Imaging system and method for field monitoring of phytoplankton

Abstract

The invention discloses an imaging system and method, belongs to the technical field of field monitoring of phytoplankton, and particularly relates to an imaging system and method for field monitoring of phytoplankton, the system comprises a water sample collection device, an image collection device, a data processing device, a sample analysis platform, a communication device and an auxiliary device, the device is designed for realizing field monitoring, real-time and continuous photographing of phytoplankton (algae), automatic three-axis (X, Y and Z) focusing photographing and rapid imaging can be realized, high-definition pictures can be automatically selected and uploaded, and an effective and reliable basis is provided for algae analysis data, so that real-time unmanned continuous detection and analysis of a water sample can be realized, and the detection efficiency is improved. According to the method, automatic and rapid focusing and photographing are achieved, pictures meeting requirements are automatically selected, effective and reliable image bases are provided for background algorithm software, the requirement that algae can be accurately analyzed on the field water surface is met, and then the images are automatically uploaded to a server or a cloud platform.

Description

Technical Field

[0001] The invention belongs to the technical field of phytoplankton field monitoring, and in particular relates to an imaging system and method for phytoplankton field monitoring. Background Art

[0002] Phytoplankton monitoring in the field is an important ecological research task that involves the assessment of water health and biodiversity. Phytoplankton field monitoring generally includes steps such as determining monitoring targets, selecting sampling points, sample processing and analysis, and reporting and application. Through these steps, you can effectively monitor the status of phytoplankton in the field water, thereby conducting a comprehensive assessment of the water ecosystem.

[0003] Common outdoor water imaging equipment includes video surveillance, underwater cameras, and drone aerial photography. Video surveillance needs to be fixed on a surface carrier, and can only take photos and images of phytoplankton (algae) on the surface of the water. Water flow and waves have a great impact on image quality. Underwater cameras also need to be fixed somewhere on the bottom of the water, and cannot focus to continuously complete the shooting of high-definition images. Drones need to be manually operated, and there are great limitations on the time of water sample monitoring. Affected by outdoor weather, it is impossible to achieve real-time all-weather continuous monitoring. Drones can only take photos of algae on the surface of the water. The representativeness of water samples is poor and cannot meet the requirements. Summary of the invention

[0004] Purpose of the invention: To provide an imaging system for field monitoring of phytoplankton to solve the above-mentioned problems.

[0005] Technical solution:

[0006] In a first aspect, an imaging system for field monitoring of phytoplankton is provided, the imaging system comprising:

[0007] Water sampling equipment, used to collect water from the area

[0008] Image acquisition equipment, using high-definition imaging equipment to collect images of water samples, and using a three-axis adjustment bracket to control the position of the high-definition imaging equipment to achieve multi-point focus photography;

[0009] A data processing device, which is installed in the image acquisition device and receives images acquired by the image acquisition device and analyzes and classifies the captured images;

[0010] A sample analysis platform receives the image data processed by the data processing device, processes and analyzes the image data, and outputs a report;

[0011] A communication device, which establishes a specific transmission channel; used for communication between the data processing device and the sample analysis platform;

[0012] The auxiliary device is installed on the image acquisition device and cooperates with the image acquisition device to acquire images.

[0013] In a further embodiment, the three-axis adjustment bracket in the image acquisition device adopts an automatic three-axis driving bracket, which is used for rapid movement of the X-axis, Y-axis and Z-axis of the high-definition imaging device, thereby achieving fast focusing and photographing.

[0014] In a further embodiment, in a further embodiment, the image acquisition device performs three rounds of photographing each time, and the first round of photographing is performed at different heights after three-axis focusing of each point, and a threshold is calculated for the imaging picture. The focus height that meets the threshold or the highest threshold is taken from the first round of imaging pictures, and the second round of photographing is performed at the point where the focus height meets the threshold or the highest threshold is located. The third round of photographing is performed at the point where the focus height meets the threshold or the highest threshold is taken from the second round of imaging pictures.

[0015] In a further embodiment, the data processing device is used to preliminarily analyze and classify the captured images on site to help quickly screen out valuable ones; the data processing device evaluates the quality of the captured images by constructing an image quality assessment model, and eliminates images that do not meet the quality requirements.

