Automatic identification method and system for water ecology monitoring shoreside station based on artificial intelligence

Abstract

The invention provides an artificial intelligence-based automatic identification method and system for a water ecology monitoring shoreside station, and the method comprises the steps: carrying out the enhancement processing of the data of a sample, carrying out the classification and counting based on an improved small target recognition deep learning model, and carrying out the early warning analysis of a monitoring position through data statistics and fusion sensing data, thereby achieving the automatic identification of a water ecology monitoring shoreside station. Current data are displayed for a user in a visual display mode, plankton in a target area is pre-warned through a pre-warning means, and fine support is provided for accurate monitoring and pre-warning of algal blooms and subsequent processing.

Description

Technical Field

[0001] The present invention relates to the technical field of water body data processing and intelligent monitoring, and in particular to an automatic identification method and system for a water ecological monitoring shore station based on artificial intelligence. Background Art

[0002] With the rapid development of industrialization and urbanization, water pollution is becoming increasingly serious, posing a huge threat to human health and the ecosystem. Traditional water ecological monitoring methods mainly rely on manual sampling and laboratory analysis, which have problems such as complex operation, low efficiency, long survey cycle, and high professional requirements for investigators. In addition, the sample data comes from biological individuals captured by traditional survey methods, and the biological species are identified from the morphology through the knowledge and experience of taxonomists, which leads to low accuracy caused by human errors and system errors.

[0003] Common floating algae in-situ monitoring equipment is mostly algae classification products based on fluorescence spectroscopy. By capturing the fluorescence of living algae, they can obtain information on chlorophyll and algae density in the water. The analysis of algae classification is limited to the phylum level, and it is impossible to obtain classification information at the genus and species level. It is difficult to provide refined support for accurate monitoring, early warning and subsequent treatment of cyanobacteria blooms. Summary of the invention

[0004] In the first aspect, the present invention provides an automatic identification method for aquatic ecological monitoring shore station based on artificial intelligence to solve the technical problems existing in the prior art, specifically including:

[0005] Step 1: The sampling device obtains water at a certain depth below the water surface and collects it into a sample water cup; specifically, the online control system automatically collects water into the sample water cup.

[0006] Step 2, using a micro peristaltic pump to extract a certain amount of water sample from the sample cup and send it into the detection slide (sample pool);

[0007] Step 3, the biological microscope automatically focuses and takes photos of the phytoplankton and zooplankton in the water sample on the test slide. Specifically, the electric stage moves the field of view in the XY direction, while the microscope objective lens focuses vertically, and the clear microscopic image of the tissue on the slide is converted into a digital image through the electronic camera;

[0008] Step 4, using image preprocessing and image enhancement and classification recognition of the above digital image based on a deep learning model, and obtaining the phytoplankton and zooplankton identification and counting results on the membrane according to the output of the deep learning model;

[0009] Step 5, statistically obtain the classification results of phytoplankton and zooplankton in the water body, provide the name, cell number, and location information of the species in each analysis field of view; summarize the Chinese name, Latin name, and species classification status of the identified species; statistically calculate the average single cell length, single cell width, single cell height, single cell diameter, single cell area, single cell volume, cell density, biomass, etc. of the species, and display the species diversity of the samples in the form of charts; calculate the Shannon-Wienner index, Margalef richness index, Pielou evenness index, Simpson ecological dominance index and water quality evaluation.

[0010] Step 6: extract the processed data according to the results of steps 3-5 to obtain the data to be displayed, transmit the data to be displayed to the data center, and the data center sends the corresponding display data and displays it to the user on the terminal.

[0011] Optionally, the present invention further comprises step 7, pre-setting a monitoring warning value, and when the corresponding data exceeds the warning value, a text message or a data prompt on the platform can be used to remind the user to start the emergency plan.

[0012] In a second aspect, the present invention provides an automatic identification system for water ecological monitoring shore stations based on artificial intelligence.

[0013] The sample acquisition unit, the sampling device acquires water at a certain depth below the water surface and collects it into the sample pool; specifically, the online control system automatically collects water into the water sample pool, and then the self-priming pump draws the water sample into the sample pool;

[0014] In the sample preparation unit, the biological microscope automatically focuses and takes photos of phytoplankton and zooplankton in the water sample on the test slide. The motorized stage moves the field of view in the XY direction, while the microscope objective lens focuses vertically, and the electronic camera converts the clear microscopic image of the tissue on the slide into a digital image;

[0015] Imaging unit, biological microscope automatically focuses on taking photos of phytoplankton and zooplankton on the membrane. The motorized stage moves the field of view in the XY direction, while the microscope objective lens focuses vertically, and the electronic camera converts the clear microscopic image of the tissue on the slide into a digital image;

