An automatic monitoring device for phytoplankton in water and a method for monitoring phytoplankton in water

Abstract

The application relates to the technical field of water ecological environment monitoring, and discloses an automatic monitoring device for planktonic algae in water and a monitoring method for planktonic algae in water. The automatic monitoring device for planktonic algae in water comprises a water sample preparation module, a water sample slice preparation module and a water sample detection module. The water sample preparation module is used for collecting a water sample and obtaining an algae concentrate of the water sample. The water sample slice preparation module is used for preparing a slice from the algae concentrate and transferring the slice to the water sample detection module. The water sample detection module is used for taking a picture of the planktonic algae in the slice and identifying and counting the algae in the picture by using a preset image visual algorithm. The device realizes automatic sampling, automatic precipitation collection, automatic slice preparation, automatic picture taking, automatic identification and counting of the planktonic algae, does not require manual attendance, saves labor cost, effectively avoids differences in identification results caused by differences in identification abilities of different laboratories and detection personnel, improves the standardization degree of algae detection, and facilitates popularization and application.

Description

Technical Field

[0001] This application relates to the field of aquatic ecological environment monitoring technology, and in particular to an automatic monitoring device and method for phytoplankton in water. Background Technology

[0002] Phytoplankton in water bodies are diverse and abundant, and their species, density, and community composition are closely related to water quality. Classification, measurement, and counting of algae in water bodies play a crucial role in analyzing the causes and mechanisms of algal blooms, as well as in the monitoring, early warning, and management of eutrophic lakes and reservoirs. It is an important foundation for conducting health diagnosis of aquatic ecosystems, water environment management, and protection.

[0003] Currently, phytoplankton detection mainly relies on professional technicians identifying and counting algal cell morphology under a microscope. This method involves manual sampling, sample preparation, slide preparation, and microscopic observation for classification and counting. It is time-consuming, labor-intensive, and heavily dependent on the expertise and experience of the personnel. Differences in the identification capabilities of different laboratories and personnel can lead to variations in results, resulting in incomparability of phytoplankton detection results from different sources and limiting the standardized promotion and widespread application of phytoplankton detection.

[0004] Therefore, how to save labor costs and improve the standardization of algae detection is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] The purpose of this application is to provide an automatic monitoring device and method for phytoplankton in water, which can save labor costs and improve the standardization of algae detection.

[0006] To address the aforementioned technical problems, this application provides an automatic monitoring device for phytoplankton in water, comprising a water sample preparation module, a water sample preparation module, and a water sample detection module;

[0007] The water sample preparation module is used to collect water samples and obtain algae concentrate from the water samples;

[0008] The water sample preparation module is used to prepare the algae concentrate into a sample and transfer the sample to the water sample detection module;

[0009] The water sample detection module is used to photograph the planktonic algae in the sample and to identify and count the algae in the photographed image using a preset image vision algorithm.

[0010] Optionally, the water sample preparation module includes: a metering pump, a multi-way valve, a sample sedimentation treatment channel, and a magnetic stirring device;

[0011] The metering pump is connected to the multi-way valve and the sample precipitation treatment channel respectively. The metering pump is used to draw the water sample and algae fixative into the sample precipitation treatment channel through the multi-way valve.

[0012] The sample sedimentation treatment channel is equipped with a waste liquid outlet and a supernatant outlet. The sample sedimentation treatment channel is used to fix and precipitate algae in the water sample and discharge the supernatant from the supernatant outlet to obtain the algae concentrate.

[0013] The magnetic stirring device is located at the bottom of the sample precipitation treatment channel and is used to mix the algae concentrate at the bottom of the sample precipitation treatment channel evenly.

[0014] Optionally, the water sample preparation module includes: a microfluidic chip, a chip storage device, a chip sample introduction device, and a chip feeding device;

[0015] The microfluidic chip is provided with an inlet port, a sample chamber, and an exhaust port;

[0016] The chip storage device includes a chip storage frame, a first sensor, and a feeding component. The chip storage frame is used to store the microfluidic chip, and the feeding component is used to transfer the microfluidic chip in the storage frame to the chip injection device. The first sensor is used to detect the number of chips and generate an alarm when the number of chips is less than a lower limit.

[0017] The chip injection device includes a peristaltic pump, an injection needle lifting motor, and a floating injection needle. The peristaltic pump is connected to the sample precipitation treatment channel and the floating injection needle. The peristaltic pump is used to draw the algae concentrate into the floating injection needle. The injection needle lifting motor is used to control the movement of the floating injection needle. The floating injection needle is used to inject the algae concentrate into the sample chamber through the injection port of the microfluidic chip.

