Polarization imaging system, biological sample detection method, electronic equipment and storage medium

Abstract

The invention provides a polarization imaging system, a biological sample detection method, electronic equipment and a storage medium, and belongs to the technical field of imaging detection. The method comprises the following steps: controlling a polarization imaging system to image a to-be-measured biological sample to obtain a to-be-measured light intensity image matrix; correcting a polarization analyzer contained in the polarization imaging system by using the standard sample in combination with a given measurement matrix corresponding to a polarizer contained in the polarization imaging system to obtain a measurement matrix corresponding to the polarization analyzer; according to a given measurement matrix corresponding to the polarizer and a measurement matrix corresponding to the analyzer, converting the to-be-measured light intensity image matrix to obtain a to-be-measured Mueller matrix, and further determining a polarization image set and a polarization feature image set; and processing the polarization image set and the polarization feature image set by using the target detection model to obtain a biological target detection result. According to the invention, deep learning can be carried out on the Mueller matrix generated under the assistance of an equipment calibration mode and the related polarization characteristics reflected by the Mueller matrix, so that reliable target detection is realized.

Description

Technical Field

[0001] The present application relates to the field of imaging detection technology, and in particular to a polarization imaging system and a biological sample detection method, an electronic device and a storage medium. Background Art

[0002] In the existing target polarization imaging detection technology, the polarization imaging system obtains multiple different light intensity images corresponding to multiple different polarization directions by imaging the biological sample to be tested. Usually, the Mueller matrix is ​​determined by combining multiple different light intensity images, and then the Mueller matrix is ​​further analyzed and interpreted based on the experience of technicians to achieve biological target detection. However, since the polarization imaging system is easily affected by environmental factors, component installation errors, and the quality of the components themselves during operation, the imaging quality is poor, making the Mueller matrix obtained directly through image calculation not accurate enough, and the target detection results obtained by subsequent manual intervention are also unreliable. Summary of the invention

[0003] The main purpose of this application is to propose a polarization imaging system and a biological sample detection method, an electronic device and a storage medium, which can be combined with an equipment calibration method to generate a more accurate Mueller matrix, and then use deep learning technology to comprehensively analyze the Mueller matrix and the related polarization characteristics it reflects to quickly obtain more reliable target detection results.

[0004] To achieve the above-mentioned object, one aspect of the present application provides a polarization imaging system, which includes a light source, a collimating lens, a polarizer, a stage, an imaging lens, a non-polarizing beam splitter prism, an analyzer, a relay lens and a camera;

[0005] When a sample is placed on the stage, the light emitted by the light source forms a collimated light beam through the collimating lens, the collimated light beam forms polarized light through the polarizer, the polarized light is incident on the semi-transparent and semi-reflective film in the non-polarizing beam splitter prism to form a 90-degree reflected light, the 90-degree reflected light is incident on the sample placed on the stage through the imaging lens to generate sample reflected light, the sample reflected light is incident on the semi-transparent and semi-reflective film in the non-polarizing beam splitter prism through the imaging lens to form transmitted light, the transmitted light enters the analyzer for polarization analysis, and then is imaged into the camera through the relay lens.

[0006] Further, the polarizer includes a first linear polarizer, a first liquid crystal variable phase retarder and a second liquid crystal variable phase retarder arranged along the light path direction, and the analyzer includes a third liquid crystal variable phase retarder, a fourth liquid crystal variable phase retarder and a second linear polarizer arranged along the light path direction;

[0007] The arrangement direction of the liquid crystal molecules included in the first liquid crystal variable phase retarder is at an angle of 45 degrees to the polarization direction of the first linear polarizer, and the arrangement direction of the liquid crystal molecules included in the second liquid crystal variable phase retarder is parallel to the polarization direction of the first linear polarizer;

[0008] The arrangement direction of the liquid crystal molecules included in the third liquid crystal variable phase retarder is parallel to the polarization direction of the second linear polarizer, and the arrangement direction of the liquid crystal molecules included in the fourth liquid crystal variable phase retarder is at an angle of 45 degrees to the polarization direction of the second linear polarizer;

[0009] The polarization direction of the first linear polarizer is parallel to the polarization direction of the second linear polarizer.

[0010] To achieve the above object, another aspect of the present application provides a biological sample detection method, which is applied to the above polarization imaging system, and the method comprises:

[0011] Controlling the polarization imaging system to image the biological sample to be measured, and obtaining a light intensity image matrix to be measured corresponding to the biological sample to be measured;

[0012] In combination with a given measurement matrix corresponding to the polarizer included in the polarization imaging system, a polarizer included in the polarization imaging system is calibrated using a standard sample to obtain a measurement matrix corresponding to the polarizer;

[0013] According to the given measurement matrix corresponding to the polarizer and the measurement matrix corresponding to the analyzer, the light intensity image matrix to be measured is converted to obtain the Mueller matrix to be measured corresponding to the biological sample to be measured;

[0014] Determining a polarization image set and a polarization feature image set according to the Mueller matrix to be measured;

[0015] The polarization image set and the polarization feature image set are processed using a target detection model to obtain a biological target detection result.

[0016] Furthermore, the controlling the polarization imaging system to image the biological sample to be measured to obtain a light intensity image matrix to be measured includes:

[0017] Determining a plurality of combined polarization states corresponding to the polarization imaging system according to a plurality of different first polarization states corresponding to the polarizer and a plurality of different second polarization states corresponding to the analyzer;

[0018] According to the plurality of combined polarization states, the polarization imaging system is controlled to image the biological sample to be measured, and a plurality of corresponding light intensity images to be measured are obtained to form the light intensity image matrix to be measured.

[0019] Furthermore, the controlling the polarization imaging system to image the biological sample to be measured according to the plurality of combined polarization states to obtain the corresponding plurality of light intensity images to be measured comprises:

[0020] For any one of the combined polarization states, the combined polarization state includes a first polarization state corresponding to the polarizer and a second polarization state corresponding to the analyzer;

[0021] Determining the phase retardation amounts required to be achieved by the first liquid crystal variable phase retarder and the second liquid crystal variable phase retarder included in the polarizer according to the first polarization state corresponding to the polarizer;

[0022] According to the second polarization state corresponding to the polarizer, determining the phase delay amount required to be achieved by the third liquid crystal variable phase retarder and the fourth liquid crystal variable phase retarder included in the polarizer;

[0023] According to the phase retardation amounts required to be achieved by the first liquid crystal variable phase retarder and the second liquid crystal variable phase retarder, adjusting the driving voltages of the first liquid crystal variable phase retarder and the second liquid crystal variable phase retarder;

[0024] According to the phase retardation amounts required to be achieved by the third liquid crystal variable phase retarder and the fourth liquid crystal variable phase retarder, adjusting the driving voltages of the third liquid crystal variable phase retarder and the fourth liquid crystal variable phase retarder;

[0025] The polarization imaging system is controlled to image the biological sample to be measured, so as to obtain the light intensity image to be measured corresponding to the combined polarization state.

[0026] Furthermore, the given measurement matrix corresponding to the polarizer included in the polarization imaging system is combined with the standard sample to calibrate the analyzer included in the polarization imaging system to obtain the measurement matrix corresponding to the analyzer, which includes:

[0027] Controlling the polarization imaging system to image the standard sample to obtain a standard light intensity image matrix;

[0028] The measurement matrix corresponding to the analyzer is determined according to the given measurement matrix corresponding to the polarizer, the given standard Mueller matrix corresponding to the standard sample, and the standard light intensity image matrix.