[0016] In a further embodiment, the sample analysis platform acquires the collected image on the slide through a microscope at a constant magnification, and performs statistics on the number and type of phytoplankton in the collected image, and stores the collected image and analysis structure in a computer for later processing and statistical use, and is externally connected to various platforms for data uploading and sharing.

[0017] In a further embodiment, the auxiliary equipment includes an auxiliary lighting device, which ensures that sufficient and uniform light is provided for imaging to highlight the characteristics of phytoplankton by installing appropriate lighting equipment on the image acquisition device.

[0018] In a second aspect, an imaging method for field monitoring of phytoplankton is implemented by the imaging system for field monitoring of phytoplankton described above, and the imaging method comprises:

[0019] Step 1: Take three rounds of photos through the image acquisition device at regular intervals or when the user initiates the process. The three rounds of photos are adjusted by the automatic three-axis driving bracket to adjust the positions of the X-axis, Y-axis and Z-axis. The focus height that meets the threshold or the highest threshold is selected to take photos.

[0020] Step 2: The data processing device analyzes, classifies and quickly screens the images collected by the image acquisition device, and transmits the images that meet the requirements through the communication device;

[0021] Step 3: The communication device establishes an independent transmission channel according to the work requirements and communicates with the sample analysis platform to transmit images;

[0022] Step 4: The sample analysis platform collects, counts, and identifies the types of phytoplankton in the phytoplankton image and stores and displays the collected images.

[0023] In a further embodiment, in step 1, the working method of the image acquisition device is as follows:

[0024] Step 11: After the photo process is triggered, the image acquisition device controls the objective lens to perform multi-point focusing on the water sample;

[0025] Step 12: After three-axis focusing of each point, take the first round of photos at different heights, and calculate the imaged pictures to obtain a threshold;

[0026] Step 13: Take the focus height that meets the threshold or the highest threshold in the first round of imaging pictures, and take the second round of pictures at the point where the focus height meets the threshold or the highest threshold;

[0027] Step 14: Take the focus height that meets the threshold or the highest threshold in the second round of imaging pictures, and take the third round of photos at the point where it is located;

[0028] Step 15: After taking three rounds of photos at each point, select a few high-definition photos with the highest threshold and upload them;

[0029] Step 16: After all images are uploaded, the process ends and the system goes into sleep mode to save power.

[0030] Step 17: Trigger the photo-taking process again at the next time node.

[0031] In a further embodiment, in step 2, the data processing device processes the input image using a Log-Gabor filter to extract the entropy, gradient components and gradient values ​​of its response as SoS features in the frequency domain; extracts weighted local binary pattern histogram features from the MLV map of the multi-resolution input image as SoS features in the time domain; finally, in order to map the extracted features into an objective prediction score, a model is established using SVR for image quality evaluation.

[0032] In a further embodiment, in step 4, the sample analysis platform receives the image filtered by the data processing device, and first converts the color image into a grayscale image for processing, then performs image denoising, performs wavelet transform on the signal, then performs threshold shrinkage on the wavelet coefficients, and finally performs inverse wavelet transform, then uses an expansion algorithm to enhance the denoised image, then performs segmentation and calculates the number and type of phytoplankton, and finally saves the image and uploads it for display.

[0033] Beneficial effects: The present invention is designed to realize real-time continuous photography of phytoplankton (algae) in the field. It can automatically focus and take photos in three axes (X, Y, Z) for fast imaging, automatically select high-definition pictures for uploading, and provide effective and reliable basis for algae analysis data. Therefore, the present invention can realize unmanned continuous detection and analysis of water samples in real time, automatically focus and take photos quickly, automatically select pictures that meet the requirements, and provide effective and reliable image basis for the background algorithm software, meeting the needs of accurate analysis of algae on the water surface in the wild, and then automatically uploading to the server or cloud platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 Schematic diagram of the imaging system of the present invention.

[0035] Figure 2 It is a flow chart of the imaging method of the present invention.

[0036] Figure 3 It is a flow chart of the working method of the image acquisition device of the present invention.