[0018] The chip feeding device includes a chip lifting motor, a rotary motor, and a rotary arm. The chip lifting motor and the rotary motor are respectively connected to the rotary arm. The chip lifting motor and the rotary motor are used to control the rotary arm to transfer the microfluidic chip with added algae concentrate to the water sample detection module.

[0019] Optionally, the water sample detection module includes: a light source, a microscope lens, a right-angle imaging device, a CCD camera, an industrial control computer, an x-axis platform, an x-axis motion motor, a y-axis platform, a y-axis motion motor, a z-axis platform, and a z-axis motion motor;

[0020] The microscope lens is used to magnify planktonic algae;

[0021] The right-angle imaging device and the CCD camera are used for capturing images of phytoplankton;

[0022] The industrial control computer is connected to the x-axis motion motor, the y-axis motion motor, the z-axis motion motor and the CCD camera respectively. The industrial control computer is used to acquire images captured by the CCD camera and to identify and count algae in the images using a preset image vision algorithm. The industrial control computer is also used to drive the z-axis motion motor to move the z-axis platform for microscopic focusing using an autofocus algorithm.

[0023] The x-axis motion motor is used to drive the x-axis platform, the y-axis motion motor is used to drive the y-axis platform, and the x-axis platform and the y-axis platform are used for positioning the chip and switching the field of view during shooting.

[0024] Optionally, it also includes a chip collection module, which includes a collection frame and a second sensor. The second sensor is used to detect whether the collection frame is full of microfluidic chips and to generate an alarm when the number of chips in the collection frame exceeds the upper limit.

[0025] This application also provides a method for monitoring phytoplankton in water, applied to the aforementioned phytoplankton monitoring device, comprising:

[0026] Water samples were collected, and an algae concentrate was obtained from the water samples using an algae fixative.

[0027] Prepare sample pieces from the algae concentrate;

[0028] Images of planktonic algae in the sample are acquired, and algae are identified and counted using a preset image vision algorithm.

[0029] Optionally, the step of preparing the algae concentrate into sample tablets includes:

[0030] The algae concentrate is injected into the microfluidic chip by controlling a floating injection needle to obtain the sample.

[0031] Optionally, acquiring the image of phytoplankton in the sample includes:

[0032] The z-axis motor is controlled to drive the z-axis platform on which the sample is placed to return to a preset point in the direction away from the microscope head. The z-axis motor is then controlled to move the z-axis platform in the direction closer to the microscope head to a specified position according to the set number of steps and the set speed to complete the coarse focus adjustment.

[0033] After coarse focusing is completed, the z-axis motor is controlled to move the z-axis platform toward the microscope head according to the set number of movement steps. After the z-axis platform moves, the CCD camera is controlled to take pictures of the sample and the shooting position of each picture is recorded.

[0034] Each captured image is converted to grayscale, and the grayscale variance between two adjacent pixels in each image is calculated.

[0035] The target image with a grayscale variance greater than a threshold is identified, and the z-axis motor is controlled to move the z-axis platform to the target shooting position corresponding to the target image.

[0036] The image of phytoplankton in the sample taken by the CCD camera at the target shooting position is acquired.

[0037] Optionally, the step of using a preset image vision algorithm to identify and count algae in the image includes:

[0038] The image is input into a preset phytoplankton identification algorithm model, and the phytoplankton identification algorithm model is used to identify the species of algae in the image;

[0039] For non-aggregated single-cell individuals in the image, count them based on the number of individual cells;

[0040] For the aggregated cells and clumps of phytoplankton in the image, algal cell segmentation and counting are performed according to algal species.

[0041] Optionally, before inputting the image into the preset phytoplankton identification algorithm model, the method further includes:

[0042] First images of algae of known species at various growth stages are obtained, and the first images are filtered according to preset requirements to obtain target images;

[0043] The target images are bounding boxes and their species names are noted to obtain a sample database;

[0044] A second image of impurities in the water body is obtained, and the second image is labeled with a rectangular box and the impurity name is noted to obtain an impurity database;

[0045] Import the sample database and the impurity database into the model training platform, establish a target detection model training task, and select the sample database and the impurity database for training to build a basic algorithm model;

[0046] The image database is optimized, and the basic algorithm model is incrementally trained using the optimized sample database and the impurity database. The model whose recall and accuracy meet the preset standards is used as the phytoplankton identification algorithm model.