[0029] Further, determining the polarization image set and the polarization feature image set according to the Mueller matrix to be measured includes:

[0030] Performing pseudo-color imaging on the values ​​of the elements in the first row and the first column contained in the Mueller matrix to be tested to obtain an intensity image;

[0031] Normalizing the Mueller matrix to be tested to obtain a first Mueller matrix to be tested;

[0032] The first Mueller matrix to be measured includes normalized values ​​of a plurality of elements, and pseudo-color imaging is performed on the normalized values ​​of the plurality of elements to obtain a plurality of corresponding polarization images to form the polarization image set;

[0033] Combining and analyzing the normalized values ​​of the plurality of elements to obtain values ​​of a plurality of polarization characteristics;

[0034] Perform pseudo-color imaging on the values ​​of the plurality of polarization characteristics to obtain a plurality of corresponding polarization characteristic images, which are then combined with the intensity image to form the polarization characteristic image set.

[0035] Further, the target detection model is trained based on an improved YOLOv8 model, the improved YOLOv8 model includes an improved backbone network, an improved neck network and an original head network connected in sequence, the improved backbone network is constructed based on an improved CSPDarknet framework, the improved CSPDarknet framework is obtained by introducing a dual-scale attention module into each residual block contained in the original CSPDarknet framework, and the improved neck network includes a sequentially connected spatial pyramid pooling network based on local attention enhancement, a feature pyramid network and a path aggregation network; the target detection model is used to process the polarization image set and the polarization feature image set, and the biological target detection results obtained include:

[0036] Preprocessing the polarization image set and the polarization feature image set to obtain a corresponding first polarization image set and a first polarization feature image set;

[0037] Inputting the first polarization image set and the first polarization feature image set into the improved backbone network for feature extraction to obtain key visual feature information;

[0038] Inputting the key visual feature information into the improved neck network to fuse and encode features of different scales to obtain visual encoding feature information;

[0039] The visual coding feature information is input into the original head network for feature mapping to obtain the biological target detection result.

[0040] To achieve the above objective, another aspect of the present application provides an electronic device, the electronic device comprising a memory and a processor, the memory storing a computer program, and the processor implementing the above method when executing the computer program.

[0041] To achieve the above objective, another aspect of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program implements the above method when executed by a processor.

[0042] The present application includes at least the following beneficial effects: by integrating a polarizer and an analyzer in a polarization imaging system to modulate the polarization state, the polarization imaging system can more efficiently and flexibly obtain multiple images of different light intensities corresponding to multiple polarization directions when imaging a biological sample to be tested. By combining a given measurement matrix corresponding to the polarizer and adopting a device calibration method based on a standard sample to determine the measurement matrix corresponding to the analyzer, the multiple images of different light intensities obtained by the polarization imaging system are further converted to obtain a more accurate Mueller matrix, and then the Mueller matrix and the related polarization characteristics reflected by it are comprehensively analyzed using deep learning technology, so that more reliable target detection results can be quickly obtained with greatly reduced manual intervention. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 is a schematic structural diagram of a polarization imaging system provided in an embodiment of the present application;

[0044] Figure 2 It is a schematic diagram of a biological sample detection method provided in an embodiment of the present application;

[0045] Figure 3 is a schematic diagram of a microsphere polarization image set provided in an embodiment of the present application;

[0046] Figure 4 is a schematic diagram of a set of microsphere polarization characteristic images provided in an embodiment of the present application;

[0047] Figure 5 is a schematic diagram of the microsphere detection results provided in the examples of the present application;

[0048] Figure 6 It is a schematic diagram of the hardware structure of the electronic device provided in the embodiment of the present application. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application is further described in detail below in conjunction with the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the embodiments of the present application. They are only examples of systems and methods consistent with some aspects of the embodiments of the present application as detailed in the attached claims.

[0050] It is understood that the terms "first", "second", etc. used in this application can be used to describe various concepts in this article, but unless otherwise specified, these concepts are not limited by these terms. These terms are only used to distinguish one concept from another concept. For example, without departing from the scope of the embodiment 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 words "if" and "if" as used herein can be interpreted as "at the time of" or "when" or "in response to determination".

[0051] The terms "at least one", "multiple", "each", "any", etc. used in this application, at least one includes one, two or more, multiple includes two or more, each refers to each of the corresponding multiple, and any refers to any one of the multiple.

[0052] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0053] Polarization measurement is an advanced optical technology that uses the polarization state of light waves to obtain information about the structure of an object. Its uniqueness lies in its non-destructive, non-labeling, non-contact and easy-to-expand characteristics, which makes it show significant application advantages in the field of sample observation and analysis. In particular, polarization optics provides a unique solution when analyzing the anisotropic structure of an object.

[0054] In the existing target polarization imaging detection technology, the polarization imaging system obtains multiple different light intensity images corresponding to multiple different polarization directions by imaging the biological sample to be tested. Usually, the Mueller matrix is ​​determined by combining multiple different light intensity images, and then the Mueller matrix is ​​further analyzed and interpreted based on the experience of technicians to achieve biological target detection. However, since the polarization imaging system is easily affected by environmental factors, component installation errors, and the quality of the components themselves during operation, the imaging quality is poor, making the Mueller matrix obtained directly through image calculation not accurate enough, and the target detection results obtained by subsequent manual intervention are also unreliable.

[0055] In the field of polarization optics, the Mueller matrix is ​​an important tool for describing all polarization properties of the measurement target. It contains complete polarization information, thus showing its powerful analytical ability in the study of anisotropic characteristics of micron-level samples. As a method of using the Mueller matrix to explore sample characteristics in detail, Mueller matrix microscopy technology has been widely used in fields such as biomedicine and materials science. At the same time, in the field of modern image processing and pattern recognition, especially in the study of target recognition using neural networks, how to incorporate polarization information into the design of convolutional neural networks to optimize the learning process has become a hot research direction. By decomposing various polarization information such as degree of polarization, polarization purity, and dichroism from the Mueller matrix and using them as important information for analyzing the anisotropic structure of samples and marking samples, complex image information can be simplified, the contrast of similar information can be enhanced, and invisible information can be visualized, thereby greatly enhancing the ability of convolutional neural networks based on polarization information in data mining and feature recognition. It can be seen that convolutional neural networks based on polarization information have great development potential.

[0056] In view of this, the embodiments of the present application provide a polarization imaging system and a biological sample detection method, an electronic device and a storage medium. The scheme integrates a polarizer and an analyzer in the polarization imaging system to modulate the polarization state, so that the polarization imaging system can more efficiently and flexibly obtain multiple images of different light intensities corresponding to multiple polarization directions when imaging the biological sample to be tested. By combining the given measurement matrix corresponding to the polarizer and adopting the equipment calibration method based on the standard sample to determine the measurement matrix corresponding to the analyzer, the multiple images of different light intensities obtained by the polarization imaging system are further converted to obtain a more accurate Mueller matrix, and then the Mueller matrix and the related polarization characteristics reflected by it are comprehensively analyzed using deep learning technology, so that more reliable target detection results can be quickly obtained while greatly reducing manual intervention.

[0057] Figure 1 It is an optional structural schematic diagram of a polarization imaging system provided in an embodiment of the present application, and the polarization imaging system includes a light source 101, a collimating lens 102, a polarizer 103, a non-polarizing beam splitter prism 104, a stage 105, an imaging lens 106, an analyzer 107, a relay lens 108 and a camera 109; wherein the imaging lens 106 preferably adopts a microscope objective lens, and the relay lens 108 preferably adopts a tube lens.

