A method and device for identifying marine particles based on Mueller matrix hyperspectral microscopy

Abstract

The present application provides a method and device for identifying marine particles based on Mueller matrix hyperspectral microscopy. It includes: adjusting the flow speed of the marine particles to be identified; determining the optimal polarization state to be measured based on the optical system; the Mueller matrix hyperspectral microscopy system generates corresponding polarization state modulated light based on multiple polarization states to be measured, irradiates the marine particles to be identified, and obtains the Mueller matrix hyperspectral image of the marine particles to be identified; extracts the physical, chemical, and morphological characteristics of the marine particles to be identified based on the Mueller matrix hyperspectral image of the marine particles to be identified; identifies the marine particles to be identified based on the Mueller matrix hyperspectral image. The method and device for identifying marine particles based on Mueller matrix hyperspectral microscopy provided in the present application are used to quickly and accurately identify marine particles.

Description

Technical Field

[0001] The present application relates to the technical field of marine particle detection, and in particular to a method and device for identifying marine particles based on Mueller matrix hyperspectral microscopy imaging. Background Art

[0002] With the in-depth development of marine science, the monitoring and research of the marine environment has received increasing attention. As an important part of the marine ecosystem, the types, distribution and dynamic changes of marine particulate matter have a significant impact on marine ecology, climate and even global environmental changes. Therefore, the identification of marine particulate matter is of great significance for understanding the function of marine ecosystems, assessing the quality of the marine environment and predicting marine disasters.

[0003] At present, the classification of marine particles is still mainly based on manual identification, which is not only time-consuming and labor-intensive, but also difficult to simultaneously identify microalgae, sediment, microplastics and other particles through a single instrument or method. Even biological and non-biological particles are prone to errors when being identified. The method of observing marine particles through a high-precision microscope is not only costly and prone to errors, but also limited by the field of view and operational complexity, making it difficult to quickly and efficiently measure and analyze marine particles. Summary of the invention

[0004] In view of this, the present application provides a method and device for identifying marine particles based on Mueller matrix hyperspectral microscopy imaging, which are used to quickly and accurately identify marine particles.

[0005] Specifically, the present application is implemented through the following technical solutions:

[0006] The first aspect of the present application provides a method for identifying marine particles based on Mueller matrix hyperspectral microscopy imaging, the method comprising:

[0007] Adjusting the flow speed of the marine particles to be identified; based on the adjusted flow speed, the duration for which the marine particles to be identified are irradiated by the polarization-modulated light is longer than the light modulation time of the light irradiation;

[0008] Determining the optimal polarization state to be measured based on the optical system;

[0009] The Mueller matrix hyperspectral microscopic imaging system generates corresponding polarization state modulated light based on the optimal polarization state to be measured, illuminates the marine particles to be identified, and obtains the Mueller matrix hyperspectral image of the marine particles to be identified;

[0010] The polarization state of the polarization-state modulated light includes the optimal polarization state to be measured, and for each pixel, the Mueller matrix images of each band are fused to obtain the Mueller matrix hyperspectral image of the marine particles to be identified, and the Mueller matrix hyperspectral image includes the physical, chemical, and morphological characteristics of the marine particles to be identified;

[0011] The marine particles to be identified are identified based on the Mueller matrix hyperspectral image.

[0012] The second aspect of the present application provides a marine particle identification device based on Mueller matrix hyperspectral microscopy imaging, the device comprising an adjustment module, a determination module, a processing module and an identification module; wherein,

[0013] The adjustment module is used to adjust the flow speed of the marine particles to be identified; based on the adjusted flow speed, the duration for which the marine particles to be identified are irradiated by the polarization-modulated light is longer than the light modulation time of the light irradiation;

[0014] The determination module is used to determine the optimal polarization state to be measured based on the optical system;

[0015] The processing module is used for the Mueller matrix hyperspectral microscopic imaging system to generate corresponding polarization state modulated light based on the optimal polarization state to be measured, illuminate the marine particles to be identified, and obtain the Mueller matrix hyperspectral image of the marine particles to be identified;

[0016] The polarization state of the polarization-state modulated light includes the optimal polarization state to be measured, and for each pixel, the Mueller matrix images of each band are fused to obtain the Mueller matrix hyperspectral image of the marine particles to be identified, and the Mueller matrix hyperspectral image includes the physical, chemical, and morphological characteristics of the marine particles to be identified;

[0017] The identification module is used to identify the marine particles to be identified based on the Mueller matrix hyperspectral image.

[0018] The marine particle identification method and device based on Mueller matrix hyperspectral microscopy provided in the present application simultaneously obtains the polarization characteristics and spectral characteristics of the marine particles to be identified through the Mueller matrix hyperspectral microscopy system, and simultaneously obtains high-dimensional information including the morphology, physical and chemical characteristics of the particles through the high-dimensional Mueller matrix hyperspectral image, so as to accurately distinguish biological particles from non-biological particles, so that the fine observation and identification of marine particles can be achieved under low-power microscopes. Further, by using the Mueller matrix hyperspectral microscopy system, the optical differences of the marine particles to be identified can be fully and accurately captured, thereby obtaining massive information. This information can be directly input into the machine learning model to identify the unique optical characteristics presented in the marine particle image. It is worth noting that the optical characteristics of the Mueller matrix hyperspectral are affected by the comprehensive influence of multiple properties of marine particles, including pigment content, microstructure, refractive index, morphological appearance, etc. These properties comprehensively form the unique Mueller matrix hyperspectral images of different marine particles. Therefore, when classifying marine particles, there is no need to quantitatively extract and calculate the relationship between properties and optical features. Instead, the identification and classification results of marine particles can be obtained directly based on the Mueller matrix hyperspectral optical characteristics. The advantage of this method is that even under low magnification, there is no need to quantitatively calculate and obtain the microstructure and material composition information of marine particles. Different types of marine particles to be identified can be accurately distinguished, greatly improving the speed and accuracy of marine particle identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 A flow chart of Example 1 of a method for identifying marine particles based on Mueller matrix hyperspectral microscopy provided in this application;

[0020] Figure 2 A flow chart of Example 2 of a method for identifying marine particles based on Mueller matrix hyperspectral microscopy provided in this application;

[0021] Figure 3 A schematic diagram of the structure of a first embodiment of a marine particle identification device based on Mueller matrix hyperspectral microscopy imaging provided by the present application;

[0022] Figure 4 A hardware structure diagram of a marine particle identification device based on Mueller matrix hyperspectral microscopy imaging in which the marine particle identification device based on Mueller matrix hyperspectral microscopy imaging of the present application is located;

[0023] Figure 5 This is a schematic diagram of the structure of Example 2 of the marine particle identification device based on Mueller matrix hyperspectral microscopy imaging provided in this application.