[0037] Figure 4 It is a flow chart of the working method of the data processing device of the present invention.

[0038] Figure 5 It is a flow chart of the working method of the sample analysis platform of the present invention. DETAILED DESCRIPTION

[0039] 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.

[0040] This application aims to improve the problems of the prior art. The traditional phytoplankton field monitoring equipment needs to be improved. The structure and installation of this application are as follows;

[0041] 1. Common high-definition imaging equipment includes:

[0042] 1. High-definition camera: widely used in film and television production, news reporting, documentary shooting and other fields, and can capture high-quality and clear images.

[0043] 2. High-definition digital camera: used for photography creation, commercial shooting and daily photography, with high pixels and excellent image quality.

[0044] 3. High-definition microscope: It plays an important role in scientific research, medicine, biology and other fields, and can clearly present the details of tiny objects.

[0045] 4. High-definition endoscope: commonly used in medical examinations, such as observation and diagnosis of the gastrointestinal tract, respiratory tract, etc.

[0046] 5. High-definition satellite remote sensing equipment: used for earth observation, resource exploration, environmental monitoring, etc., and can obtain high-definition surface images over a wide range.

[0047] 6. High-definition industrial cameras: used in industrial inspection, quality control and other links to perform high-precision imaging and inspection of products.

[0048] Second, the following methods can be considered for installing high-definition imaging equipment in wild waters:

[0049] 1. Buoy installation: Use a special buoy device to fix the imaging device on the buoy. The buoy should have good stability and buoyancy to ensure that the device remains balanced and stable on the water surface.

[0050] 2. Shore column installation: Install a solid column on the shore near the water, and extend the imaging equipment to a suitable position above the water surface through a bracket. This method requires ensuring that the column is stable and corrosion-resistant.

[0051] 3. Underwater mount: If underwater imaging is required, a specially designed underwater mount can be used to fix the device at or near the bottom of the water. However, attention should be paid to waterproofing and anti-corrosion treatment.

[0052] 4. Drone: Use a drone with high-definition imaging capabilities to fly over the water area for filming when necessary. However, this method is limited by battery life and flight conditions.

[0053] 3. When implementing the installation, the following factors also need to be considered:

[0054] 1. Waterproof and moisture-proof: Make sure the imaging device has sufficient waterproof performance to cope with humid environments on water and possible water splashes.

[0055] 2. Power supply issue: Choose a suitable power supply method, such as solar panels, battery packs, or connecting to a nearby power source via cables.

[0056] 3. Data transmission: Determine how to transmit the captured high-definition image data to the receiving end stably and quickly, such as through wireless transmission or wired connection.

[0057] 4. Protective measures: Consider installing protective devices to prevent the equipment from being hit and damaged by wild animals, floating objects, etc.

[0058] Here are some more specific suggestions and points to consider when looking for phytoplankton species:

[0059] 1) Installation method:

[0060] 1. Three-axis adjustment bracket combined with waterproof transparent cover: Use a special three-axis adjustment bracket to install the high-definition imaging device at a specific depth underwater. In order to protect the device and provide a clear view, a waterproof transparent cover can be installed on the front of the device to reduce the interference of water flow and impurities on imaging.

[0061] 2. Remotely operated underwater vehicles (ROVs): If budget permits, use remotely operated underwater vehicles (ROVs) to carry high-definition imaging equipment. ROVs can provide more flexibility in observing at different depths and locations.

[0062] 2) Equipment selection:

[0063] 1. High-resolution microscope lens: helps to clearly distinguish the fine structure of phytoplankton.

[0064] 2. Equipment with good low-light performance: Because the light underwater is usually dim, good low-light performance can ensure clear images in low-light environments.

[0065] Auxiliary facilities:

[0066] 1. Lighting system: Install appropriate lighting equipment to ensure sufficient and uniform light for imaging to highlight the characteristics of phytoplankton.

[0067] 2. Sample collection device: In conjunction with the imaging equipment, a simple sample collection device is set up to facilitate more detailed analysis in the laboratory later.

[0068] 3) Data processing:

[0069] 1. Real-time image analysis software: It can initially analyze and classify the captured images on site, helping to quickly screen valuable information.