[0058] In a specific implementation process, the light source 101, the collimating lens 102, the polarizer 103, the non-polarizing beam splitter prism 104, the stage 105, the imaging lens 106, the analyzer 107, the relay lens 108 and the camera 109 form a reflective imaging optical path. When a sample is placed on the stage 105, the sample can be a transparent sample or a non-transparent sample. The light emitted by the light source 101 forms a collimated light beam through the collimating lens 102. The light emitted by the light source 101 can be high-brightness quasi-monochromatic light. The straight light beam is formed into polarized light through the polarizer 103, and the polarized light is incident on the semi-transparent and semi-reflective film in the non-polarized beam splitter prism 104 to form a 90-degree reflected light. The 90-degree reflected light is incident on the sample placed on the stage 105 through the imaging lens 106 to generate sample reflected light. The sample reflected light is incident on the semi-transparent and semi-reflective film in the non-polarized beam splitter prism 104 through the imaging lens 106 to form transmitted light. After the transmitted light enters the analyzer 107 for polarization analysis, it is imaged into the camera 109 through the relay lens 108. It can be understood that the normal incidence operation refers to the light incident perpendicular to the component plane.

[0059] By integrating a polarizer and an analyzer in a polarization imaging system to modulate the polarization state, the polarization imaging system can more efficiently and flexibly obtain multiple images of different light intensities corresponding to multiple polarization directions when imaging a biological sample to be measured.

[0060] In some embodiments, the polarizer 103 includes a first linear polarizer, a first liquid crystal variable phase retarder, and a second liquid crystal variable phase retarder arranged along the direction of the optical path. The arrangement direction of the liquid crystal molecules contained in the first liquid crystal variable phase retarder is 45 degrees to the polarization direction of the first linear polarizer, and the arrangement direction of the liquid crystal molecules contained in the second liquid crystal variable phase retarder is parallel to the polarization direction of the first linear polarizer, that is, the arrangement direction of the liquid crystal molecules contained in the second liquid crystal variable phase retarder is 0 degrees to the polarization direction of the first linear polarizer. In actual application, when the polarizer 103 receives the collimated light beam output by the collimating lens 102, the collimated light beam forms linear polarized light through the first linear polarizer, and the linear polarized light changes its own polarization direction or polarization state through the first liquid crystal variable phase retarder and the second liquid crystal variable phase retarder in turn to form the polarized light.

[0061] In some embodiments, the analyzer 107 includes a third liquid crystal variable phase retarder, a fourth liquid crystal variable phase retarder and a second linear polarizer arranged along the optical path direction, the arrangement direction of the liquid crystal molecules contained in the third liquid crystal variable phase retarder is parallel to the polarization direction of the second linear polarizer, that is, the arrangement direction of the liquid crystal molecules contained in the third liquid crystal variable phase retarder is parallel to the polarization direction of the second linear polarizer at an angle of 0 degree, the arrangement direction of the liquid crystal molecules contained in the fourth liquid crystal variable phase retarder is parallel to the polarization direction of the second linear polarizer at an angle of 45 degrees, and the polarization direction of the second linear polarizer is parallel to the polarization direction of the first linear polarizer, that is, the polarization direction of the second linear polarizer is at an angle of 0 degree to the polarization direction of the first linear polarizer; in actual application, when the analyzer 107 receives the transmitted light output by the non-polarizing beam splitter prism 104, the transmitted light is sequentially analyzed by the third liquid crystal variable phase retarder, the fourth liquid crystal variable phase retarder and the second linear polarizer.

[0062] It should be noted that the first liquid crystal variable phase retarder, the second liquid crystal variable phase retarder, the third liquid crystal variable phase retarder and the fourth liquid crystal variable phase retarder are the most common nematic liquid crystal variable phase retarders, which are adapted to the adjustment of any polarization state by setting the arrangement direction of the liquid crystal molecules inside them; and the modulation of different polarization states can be achieved by specifically driving the various liquid crystal variable phase retarders included in the polarizer 103 and the analyzer 107 in an electrically controlled manner, without the need to manually adjust the angles of the polarizer 103 and the analyzer 107 or to configure a stepping motor, thereby providing strong technical support for high-speed polarization imaging and also being beneficial to improving imaging quality.

[0063] In the present application, the polarizer 103 and the analyzer 107 are not limited to being constructed by a linear polarizer and two liquid crystal variable phase retarders with different liquid crystal arrangement directions, but can also be constructed using components such as a rotatable polarizer, a half-wave plate, a quarter-wave plate, a photoelastic modulator, etc.

[0064] In some embodiments, the polarization imaging system may further include a first light source 110, a first collimating lens 111 and a first polarizer 112. Figure 1 As shown, the first light source 110, the first collimating lens 111, the first polarizer 112, the stage 105, the imaging lens 106, the non-polarizing beam splitter prism 104, the analyzer 107, the relay lens 108 and the camera 109 form a transmission imaging optical path; it should be noted that the internal structure and implementation principle of the first polarizer 112 are the same as the internal structure and implementation principle of the polarizer 103, which will not be repeated here.

[0065] In a specific implementation process, when a transparent sample is placed on the stage 105, the light emitted by the first light source 110 forms a first collimated light beam through the first collimating lens 111. The light emitted by the first light source 110 can be high-brightness quasi-monochromatic light. The first collimated light beam forms a first polarized light through the first polarizer 112. The first polarized light is incident on the transparent sample placed on the stage 105 to generate sample transmitted light. The sample transmitted light is incident on the semi-transparent and semi-reflective film in the non-polarizing beam splitter prism 104 through the imaging lens 106 to form a first transmitted light. The first transmitted light enters the analyzer 107 for polarization analysis and is imaged into the camera 109 through the relay lens 108.

[0066] Optionally, before the 90-degree reflected light or the first polarized light is incident on the stage 105 , it is preferentially passed through frosted glass to eliminate some interference.

[0067] By arranging a reflective imaging light path and a transmissive imaging light path in the polarization imaging system, technicians can select the imaging light path according to the properties of the sample, so that the polarization imaging system can be flexibly adapted to different measurement requirements, that is, the scope of application of the polarization imaging system can be expanded.

[0068] In some embodiments, the polarization imaging system may further include an external control device (in Figure 1 ), the external control device is respectively connected to the polarizer 103, the first polarizer 112, the stage 105, the analyzer 107 and the camera 109, and the external control device is used to realize at least the following functions: control the polarization imaging system to image the biological sample to be measured to obtain a light intensity image matrix to be measured; in combination with a given measurement matrix corresponding to the polarizer contained in the polarization imaging system, calibrate the analyzer contained in the polarization imaging system using a standard sample to obtain a measurement matrix corresponding to the analyzer; according to the given measurement matrix corresponding to the polarizer and the measurement matrix corresponding to the analyzer, convert the light intensity image matrix to be measured to obtain a Mueller matrix to be measured; according to the Mueller matrix to be measured, determine a polarization image set and a polarization feature image set; process the polarization image set and the polarization feature image set through a target detection model to obtain a biological target detection result.