[0024] Description of reference numerals:

[0025] 1: Lighting source;

[0026] 2: first polarizer;

[0027] 3: The first photoelastic modulator;

[0028] 4: Condenser;

[0029] 5: Microfluidic system;

[0030] 6: Sample stage;

[0031] 7: Microscope objective lens;

[0032] 8: second photoelastic modulator;

[0033] 9: second polarizer;

[0034] 10: Hyperspectral camera. DETAILED DESCRIPTION

[0035] Here, exemplary embodiments are described in detail, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application.

[0036] The terms used in this application are only for the purpose of describing specific embodiments and are not intended to limit this application. The singular forms of "a", "said" and "the" used in this application are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used in this article refers to and includes any or all possible combinations of one or more associated listed items.

[0037] It should be understood that although the terms first, second, third, etc. may be used in the present application to describe various information, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0038] The present application provides a method and device for identifying marine particles based on Mueller matrix hyperspectral microscopy imaging, which are used to quickly and accurately identify marine particles.

[0039] Specific embodiments are given below to introduce the technical solution of the present application in detail.

[0040] The marine particle identification method based on Mueller matrix hyperspectral microscopy provided in this embodiment fully captures the optical differences of the marine particles to be identified through the Mueller matrix hyperspectral microscopy system, obtains a large amount of rich information, and directly identifies the optical absorption characteristics reflected in the entire image through machine learning. The optical absorption characteristics are affected by the Mueller characteristics and spectral characteristics, and there is no need for quantitative extraction and calculation. The identification result is directly obtained through the optical absorption characteristics. In this way, even under a low-power microscope, there is no need to quantitatively calculate the microstructure and material composition information, and different types of marine particles to be identified can be accurately distinguished, thereby improving the precision and accuracy of marine particle identification.

[0041] Figure 1 This is a flow chart of the first embodiment of the method for identifying marine particles based on Mueller matrix hyperspectral microscopy provided in this application. Please refer to Figure 1 , the method provided in this embodiment may include:

[0042] S101, adjusting the flow speed of the marine particles to be identified; based on the adjusted flow speed, the duration for which the marine particles to be identified are irradiated by polarization-modulated light is longer than the light modulation time of the light irradiation.

[0043] Specifically, the marine particulate matter to be identified refers to tiny solid or liquid particles suspended or dissolved in the ocean water. The particle size range of the marine particulate matter to be identified ranges from nanometers to millimeters.

[0044] It should be noted that the marine particles to be identified can come from the remains of marine organisms, metabolites, silt, aerosol sedimentation, human emissions and other sources. For example, in one embodiment, the marine particles to be identified may include microalgae, bacteria, viruses, silt, mineral particles, and organic and inorganic particles. There are many types of marine particles to be identified, including non-biological marine particles, such as algae, etc., and biological marine particles. Marine particles include rich features, such as pigment content, chemical features such as carbohydrates, and physical properties such as refractive index and microstructure. A single Mueller matrix or optical image cannot complete the identification of the above-mentioned complex marine particles. For example, biological and non-biological marine particles cannot be effectively and specifically identified simply through the Mueller matrix image.

[0045] Furthermore, when adjusting the flow rate of the marine particles to be identified, the marine particles to be identified can be made to flow through the sample stage or the observation area, thereby ensuring that the marine particles to be identified stably pass through the sample stage.

[0046] It should be noted that by controlling the flow speed, the marine particles to be identified can be continuously illuminated by the polarization-modulated light within the light modulation time of the light irradiation. For example, in one embodiment, the light modulation time is 0.1 seconds, and the flow speed of the marine particles to be identified is adjusted so that the time for which the marine particles to be identified are illuminated by the polarization-modulated light exceeds 0.1 seconds.

[0047] Furthermore, the light irradiation has an irradiation range, and the flow rate is used to control the movement time of the marine particles in the irradiation range to exceed the modulation time of the light.

[0048] Furthermore, when the marine particles to be identified flow through the sample stage, the movement distance of the marine particles to be identified can be basically ignored. Due to the extremely short shooting time, different illumination results of the same marine particle to be identified can be approximately considered to be at the same position. In specific implementation, the flow speed of the marine particles to be identified is much greater than the modulation speed of light.

[0049] Furthermore, the sample stage has a stable structure and design to ensure that there will be no shaking or deviation during the identification process of the marine particles to be identified, thereby ensuring the stability of imaging. At the same time, the sample stage can be made of a material with good light transmittance so that light can pass through the sample stage and shine on the marine particles to be identified, thereby facilitating the clear capture of the image of the marine particles to be identified.

[0050] It should be noted that when locating the marine particles to be identified, the microfluidic system can be used to control the marine particles to be identified to continuously pass through the sample stage. Furthermore, the entire time that the marine particles to be identified are irradiated by light is the duration of irradiation by polarization-modulated light. In specific implementation, for example, in one embodiment, the marine particles to be identified can be irradiated from the entrance to the exit, and this period of time is determined as the duration of irradiation by polarization-modulated light. By matching the flow speed with the light modulation time, it is ensured that the marine particles to be identified can be effectively illuminated by light of different polarization states, thereby improving the measurement accuracy of the Mueller matrix hyperspectral image of the marine particles to be identified.

[0051] A specific embodiment is given below to introduce in detail the process of determining the flow velocity of the marine particles to be identified:

[0052] (1) Calculating the light modulation time according to the historical imaging time of the polarization state modulated light of the Mueller matrix hyperspectral microscopy imaging system.

[0053] Furthermore, the length of time that the sample is circulated on the sample stage directly affects the data acquisition efficiency of the imaging system. If the time is too short, the polarization-modulated light will not be able to fully illuminate the marine particles to be identified, and thus the complete and accurate Mueller matrix optical information cannot be obtained. Therefore, the method of matching the flow speed with the polarization-modulated light speed provided by the present invention ensures the acquisition of the Mueller matrix hyperspectral information and improves the recognition accuracy of the marine particles to be identified. In specific implementation, some marine particles to be identified can be marked, and the time they are irradiated by light can be measured, and the time is determined as the duration of irradiation by the polarization-modulated light.

[0054] For the same type of marine particles to be identified, the required light is consistent. The light modulation time is preliminarily determined according to the identification history of the same type, and the flow of the marine particles to be identified is controlled according to the above time.

[0055] (2) Calculating the flow velocity of the marine particles based on the light modulation time, wherein the light modulation time is proportional to the flow velocity.