[0047] This application provides an automatic monitoring device for phytoplankton in water, comprising: a water sample preparation module, a water sample slide preparation module, and a water sample detection module. The water sample preparation module is used to collect water samples and obtain algae concentrate from the water samples. The water sample slide preparation module is used to prepare sample slides from the algae concentrate and transfer the sample slides to the water sample detection module. The water sample detection module is used to photograph the phytoplankton in the sample slides and to identify and count the algae in the photographed images using a preset image vision algorithm. This device achieves automatic sampling, automatic sedimentation and collection, automatic slide preparation, automatic photographing, and automatic identification and counting of phytoplankton, eliminating the need for manual operation, saving labor costs, effectively avoiding differences in identification results caused by variations in the identification capabilities of different laboratories and testing personnel, improving the standardization of algae detection, and facilitating widespread application.

[0048] The beneficial effects of the method for monitoring phytoplankton in water provided in this application correspond to those of the automatic monitoring device for phytoplankton in water, as described above. Attached Figure Description

[0049] To more clearly illustrate the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 A structural diagram of an automatic monitoring device for phytoplankton in water provided in an embodiment of this application;

[0051] Figure 2 A structural diagram of a water sample preparation module provided in an embodiment of this application;

[0052] Figure 3 A structural diagram of a water sample preparation module and a water sample slide preparation module provided in the embodiments of this application;

[0053] Figure 4 A structural diagram of a water sample detection module provided in an embodiment of this application;

[0054] Figure 5 This is a structural diagram of a chip collection module provided in an embodiment of this application;

[0055] Figure 6 A flowchart illustrating a method for monitoring phytoplankton in water, provided in an embodiment of this application;

[0056] The attached diagram is labeled as follows: 1 is the water sample preparation module, 2 is the water sample preparation module, 3 is the water sample detection module, 4 is the chip collection module, 11 is the quantitative pump, 12 is the multi-way valve, 13 is the sample precipitation treatment channel, 14 is the magnetic stirring device, 15 is the waste liquid discharge port, 16 is the supernatant discharge port, 17 is the first pipe, 18 is the second pipe, 19 is the third pipe, 110 is the fourth pipe, 21 is the chip storage frame, 22 is the first sensor, 23 is the feeding component, and 24 is the peristaltic... 25 is a pump, 26 is a floating injection needle, 27 is a chip lifting motor, 28 is a rotary motor, 29 is a rotating arm, 31 is a CCD camera, 32 is an industrial computer, 33 is a detection module mounting plate, 34 is an x-axis platform and x-axis motion motor, 35 is a y-axis platform and y-axis motion motor, 36 is a light source, 37 is a z-axis platform and z-axis motion motor, 38 is a microscope lens, 39 is a right-angle imaging device, 41 is a collection frame, and 42 is a second sensor. Detailed Implementation

[0057] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of this application.

[0058] The core of this application is to provide an automatic monitoring device and a method for monitoring phytoplankton in water.

[0059] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0060] Figure 1 A structural diagram of an automatic monitoring device for phytoplankton in water provided in this application embodiment is shown below. Figure 1 As shown, the automatic monitoring device for phytoplankton in water includes a water sample preparation module 1, a water sample preparation module 2, and a water sample detection module 3. The water sample preparation module 1 is used to collect water samples and obtain algae concentrate from the water samples. The water sample preparation module 2 is used to prepare algae concentrate into sample sheets and transfer the sample sheets to the water sample detection module 3. The water sample detection module 3 is used to photograph the phytoplankton in the sample sheets and use a preset image vision algorithm to identify and count algae in the photographed images.