[0069] During the process of the external control device controlling the polarization imaging system to image the biological sample to be tested, the external control device needs to complete at least the following tasks: first, select and enable the reflective imaging light path or the transmissive imaging light path according to the attributes of the biological sample to be tested input by the technician. In general, it is preferred to use the reflective imaging light path with a wider range of applications; second, adjust the stage 105 so that the biological sample to be tested can be clearly imaged on the target surface of the camera 109; third, adjust the driving voltage of each liquid crystal variable phase retarder included in the polarizer 103 (or the first polarizer 112) and the analyzer 107 to receive the corresponding light intensity image under different polarization state combinations acquired by the camera 109.

[0070] Figure 2 is an optional flow chart of a biological sample detection method provided in an embodiment of the present application, which is mainly applied to the above-mentioned polarization imaging system. Figure 2 The method may include but is not limited to steps S201 to S205:

[0071] Step S201: Control the polarization imaging system to image the biological sample to be measured, and obtain a light intensity image matrix to be measured corresponding to the biological sample to be measured.

[0072] In this step, the reflective imaging light path or the transmissive imaging light path set in the polarization imaging system is selected and enabled according to the properties of the biological sample to be tested: when the biological sample to be tested is a transparent sample, the reflective imaging light path or the transmissive imaging light path can be used for imaging; when the biological sample to be tested is a non-transparent sample, only the reflective imaging light path can be used for imaging.

[0073] Step S202 : In combination with a given measurement matrix corresponding to the polarizer included in the polarization imaging system, the analyzer included in the polarization imaging system is calibrated using a standard sample to obtain a measurement matrix corresponding to the analyzer.

[0074] Among them, the given measurement matrix corresponding to the polarizer can be understood as the Mueller matrix of the polarizer, which is used to describe how the polarizer converts non-polarized light or partially polarized light into light of a specific polarization state; the measurement matrix corresponding to the analyzer can be understood as the Mueller matrix of the analyzer, which is used to assist the analyzer in analyzing the state of the polarized light obtained by processing by the polarizer.

[0075] Step S203: converting the light intensity image matrix to be measured according to the given measurement matrix corresponding to the polarizer and the measurement matrix corresponding to the analyzer to obtain the Mueller matrix to be measured corresponding to the biological sample to be measured.

[0076] Step S204: Determine a polarization image set and a polarization feature image set according to the Mueller matrix to be measured.

[0077] Step S205: Use the target detection model to process the polarization image set and the polarization feature image set to obtain a biological target detection result.

[0078] Steps S201 to S205 shown in the embodiment of the present application can be combined with the device calibration method to generate a more accurate Mueller matrix, and then use deep learning technology to comprehensively analyze the Mueller matrix and the related polarization characteristics it reflects to quickly obtain more reliable target detection results.

[0079] In some embodiments, the above step S201 may include but is not limited to steps S301 to S302:

[0080] Step S301, determining a plurality of combined polarization states corresponding to the polarization imaging system according to a plurality of different first polarization states corresponding to the polarizer and a plurality of different second polarization states corresponding to the analyzer.

[0081] In this step, when the polarizer is allowed to operate in N polarization states and the analyzer is allowed to operate in M ​​polarization states, N and M are both positive integers greater than 1, then the polarization imaging system is allowed to operate in N×M combined polarization states, wherein each combined polarization state includes a single first polarization state corresponding to the polarizer and a single second polarization state corresponding to the analyzer, and the single first polarization state corresponding to the polarizer and the single second polarization state corresponding to the analyzer may be the same.

[0082] Exemplarily, the multiple different first polarization states corresponding to the polarizer include a horizontal linear polarization state, a vertical linear polarization state, a 135° linear polarization state, a 45° linear polarization state, a right-handed circular polarization state, and a left-handed circular polarization state, and the multiple different second polarization states corresponding to the analyzer also include a horizontal linear polarization state, a vertical linear polarization state, a 135° linear polarization state, a 45° linear polarization state, a right-handed circular polarization state, and a left-handed circular polarization state. Then, the several combined polarization states corresponding to the polarization imaging system include HH, HV, HP, HM, HR, HL, VH, VV, V P, VM, VR, VL, PH, PV, PP, PM, PR, PL, MH, MV, MP, MM, MR, ML, RH, RV, RP, RM, RR, RL, LH, LV, LP, LM, LR and LL; wherein the left side of the horizontal line represents the first polarization state corresponding to the polarizer, the right side of the horizontal line represents the second polarization state corresponding to the analyzer, H refers to the horizontal linear polarization state, V refers to the vertical linear polarization state, P refers to the 135° linear polarization state, M refers to the 45° linear polarization state, R refers to the right-handed circular polarization state, and L refers to the left-handed circular polarization state.

[0083] Step S302: According to the plurality of combined polarization states corresponding to the polarization imaging system, the polarization imaging system is controlled to image the biological sample to be measured, and a plurality of corresponding light intensity images to be measured are obtained to form a light intensity image matrix to be measured.

[0084] In this step, the polarization imaging system is controlled to operate in the i-th combined polarization state, and the biological sample to be tested is imaged to obtain the corresponding i-th light intensity image to be tested, i=1, 2, ..., N×M; according to this implementation method, N×M imaging acquisitions are performed, and then the acquired N×M light intensity images to be tested are combined into the light intensity image matrix to be tested for output.

[0085] In steps S301 to S302 shown in the embodiment of the present application, by adjusting the first polarization state corresponding to the polarizer and the second polarization state corresponding to the analyzer, the polarization imaging system performs diversified imaging of the biological sample to be tested, which is beneficial for the Mueller matrix to be tested obtained by subsequent calculation to cover more structural information about the biological sample to be tested.

[0086] In step S302 of some embodiments, taking any one combined polarization state as an example, the combined polarization state includes the first polarization state corresponding to the polarizer and the second polarization state corresponding to the analyzer, the polarization imaging system is controlled to operate in the combined polarization state, and the biological sample to be measured is imaged to obtain a corresponding light intensity image to be measured. The corresponding implementation process may include but is not limited to steps S401 to S405:

[0087] Step S401: determining the phase retardation amounts required to be achieved by the first liquid crystal variable phase retarder and the second liquid crystal variable phase retarder included in the polarizer according to the first polarization state corresponding to the polarizer.

[0088] In this step, all phase delays corresponding to the polarizer can be obtained by directly querying the first data statistical table. The first data statistical table is used to record the phase delays required to be achieved by all liquid crystal variable phase retarders contained in the polarizer when the polarizer works in different polarization states. The first data statistical table is formulated by technicians based on their own experience. For details, see Table 1. The polarizer and the analyzer are collectively referred to as the polarizer, the first liquid crystal variable phase retarder contained in the polarizer and the fourth liquid crystal variable phase retarder contained in the analyzer are collectively referred to as 45° oriented liquid crystal variable phase retarder, and the second liquid crystal variable phase retarder contained in the polarizer and the third liquid crystal variable phase retarder contained in the analyzer are collectively referred to as 0° oriented liquid crystal variable phase retarder.

[0089] Table 1 The first statistical table

[0090]

[0091] Step S402: determining the phase retardation amounts required to be achieved by the third liquid crystal variable phase retarder and the fourth liquid crystal variable phase retarder included in the analyzer according to the second polarization state corresponding to the analyzer.

[0092] In this step, all phase delays corresponding to the analyzer can also be obtained by directly querying the first data statistical table.

[0093] Step S403: adjusting the driving voltages of the first liquid crystal variable phase retarder and the second liquid crystal variable phase retarder according to the phase retardation amounts required to be achieved by the first liquid crystal variable phase retarder and the second liquid crystal variable phase retarder included in the polarizer.