[0056] Specifically, the shorter the light modulation time is, the faster the flow speed can be, and the light modulation is completed when the marine particles to be identified flow out. Further, the distance of the section is determined according to the time required for the marine particles to be identified to move from the inlet to the sample stage, and the lower limit of the flow speed is determined according to the ratio of the distance to the light modulation time.

[0057] It can be understood that the longer the light modulation time is, the slower the flow speed of the marine particles is, ensuring that the marine particles can be fully illuminated and detected by light of different polarization states.

[0058] The method provided by the present invention is generated by matching the flow speed with the light modulation time. On the one hand, it controls the marine particles to be identified to pass through the sample stage quickly to complete the rapid identification of large quantities; on the other hand, it ensures that the polarization modulated light can be generated and focused to irradiate the marine particles to be identified, so that the marine particles are completely irradiated, thereby obtaining complete Mueller matrix hyperspectral information, realizing the distinction and identification of various types of particles, and improving the accuracy of identification.

[0059] (3) The marine particles to be identified flow into the sample stage at the flow rate.

[0060] Specifically, according to the calculated speed, the microfluidic driven pump controls the marine particles to be identified to move at the flow speed.

[0061] (4) Optimizing the real-time flow rate based on the real-time light modulation time of the Mueller matrix hyperspectral microscopy imaging system so that the time for which the marine particles to be identified are irradiated by the polarization-modulated light is longer than the light modulation time.

[0062] Since the calculation of the flow rate is determined based on historical conditions, in order to further improve the accuracy of control and ensure the acquisition of the Mueller matrix hyperspectral image, the present invention can also adjust the real-time flow rate according to the real-time light modulation time. When the light modulation time is shorter than the estimated time, the real-time flow rate is increased; conversely, the flow rate can be reduced to increase the length of time that the marine particles to be identified are irradiated by the polarization modulated light on the sample stage.

[0063] Furthermore, the microchannel depth design of the microfluidic system needs to be combined with the parameters of the objective lens so that the particles are clearly focused. Specifically, the focusing distance is calculated according to the parameters of the objective lens, and the focusing distance is used as the microchannel depth value, so that the particles can capture clear images of the marine particles to be identified under the irradiation of polarized modulated light.

[0064] Furthermore, Mueller matrix hyperspectral image is an image representation method that combines Mueller matrix measurement technology and hyperspectral imaging technology. In specific implementation, the Mueller matrix describes the change in the polarization state of light when it passes through the marine particles to be identified. The Mueller matrix contains all the polarization characteristics of the marine particles. At the same time, hyperspectral imaging is used to obtain information about the object in multiple spectral bands, so that the Mueller matrix hyperspectral image can be quickly calculated. Therefore, the Mueller matrix hyperspectral image can reveal the physical, chemical and morphological characteristics of the marine particles to be identified, including refractive index, microstructure, morphological appearance, pigment content, optical rotation, etc.

[0065] Furthermore, the flow rate and the recognition effect of the marine particles to be identified influence each other. If the flow rate is too fast, the polarization information of the marine particles to be identified cannot be fully collected, thereby affecting the recognition effect of the marine particles to be identified; conversely, if the flow rate is too slow, the recognition effect of the marine particles to be identified can be improved, but the recognition efficiency will be reduced.

[0066] Furthermore, the optimal flow rate is determined based on the duration of the polarization-modulated light irradiation of the marine particles to be identified and the light modulation time. In specific implementation, the flow rate can be determined to ensure that sufficient polarization information is collected within the duration of the polarization-modulated light irradiation to meet the requirements of the light modulation time, while maintaining a high measurement efficiency (i.e., shortening the duration of the polarization-modulated light irradiation while ensuring imaging).

[0067] The marine particle identification method based on Mueller matrix hyperspectral microscopy provided in this embodiment can ensure that the imaging system fully collects the polarization information of the particles when they flow through the sample stage by accurately controlling the flow speed of the marine particles. At the same time, it takes into account the efficiency of measurement and ensures the stability and reliability of the identification process. In this way, the identification effect of the marine particles to be identified can be improved and the imaging quality can be stabilized.

[0068] S102: Determine an optimal polarization state to be measured based on the optical system.

[0069] Specifically, the optical system is an optical system for identifying marine particles to be identified.

[0070] In specific implementation, the optimal polarization state to be measured can be determined by the measurement accuracy of the optical system. The measurement accuracy of the optical system is the recognition ability of the optical system to identify the marine particles to be identified.

[0071] Furthermore, the amount of polarization modulated light can be adjusted to change the measurement accuracy of the optical system, or the illumination polarization state can be adjusted to change the measurement accuracy of the optical system while keeping the amount unchanged. It should be noted that when the measurement accuracy is different, the devices used and the measurement schemes can be different.

[0072] Furthermore, the Mueller matrix is ​​a mathematical matrix used to describe the change in the polarization state of light when it passes through the marine particles to be identified. The Mueller matrix can reflect the polarization optical properties and microscopic structure information of the marine particles to be identified.

[0073] Furthermore, polarization-state modulated light corresponding to the optimal polarization state to be measured is generated. In this way, a more accurate Mueller matrix can be calculated based on the polarization-state modulated light to detect the marine particles to be identified, thereby ensuring the accuracy of the identification result.

[0074] Further, the optimal polarization state to be measured is used to describe the polarization state information of the light signal irradiating the marine particles to be identified. As an optional embodiment, the required information quantity fineness can be inferred according to the accuracy to be identified, and the optimal polarization state to be measured can be determined based on the information quantity fineness.

[0075] A specific embodiment is given below to introduce in detail the steps of determining the optimal polarization state to be measured:

[0076] (1) Obtaining an optimal polarization measurement scheme for the optical system based on optical parameters and configuration of the optical system.

[0077] Specifically, the polarization state parameter is the value of the polarization state of the detected light. The specific data type included in the polarization state parameter is determined according to actual needs, and is not limited in this embodiment. For example, in one embodiment, the polarization state parameter may include polarization degree, polarization state, polarization direction, parameters of polarization ellipse, and Stokes parameters.

[0078] In specific implementation, different degrees of interference noise will appear under the measurement accuracy of different optical systems, and serious interference noise will affect the accuracy of recognition.

[0079] In this step, an optimal polarization measurement solution for the optical system may be obtained based on the optical parameters and configuration corresponding to the optical system with the minimum noise.

[0080] (2) Determining the optimal polarization state to be measured based on the polarization state corresponding to the optimal polarization measurement scheme.