[0061] Figure 2 A structural diagram of a water sample preparation module provided in an embodiment of this application is shown below. Figure 2As shown, the water sample preparation module includes a metering pump 11, a multi-way valve 12, a sample precipitation treatment channel 13, and a magnetic stirring device 14. The metering pump 11 is connected to the multi-way valve 12 and the sample precipitation treatment channel 13. The metering pump 11 is used to pump the water sample and algae fixative into the sample precipitation treatment channel 13 through the multi-way valve 12. The sample precipitation treatment channel 13 is provided with a waste liquid outlet 15 and a supernatant outlet 16. The sample precipitation treatment channel 13 is used to fix and precipitate the algae in the water sample and discharge the supernatant from the supernatant outlet 16 to obtain an algae concentrate. The magnetic stirring device 14 is located at the bottom of the sample precipitation treatment channel 13 and is used to mix the algae concentrate at the bottom of the sample precipitation treatment channel 13 evenly. The multi-way valve 12 connects to the first pipe 17, the second pipe 18, the third pipe 19, and the fourth pipe 110 respectively. Water samples enter the sample sedimentation treatment channel 13 through the first pipe 17, and algae fixatives enter the sample sedimentation treatment channel 13 through the second pipe 18. In the sample sedimentation treatment channel 13, the algae in the water sample are fixed and precipitated by the algae fixatives, and the supernatant is discharged from the supernatant drain 16. The algae settle to the bottom of the sample sedimentation treatment channel 13, where a magnetic stirrer 14 at the bottom of the sample sedimentation treatment channel 13 evenly mixes the algae concentrate. Multiple sample sedimentation treatment channels can be set up to facilitate multiple sampling and testing. Sodium hypochlorite solution flows into the sample sedimentation treatment channel 13 through the third pipe 19, and pure water enters the sample sedimentation treatment channel 13 through the fourth pipe 110. Sodium hypochlorite solution is used to disinfect the pipes and channels, and pure water is used to clean the pipes and channels.

[0062] Figure 3 A structural diagram of a water sample preparation module and a water sample slide preparation module provided in the embodiments of this application is shown below. Figure 3As shown, the water sample preparation module includes: a microfluidic chip, a chip storage device, a chip injection device, and a chip feeding device; the microfluidic chip has an injection port, a sample chamber, and an exhaust port; the chip storage device includes a chip storage frame 21, a first sensor 22, and a feeding assembly 23. The chip storage frame 21 is used to store the microfluidic chip, and the feeding assembly 23 is used to transfer the microfluidic chip in the storage frame to the chip injection device. The first sensor 22 is used to detect the number of chips and generate an alarm when the number of chips is less than a lower limit; the chip injection device includes a peristaltic pump 24, an injection needle lifting motor 25, and a floating injection needle 26. The sample precipitation processing channel 13 and the floating injection needle 26 are connected respectively. A peristaltic pump 24 is used to draw algae concentrate into the floating injection needle 26, and an injection needle lifting motor 25 is used to control the movement of the floating injection needle 26. The floating injection needle 26 is used to inject the algae concentrate into the sample chamber through the sample inlet of the microfluidic chip. The chip feeding device includes a chip lifting motor 27, a rotary motor 28, and a rotating arm 29. The chip lifting motor 27 and the rotary motor 28 are respectively connected to the rotating arm 29, and are used to control the rotating arm 29 to transfer the microfluidic chip with added algae concentrate to the water sample detection module 3. This application uses an integrated microfluidic chip to replace the planktonic algae counting frame and coverslip. The microfluidic chip includes a sample inlet, a sample chamber, and an vent. The length and width of the sample chamber are both not less than 20 mm, and the liquid thickness does not exceed 0.2 mm. The shape of the sample chamber is narrow at both ends and wide in the middle, which facilitates the sample filling the entire chamber. An injection needle is used to inject excess sample into the sample chamber to prevent air bubbles from forming. Because algae readily adhere to the sample chamber and are difficult to clean, the chip is disposable and requires no cleaning, thus avoiding cross-contamination of the water sample. The chip storage device consists of a chip storage frame 21, a first sensor 22, and a feeding assembly 23. The feeding assembly 23 includes a motor and a robotic arm. The chip is placed in the chip storage frame 21, and the robotic arm is responsible for delivering the chip from the chip storage frame 21 to the sample introduction device. The first sensor 22 is an optocoupler that can detect the number of chips, and an alarm is triggered when the number of chips is less than a lower limit. The sample introduction device consists of a peristaltic pump 24, an injection needle lifting motor 25, and a floating injection needle 26. The precision peristaltic pump extracts algae concentrate, which is then injected into the microfluidic chip via the injection needle lifting motor 25 and the floating injection needle 26. The chip feeding device consists of a chip lifting motor 27, a rotary motor 28, and a rotating arm 29, responsible for transferring the sampled microfluidic chip to the water sample detection module 3, or transferring the detected microfluidic chip to the chip collection module 4 (described below).