[0094] In this step, all driving voltages corresponding to the polarizer can be obtained by directly querying the second data statistical table. The second data statistical table is used to record the driving voltage required for each liquid crystal variable phase retarder contained in the polarizer (whether it is a 45° oriented liquid crystal variable phase retarder or a 0° oriented liquid crystal variable phase retarder) when it reaches different phase delay amounts. The second data statistical table is developed by technicians through preliminary experiments, see Table 2 for details.

[0095] Table 2 Second data statistics table

[0096] Phase delay Driving voltage π / 2 3.12V π 2.6V 2π 3.64V

[0097] Step S404: adjusting the driving voltages of the third liquid crystal variable phase retarder and the fourth liquid crystal variable phase retarder included in the analyzer according to the phase retardation amounts required to be achieved by the third liquid crystal variable phase retarder and the fourth liquid crystal variable phase retarder.

[0098] In this step, all driving voltages corresponding to the analyzer can also be obtained by directly querying the second data statistical table.

[0099] Step S405: when the polarizer operates in the first polarization state and the analyzer operates in the second polarization state, control the polarization imaging system to image the biological sample to be measured, and obtain a light intensity image to be measured corresponding to the combined polarization state.

[0100] It should be noted that the phase delay amount required to be achieved by all the liquid crystal variable phase retarders contained in the polarizer and the analyzer can be determined first, and then the driving voltage of all the liquid crystal variable phase retarders contained in the polarizer and the analyzer can be adjusted. Alternatively, the phase delay amount required to be achieved by all the liquid crystal variable phase retarders contained in the polarizer can be determined first and the driving voltage can be adjusted, and then the phase delay amount required to be achieved by all the liquid crystal variable phase retarders contained in the analyzer can be determined and the driving voltage can be adjusted. Alternatively, the phase delay amount required to be achieved by all the liquid crystal variable phase retarders contained in the analyzer can be determined first and the driving voltage can be adjusted, and then the phase delay amount required to be achieved by all the liquid crystal variable phase retarders contained in the analyzer can be determined and the driving voltage can be adjusted. The present application does not make any limitation on this.

[0101] In steps S401 to S405 shown in the embodiment of the present application, the driving voltage of each liquid crystal variable phase retarder contained in the polarizer and the analyzer is adjusted to achieve modulation of different polarization states. There is no need to manually adjust the angles of the polarizer and the analyzer or to configure a stepper motor. This electrical control method can increase the speed of polarization imaging.

[0102] In some embodiments, the source of the second data statistics table is described as follows:

[0103] An experimental optical path is constructed, which includes a second light source, a 0° polarizer, a 90° polarizer and a spectrometer, and a liquid crystal variable phase retarder is placed between the 0° polarizer and the 90° polarizer, and the liquid crystal variable phase retarder can be a 45° oriented liquid crystal variable phase retarder or a 0° oriented liquid crystal variable phase retarder, and the arrangement direction of the liquid crystal molecules contained in the liquid crystal variable phase retarder is at a 45 degree angle to the vibration transmission axis of the 0° polarizer, and similarly, the arrangement direction of the liquid crystal molecules contained in the liquid crystal variable phase retarder is at a 45 degree angle to the vibration transmission axis of the 90° polarizer; in a specific implementation process, the light emitted by the second light source is processed by the 0° polarizer, the liquid crystal variable phase retarder and the 90° polarizer in sequence, and then enters the spectrometer for light intensity measurement;

[0104] By applying different driving voltages to the liquid crystal variable phase retarder, and then using the spectrometer to measure different light intensities, finally generating a first relational expression between light intensity and driving voltage by means of data aggregation and curve fitting, and then obtaining a predetermined second relational expression between phase delay amount and light intensity, and mathematically converting the first relational expression and the second relational expression to obtain a third relational expression between phase delay amount and driving voltage, and finally substituting the phase delay amount required to be achieved by the liquid crystal variable phase retarder into the third relational expression, the driving voltage corresponding to the liquid crystal variable phase retarder can be solved; wherein, the second relational expression is:

[0105]

[0106] In the formula, is the phase delay, and T is the light intensity.

[0107] In the prior art, it is usually considered to determine the Mueller matrix by combining all the light intensity images to be measured contained in the light intensity image matrix after the light intensity image matrix to be measured is acquired. The corresponding implementation method can be described by the following expression:

[0108]

[0109] Where M 0 is the Mueller matrix, YZ is a single light intensity image to be measured contained in the light intensity image matrix to be measured, which can be understood as the light intensity image to be measured obtained by imaging the biological sample to be measured by the polarization imaging system when the polarizer works in the first polarization state Y and the analyzer works in the second polarization state Z, Y=H,V,L,P,M,R, Z=H,V,L,P,M,R;

[0110] However, since the polarization imaging system is easily affected by environmental factors, component installation errors and the quality of the components themselves during operation, the imaging quality is poor, making the Mueller matrix obtained directly through image calculation not accurate enough. Therefore, it is proposed to combine the equipment calibration method to generate a more accurate Mueller matrix, see the above step S202 and the above step S203 for details.

[0111] In some embodiments, the above step S202 may include but is not limited to steps S501 to S502:

[0112] Step S501, control the polarization imaging system to image the standard sample to obtain a standard light intensity image matrix; wherein, the standard sample is preferably air, and the given standard Mueller matrix corresponding to the air is generally a 4×4 unit matrix. The entire imaging acquisition process is similar to the above step S201 and will not be repeated here.

[0113] Step S502, determining the measurement matrix corresponding to the analyzer according to the given measurement matrix corresponding to the polarizer, the given standard Mueller matrix corresponding to the standard sample and the standard light intensity image matrix.

[0114] In this step, we first need to understand the basic imaging principle of the polarization imaging system: natural light After being incident on the polarizer and modulated with different polarization states for multiple times, the sample on the stage is irradiated, and the outgoing light is modulated by the analyzer for multiple times corresponding to different polarization states to obtain multiple light intensity images to form a light intensity image matrix. The corresponding implementation method can be described by the following first expression:

[0115]

[0116] In the formula, I 0 is the intensity image matrix, A 0 is the initial measurement matrix corresponding to the polarization analyzer, M 0 is the Mueller matrix corresponding to the sample, P 0 is the initial measurement matrix corresponding to the polarizer, T represents the matrix transpose symbol;

[0117] Then, the given measurement matrix corresponding to the polarizer, the given standard Mueller matrix corresponding to the standard sample, and the standard light intensity image matrix corresponding to the standard sample are substituted into the above first expression for conversion, and the measurement matrix corresponding to the analyzer can be solved as follows:

[0118]

[0119] In the formula, A 1 is the measurement matrix corresponding to the polarizer, M 1 is the given standard Mueller matrix corresponding to the standard sample, P 1is the given measurement matrix corresponding to the polarizer, I 1 is the standard light intensity image matrix corresponding to the standard sample, and -1 represents the inverse sign of the matrix.

[0120] In step S203 of some embodiments, the given measurement matrix corresponding to the polarizer, the measurement matrix corresponding to the analyzer, and the measured light intensity image matrix corresponding to the biological sample to be tested are substituted into the above first expression for conversion, and the Mueller matrix to be tested corresponding to the biological sample to be tested can be solved as follows:

[0121]

[0122] Where M 2 is the Mueller matrix corresponding to the biological sample to be tested, I 2 is the light intensity image matrix corresponding to the biological sample to be tested, M ij are the elements contained in the Mueller matrix to be tested corresponding to the biological sample to be tested, i=1,2,3,4, j=1,2,3,4. It can be seen that the Mueller matrix to be tested corresponding to the biological sample to be tested contains 16 elements.