[0081] When the optimal polarization state to be measured is specifically implemented, the optimal polarization state to be measured is obtained based on the optimal polarization measurement scheme. For example, in one embodiment, the optimal polarization state to be measured includes four polarization state characteristics, which can be expressed by the following formula:

[0082] ;

[0083] Furthermore, in another optional embodiment, the optimal polarization state to be measured may also be 6 polarization state characteristics (0 ° Linear polarized light, 45 ° Linear polarized light, 90 ° Linear polarized light, 135 ° Linearly polarized light, left-handed circularly polarized light, right-handed circularly polarized light), generates light for detection based on the optimal polarization state to be measured. It should be noted that when more polarization states of light are needed for measurement, the light modulation time needs to be lengthened and the flow speed needs to be reduced.

[0084] In specific implementation, the number of polarization state characteristics may be changed to adjust the polarization state signal, or the value of the signal included in the polarization state characteristics may be changed to adjust the detected light.

[0085] The marine particle identification method based on Mueller matrix hyperspectral microscopy provided in this embodiment can better suppress Gaussian noise and Poisson noise in the measurement process by determining the optimal polarization measurement scheme, improve the signal-to-noise ratio of the data, and thus enhance the imaging quality. Furthermore, the optimal polarization state to be measured is obtained through the optimal polarization measurement scheme, which ensures the accuracy of in-depth understanding of the optical characteristics of marine particles and improves the efficiency and accuracy of marine particle identification.

[0086] S103, the Mueller matrix hyperspectral microscopic imaging system generates corresponding polarization state modulated light based on the optimal polarization state to be measured, illuminates the marine particles to be identified, and obtains a Mueller matrix hyperspectral image of the marine particles to be identified;

[0087] Among them, the polarization state of the polarization-state modulated light includes the optimal polarization state to be measured. For the marine particles to be identified, the Mueller matrix images of each band are fused to obtain the Mueller matrix hyperspectral image of the marine particles to be identified, and the Mueller matrix hyperspectral image includes the physical, chemical and morphological characteristics of the marine particles to be identified.

[0088] The specific polarization-state modulated light includes the optimal polarization state to be measured.

[0089] Furthermore, when the polarization modulated light illuminates the marine particles to be identified, there are multiple images corresponding to the marine particles to be identified, and each image contains multiple pixel points.

[0090] In specific implementation, the marine particles to be identified have 16 Mueller matrix images of different bands. The images of the marine particles to be identified in different bands are fused to obtain the Mueller matrix hyperspectral image of the marine particles to be identified.

[0091] It should be noted that the more bands a hyperspectral camera has, the more Mueller matrix images there are on one pixel, and the more accurate the results of identifying the marine particles to be identified will be.

[0092] A specific embodiment is given below to introduce the generation process of polarization modulated light in detail:

[0093] (1) Adjusting the white light source generated by the illumination light source based on the first polarizer to obtain polarized illumination white light.

[0094] Specifically, the illumination light source is used to generate continuous illumination white light. It should be noted that since the illumination white light has a wide spectral range, it can cover visible light and part of the ultraviolet and infrared regions, and can provide rich information for subsequent spectral analysis.

[0095] Furthermore, the first polarizer is an optical element that allows only light waves in a certain direction to pass through, while absorbing or reflecting light waves in other directions. In specific implementation, the first polarizer is used to screen a wide spectrum white light source to obtain polarized white light with a wide spectrum.

[0096] (2) Modulating the polarized illumination white light based on the first photoelastic modulator to obtain the polarization-state modulated light.

[0097] Specifically, the number of the first photoelastic modulators is determined according to actual needs, and is not limited in this embodiment. For example, in one embodiment, the rapid modulation of polarized illumination white light can be completed based on two first photoelastic modulators.

[0098] In a specific implementation, the polarized white light is passed through the first photoelastic modulator in sequence, and the polarized white light is quickly modulated by a specific polarization state sequence generated by the first photoelastic modulator, so that the polarization state of the output polarization state modulated light is consistent with the polarization state required for calculating the Mueller matrix.

[0099] The marine particle identification method based on Mueller matrix hyperspectral microscopy provided in this embodiment can realize polarization measurement by converting a wide-spectrum white light source into polarized illumination white light through a first polarizer. Furthermore, by modulating the polarization state modulated light through the first photoelastic modulator, the Gaussian noise and Poisson noise in the polarization state modulated light can be ensured to be low, thereby improving the signal-to-noise ratio and reliability of the measurement result.

[0100] A specific embodiment is given below to introduce the generation process of polarization modulated light:

[0101] Step 1: Select a preset type of condenser based on the parameter to be measured, and different types of condensers have different refraction angles and performances.

[0102] Specifically, the direction to be measured is the focus identification direction of the marine particles to be identified. In other words, when identifying the marine particles to be identified, the focus direction of the identification result is the direction to be measured. In specific implementation, the direction to be measured can be the overall shape of the marine particles to be identified, or the direction to be measured can also be the information of the edge and internal structure of the marine particles to be identified.

[0103] Furthermore, different types of condensers correspond to different directions to be measured. It should be noted that the specific types included in the preset types are determined according to actual needs, and are not limited in this embodiment.

[0104] In this step, when the direction to be measured is to identify the overall shape of the marine particles to be identified, the condenser can be determined as an Abbe condenser; when the direction to be measured is to identify the edge and internal structure of the marine particles to be identified, the condenser can be determined as a dark field condenser. The condenser provided by the present invention only needs to meet the qualitative analysis, that is, to extract the characteristics and classify and identify the particles, so a dark field condenser can also be used, and the angle of the condenser can be adjusted in a dark field environment to meet the goal of observing different contents.

[0105] Step 2: input the polarized illumination white light modulated by the photoelastic modulator into the condenser, and adjust the polarization measurement angle based on the condenser to obtain the polarization state modulated light.

[0106] In specific implementation, the modulated polarized illumination white light is optimized and focused by a condenser to obtain polarization-modulated light.

[0107] The marine particle identification method based on Mueller matrix hyperspectral microscopy provided in this embodiment can effectively focus light on marine particles through a condenser to ensure that the polarization modulated light can illuminate the marine particles to be identified, providing conditions for subsequent microscopic imaging. Furthermore, by selecting a suitable condenser type according to specific research needs, the best information extraction effect can be obtained.

[0108] S104: Identify the marine particles to be identified based on the Mueller matrix hyperspectral image.

[0109] In the specific implementation, the generated Mueller matrix hyperspectral image is analyzed to obtain the category of the marine particles to be identified.