[0063] Figure 4 This is a structural diagram of a water sample detection module provided in an embodiment of this application, as shown below. Figure 4As shown, the water sample detection module 3 includes a light source 36, a microscope lens 38, a right-angle imaging device 39, a CCD camera 31, an industrial control computer 32, a detection module fixing plate 33, an x-axis platform and x-axis motion motor 34, a y-axis platform and y-axis motion motor 35, and a z-axis platform and z-axis motion motor 37. The microscope lens 38 is used to magnify the phytoplankton. The right-angle imaging device 39 and the CCD camera 31 are used to capture images of the phytoplankton. The industrial control computer 32 is connected to the x-axis motion motor, y-axis motion motor, z-axis motion motor, and CCD camera 31. The industrial control computer 32 is used to acquire images captured by the CCD camera 41 and to identify and count the algae in the images using a preset image vision algorithm. The industrial control computer 32 is also used to drive the z-axis motion motor to move the z-axis platform for microscopic focusing using an autofocus algorithm. The x-axis motion motor is used to drive the x-axis platform, and the y-axis motion motor is used to drive the y-axis platform. The x-axis platform and y-axis platform are used for chip positioning and field of view switching during the shooting process. This application employs a right-angle imaging device 39 to transform the imaging optical path, preventing dust and other contaminants from falling into the imaging system during use and maintenance. The sampled microfluidic chip is transferred to the z-axis platform by the chip feeding device. The industrial control computer 32 uses an autofocus algorithm to drive the z-axis motion motor for microscopic focusing. A 40X microscope lens magnifies the planktonic algae, and the right-angle imaging device 39 and CCD camera 31 capture images of the planktonic algae. The x-axis and y-axis platforms are responsible for chip positioning and field-of-view switching during imaging. After acquiring data from the CCD camera 31, the industrial control computer 32 uses a built-in image vision algorithm to identify and count the algae.

[0064] Figure 5 This application provides a structural diagram of a chip collection module, as shown in the embodiment. Figure 5 As shown, the chip collection module 4 includes a collection frame 41 and a second sensor 42. The second sensor 42 is used to detect whether the collection frame 41 is full of microfluidic chips, and generates an alarm when the number of chips in the collection frame 41 exceeds the upper limit. The collected chips are then processed uniformly.

[0065] This application provides an automatic monitoring device for phytoplankton in water, comprising: a water sample preparation module, a water sample slide preparation module, and a water sample detection module. The water sample preparation module is used to collect water samples and obtain algae concentrate from the water samples. The water sample slide preparation module is used to prepare sample slides from the algae concentrate and transfer the sample slides to the water sample detection module. The water sample detection module is used to photograph the phytoplankton in the sample slides and to identify and count the algae in the photographed images using a preset image vision algorithm. This device achieves automatic sampling, automatic sedimentation and collection, automatic slide preparation, automatic photographing, and automatic identification and counting of phytoplankton, eliminating the need for manual supervision, saving labor costs, effectively avoiding differences in identification results caused by differences in the identification capabilities of different laboratories and testing personnel, improving the standardization of algae detection, and facilitating widespread application.

[0066] Based on the aforementioned automatic monitoring device for phytoplankton in water Figure 6 A flowchart of a method for monitoring phytoplankton in water provided in this application embodiment is shown below. Figure 6 As shown, methods for monitoring phytoplankton in water include:

[0067] S10: Collect water samples and use algae fixatives to obtain algae concentrate from the water samples.

[0068] S11: Prepare sample pieces from algal concentrate.

[0069] S12: Acquire images of planktonic algae in the sample and use a preset image vision algorithm to identify and count algae in the images.

[0070] To better understand this application, the detection process is described below based on the automatic monitoring device for phytoplankton in water. A quantitative pump 11 draws 1L of water sample and 15ml of algae fixative into the sample sedimentation channel 13. After 24 hours of fixation and sedimentation, the supernatant is discharged from the supernatant outlet 16, leaving 30-50ml of concentrated algae at the bottom. A magnetic stirrer 14 mixes the concentrated algae at the bottom of the channel. A robotic arm delivers the microfluidic chip from the chip storage frame 21 to the chip injection device. A peristaltic pump 24 draws 500ul of concentrated algae, and the excess concentrated algae is injected into the chip's sample chamber through the floating injection needle 26 via the injection port. The chip, after sample addition, is transferred to the z-axis platform of the water sample detection module 3 via a chip lifting motor 27, a rotary motor 28, and a rotating arm 29. With the x / y / z axes zeroed, a 40X microscope lens magnifies the phytoplankton. An autofocus algorithm drives the z-axis motor for microscopic focusing. The CCD camera 31 captures images of the phytoplankton. The x-axis and y-axis platforms are responsible for chip positioning and field-of-view switching during image capture, resulting in 100-400 images. The industrial computer 32 acquires data from the CCD camera 31 and uses its built-in image vision algorithm to identify and count the algae, ultimately producing a report including the dominant algae species and their concentrations, as well as the genera and concentrations of other algae.