[0123] In some embodiments, the above step S204 may include but is not limited to steps S601 to S605:

[0124] Step S601: Perform pseudo-color imaging on the values ​​of the elements that are included in the Mueller matrix to be tested and that fall simultaneously in the first row and the first column, corresponding to the biological sample to be tested, to obtain an intensity image.

[0125] In this step, the element M contained in the Mueller matrix to be tested corresponding to the biological sample to be tested 11 The numerical value of is converted into the intensity image represented by a specific color through a first pseudo color mapping bar, and the first pseudo color mapping bar represents the transition from black to white.

[0126] Step S602: normalize the Mueller matrix to be tested corresponding to the biological sample to be tested to obtain a first Mueller matrix to be tested.

[0127] In this step, the values ​​of all elements contained in the Mueller matrix to be tested corresponding to the biological sample to be tested are compared with the values ​​of the elements contained in the Mueller matrix to be tested that fall in the first row and the first column at the same time (i.e., the elements M 11 The normalization operation is completed by dividing the value of the polarization information corresponding to the biological sample to be tested, so as to facilitate the subsequent extraction of the relevant polarization information corresponding to the biological sample to be tested. The normalization formula used in this implementation process is as follows:

[0128]

[0129] Where M 3is the first Mueller matrix to be tested, that is, the normalized result of the Mueller matrix to be tested corresponding to the biological sample to be tested, m ij is the normalized value of the elements contained in the first Mueller matrix to be tested.

[0130] Step S603: performing pseudo-color imaging on the normalized values ​​of a plurality of elements included in the first Mueller matrix to be measured, and obtaining a plurality of corresponding polarization images to form a polarization image set.

[0131] In this step, for the normalized value of any element contained in the first Mueller matrix to be measured, the normalized value of the element is converted into a polarization image represented by a specific color through a second pseudo color mapping bar, and the second pseudo color mapping bar represents the transition from blue to red.

[0132] Step S604: performing combined analysis on normalized values ​​of a plurality of elements contained in the first Mueller matrix to be measured to obtain values ​​of a plurality of polarization characteristics.

[0133] In this step, the multiple polarization characteristics obtained by analyzing the first Mueller matrix to be measured may include, but are not limited to, the first type of anisotropy angular orientation, the second type of anisotropy angular orientation, the standardized anisotropy, the extinction and the polarization degree. The calculation formulas for each polarization characteristic are as follows:

[0134]

[0135]

[0136] In the formula, X 3 is the first type of anisotropic angular orientation, X is the second type of anisotropic angular orientation, D is the extinction, P is the polarization degree, A is the standardized anisotropy, b and t 1 It is a reference parameter set for the convenience of description.

[0137] Step S605 , performing pseudo-color imaging on the multiple polarization feature values ​​obtained by analysis to obtain multiple corresponding polarization feature images, and then forming a polarization feature image set together with the intensity image.

[0138] In this step, for any polarization feature value obtained through analysis, the polarization feature value is converted into a polarization feature image represented by a specific color through the second pseudo color mapping bar.

[0139] Steps S601 to S605 shown in the embodiment of the present application are to decompose the normalized Mueller matrix to be tested corresponding to the biological sample to be tested to quickly extract the key polarization features, and then convert the normalized Mueller matrix to be tested and the key polarization features into image form for subsequent input into the deep learning model, which is beneficial to enhancing the multidimensionality and depth of the model analysis.

[0140] In step S205 of some embodiments, the target detection model is trained based on an improved YOLOv8 (You Only LookOnce v8) model. Since the original YOLOv8 model is not specially optimized for a specific target object, it is unable to fully explore and utilize the unique morphological features of the target object, such as shape, texture, edge details, etc., resulting in a low accuracy rate of the final detection result. Therefore, the present application proposes to make some improvements to the original YOLOv8 model. The improved YOLOv8 model includes an improved backbone network, an improved neck network and an original head network connected in sequence. The structural composition of each network is described below.

[0141] Specifically, the improved backbone network is constructed based on the improved CSPDarknet (Cross Stage PartialDarknet) framework. It is known that the original CSPDarknet framework is composed of multiple convolutional layers and multiple residual blocks connected alternately. The improved CSPDarknet framework is generated by introducing a bi-scale attention module (BSAM, Bi-Scale Attention Module) in each residual block contained in the original CSPDarknet framework. Through the improved backbone network, key features can be efficiently extracted from the input image, while reducing the network parameters and computational complexity.

[0142] Taking any residual block as an example, a dual-scale attention module is connected to the output end of the residual block to complete the improvement of the residual block. The dual-scale attention module includes a first convolutional layer, a dual-scale attention mechanism layer and a weighted layer. The first convolutional layer is used to process the output of the residual block to obtain first feature information, and the dual-scale attention mechanism layer is used to process the output of the residual block to obtain second feature information. The weighted layer is used to combine the first feature information and the second feature information by addition or splicing and then output.

[0143] By introducing the dual-scale attention module, the model can pay more attention to the feature differences between small and large targets when processing images, thereby effectively improving the model's ability to recognize objects of various sizes. Especially in biological sample image analysis, the use of the dual-scale attention module can help the model more accurately distinguish biological samples of different sizes and materials, and maintain a high recognition rate even in dim light or complex background conditions.

[0144] Specifically, the improved neck network includes a sequentially connected spatial pyramid pooling network (SPPELAN, Spatial Pyramid Pooling Enhanced with Local Attention Network), a feature pyramid network (FPN, Feature Pyramid Network) and a path aggregation network (PAN, Path Aggregation Network). By combining spatial pyramid pooling (SPP) and local attention mechanism in the spatial pyramid pooling network based on local attention enhancement, it is possible to extract feature maps of fixed size in windows of different scales, and effectively retain spatial information and scale invariance; by combining the feature pyramid network and the path aggregation network, the features of different scales can be fused and encoded using bottom-up and top-down paths. The improved neck network can enhance the model's detection ability for targets of different scales, so that the model can more accurately handle targets of various sizes.

[0145] By introducing a spatial pyramid pooling network based on local attention enhancement, the model can perform feature analysis at different levels from local to global, thereby enhancing the control ability of fine-grained features. Especially when processing high-density biological sample images, such as densely distributed microspheres in water samples, the spatial pyramid pooling network based on local attention enhancement can accurately identify and separate multiple closely adjacent microspheres through the local attention mechanism.

[0146] Specifically, the original head network refers to the head network included in the original YOLOv8 model, which includes multiple convolutional layers and is mainly responsible for using an adaptive anchor box generation strategy to map the features output by the improved neck network encoding into specific target detection results. The target detection results include an intensity image annotated with relevant detection information, that is, the target bounding box, the position information of the target bounding box, the category information of the target object contained in the target bounding box, and the category prediction probability are annotated on the intensity image.

[0147] In some embodiments, the above step S205 may include but is not limited to steps S701 to S702:

[0148] Step S701, preprocessing the polarization image set and the polarization feature image set converted by the Mueller matrix to be tested corresponding to the biological sample to be tested, to obtain a corresponding first polarization image set and a first polarization feature image set.