[0110] In the specific implementation, in this embodiment, white light is used as the detection light beam to detect the marine particles to be identified. Since the wide-spectrum white light contains rich wavelengths, there is no need to replace the light source multiple times. The marine particles to be identified can be identified through the Mueller matrix under the high spectrum through a single irradiation.

[0111] It should be noted that the hyperspectral spectrum is a spectrum containing multiple wavelengths.

[0112] A specific embodiment is given below to introduce in detail the process of identifying the marine particles to be identified:

[0113] Receiving the signal of the hyperspectral camera and then calculating the Mueller matrix hyperspectral image of the marine particles to be identified;

[0114] For each pixel of the Mueller matrix hyperspectral image, a Mueller matrix hyperspectral characteristic curve is obtained, wherein the abscissa of the Mueller matrix hyperspectral characteristic curve is the band of the spectrum, and the ordinate is the light intensity value of the corresponding pixel under the abscissa band.

[0115] Specifically, the mean of multiple Mueller matrix hyperspectral characteristic curves of all pixels of the Mueller matrix hyperspectral image can be calculated to obtain the value corresponding to each pixel, and then the recognition result is determined based on the mean corresponding to these pixels. It can be understood that each pixel point corresponds to 16× the number of camera spectral channels.

[0116] The Mueller matrix hyperspectral characteristic curve is identified based on machine learning, and the characteristics of the marine particles to be identified are determined based on the identification results.

[0117] Furthermore, the recognition result can be obtained based on the recognition method of the neural network. In specific implementation, for example, in one embodiment, there are 100 pixels in the Mueller matrix hyperspectral image of a marine particle, corresponding to 16×100×the number of camera spectral channels data points, representing all the information dimensions of the marine particle. After averaging the 100 pixels, the overall Mueller matrix hyperspectral characteristic curve can be obtained, and these curves are input into the pre-trained recognition model to obtain the corresponding recognition results.

[0118] Furthermore, the Mueller matrix hyperspectral characteristic curve includes the spectral absorption characteristic curve, and the amount of information of the Mueller matrix hyperspectral characteristic curve far exceeds that of the spectral absorption characteristic curve.

[0119] The method of identifying the Mueller matrix hyperspectral characteristic curve based on machine learning to obtain the identification result of each pixel includes:

[0120] Locating specific features in the Mueller matrix hyperspectral characteristic curve based on the machine learning, wherein the specific features are determined by the physical and chemical properties of each pixel;

[0121] The type of the marine particulate matter to be identified is determined based on the specific characteristics.

[0122] Specifically, the Mueller matrix hyperspectral image is processed to analyze the shape, internal arrangement and anisotropy of the marine particles to be identified.

[0123] It should be noted that, since the Mueller matrix hyperspectral image contains rich information, if necessary, the chemical characteristics, microstructure and physical property information of marine particles can be extracted from the Mueller matrix hyperspectral image. The Mueller matrix hyperspectral image is processed to analyze the chemical characteristics of the marine particles to be identified. In specific implementation, spectral characteristic parameters such as absorption peaks, reflectivity curves, etc. can be obtained to determine the composition of the marine particles to be identified. For example, in one embodiment, the chemical composition, pigment content and other data of the marine particles to be identified are determined based on the characteristics corresponding to the Mueller matrix hyperspectral image. The type information of the marine particles to be identified is determined based on the microstructure and the composition. In specific implementation, after obtaining the microstructure information and composition information of the marine particles to be identified, the two types of information are comprehensively judged. In specific implementation, the clues of the morphology and physical properties of the marine particles to be identified provided by the microstructure information, combined with the chemical composition and chemical properties of the marine particles to be identified revealed by the composition information, and then the type information of the marine particles to be identified is determined.

[0124] Furthermore, the curve graph corresponding to the marine particles to be identified can be determined through the Mueller matrix hyperspectral image. According to the marine particles curve graph, the properties of the marine particles to be identified can be directly determined according to the numerical value of the target part, and then the characteristics of the marine particles to be identified can be determined.

[0125] It should be noted that each marine particle to be identified contains several images of the spectral channels of the 16× camera, and the obtained Mueller matrix hyperspectral image contains rich information, which is sufficient to complete the detection of marine particles to be identified at low pixels.

[0126] It can be understood that with the support of a large number of Mueller matrix hyperspectral images, there is no need to clearly see indicators such as the outline of the marine particles to be identified, nor is there any need to quantitatively calculate the physical and chemical characteristics of the pixels to be identified. The detection of the marine particles to be identified can be accurately completed through the specific features in the Mueller matrix hyperspectral images.

[0127] The marine particle identification method based on Mueller matrix hyperspectral microscopy provided in this embodiment can obtain ultra-high dimensional information of the marine particles to be identified through the Mueller matrix hyperspectral microscopy system, enrich the characteristic data of the marine particles to be identified through a comprehensive information acquisition method, and provide a solid foundation for subsequent classification and identification. Further, through the combination with the microfluidic system, it is realized that the marine particles to be identified can be continuously transported to the sample stage and flow out after observation and measurement, which greatly improves the measurement efficiency. Further, the Mueller matrix hyperspectral microscopic features of the marine particles to be identified are received and analyzed at the same time, which greatly increases the information dimension and makes the identification results more accurate and reliable. In this way, the microstructure, pigment content, chemical composition and other components of the particles are deeply understood through the Mueller matrix hyperspectral features, which improves the accuracy and efficiency of the identification of the marine particles to be identified.

[0128] Figure 2 This is a flow chart of Example 2 of the marine particle identification method based on Mueller matrix hyperspectral microscopy provided in this application. Please refer to Figure 2 Based on the above embodiment, the step of obtaining the Mueller matrix hyperspectral image of the marine particles to be identified includes:

[0129] S201, obtaining, based on a microscope objective, an objective light signal after the polarization-modulated light irradiates the marine particle to be identified; wherein the focal length of the microscope objective matches the position of the marine particle to be identified on the sample stage.

[0130] Specifically, the microscope objective is used to amplify the result after being irradiated with polarization-modulated light, so as to clearly obtain its morphological, physical and chemical characteristic information.

[0131] Furthermore, the focal length of the microscope objective lens and the distance between the microscope objective lens and the sample stage can be set to the same value to ensure accurate focusing of the microscope objective lens. The focal length matching of the microscope objective lens ensures the high quality and high resolution of the optical signal, providing a reliable data basis for subsequent feature extraction and classification recognition.

[0132] S202, modulating the objective light signal based on the polarization analyzer system, inputting the modulated objective light signal into the hyperspectral camera, calculating and obtaining the Mueller matrix image of each band, and the set of the Mueller matrix images of all bands is the Mueller matrix hyperspectral image of each marine particle to be identified.