[0071] This application provides a method for monitoring phytoplankton in water, comprising: collecting water samples and obtaining an algae concentrate from the water samples using an algae fixative; preparing the algae concentrate into a sample slide; acquiring images of phytoplankton in the sample slide; and using a preset image vision algorithm to identify and count the algae in the images. By achieving automatic sampling, automatic sedimentation and collection, automatic slide preparation, automatic imaging, and automatic identification and counting of phytoplankton, no manual intervention is required, saving labor costs. The use of image vision algorithms for automatic identification and counting of phytoplankton effectively avoids differences in identification results caused by variations in the identification capabilities of different laboratories and testing personnel, improving the standardization of algae detection and facilitating widespread application.

[0072] Based on the above embodiments, this application embodiment acquires images of phytoplankton in a sample, including: controlling a z-axis motion motor to drive a z-axis platform on which the sample is placed to reset to a preset point in a direction away from the microscope lens, and controlling the z-axis motion motor to move the z-axis platform towards a designated position in a direction closer to the microscope lens according to a set number of steps and a set speed to complete coarse focusing; after coarse focusing is completed, controlling the z-axis motion motor to move the z-axis platform towards a direction closer to the microscope lens according to a set number of movement steps, and controlling a CCD camera to capture images of the sample after the z-axis platform moves, recording the shooting position of each captured image; performing grayscale processing on each captured image, and calculating the grayscale variance between two adjacent pixels in each image; determining a target image with a grayscale variance greater than a threshold, and controlling the z-axis motion motor to move the z-axis platform to the target shooting position corresponding to the target image; acquiring an image of phytoplankton in the sample captured by the CCD camera at the target shooting position.

[0073] The specific autofocus process is as follows: After the sample image is produced and transferred, the autofocus function is activated. Based on the data accumulated from previous experiments, the number of coarse adjustment steps and speed are set. The Z-axis motion motor resets to zero point 1, moving away from the lens. Then, according to the system-set number of steps and speed, it moves towards the lens to the designated position to complete the coarse focus adjustment. After the coarse adjustment, the CCD camera takes an image at the current position and records the current position as zero point 2. The Z-axis motion motor drives the Z-axis platform to move towards the lens, with the same number of steps each time. The maximum movement must not exceed the limit protection switch; if this position is reached, the motor will stop moving to prevent collision between the lens and the chip. After each movement, the CCD camera takes an image and records the current relative position with respect to zero point. Each captured image is converted to grayscale, and the grayscale variance between two adjacent pixels is calculated to evaluate the image sharpness. A threshold T is set, which can be determined based on the grayscale variance values ​​of existing sharp images. The formula for calculating F is as follows:

[0074]

[0075] Where x and y are the pixel positions, and f(x, y) is the grayscale value of each pixel. If F > T, the image clarity meets the requirements, and the z-axis motion motor drives the z-axis platform back to the image shooting position, completing the autofocus process. The z-axis motion motor remains stationary, while the x-axis and y-axis motion motors move to switch the lens field of view and capture images.

[0076] Based on the above embodiments, the present application embodiment uses a preset image vision algorithm to identify and count algae in an image, including: inputting the image into a preset phytoplankton identification algorithm model, using the phytoplankton identification algorithm model to identify the species of algae in the image; counting non-aggregated single cells in the image based on the number of individual cells; and segmenting and counting algal cells in the image based on the algal species for aggregated cells and clumps of phytoplankton.

[0077] The specific algae counting method is as follows: For non-aggregate single-cell individuals, they are directly identified and counted as individual cells. For phytoplankton samples containing cell aggregations or clumps, such as *Aerophyta*, *Panthera*, and filamentous algae, the algae identification process is based on the entire population as the identification object. Algae cell segmentation and counting are performed after algae species identification. The captured image is converted to grayscale to obtain grayscale image A. The grayscale values ​​of each pixel in grayscale image A are swapped to obtain the converted grayscale image B. The background image of grayscale image B is obtained through computational processing. The background image is then subtracted from grayscale image B to obtain grayscale image C. The optimal threshold T for grayscale image C is calculated, and the image is converted to a binary image based on the obtained threshold. A label matrix is ​​created through connected region analysis to mark the algae outlines. The total area of ​​the algae population and the area of ​​a single cell are determined based on the algae outlines. The ratio between the two yields the number of individual algae cells in the image.