[0016] The classification and recognition unit uses image preprocessing and deep learning model to perform image enhancement and classification recognition on the digital image, and obtains the phytoplankton and zooplankton recognition and counting results on the membrane according to the deep learning model output;

[0017] The statistical calculation unit obtains the classification results of phytoplankton and zooplankton in the water body, and provides the name, cell number, and location information of the species in each analysis field of view; summarizes the Chinese name, Latin name, and species classification status of the identified species; calculates the average single cell length, single cell width, single cell height, single cell diameter, single cell area, single cell volume, cell density, biomass, etc. of the species, and displays the species diversity of the samples in the form of charts; calculates the Shannon-Wienner index, Margalef richness index, Pielou evenness index, Simpson ecological dominance index and water quality evaluation.

[0018] The sensing data acquisition unit,the sampling device includes a float, which is also equipped with a positioning system to record the sampling coordinates in real time.

[0019] The buoy is also equipped with a variety of sensors such as water temperature, pH, dissolved oxygen, dissolved oxygen saturation, conductivity, turbidity, chlorophyll a, and blue-green algae density to obtain water quality parameter information regularly;

[0020] The buoy is also equipped with equipment for detecting meteorological parameters such as temperature and humidity, wind speed, wind direction, sunshine, light, depth, and precipitation.

[0021] The data fusion display unit extracts the data obtained by the above processing units to obtain the data to be displayed, and transmits the data to be displayed to the data center. The data center sends the corresponding display data and displays it to the user on the terminal.

[0022] Optionally, the present invention further comprises an early warning unit, which pre-sets a monitoring early warning value, and when the corresponding data exceeds the early warning value, a text message or a data prompt on the platform can be used to remind the user to start the emergency plan.

[0023] Beneficial effects of the present invention: an automatic identification method for shore stations for water ecological monitoring based on artificial intelligence is proposed, which enhances the data of samples and performs classification and counting based on an improved deep learning model for small target recognition, and performs early warning analysis of the monitoring location through data statistics and fusion of sensor data. The current data is displayed to users in an intuitive manner, and early warning means are used to warn plankton in the target area, providing refined support for accurate monitoring, early warning and subsequent processing of algal blooms.

[0024] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the related technologies, the drawings required for use in the specific embodiments or the related technical descriptions will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0026] Figure 1 Flowchart of the method for automatic identification of shore stations for water ecological monitoring based on artificial intelligence in this application;

[0027] Figure 2 The digital image enhancement result of one of the embodiments of the present application

[0028] Figure 3 A diagram of the deep learning classification and recognition architecture of one of the embodiments of the present application;

[0029] Figure 4 An example of one of the embodiments of the present application being displayed to a user on a terminal. DETAILED DESCRIPTION

[0030] The technical solution of the present invention will be described clearly and completely below 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. 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.

[0031] In the description of the present invention, it should be noted that the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. In addition, the terms "first", "second", and "third" are only used to describe the difference in names, and cannot be understood as indicating or implying relative importance. The physical quantities in the formula, if not separately marked, should be understood as basic quantities of the basic units of the International System of Units, or derived quantities derived from the basic quantities through mathematical operations such as multiplication, division, differentiation or integration.

[0032] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0033] like Figure 1 As shown, the embodiment of the present invention provides an automatic identification method for aquatic ecological monitoring shore station based on artificial intelligence to solve the technical problems existing in the prior art, specifically including:

[0034] Step 1, the sampling device obtains water at a certain depth below the water surface and collects it into a sample water cup; specifically, the water is automatically sampled and fed into the sample water cup through an online control system; wherein the sampling device includes a water sampling pump, a water sampling pipeline, a float, a coarse filter screen and a sample feed filter head.

[0035] Step 2, using a micro peristaltic pump to extract a certain amount of water sample from the sample cup and send it into the detection slide (sample pool);

[0036] Step 3, the biological microscope automatically focuses and takes photos of the phytoplankton and zooplankton in the water sample on the test slide. Specifically, the electric stage moves the field of view in the XY direction, while the microscope objective lens focuses vertically, and the clear microscopic image of the tissue on the slide is converted into a digital image through the electronic camera;

[0037] Step 4, using image preprocessing and image enhancement and classification recognition of the above digital image based on a deep learning model, and obtaining the phytoplankton and zooplankton identification and counting results on the membrane according to the output of the deep learning model;

[0038] In this embodiment, see the attached Figure 2 This is the result after digital image enhancement. See also Figure 3 A deep learning model is used to obtain recognition results and count according to classification. Specifically, in this embodiment, the following image enhancement and classification recognition method is used to classify phytoplankton and zooplankton; the specific image enhancement algorithm can be a histogram equalization image enhancement algorithm, a wavelet transform image enhancement algorithm, a partial differential equation image enhancement algorithm, an image enhancement algorithm based on Retinex theory, or an algorithm based on deep learning;