[0133] Optionally, the polarization analysis system includes a second polarizer and a plurality of second photoelastic modulators.

[0134] Specifically, the polarization analyzer system further modulates the received optical signal through the internal second photoelastic modulator and the second polarizer. A specific embodiment is given below to introduce the working process of the polarization analyzer system in detail:

[0135] The polarization analyzer system includes a second polarizer and a plurality of second photoelastic modulators; the polarization analyzer system modulates the objective lens light signal, and the modulated objective lens light signal is input into a hyperspectral camera to obtain the Mueller matrix hyperspectral image of each band.

[0136] The objective lens light signal is polarized based on the second photoelastic modulator and the second polarizer, and the Mueller matrix image of each band is calculated by the hyperspectral camera, including:

[0137] Specifically, the second photoelastic modulator and the second polarizer are used to modulate the target light signal for polarization analysis, obtain the modulated target light signal in different polarization states, and provide a data basis for subsequent Mueller matrix imaging. Then, the hyperspectral camera receives the modulated target light signal from the polarization analysis system and generates a Mueller matrix image for each band.

[0138] The marine particle identification method based on Mueller matrix hyperspectral microscopy provided in this embodiment first modulates the light signal obtained by the objective lens based on the polarization analyzer system, and then calculates the Mueller matrix image of each band based on the hyperspectral camera. In this way, the high resolution and spectral coverage of the hyperspectral camera enable researchers to have a deeper understanding of the microstructure and chemical composition of marine particles. Furthermore, the hyperspectral camera can continuously receive and process the particle light signals from the microfluidic system to achieve high-throughput measurement and imaging of marine particles. Furthermore, the polarization analyzer system can effectively extract the Mueller matrix hyperspectral features of the marine particles to be identified from the objective lens light signal. After these feature data are processed and analyzed by complex algorithms, the type and properties of the particles can be accurately obtained, thereby improving the accuracy and reliability of identification.

[0139] Corresponding to the aforementioned embodiment of a method for identifying marine particles based on Mueller matrix hyperspectral microscopy, the present application also provides an embodiment of a device for identifying marine particles based on Mueller matrix hyperspectral microscopy.

[0140] A specific device embodiment is given below to introduce the process of marine particle identification in detail:

[0141] Figure 3 This is a schematic diagram of the structure of the first embodiment of the marine particle identification device based on Mueller matrix hyperspectral microscopy provided by this application. Figure 3The marine particle identification device based on Mueller matrix hyperspectral microscopy imaging includes an illumination light source 1, a first polarizer 2, a first photoelastic modulator 3, a first photoelastic modulator 3, a condenser 4, a microfluidic system 5, a sample stage 6, a microscope objective 7, a second photoelastic modulator 8, a second photoelastic modulator 8, a second polarizer 9, and a hyperspectral camera 10.

[0142] Please continue to refer to Figure 3 , the illumination light source 1 generates a white light source, and the white light is input into the first polarizer 2. It should be noted that the first polarizer 2 and the two first photoelastic modulators 3 constitute a polarization system. When the white light source is input into the polarization system, it is rapidly modulated to become polarization modulated light.

[0143] In specific implementation, the first polarizer 2 in the polarizing system adjusts the illumination white light into polarized illumination white light and inputs it into two first photoelastic modulators 3. The two first photoelastic modulators 3 modulate the polarized illumination white light twice to obtain polarization-modulated light.

[0144] Furthermore, the polarization system inputs the polarization-state modulated light into the condenser 4 , and the condenser 4 focuses the polarization-state modulated light onto the marine particles.

[0145] It should be noted that the flow rate of the marine particles is controlled by the microfluidic system 5, so that the marine particles flow into the sample stage 6 at a suitable flow rate. Figure 3 , the marine particles flow out of the sample stage 6 from direction a.

[0146] Furthermore, after the polarization modulated light irradiates the marine particles, the light signal transmitted by the marine particles to be inspected is captured by the microscope objective 7 to generate an objective light signal, and the microscope objective 7 inputs the objective light signal into the second photoelastic modulator 8.

[0147] It should be noted that the two second photoelastic modulators 8 and the second polarizer 9 constitute a polarization analyzer system. When the objective lens light signal is input into the polarization analyzer system, it is processed and then input into the hyperspectral camera 10, and finally the Mueller matrix hyperspectral image is calculated.

[0148] In a specific implementation, the objective lens light signal is input into the second photoelastic modulator 8 , and after passing through the two second photoelastic modulators 8 and the second polarizer 9 in sequence, a target light signal is obtained, which is input into the hyperspectral camera 10 .

[0149] Furthermore, the hyperspectral camera 10 obtains the corresponding Mueller matrix image of each band, and fuses the Mueller matrix images of all bands to obtain the Mueller matrix hyperspectral image.

[0150] The marine particle identification device based on Mueller matrix hyperspectral microscopy provided in this embodiment generates polarization modulated light to detect marine particles through an illumination light source 1, a first polarizer 2, a first photoelastic modulator 3, a first photoelastic modulator 3 and a condenser 4, and allows the marine particles to pass through and be detected in sequence through a microfluidic system 5 and a sample stage 6, and then forms a Mueller matrix hyperspectral image through a microscope objective 7, a second photoelastic modulator 8, a second photoelastic modulator 8, a second polarizer 9 and a hyperspectral camera 10. In this way, the Mueller matrix hyperspectral image of a single marine particle is received and analyzed at the same time, which greatly increases the information dimension, enriches the characteristic data of marine particles, and deeply understands the microstructure, pigment content and chemical composition of marine particles, so as to achieve further interpretation, and greatly improves the accuracy of marine particle identification. In this way, high-throughput measurement and identification of marine particles is achieved.

[0151] The present application provides an embodiment of a marine particle identification device based on Mueller matrix hyperspectral microscopy, which can be applied to a marine particle identification device based on Mueller matrix hyperspectral microscopy. The device embodiment can be implemented by software, or by hardware or a combination of software and hardware. Taking software implementation as an example, as a device in a logical sense, the processor of the marine particle identification device based on Mueller matrix hyperspectral microscopy in which it is located reads the corresponding computer program instructions in the non-volatile memory into the memory and runs them. From a hardware perspective, if Figure 4 As shown, it is a hardware structure diagram of the marine particle identification device based on Mueller matrix hyperspectral microscopy imaging of the present application. Figure 4 In addition to the processor, memory, network interface, and non-volatile memory shown, the marine particle identification device based on Mueller matrix hyperspectral microscopy in which the device in the embodiment is located may also include other hardware according to the actual function of the marine particle identification device based on Mueller matrix hyperspectral microscopy, which will not be described in detail.