[0078] Based on the above embodiments, before inputting the image into the preset phytoplankton recognition algorithm model, this application embodiment further includes: acquiring first images of algae of known species at various growth stages; filtering the first images according to preset requirements to obtain target images; labeling the target images with rectangular boxes and noting the species names to obtain a sample database; acquiring second images of impurities in the water body, labeling the second images with rectangular boxes and noting the impurity names to obtain an impurity database; importing the sample database and impurity database into a model training platform, establishing a target detection model training task and selecting the sample database and impurity database for training to construct a basic algorithm model; optimizing the image database; incrementally training the basic algorithm model using the optimized sample database and impurity database; and using the model whose recall and accuracy meet preset standards as the phytoplankton recognition algorithm model.

[0079] The specific process for establishing the phytoplankton identification algorithm model is as follows:

[0080] (1) Various algae were isolated and extracted from natural water bodies for cultivation, and the species and genera of algae were determined by morphological and molecular identification.

[0081] (2) Take pictures of algae of known species at each growth stage, select high-resolution pictures with an algae density of about 10 per picture, label them with rectangular boxes and indicate the species name, and form a sample database of multiple labeled algae pictures, with no less than 100 pictures of each species, and ensure that the number of label boxes for each species is basically similar.

[0082] (3) Take pictures of impurities in the water body, mark them with rectangular boxes and label them as impurity 1 to impurity n. A database of impurity images with multiple labels is formed.

[0083] (4) Import the established database (sample database and impurity database) into the model training platform, establish the target detection model training task and select the above database (sample database and impurity database) for training, and establish the basic algorithm model through the initial training.

[0084] (5) For the part where the basic algorithm model has weak recognition ability, the image database is optimized. The optimized database is used to carry out multiple incremental training on the basis of the basic algorithm model. Finally, a model with a recall rate and accuracy of over 99% is established. This model is used as the phytoplankton recognition algorithm model for subsequent algae recognition.

[0085] A general algae identification model was established based on the distribution of algae in freshwater bodies, and the algae identification model was extended to suit the distribution characteristics of algae in various lakes and reservoirs.

[0086] The above provides a detailed description of an automatic monitoring device and method for phytoplankton in water, as provided in this application. The various embodiments in the specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the device disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section. It should be noted that those skilled in the art can make several improvements and modifications to this application without departing from the principles of this application, and these improvements and modifications also fall within the protection scope of the claims of this application.

[0087] It should also be noted that, in this specification, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

Claims

1. An automatic monitoring device for phytoplankton in water, characterized in that, It includes a water sample preparation module, a water sample film preparation module, and a water sample testing module; The water sample preparation module is used to collect water samples and obtain algae concentrate from the water samples; The water sample preparation module is used to prepare the algae concentrate into a sample and transfer the sample to the water sample detection module; The water sample detection module is used to photograph the planktonic algae in the sample and to identify and count the algae in the photographed image using a preset image vision algorithm. The water sample preparation module includes: a metering pump, a multi-way valve, a sample sedimentation treatment channel, and a magnetic stirring device; The metering pump is connected to the multi-way valve and the sample precipitation treatment channel respectively. The metering pump is used to draw the water sample and algae fixative into the sample precipitation treatment channel through the multi-way valve. The sample sedimentation treatment channel is equipped with a waste liquid outlet and a supernatant outlet. The sample sedimentation treatment channel is used to fix and precipitate algae in the water sample and discharge the supernatant from the supernatant outlet to obtain the algae concentrate. The magnetic stirring device is located at the bottom of the sample precipitation treatment channel and is used to mix the algae concentrate at the bottom of the sample precipitation treatment channel evenly. The water sample preparation module includes: a microfluidic chip, a chip storage device, a chip sample introduction device, and a chip feeding device; The microfluidic chip is provided with an inlet port, a sample chamber, and an exhaust port; The chip storage device includes a chip storage frame, a first sensor, and a feeding component. The chip storage frame is used to store the microfluidic chip, and the feeding component is used to transfer the microfluidic chip in the storage frame to the chip injection device. The first sensor is used to detect the number of chips and generate an alarm when the number of chips is less than a lower limit. The chip injection device includes a peristaltic pump, an injection needle lifting motor, and a floating injection needle. The peristaltic pump is connected to the sample precipitation treatment channel and the floating injection needle. The peristaltic pump is used to draw the algae concentrate into the floating injection needle. The injection needle lifting motor is used to control the movement of the floating injection needle. The floating injection needle is used to inject the algae concentrate into the sample chamber through the injection port of the microfluidic chip. The chip feeding device includes a chip lifting motor, a rotary motor, and a rotary arm. The chip lifting motor and the rotary motor are respectively connected to the rotary arm. The chip lifting motor and the rotary motor are used to control the rotary arm to transfer the microfluidic chip with added algae concentrate to the water sample detection module.