[0149] In this step, the pixel values ​​of each polarization image contained in the polarization image set and each polarization feature image contained in the polarization feature image set can be normalized to improve the stability of the numerical calculation. The visual parameters of each polarization image contained in the polarization image set and each polarization feature image contained in the polarization feature image set can also be adjusted to enhance the key biological morphological features such as edges and textures contained in the images. The visual parameters include at least one of contrast parameters, brightness parameters, and sharpness parameters. The present application does not make specific limitations on the image preprocessing method here. Only basic pixel value normalization can be performed, and visual parameter adjustment can continue.

[0150] Step S702, input the first polarization image set and the first polarization feature image set into the target detection model for processing, specifically: input the first polarization image set and the first polarization feature image set into the improved backbone network for feature extraction to obtain key visual feature information; input the key visual feature information into the improved neck network to fuse and encode features of different scales to obtain visual coding feature information; input the visual coding feature information into the original head network for feature mapping to obtain biological target detection results.

[0151] In this application, the improved YOLOv8 model is trained to obtain the target detection model, and the corresponding implementation process may include but is not limited to the following steps:

[0152] Acquire multiple training samples corresponding to multiple tested biological samples, each of which carries biological individuals of known categories, and each training sample corresponding to the tested biological sample includes a set of polarization images to be analyzed and a set of polarization feature images to be analyzed, and each training sample corresponding to the tested biological sample carries a real biological individual category label;

[0153] Preprocessing and data enhancement are performed on the multiple training samples. The corresponding preprocessing operations include basic image pixel value normalization and further image visual parameter adjustment. The corresponding data enhancement operations include image rotation, scaling, cropping and flipping, etc., to adapt to practical application scenarios such as biological individuals of different sizes carried by the biological sample itself and the position of the biological individuals carried by the biological sample changes when the biological sample is imaged, thereby obtaining a larger number of training samples;

[0154] The training samples are divided into a training set, a validation set and a test set in a ratio of 8:1:1, and the training samples contained in the training set, the validation set and the test set are not repeated;

[0155] The improved YOLOv8 model is trained multiple times using the training set, with the initial learning rate set to 0.001 and the training batch size set to 32. In each iterative training process, the output prediction result is calculated by forward propagation, and the Focal Loss function is used as the loss function to measure the error between the prediction result and the true label. The model parameters are then updated by the Ranger optimizer to minimize the loss function. The validation set is used to adjust the hyperparameters of the improved YOLOv8 model by back propagation after each iterative training to prevent the model from overfitting on the training set. The test set is used to evaluate the performance of the current improved YOLOv8 model after each iterative training. The model detection performance can be quantified through multiple evaluation indicators such as accuracy, recall, F1 score, mean average precision (mAP), etc., thereby obtaining the final target detection model.

[0156] The biological sample detection method proposed in the embodiment of the present application combines the given measurement matrix corresponding to the polarizer and adopts the equipment calibration method based on the standard sample to determine the measurement matrix corresponding to the analyzer, so as to further convert the multiple different light intensity images obtained by the polarization imaging system to obtain a more accurate Mueller matrix, and then use deep learning technology to comprehensively analyze the Mueller matrix and the related polarization characteristics it reflects, so as to quickly obtain more reliable target detection results while greatly reducing human intervention.

[0157] The following is a further description of the solution provided in the embodiment of the present application in conjunction with a specific application example, which specifically includes the following steps:

[0158] The polarization imaging system is controlled to image a biological sample prepared in advance, and the biological sample carries microsphere individuals of unknown categories to obtain a light intensity image matrix; the analyzer included in the polarization imaging system is calibrated using a standard sample in combination with a given measurement matrix corresponding to the polarizer included in the polarization imaging system to obtain a measurement matrix corresponding to the analyzer; the light intensity image matrix is ​​converted according to the given measurement matrix corresponding to the polarizer and the measurement matrix corresponding to the analyzer to obtain a microsphere Mueller matrix; the values ​​of the elements included in the microsphere Mueller matrix that fall in the first row and the first column at the same time are subjected to pseudo-color imaging to obtain an intensity image; the microsphere Mueller matrix is ​​normalized to obtain a first microsphere Mueller matrix; the first microsphere Mueller matrix is ​​subjected to pseudo-color imaging to obtain the following: Figure 3The microsphere polarization image set shown; the first microsphere Mueller matrix is ​​transformed and analyzed to obtain the values ​​of multiple microsphere polarization characteristics and further complete the pseudo-color imaging, and then together with the intensity image to form as shown Figure 4 The microsphere polarization feature image set shown in FIG. 1 is processed by the target detection model to obtain the microsphere polarization feature image set and the microsphere polarization feature image set. Figure 5 In the microsphere detection result shown, PS indicates that the model predicts that the microsphere in the bounding box belongs to the category, PS is a pre-set category code, and the number after PS indicates the probability value that the model predicts that the microsphere in the bounding box belongs to the category PS.

[0159] The embodiment of the present application also provides an electronic device, which includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the above-mentioned biological sample detection method when executing the computer program. The electronic device may include any intelligent terminal such as a tablet computer and a car computer.

[0160] It can be understood that the contents of the above method embodiments are all applicable to the present device embodiments, the functions specifically implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0161] See also Figure 6 , Figure 6 The hardware structure of an electronic device according to another embodiment is shown. The electronic device includes:

[0162] The processor 801 may be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present application;

[0163] The memory 802 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 802 can store an operating system and other application programs. When the technical solution provided in the embodiment of the present application is implemented by software or firmware, the relevant program code is stored in the memory 802, and the processor 801 calls and executes the technical solution provided in the embodiment of the present application;

[0164] Input / output interface 803, used to implement information input and output;

[0165] The communication interface 804 is used to realize the communication interaction between the device and other devices. The communication can be realized through a wired manner (such as USB, network cable, etc.) or a wireless manner (such as mobile network, WIFI, Bluetooth, etc.);

[0166] A bus 805 that transmits information between the various components of the device (e.g., the processor 801, the memory 802, the input / output interface 803, and the communication interface 804);

[0167] The processor 801 , the memory 802 , the input / output interface 803 and the communication interface 804 are connected to each other in communication within the device via a bus 805 .

[0168] An embodiment of the present application further provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the above-mentioned biological sample detection method is implemented.

[0169] It can be understood that the contents of the above method embodiments are all applicable to the present storage medium embodiments, the functions specifically implemented by the present storage medium embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0170] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0171] The embodiments described in the embodiments of the present application are intended to more clearly illustrate the technical solutions of the embodiments of the present application and do not constitute a limitation on the technical solutions provided in the embodiments of the present application. Those skilled in the art will appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0172] Those skilled in the art will appreciate that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present application, and may include more or fewer steps than shown in the figures, or a combination of certain steps, or different steps.

[0173] The system embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place or 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 embodiment.

[0174] Those skilled in the art will appreciate that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices may be implemented as software, firmware, hardware, or a suitable combination thereof.

[0175] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0176] It should be understood that in the present application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the objects associated before and after are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0177] In the several embodiments provided in the present application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are only schematic. For example, the division of the above units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of systems or units, which can be electrical, mechanical or other forms.

[0178] The units described above as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0179] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0180] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including multiple instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, referred to as ROM), random access memory (Random Access Memory, referred to as RAM), disk or optical disk and other media that can store programs.

[0181] The preferred embodiments of the present invention are described above with reference to the accompanying drawings, but the scope of the rights of the present invention is not limited thereto. Any modification, equivalent substitution and improvement made by a person skilled in the art without departing from the scope and essence of the present invention should be within the scope of the rights of the present invention.