[0070] 2. Large-capacity storage devices: used to store large amounts of high-definition image data for subsequent in-depth research.

[0071] 4) Environmental adaptation:

[0072] 1. Anti-corrosion treatment: Since the underwater environment is corrosive, anti-corrosion treatment is performed on the equipment and installation components to extend their service life.

[0073] 2. Water flow and water pressure response: Choose equipment and installation methods that can withstand a certain amount of water flow impact and water pressure.

[0074] The structure and working method of this application are described in detail below:

[0075] Example 1

[0076] An imaging system for field monitoring of phytoplankton, such as Figure 1 As shown, including:

[0077] Water sampling equipment, used to collect water from the area

[0078] Image acquisition equipment, using high-definition imaging equipment to collect images of water samples, and using a three-axis adjustment bracket to control the position of the high-definition imaging equipment to achieve multi-point focus photography;

[0079] A data processing device, which is installed in the image acquisition device and receives images acquired by the image acquisition device and analyzes and classifies the captured images;

[0080] A sample analysis platform receives the image data processed by the data processing device, processes and analyzes the image data, and outputs a report;

[0081] A communication device, which establishes a specific transmission channel; used for communication between the data processing device and the sample analysis platform;

[0082] The auxiliary device is installed on the image acquisition device and cooperates with the image acquisition device to acquire images.

[0083] In one embodiment, the three-axis adjustment bracket in the image acquisition device adopts an automatic three-axis driving bracket, which is used for rapid movement of the X-axis, Y-axis and Z-axis of the high-definition imaging device, thereby achieving fast focusing and photographing.

[0084] In one embodiment, the image acquisition device performs three rounds of photographing each time. After each point is focused on three axes, the first round of photographing is performed at different heights. A threshold is calculated for the imaged pictures. The focus height that meets the threshold or the highest threshold is taken from the first round of imaged pictures, and the second round of photographing is performed at the point where the focus height meets the threshold or the highest threshold is taken from the second round of imaged pictures. The third round of photographing is performed at the point where the focus height meets the threshold or the highest threshold is taken from the second round of imaged pictures.

[0085] In one embodiment, the data processing device is used to preliminarily analyze and classify the captured images on site to help quickly screen out valuable ones; the data processing device evaluates the quality of the captured images by constructing an image quality assessment model, and eliminates images that do not meet the quality requirements.

[0086] In one embodiment, the sample analysis platform acquires the collected image on the slide through a microscope at a constant magnification, and performs statistics on the number and type of phytoplankton in the collected image. At the same time, the collected image and analysis structure are stored in a computer for later processing and statistical use, and various platforms are externally connected to upload and share data.

[0087] In one embodiment, the auxiliary equipment includes an auxiliary lighting device, which ensures that sufficient and uniform light is provided for imaging to highlight the characteristics of phytoplankton by installing appropriate lighting equipment on the image acquisition device.

[0088] The present invention utilizes a self-developed scanning platform to automatically focus on designated water samples, automatically take photos and images, and automatically select high-definition pictures for uploading, providing effective and reliable image basis for the background algorithm software, meeting the needs of accurate algae analysis on wild water surfaces.

[0089] Embodiment 2:

[0090] An imaging method for field monitoring of phytoplankton is implemented by the imaging system for field monitoring of phytoplankton described above, such as Figure 2 As shown, the imaging method comprises:

[0091] Step 1: Take three rounds of photos through the image acquisition device at regular intervals or when the user initiates the process. The three rounds of photos are adjusted by the automatic three-axis driving bracket to adjust the positions of the X-axis, Y-axis and Z-axis. The focus height that meets the threshold or the highest threshold is selected to take photos.

[0092] Step 2: The data processing device analyzes, classifies and quickly screens the images collected by the image acquisition device, and transmits the images that meet the requirements through the communication device;

[0093] Step 3: The communication device establishes an independent transmission channel according to the work requirements and communicates with the sample analysis platform to transmit images;

[0094] Step 4: The sample analysis platform collects, counts, and identifies the types of phytoplankton in the phytoplankton image and stores and displays the collected images.