[0039] The result after digital image enhancement is used as the input of the deep learning model. The deep learning model of the present invention is specifically composed of four consecutive integrated convolutional layers composed of separable convolution ReLU and maximum pooling, a convolutional residual layer, and three consecutive integrated convolutional layers and classification layers. After the convolution residual layer outputs the features through the first layer of convolution, the network divides the features into multiple groups according to the number of channels, each group of features is Ri, and then convolution operation is performed on each group of features except Ri; the input of each calculation is composed of the previous group of residuals connected to the current group of features Ri; finally, all the obtained multi-scale features are spliced ​​and input into the next convolutional layer to obtain the output of the residual layer.

[0040] First, perform a deep convolution operation on the data in each channel of the input feature; then perform a point-by-point convolution operation using a convolution kernel of size 1×1×M, where M is the number of feature channels in the previous layer, and combine the outputs of different channels to obtain the final output result;

[0041] The present invention proposes that for the microscopic targets of phytoplankton and zooplankton, it is necessary to further improve the feature extraction capability on the basis of the traditional deep network. Therefore, a residual convolution layer is introduced to realize the downsampling features transmitted by the skip connection layer and realize the refined extraction of microscopic image information. At the same time, in order to reduce and reduce the number of parameters, all convolution layers are adjusted to deep separable convolution layers with a kernel size of 3X3. After the interval between the two is set, both the feature extraction capability of the network and the training efficiency of the network are satisfied.

[0042] At the same time, in the output layer, the present invention uses three consecutive full convolution and activation layers to further enhance the segmentation and recognition capabilities of tiny targets, and finally outputs them to the classification SoftMax layer to obtain the classification results of phytoplankton and zooplankton.

[0043] The acquisition of sample data of various phytoplankton and zooplankton used in the deep learning model and the deep learning training iterations are not elaborated here.

[0044] Step 5, statistically obtain the classification results of phytoplankton and zooplankton in the water body, provide the name, cell number, and location information of the species in each analysis field of view; summarize the Chinese name, Latin name, and species classification status of the identified species; statistically calculate the average single cell length, single cell width, single cell height, single cell diameter, single cell area, single cell volume, cell density, biomass, etc. of the species, and display the species diversity of the samples in the form of charts; calculate the Shannon-Wienner index, Margalef richness index, Pielou evenness index, Simpson ecological dominance index and water quality evaluation.

[0045] Step 6: extract the processed data according to the results of steps 3-5 to obtain the data to be displayed, transmit the data to be displayed to the data center, and the data center sends the corresponding display data and displays it to the user on the terminal.

[0046] See attached Figure 3 In this embodiment, the displayed data includes the classification name, physical morphology and biomass statistics, density ranking, and proportion statistical pie chart of each algae species in the water sample.

[0047] Optionally, the present invention further comprises step 7, pre-setting a monitoring warning value, and when the corresponding data exceeds the warning value, a text message or a data prompt on the platform can be used to remind the user to start the emergency plan.

[0048] In the second embodiment, the present invention provides an automatic identification system for water ecological monitoring shore stations based on artificial intelligence.

[0049] The sample acquisition unit, the sampling device acquires water at a certain depth below the water surface and collects it into the sample pool; specifically, the online control system automatically collects water into the water sample pool, and then the self-priming pump draws the water sample into the sample pool;

[0050] In the sample preparation unit, the biological microscope automatically focuses and takes photos of phytoplankton and zooplankton in the water sample on the test slide. The motorized stage moves the field of view in the XY direction, while the microscope objective lens focuses vertically, and the electronic camera converts the clear microscopic image of the tissue on the slide into a digital image;

[0051] Imaging unit, biological microscope automatically focuses on taking photos of phytoplankton and zooplankton on the membrane. The motorized stage moves the field of view in the XY direction, while the microscope objective lens focuses vertically, and the electronic camera converts the clear microscopic image of the tissue on the slide into a digital image;

[0052] The classification and recognition unit uses image preprocessing and deep learning model to perform image enhancement and classification recognition on the digital image, and obtains the phytoplankton and zooplankton recognition and counting results on the membrane according to the deep learning model output;

[0053] The statistical calculation unit obtains the classification results of phytoplankton and zooplankton in the water body, and provides the name, cell number, and location information of the species in each analysis field of view; summarizes the Chinese name, Latin name, and species classification status of the identified species; calculates the average single cell length, single cell width, single cell height, single cell diameter, single cell area, single cell volume, cell density, biomass, etc. of the species, and displays the species diversity of the samples in the form of charts; calculates the Shannon-Wienner index, Margalef richness index, Pielou evenness index, Simpson ecological dominance index and water quality evaluation.