[0152] Figure 5 This is a schematic diagram of the structure of the second embodiment of the marine particle identification device based on Mueller matrix hyperspectral microscopy provided by this application. Figure 5 The device provided in this embodiment includes an adjustment module 510, a determination module 520, a processing module 530 and an identification module 540; wherein,

[0153] The adjustment module 510 is used to adjust the flow speed of the marine particles to be identified; based on the adjusted flow speed, the duration for which the marine particles to be identified are irradiated by the polarization-modulated light is longer than the light modulation time of the light irradiation;

[0154] The determination module 520 is used to determine the optimal polarization state to be measured based on the measurement accuracy;

[0155] The processing module 530 is used to generate corresponding polarization state modulated light based on the optimal polarization state to be measured, illuminate the marine particles to be identified, and obtain the Mueller matrix hyperspectral image of the marine particles to be identified;

[0156] The polarization state of the polarization-state modulated light includes the optimal polarization state to be measured, and for each pixel, the Mueller matrix images of each band are fused to obtain the Mueller matrix hyperspectral image of the marine particles to be identified, and the Mueller matrix hyperspectral image includes the physical, chemical, and morphological characteristics of the marine particles to be identified;

[0157] The identification module 540 is used to identify the marine particles to be identified based on the Mueller matrix hyperspectral image.

[0158] The device of this embodiment can be used to perform Figure 1 The steps, specific implementation principles and implementation processes of the method embodiment shown are similar and will not be repeated here.

[0159] Optionally, the adjustment module 510 is specifically used to calculate the light modulation time according to the historical imaging time of the polarization state modulated light of the Mueller matrix hyperspectral microscopy imaging system;

[0160] The adjustment module 510 is further specifically used to calculate the flow velocity of the marine particles based on the light modulation time, and the light modulation time is proportional to the flow velocity;

[0161] The adjustment module 510 is also specifically used for the marine particles to be identified to flow into the sample stage according to the flow speed;

[0162] The adjustment module 510 is also specifically used to optimize the real-time flow speed based on the real-time light modulation time of the Mueller matrix hyperspectral microscopy imaging system, so that the time for which the marine particles to be identified are irradiated by the polarization modulated light is longer than the light modulation time.

[0163] Optionally, the determination module 520 is specifically configured to obtain an optimal polarization measurement scheme for the optical system according to optical parameters and configuration of the optical system;

[0164] The determination module 520 is further specifically configured to determine the optimal polarization state to be measured based on the polarization state corresponding to the optimal polarization measurement solution.

[0165] Optionally, the processing module 530 is specifically configured to adjust the white light source generated by the illumination light source based on the first polarizer to obtain polarized illumination white light;

[0166] The processing module 530 is further specifically configured to modulate the polarized illumination white light based on the first photoelastic modulator to obtain the polarization state modulated light that satisfies the optimal polarization state to be measured.

[0167] Optionally, the processing module 530 is further specifically configured to select a preset type of condenser based on the parameter to be measured, and different types of condensers have different refraction angles and performances;

[0168] The processing module 530 is further specifically configured to input the polarized illumination white light modulated by the first photoelastic modulator into the condenser, and adjust the polarization measurement angle based on the condenser to obtain the polarization state modulated light.

[0169] Optionally, the processing module 530 is further specifically used to obtain, based on a microscope objective lens, an objective lens light signal after the polarization state modulated light irradiates the marine particle to be identified; wherein the focal length of the microscope objective lens matches the position of the marine particle to be identified on the sample stage;

[0170] The processing module 530 is further specifically used to modulate the objective light signal based on the polarization analyzer system, input the modulated objective light signal into the hyperspectral camera, calculate the Mueller matrix image of each band, and the set of Mueller matrix images of all bands is the Mueller matrix hyperspectral image of each marine particle to be identified.

[0171] Optionally, the processing module 530 is further specifically configured to analyze the objective lens light signal based on the second photoelastic modulator and the second polarizer, and obtain the Mueller matrix image of each band by calculating through the hyperspectral camera.

[0172] Optionally, the identification module 540 is specifically configured to receive the Mueller matrix hyperspectral image of the to-be-identified marine particles obtained by a hyperspectral camera;

[0173] The identification module 540 is specifically used to receive the signal of the hyperspectral camera and then calculate the Mueller matrix hyperspectral image of the marine particles to be identified;

[0174] The identification module 540 is further specifically used to identify the Mueller matrix hyperspectral characteristic curve based on machine learning, and determine the characteristics of the marine particles to be identified based on the identification results.

[0175] Optionally, the identification module 540 is further specifically used to locate specific features in the Mueller matrix hyperspectral characteristic curve based on the machine learning, where the specific features are determined by the physical and chemical properties of each pixel;

[0176] The identification module 540 is further specifically configured to determine the type of the marine particle to be identified based on the specific feature.

[0177] Please continue to refer to Figure 4 The present application also provides a marine particle identification device based on Mueller matrix hyperspectral microscopy imaging, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the steps of any one of the methods provided in the first aspect of the present application are implemented.

[0178] The present application also provides a computer-readable storage medium having a computer program stored thereon, and when the program is executed by a processor, the steps of any one of the methods provided in the present application are implemented.

[0179] The implementation process of the functions and effects of each unit in the above-mentioned device is specifically described in the implementation process of the corresponding steps in the above-mentioned method, and will not be repeated here.

[0180] For the device embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The device embodiment described above is only schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present application scheme. A person of ordinary skill in the art can understand and implement it without paying any creative work.