[0095] In one embodiment, Figure 3 As shown, in step 1, the working method of the image acquisition device is as follows:

[0096] Step 11: After the photo process is triggered, the image acquisition device controls the objective lens to perform multi-point focusing on the water sample;

[0097] Step 12: After three-axis focusing of each point, take the first round of photos at different heights, and calculate the imaged pictures to obtain a threshold;

[0098] Step 13: Take the focus height that meets the threshold or the highest threshold in the first round of imaging pictures, and take the second round of pictures at the point where the focus height meets the threshold or the highest threshold;

[0099] Step 14: Take the focus height that meets the threshold or the highest threshold in the second round of imaging pictures, and take the third round of photos at the point where it is located;

[0100] Step 15: After taking three rounds of photos at each point, select a few high-definition photos with the highest threshold and upload them;

[0101] Step 16: After all images are uploaded, the process ends and the system goes into sleep mode to save power.

[0102] Step 17: Trigger the photo-taking process again at the next time node.

[0103] In one embodiment, in step 2, Figure 4 As shown, the data processing device processes the input image using a Log-Gabor filter to extract the entropy, gradient component and gradient value of its response as SoS features in the frequency domain; extracts weighted local binary pattern histogram features from the MLV map of the multi-resolution input image as SoS features in the time domain; finally, in order to map the extracted features into an objective prediction score, a model is established using SVR for image quality evaluation.

[0104] Specifically, the present invention uses a Log-Gabor filter to extract the statistical characteristics of clarity perception in the frequency domain as the first layer of structural information. The Log-Gabor filter can better reflect the frequency response of textures in natural images; the definition of the Log-Gabor filter is:

[0105]

[0106] Among them G s,o represents Log-Gabor filters with different wavelet scales S and different directions o. The variable ω represents the normalized radial frequency, and θ represents the direction. The gradient-related information of each Log-Gabor response is extracted as the SoS feature for clarity evaluation. First, each response image is gradient processed, and its gradient map is calculated by the following formula:

[0107] D i,1 =[-1 0 1]*G s,o

[0108] D i,2 =[-1 0 1] T *G s,o

[0109] D i =D i,1 +D i,2

[0110] Where i = 0, 1, ..., 12, representing response images of different scales and directions. [-1 0 1] and [-1 0 1] T are the horizontal and vertical gradient operators, G s,orepresents the response image of the sth scale and oth direction. i,1 and D i,2 Represents the horizontal and vertical gradient response images, D i is the gradient value image; for each subband, the horizontal and vertical gradient images D i,1 and D i,2 The generalized Gaussian distribution (GGD) model is used for fitting; finally, the entropy of the Log-Gabor response is extracted as a feature of image quality perception.

[0111] Specifically, the original image is downsampled to different resolutions as the first layer of structural features in the time domain. Specifically, three low-resolution images are obtained by performing bicubic interpolation of 2×, 4×, and 8× in the horizontal and vertical directions on the original image. The reason for choosing bicubic interpolation is that it can produce high-fidelity images. In this way, the downsampling operation does not introduce additional distortion.

[0112] More specifically, LBP is used to perceive image quality and establish a quality evaluation model. LBP represents images of local structure and text information by converting the binary relationship between the center pixel and the binary sequence of neighboring pixels to describe the spatial relationship. Since the structure and texture of the image are distorted when the image quality decreases, LBP can be used to perceive the image quality; support vector regression is used to obtain the quality evaluation model, and finally the trained SVR model is used to predict the clarity score of the image to be tested.

[0113] In one embodiment, in step 4, Figure 5 As shown, the sample analysis platform receives the image filtered by the data processing device, and first converts the color image into a grayscale image for processing, then performs image denoising, performs wavelet transform on the signal, then performs threshold contraction on the wavelet coefficient, and finally performs inverse wavelet transform, then uses the expansion algorithm to enhance the denoised image, then performs segmentation and calculates the number and type of phytoplankton, and finally saves and uploads the image for display. .

