: Dynamic monitoring of Sheyang Estuary in Yancheng.

[0047] Seasonal adjustments were made to the Sheyang River estuary in Yancheng, with the biomass coefficient of aquaculture dams reduced by 10% in autumn. The model predicted that the biomass would exceed the threshold (>50 g / m²), triggering a red tide warning signal and guiding management departments to intervene in advance.

[0048] The above examples demonstrate that the correlation coefficient between the model's estimated results and field data is ≥0.85, significantly outperforming traditional remote sensing inversion methods (correlation coefficient ≤0.6). This method utilizes automated processing throughout the entire process, with a single 580-kilometer coastline analysis taking ≤2 hours, a 90% improvement in efficiency compared to manual sampling. This method has broad application, including early warning of green algae blooms, shoreline restoration planning, and aquaculture area management.

[0049] The embodiments of the present invention have been specifically described above, but the present invention is not limited to the embodiments. Those skilled in the art may make various equivalent modifications or substitutions without departing from the spirit of the present invention. These equivalent modifications or substitutions are all included in the scope defined by the claims of this application.

Claims

1. A method for estimating green tide algae biomass based on aerial remote sensing interpretation, characterized in that: The following steps are involved: Step S1, collecting aerial remote sensing images of the target coast and green tide algae data; Step S2, using a convolutional neural network to interpret the beach type; Step S3, constructing a biomass estimation model; Step S4: Dynamically optimize and verify the biomass estimation model.

2. The method for estimating green tide algae biomass based on aerial remote sensing interpretation according to claim 1, characterized in that: The green tide algae data includes green tide algae species, biomass, distribution location and shore type labels.

3. The method for estimating green tide algae biomass based on aerial remote sensing interpretation according to claim 1, characterized in that: In step S2, a convolutional neural network is used to automatically classify the remote sensing image, output a shore distribution map, and combine the classification results with manual visual interpretation to generate high-precision shore type vector data.

4. The method for estimating green tide algae biomass based on aerial remote sensing interpretation according to claim 1, characterized in that: The beach types include grass beach, reed salt wormwood, aquaculture dam, coastal stone embankment, estuary beach and others.

5. The method for estimating green tide algae biomass based on aerial remote sensing interpretation according to claim 1, characterized in that: Step S3 includes the following steps: Step S301, calculating the historical biomass averages of different beach types; Step S302: Establish a multiple regression model, which is expressed as follows: Where B is the total biomass, k i is the biomass coefficient of the i-th type of beach, A i is the area of this type of beach, C is the environmental correction factor, and n represents the number of beach types.

6. The method for estimating green tide algae biomass based on aerial remote sensing interpretation according to claim 1, characterized in that: The dynamic optimization described in step S4 introduces time series analysis to adjust the biomass coefficient according to seasonal changes.

7. The method for estimating green tide algae biomass based on aerial remote sensing interpretation according to claim 1, characterized in that: In step S4, a cross-validation method is used for verification, and the average error of the model is calculated based on the field sampling data.

8. The green tide algae biomass estimation system based on aerial remote sensing interpretation is characterized by: It includes acquisition module, interpretation module, modeling module, and optimization and verification module, among which: The acquisition module is used to collect aerial remote sensing images of the target coast and green tide algae data; The interpretation module is used to interpret the shore type using a convolutional neural network; The modeling module is used to build biomass estimation models; The optimization and verification module is used to dynamically optimize and verify the biomass estimation model.

Images