[0054] The data fusion display unit extracts the processed data according to the above results to obtain the data to be displayed, transmits the data to be displayed to the data center, and the data center sends the corresponding display data and displays it to the user on the terminal.

[0055] Optionally, the present invention further comprises an early warning unit, which pre-sets a monitoring early warning value, and when the corresponding data exceeds the early warning value, a text message or a data prompt on the platform can be used to remind the user to start the emergency plan.

[0056] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An automatic identification method for water ecological monitoring shore stations based on artificial intelligence, characterized in that Include: Step 1: The sampling device obtains water at a certain depth below the water surface and collects it into a sample water cup; Step 2, using a micro peristaltic pump to extract a certain amount of water sample from the sample cup and send it to the test slide; Step 3, the biological microscope automatically focuses on taking photos of the phytoplankton and zooplankton in the water sample in the test slide, and converts the clear microscopic images of the tissues on the slide into digital images through an electronic camera; Step 4, using image preprocessing and image enhancement and classification recognition of the above digital image based on a deep learning model, and obtaining the phytoplankton and zooplankton identification and counting results on the membrane according to the output of the deep learning model; Step 5, obtaining the classification results of phytoplankton and zooplankton in the water body by statistics; Step 6: extract the processed data according to the results of steps 3-5 to obtain the data to be displayed, transmit the data to be displayed to the data center, and the data center sends the corresponding display data and displays it to the user on the terminal.

2. The method according to claim 1, characterized in that In step 1, the water sampling pump is automatically used by the online control system, and then the water sample is pumped into the test slide by the micro peristaltic pump.

3. The method according to claim 2, characterized in that In step 3, the motorized stage moves the field of view in the XY direction, while the microscope objective lens focuses vertically, and the clear microscopic image of the tissue on the slide is converted into a digital image by the electronic camera.

4. The method according to any one of claim 3, characterized in that The image enhancement in step 4 specifically includes: histogram equalization image enhancement algorithm, wavelet transform image enhancement algorithm, partial differential equation image enhancement algorithm, image enhancement algorithm based on Retinex theory or algorithm based on deep learning.

5. The method according to any one of claim 4, characterized in that In step 4, the deep learning model outputs the identification, classification and counting results of phytoplankton and zooplankton in the detection glass slide, including: the deep learning model is specifically composed of four consecutive integrated convolutional layers composed of separable convolution ReLU and maximum pooling, a convolutional residual layer and three consecutive integrated convolutional layers and classification layers. After the convolution residual layer outputs the features through the first layer of convolution, the network divides the features into multiple groups according to the number of channels, each group of features is Ri, and then convolution operations are performed on each group of features except Ri; the input of each calculation is composed of the previous group of residuals connected to the current group of features Ri; finally, all the multi-scale features obtained are spliced ​​and input into the next convolutional layer to obtain the output of the residual layer.

6. The method according to claim 1, characterized in that Step 1 The sampling device includes a water sampling pump, a water sampling pipeline, a float, a coarse filter and a sampling filter head.

7. The method according to claim 5, characterized in that It also includes step 7, pre-setting the monitoring warning value, and when the corresponding data exceeds the warning value, the user can be reminded to start the emergency plan through text messages or data prompts on the platform.

8. An automatic identification of water ecological monitoring shore stations based on artificial intelligence, characterized by Contains the following processing units: The sample acquisition unit, the sampling device acquires water at a certain depth below the water surface and collects it into a sample water cup; The sample preparation unit uses a micro peristaltic pump to extract a certain amount of water sample from the sample cup and send it to the test slide; The imaging unit, the biological microscope, automatically focuses on taking photos of phytoplankton and zooplankton in the water sample in the test slide, and converts the clear microscopic images of the tissues on the slide into digital images through an electronic camera; The classification and recognition unit uses image preprocessing and deep learning model to perform image enhancement and classification recognition on the digital image, and obtains the phytoplankton and zooplankton recognition and counting results on the membrane according to the deep learning model output; A statistical calculation unit, which obtains the classification results of phytoplankton and zooplankton in the water body; The data fusion display unit extracts the data obtained by the above processing units to obtain the data to be displayed, and transmits the data to be displayed to the data center. The data center sends the corresponding display data and displays it to the user on the terminal.

9. The generation system according to claim 7, characterized in that It also includes an early warning unit, which pre-sets monitoring early warning values. When the corresponding data exceeds the early warning value, it can remind users to activate the emergency plan through text messages or data prompts on the platform.

10. The system according to claim 7, characterized in that The sampling device includes a water sampling pump, a water sampling pipeline, a float, a coarse filter and a sampling filter head.

Images