[0114] Specifically, the image obtained by the microscope and CCD digital camera is a color image. Since there are problems such as large amount of calculation and slow software running speed when processing color images, the color image is converted into a grayscale image for processing in the present invention. The present invention adopts the HLS (hue, brightness, saturation) model conversion algorithm to convert the color image into a grayscale image;

[0115] Then the image is denoised and enhanced. The noise caused by uneven illumination due to various suspended "particles" in the water environment and the resulting reflection directly affects the statistics and identification of the number of phytoplankton. In addition, the grayscale histogram of the microscopic phytoplankton image is a narrow single peak, indicating that the contrast between the background and foreground images of such images is not strong, which is not conducive to image segmentation. In order to ensure the correct statistics of the number of phytoplankton, the collected images must be denoised and enhanced; first, the image is denoised using the wavelet shrinkage method proposed by Donoho. The basic principle of this method is: first, the signal is subjected to wavelet transform, then the wavelet coefficients are threshold-shrinked, and finally the wavelet inverse transform is performed. The present invention uses a hard threshold method to shrink the wavelet coefficients to achieve the purpose of denoising; secondly, in view of the fact that the contrast between the background and foreground of the microscopic phytoplankton image is not strong, and some phytoplankton have the characteristics of translucency, the present invention applies the morphological expansion algorithm to the enhancement of the phytoplankton grayscale image, and achieves better results; the expansion algorithm is as follows:

[0116] Let A be the input image, B be the structural element, A and B are E = z 2 The set in which A is expanded by B is defined as:

[0117]

[0118] Dilation is to translate the structural element B by z to obtain B z , so that B z The set of all points z whose intersection with A is not an empty set. Dilation is a process of expansion, which makes the target limb expand and fill the small holes. This fam structure element adopts a diamond structure with a size of 8, and uses a dilation algorithm to enhance the denoised image.

[0119] The purpose of phytoplankton image segmentation is to separate the area occupied by individual plants in the image from the background, so as to facilitate the statistics of the number of phytoplankton. Image segmentation uses a method combining the maximum variance threshold and morphological opening operation to complete the segmentation of phytoplankton images. The principle of the maximum inter-class variance algorithm is to take a certain gray value as the initial threshold and divide the image into two areas. When the gray variance between the two areas is the largest, the threshold is the image segmentation threshold.

[0120] The segmented image is further refined using the opening algorithm. There may be small impurity interference blocks in the segmented image, which will cause statistical errors in the number of microalgae. Since the opening operation in morphology has a good effect in separating adhesion targets and removing background noise (especially salt and pepper noise), the opening operation is used to refine the segmented image and remove small blocks of noise. The opening operation is actually an operation that uses the same structural element to corrode the image first and then expand it;

[0121] Let A be the input image, B be the structural element, A and B are E = z 2 The set in the table uses B to perform corrosion operation on A, which is defined as:

[0122]

[0123] Erosion is to translate the structural element z to obtain B z , so that B z The set of all points z contained in A. Erosion is a process of contraction transformation. The result of erosion is a subset of A. Using B to open A, according to the definition of opening operation

[0124]

[0125] Finally, the phytoplankton in the image is counted and identified.

[0126] Obviously, the above embodiments are merely examples for the purpose of clear explanation, and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived therefrom are still within the scope of protection of the invention.

Claims

1. An imaging system for field monitoring of phytoplankton, characterized in that: The imaging system comprises: Water sampling equipment, used to collect water from the area Image acquisition equipment, using high-definition imaging equipment to collect images of water samples, and using a three-axis adjustment bracket to control the position of the high-definition imaging equipment to achieve multi-point focus photography; A data processing device, which is installed in the image acquisition device and receives images acquired by the image acquisition device and analyzes and classifies the captured images; A sample analysis platform receives the image data processed by the data processing device, processes and analyzes the image data, and outputs a report; A communication device, which establishes a specific transmission channel; used for communication between the data processing device and the sample analysis platform; The auxiliary device is installed on the image acquisition device and cooperates with the image acquisition device to acquire images.

2. An imaging system for field monitoring of phytoplankton according to claim 1, characterized in that: The three-axis adjustment bracket in the image acquisition device adopts an automatic three-axis driving bracket, which is used for rapid movement of the X-axis, Y-axis and Z-axis of the high-definition imaging device, thereby realizing fast focusing and photographing.