Claims

1. A polarization imaging system, characterized in that: The polarization imaging system comprises a light source, a collimating lens, a polarizer, a stage, an imaging lens, a non-polarizing beam splitter prism, an analyzer, a relay lens and a camera; When a sample is placed on the stage, the light emitted by the light source forms a collimated light beam through the collimating lens, the collimated light beam forms polarized light through the polarizer, the polarized light is incident on the semi-transparent and semi-reflective film in the non-polarizing beam splitter prism to form a 90-degree reflected light, the 90-degree reflected light is incident on the sample placed on the stage through the imaging lens to generate sample reflected light, the sample reflected light is incident on the semi-transparent and semi-reflective film in the non-polarizing beam splitter prism through the imaging lens to form transmitted light, the transmitted light enters the analyzer for polarization analysis, and then is imaged into the camera through the relay lens.

2. The polarization imaging system according to claim 1, characterized in that: The polarizer includes a first linear polarizer, a first liquid crystal variable phase retarder and a second liquid crystal variable phase retarder arranged along the light path direction, and the analyzer includes a third liquid crystal variable phase retarder, a fourth liquid crystal variable phase retarder and a second linear polarizer arranged along the light path direction; The arrangement direction of the liquid crystal molecules included in the first liquid crystal variable phase retarder is at an angle of 45 degrees to the polarization direction of the first linear polarizer, and the arrangement direction of the liquid crystal molecules included in the second liquid crystal variable phase retarder is parallel to the polarization direction of the first linear polarizer; The arrangement direction of the liquid crystal molecules included in the third liquid crystal variable phase retarder is parallel to the polarization direction of the second linear polarizer, and the arrangement direction of the liquid crystal molecules included in the fourth liquid crystal variable phase retarder is at an angle of 45 degrees to the polarization direction of the second linear polarizer; The polarization direction of the first linear polarizer is parallel to the polarization direction of the second linear polarizer.

3. A biological sample detection method, characterized in that: Applied to the polarization imaging system as claimed in claim 2, the method comprises: Controlling the polarization imaging system to image the biological sample to be measured, and obtaining a light intensity image matrix to be measured corresponding to the biological sample to be measured; In combination with a given measurement matrix corresponding to the polarizer included in the polarization imaging system, a polarizer included in the polarization imaging system is calibrated using a standard sample to obtain a measurement matrix corresponding to the polarizer; According to the given measurement matrix corresponding to the polarizer and the measurement matrix corresponding to the analyzer, the light intensity image matrix to be measured is converted to obtain the Mueller matrix to be measured corresponding to the biological sample to be measured; Determining a polarization image set and a polarization feature image set according to the Mueller matrix to be measured; The polarization image set and the polarization feature image set are processed using a target detection model to obtain a biological target detection result.

4. The biological sample detection method according to claim 3, characterized in that: The controlling the polarization imaging system to image the biological sample to be measured to obtain a light intensity image matrix to be measured comprises: Determining a plurality of combined polarization states corresponding to the polarization imaging system according to a plurality of different first polarization states corresponding to the polarizer and a plurality of different second polarization states corresponding to the analyzer; According to the plurality of combined polarization states, the polarization imaging system is controlled to image the biological sample to be measured, and a plurality of corresponding light intensity images to be measured are obtained to form the light intensity image matrix to be measured.

5. The biological sample detection method according to claim 4, characterized in that: According to the plurality of combined polarization states, controlling the polarization imaging system to image the biological sample to be measured to obtain the corresponding plurality of light intensity images to be measured comprises: For any one of the combined polarization states, the combined polarization state includes a first polarization state corresponding to the polarizer and a second polarization state corresponding to the analyzer; Determining the phase retardation amounts required to be achieved by the first liquid crystal variable phase retarder and the second liquid crystal variable phase retarder included in the polarizer according to the first polarization state corresponding to the polarizer; According to the second polarization state corresponding to the polarizer, determining the phase delay amount required to be achieved by the third liquid crystal variable phase retarder and the fourth liquid crystal variable phase retarder included in the polarizer; According to the phase retardation amounts required to be achieved by the first liquid crystal variable phase retarder and the second liquid crystal variable phase retarder, adjusting the driving voltages of the first liquid crystal variable phase retarder and the second liquid crystal variable phase retarder; According to the phase retardation amounts required to be achieved by the third liquid crystal variable phase retarder and the fourth liquid crystal variable phase retarder, adjusting the driving voltages of the third liquid crystal variable phase retarder and the fourth liquid crystal variable phase retarder; The polarization imaging system is controlled to image the biological sample to be measured, so as to obtain the light intensity image to be measured corresponding to the combined polarization state.

6. The biological sample detection method according to claim 3, characterized in that: The step of combining a given measurement matrix corresponding to the polarizer included in the polarization imaging system and calibrating the analyzer included in the polarization imaging system using a standard sample to obtain a measurement matrix corresponding to the analyzer includes: Controlling the polarization imaging system to image the standard sample to obtain a standard light intensity image matrix; The measurement matrix corresponding to the analyzer is determined according to the given measurement matrix corresponding to the polarizer, the given standard Mueller matrix corresponding to the standard sample, and the standard light intensity image matrix.

7. The biological sample detection method according to claim 3, characterized in that: The determining of the polarization image set and the polarization feature image set according to the Mueller matrix to be measured comprises: Performing pseudo-color imaging on the values ​​of the elements in the first row and the first column contained in the Mueller matrix to be tested to obtain an intensity image; Normalizing the Mueller matrix to be tested to obtain a first Mueller matrix to be tested; The first Mueller matrix to be measured includes normalized values ​​of a plurality of elements, and pseudo-color imaging is performed on the normalized values ​​of the plurality of elements to obtain a plurality of corresponding polarization images to form the polarization image set; Combining and analyzing the normalized values ​​of the plurality of elements to obtain values ​​of a plurality of polarization characteristics; Perform pseudo-color imaging on the values ​​of the plurality of polarization characteristics to obtain a plurality of corresponding polarization characteristic images, which are then combined with the intensity image to form the polarization characteristic image set.

8. The biological sample detection method according to claim 3, characterized in that: The target detection model is obtained by training based on an improved YOLOv8 model, wherein the improved YOLOv8 model includes an improved backbone network, an improved neck network and an original head network connected in sequence, wherein the improved backbone network is constructed based on an improved CSPDarknet framework, wherein the improved CSPDarknet framework is obtained by introducing a dual-scale attention module into each residual block contained in the original CSPDarknet framework, wherein the improved neck network includes a spatial pyramid pooling network based on local attention enhancement, a feature pyramid network and a path aggregation network connected in sequence; wherein the target detection model is used to process the polarization image set and the polarization feature image set, and the biological target detection results obtained include: Preprocessing the polarization image set and the polarization feature image set to obtain a corresponding first polarization image set and a first polarization feature image set; Inputting the first polarization image set and the first polarization feature image set into the improved backbone network for feature extraction to obtain key visual feature information; Inputting the key visual feature information into the improved neck network to fuse and encode features of different scales to obtain visual encoding feature information; The visual coding feature information is input into the original head network for feature mapping to obtain the biological target detection result.

9. An electronic device, characterized in that: The electronic device comprises a memory and a processor, the memory stores a computer program, and the processor implements the method according to any one of claims 3 to 8 when executing the computer program.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 3 to 8 is implemented.

Images