2. The automatic monitoring device for planktonic algae in water according to claim 1, characterized by The water sample detection module includes: a light source, a microscope lens, a right-angle imaging device, a CCD camera, an industrial control computer, an x-axis platform, an x-axis motion motor, a y-axis platform, a y-axis motion motor, a z-axis platform, and a z-axis motion motor; The microscope lens is used to magnify planktonic algae; The right-angle imaging device and the CCD camera are used for capturing images of phytoplankton; The industrial control computer is connected to the x-axis motion motor, the y-axis motion motor, the z-axis motion motor and the CCD camera respectively. The industrial control computer is used to acquire images captured by the CCD camera and to identify and count algae in the images using a preset image vision algorithm. The industrial control computer is also used to drive the z-axis motion motor to move the z-axis platform for microscopic focusing using an autofocus algorithm. The x-axis motion motor is used to drive the x-axis platform, the y-axis motion motor is used to drive the y-axis platform, and the x-axis platform and the y-axis platform are used for positioning the chip and switching the field of view during shooting.

3. The automatic monitoring device for planktonic algae in water according to claim 2, characterized by It also includes a chip collection module, which includes a collection frame and a second sensor. The second sensor is used to detect whether the collection frame is full of microfluidic chips and to generate an alarm when the number of chips in the collection frame exceeds the upper limit.

4. A method of monitoring algae in water, characterized by, The automatic monitoring device for phytoplankton in water according to any one of claims 1 to 3 comprises: Water samples were collected, and an algae concentrate was obtained from the water samples using an algae fixative. Prepare sample pieces from the algae concentrate; Images of planktonic algae in the sample are acquired, and algae are identified and counted using a preset image vision algorithm.

5. The method of claim 4, wherein the water is a body of water. The step of preparing the algae concentrate into sample tablets includes: The algae concentrate is injected into the microfluidic chip by controlling a floating injection needle to obtain the sample.

6. The method for monitoring phytoplankton in water according to claim 5, characterized in that, The step of acquiring images of phytoplankton in the sample includes: The z-axis motion motor is controlled to drive the z-axis platform on which the sample is placed to return to a preset point in the direction away from the microscope head. The z-axis motion motor is then controlled to drive the z-axis platform to move towards the microscope head to a specified position according to the set number of steps and the set speed to complete the coarse focus adjustment. After coarse focusing is completed, the z-axis motion motor is controlled to move the z-axis platform toward the microscope head according to the set number of movement steps. After the z-axis platform moves, the CCD camera is controlled to take pictures of the sample and the shooting position of each picture is recorded. Each captured image is converted to grayscale, and the grayscale variance between two adjacent pixels in each image is calculated. The target image with a grayscale variance greater than a threshold is identified, and the z-axis motion motor is controlled to move the z-axis platform to the target shooting position corresponding to the target image. The image of phytoplankton in the sample taken by the CCD camera at the target shooting position is acquired.

7. The method of claim 5, wherein the water is a body of water. The process of using a preset image vision algorithm to identify and count algae in the image includes: The image is input into a preset phytoplankton identification algorithm model, and the phytoplankton identification algorithm model is used to identify the species of algae in the image; For non-aggregated single-cell individuals in the image, count them based on the number of individual cells; For the aggregated cells and clumps of phytoplankton in the image, algal cell segmentation and counting are performed according to algal species.

8. The method of claim 7, wherein the water is a body of water. Before inputting the image into the preset phytoplankton identification algorithm model, the following steps are also included: First images of algae of known species at various growth stages are obtained, and the first images are filtered according to preset requirements to obtain target images; The target images are bounding boxes and their species names are noted to obtain a sample database; A second image of impurities in the water body is obtained, and the second image is labeled with a rectangular box and the impurity name is noted to obtain an impurity database; Import the sample database and the impurity database into the model training platform, establish a target detection model training task, and select the sample database and the impurity database for training to build a basic algorithm model; The image database is optimized, and the basic algorithm model is incrementally trained using the optimized sample database and the impurity database. The model whose recall and accuracy meet the preset standards is used as the phytoplankton identification algorithm model.

Images