3. An imaging system for field monitoring of phytoplankton according to claim 2, characterized in that: The image acquisition device performs three rounds of photographing each time. After each point is focused on three axes, the first round of photographing is performed at different heights. A threshold is calculated for the imaged pictures. The focus height that meets the threshold or the highest threshold is taken from the first round of imaged pictures, and the second round of photographing is performed at the point where the focus height is located. The third round of photographing is performed at the point where the focus height meets the threshold or the highest threshold is taken from the second round of imaged pictures.

4. The imaging system for field monitoring of phytoplankton according to claim 1, characterized in that: The data processing device is used to preliminarily analyze and classify the captured images on site to help quickly screen out valuable ones; the data processing device evaluates the quality of the captured images by constructing an image quality assessment model and eliminates images that do not meet the quality requirements.

5. The imaging system for field monitoring of phytoplankton according to claim 1, characterized in that: The sample analysis platform obtains the collected images on the slide at a constant magnification through a microscope, and performs statistics on the number and type discrimination of phytoplankton in the collected images. At the same time, the collected images and analysis structures are stored in a computer for later processing and statistical use, and various platforms are externally connected to upload and share data.

6. The imaging system for field monitoring of phytoplankton according to claim 1, characterized in that: The auxiliary equipment includes an auxiliary lighting device, which ensures that sufficient and uniform light is provided for imaging by installing appropriate lighting equipment on the image acquisition device to highlight the characteristics of phytoplankton.

7. An imaging method for field monitoring of phytoplankton, characterized in that: The imaging system for field monitoring of phytoplankton according to any one of claims 1 to 6 is implemented, and the imaging method comprises: Step 1: Take three rounds of photos through the image acquisition device at regular intervals or when the user initiates the process. The three rounds of photos are adjusted by the automatic three-axis driving bracket to adjust the positions of the X-axis, Y-axis and Z-axis. The focus height that meets the threshold or the highest threshold is selected to take photos. Step 2: The data processing device analyzes, classifies and quickly screens the images collected by the image acquisition device, and transmits the images that meet the requirements through the communication device; Step 3: The communication device establishes an independent transmission channel according to the work requirements and communicates with the sample analysis platform to transmit images; Step 4: The sample analysis platform collects, counts, and identifies the types of phytoplankton in the phytoplankton image and stores and displays the collected images.

8. An imaging method for field monitoring of phytoplankton according to claim 7, characterized in that: In step 1, the image acquisition device works as follows: Step 11: After the photo process is triggered, the image acquisition device controls the objective lens to perform multi-point focusing on the water sample; Step 12: After three-axis focusing of each point, take the first round of photos at different heights, and calculate the imaged pictures to obtain a threshold; Step 13: Take the focus height that meets the threshold or the highest threshold in the first round of imaging pictures, and take the second round of pictures at the point where the focus height meets the threshold or the highest threshold; Step 14: Take the focus height that meets the threshold or the highest threshold in the second round of imaging pictures, and take the third round of photos at the point where it is located; Step 15: After taking three rounds of photos at each point, select a few high-definition photos with the highest threshold and upload them; Step 16: After all images are uploaded, the process ends and the system goes into sleep mode to save power. Step 17: Trigger the photo-taking process again at the next time node.

9. The imaging method for field monitoring of phytoplankton according to claim 7, characterized in that: In step 2, the data processing device processes the input image using a Log-Gabor filter to extract the entropy, gradient components and gradient values ​​of its response as SoS features in the frequency domain; extracts weighted local binary pattern histogram features from the MLV map of the multi-resolution input image as SoS features in the time domain; finally, in order to map the extracted features into an objective prediction score, a model is established using SVR for image quality evaluation.

10. The imaging method for field monitoring of phytoplankton according to claim 7, characterized in that: In step 4, the sample analysis platform receives the image filtered by the data processing device, and first converts the color image into a grayscale image for processing, then performs image denoising, performs wavelet transform on the signal, then performs threshold shrinkage on the wavelet coefficients, and finally performs inverse wavelet transform. The denoised image is then enhanced by the expansion algorithm, and then segmented and calculated to determine the number and type of phytoplankton, and finally the image is saved and uploaded for display.

Images