[0181] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A method for identifying marine particles based on Mueller matrix hyperspectral microscopy, characterized in that: The method comprises: Adjusting the flow speed of the marine particles to be identified; based on the adjusted flow speed, the duration for which the marine particles to be identified are irradiated by the polarization-modulated light is longer than the light modulation time of the light irradiation; The adjusting the flow rate of the marine particles to be identified comprises: Calculating the light modulation time according to the historical imaging time of the polarization state modulated light of the Mueller matrix hyperspectral microscopy imaging system; Calculating the flow velocity of the marine particles based on the light modulation time, wherein the light modulation time is proportional to the flow velocity; The marine particles to be identified flow into the sample stage at the flow rate; Optimizing the real-time flow rate based on the real-time light modulation time of the Mueller matrix hyperspectral microscopic imaging system, so that the time for which the marine particles to be identified are irradiated by the polarization state modulated light is longer than the light modulation time; Determining the optimal polarization state to be measured based on the optical system; The Mueller matrix hyperspectral microscopic imaging system generates corresponding polarization state modulated light based on the optimal polarization state to be measured, illuminates the marine particles to be identified, and obtains the Mueller matrix hyperspectral image of the marine particles to be identified; The Mueller matrix hyperspectral microscopic imaging system comprises an illumination light source, a first polarizer, and a plurality of first photoelastic modulators; the Mueller matrix hyperspectral microscopic imaging system generates corresponding polarization state modulated light based on the optimal polarization state to be measured, and illuminates the marine particles to be identified, including: Adjusting the white light source generated by the illumination light source based on the first polarizer to obtain polarized illumination white light; Modulating the polarized illumination white light based on the first photoelastic modulator to obtain the polarization state modulated light that satisfies the optimal polarization state to be measured; The method of modulating the polarized illumination white light based on the first photoelastic modulator to obtain the polarization state modulated light includes: Selecting a preset type of condenser based on the parameter to be measured, wherein different types of condensers have different refraction angles and performances; Inputting the polarized illumination white light modulated by the first photoelastic modulator into the condenser, adjusting the polarization measurement angle based on the condenser, and obtaining the polarization state modulated light; The polarization state of the polarization-state modulated light includes the optimal polarization state to be measured, and for each pixel, the Mueller matrix images of each band are fused to obtain the Mueller matrix hyperspectral image of the marine particles to be identified, and the Mueller matrix hyperspectral image includes the physical, chemical, and morphological characteristics of the marine particles to be identified; The marine particles to be identified are identified based on the Mueller matrix hyperspectral image.

2. The method according to claim 1, characterized in that The method of determining the optimal polarization state to be measured based on the optical system comprises: Obtaining an optimal polarization measurement scheme for the optical system according to optical parameters and configuration of the optical system; The optimal polarization state to be measured is determined based on the polarization state corresponding to the optimal polarization measurement solution.

3. The method according to claim 1, characterized in that The step of obtaining the Mueller matrix hyperspectral image of the marine particles to be identified comprises: Obtaining an objective lens light signal after the polarization-modulated light irradiates the marine particle to be identified based on a microscope objective lens; wherein the focal length of the microscope objective lens matches the position of the marine particle to be identified on the sample stage; The objective lens light signal is modulated based on the polarization analyzer system, and the modulated objective lens light signal is input into the hyperspectral camera to calculate the Mueller matrix image of each band. The set of the Mueller matrix images of all bands is the Mueller matrix hyperspectral image of each marine particle to be identified.

4. The method according to claim 3, characterized in that The polarization analyzer system comprises a plurality of second photoelastic modulators and a second polarizer; the polarization analyzer system is used to modulate the objective lens light signal, and the modulated objective lens light signal is input into the hyperspectral camera to calculate and obtain the Mueller matrix image of each band, including: The objective lens light signal is polarized based on the second photoelastic modulator and the second polarizer, and the Mueller matrix image of each band is calculated by the hyperspectral camera.

5. The method according to claim 4, characterized in that The step of calculating the Mueller matrix image of each band by the hyperspectral camera includes: Receiving the signal of the hyperspectral camera and then calculating the Mueller matrix hyperspectral image of the marine particles to be identified; For each pixel of the Mueller matrix hyperspectral image, a Mueller matrix hyperspectral characteristic curve is obtained, wherein the abscissa of the Mueller matrix hyperspectral characteristic curve is the band of the spectrum, and the ordinate is the light intensity value of the corresponding pixel under the abscissa band; The Mueller matrix hyperspectral characteristic curve is identified based on machine learning, and the characteristics of the marine particles to be identified are determined based on the identification results.

6. The method according to claim 5, characterized in that The method of identifying the Mueller matrix hyperspectral characteristic curve based on machine learning and determining the characteristics of the to-be-identified marine particles based on the identification result includes: Locating specific features in the Mueller matrix hyperspectral characteristic curve based on the machine learning, wherein the specific features include polarization and spectral features, and the specific features are determined by the physical and chemical properties of each pixel; The type of the marine particulate matter to be identified is determined based on the specific characteristics.

7. A marine particle identification device based on Mueller matrix hyperspectral microscopy imaging, characterized in that: The device includes an adjustment module, a determination module, a processing module and an identification module; wherein, The adjustment module is used to adjust the flow speed of the marine particles to be identified; based on the adjusted flow speed, the duration for which the marine particles to be identified are irradiated by the polarization-modulated light is longer than the light modulation time of the light irradiation; The adjusting the flow rate of the marine particles to be identified comprises: Calculating the light modulation time according to the historical imaging time of the polarization state modulated light of the Mueller matrix hyperspectral microscopy imaging system; Calculating the flow velocity of the marine particles based on the light modulation time, wherein the light modulation time is proportional to the flow velocity; The marine particles to be identified flow into the sample stage at the flow rate; Optimizing the real-time flow rate based on the real-time light modulation time of the Mueller matrix hyperspectral microscopic imaging system, so that the time for which the marine particles to be identified are irradiated by the polarization state modulated light is longer than the light modulation time; The determination module is used to determine the optimal polarization state to be measured based on the optical system; The processing module is used to generate corresponding polarization state modulated light based on the optimal polarization state to be measured, illuminate the marine particles to be identified, and obtain the Mueller matrix hyperspectral image of the marine particles to be identified; The Mueller matrix hyperspectral microscopic imaging system comprises an illumination light source, a first polarizer, and a plurality of first photoelastic modulators; the Mueller matrix hyperspectral microscopic imaging system generates corresponding polarization state modulated light based on the optimal polarization state to be measured, and illuminates the marine particles to be identified, including: Adjusting the white light source generated by the illumination light source based on the first polarizer to obtain polarized illumination white light; Modulating the polarized illumination white light based on the first photoelastic modulator to obtain the polarization state modulated light that satisfies the optimal polarization state to be measured; The method of modulating the polarized illumination white light based on the first photoelastic modulator to obtain the polarization state modulated light includes: Selecting a preset type of condenser based on the parameter to be measured, wherein different types of condensers have different refraction angles and performances; Inputting the polarized illumination white light modulated by the first photoelastic modulator into the condenser, adjusting the polarization measurement angle based on the condenser, and obtaining the polarization state modulated light; The polarization state of the polarization-state modulated light includes the optimal polarization state to be measured, and for each pixel, the Mueller matrix images of each band are fused to obtain the Mueller matrix hyperspectral image of the marine particles to be identified, and the Mueller matrix hyperspectral image includes the physical, chemical, and morphological characteristics of the marine particles to be identified; The identification module is used to identify the marine particles to be identified based on the Mueller matrix hyperspectral image.

Images