A microscopic imaging focusing method and an automatic focusing device

Abstract

This invention discloses a microscopic imaging focusing method and an automatic focusing device, belonging to the field of microscopic optical imaging technology. The microscopic imaging focusing method sequentially includes: preparation, coarse focusing, and fine focusing; the preparation includes calibrating the relationship between the image sharpness of chip markers and the driving current of the liquid lens. This invention is applied to the imaging focusing of complex biological samples containing a large number of impurities, solving the technical problem that existing automatic focusing methods cannot achieve fully automatic, high-speed focusing on complex biological samples within microfluidic chips, enabling fully automatic and rapid focusing for highly complex biological samples.

Description

Technical Field

[0001] This invention belongs to the field of microscopic optical imaging technology, and particularly relates to a microscopic imaging focusing method and an automatic focusing device. Background Technology

[0002] In microscopic optical imaging technology and systems, high-quality images with clear focus not only visually display the morphology of samples, but also serve as the foundation and guarantee for many analyses such as sample identification and classification based on morphological features. However, high-magnification microscopic optical systems have a small depth of field (approximately 1-10 μm), which can easily lead to blurry images due to mechanical micro-vibrations, environmental conditions, and other reasons. As the degree of automation and the demand for clear imaging efficiency continue to increase, higher requirements are placed on the automatic focusing of the system.

[0003] In existing technologies, one automatic focusing method involves controlling the sample stage to move axially within a certain range, acquiring an image corresponding to the current axial position at each step, and using an algorithm to establish a sharpness model to predict the optimal focal plane. Ultimately, the platform moves to the position with the best sharpness as the focal plane to complete focusing. Another automatic focusing method involves keeping the relative positions of the sample and the objective lens constant, inserting a liquid lens between the objective lens and the camera focusing lens. The curvature of the liquid lens changes with the driving current. A predictive model of the relationship between image sharpness and driving current is established through certain means. Based on the current sharpness, the driving current is changed to complete focusing.

[0004] However, current technologies cannot achieve fully automated, high-speed focusing of complex biological samples within microfluidic chips. The reasons are as follows: First, the samples are highly complex. In addition to the sample, the suspension contains a large number of impurities of the same size as the sample that cannot be filtered out. Due to their different physical properties, these impurities and the sample are located on different focal planes. When using the above method for image sharpness analysis, the impurities often dominate the image, making the image sharpness of the impurity focal plane greater than that of the sample focal plane. Ultimately, the focal plane is calculated as the impurity focal plane, and the automatic focusing of the sample fails. Second, due to the diversity of the samples to be tested, the preset image prediction model may have low accuracy and recall or even fail due to sample differences. Therefore, it is necessary to replace the sample and remodel, which is time-consuming, labor-intensive, and requires too much manual intervention, making it impossible to achieve full automation. Finally, because high chip thickness uniformity will lead to a sharp increase in chip cost, when applying the existing scheme one, the total step range needs to be set to be large (>50μm) to cover different chip thickness errors, but this greatly affects efficiency. When applying the existing scheme two, the maximum adjustable focusing range of the 50x microscopic imaging liquid lens is only about 20μm, which cannot cover chip thickness errors. Therefore, neither of the above two schemes can be applied. Summary of the Invention

[0005] Details of one or more embodiments of the present invention are set forth in the following drawings and description to make other features, objects and advantages of the present application more readily apparent.

[0006] This invention proposes a microscopic imaging focusing method and an automatic focusing device, which solves the technical problem that existing automatic focusing methods cannot achieve fully automatic and high-speed focusing on complex biological samples in microfluidic chips, and can perform fully automatic and rapid focusing on highly complex biological samples.

[0007] This invention discloses a microscopic imaging focusing method, comprising: preparation, coarse focusing, and fine focusing. The preparation includes calibrating the relationship between the image sharpness of a chip marker and the driving current of a liquid lens. This calibration involves: positioning the initial reset position of the liquid lens, moving the objective lens one step at a time, recording the objective lens position, and obtaining a defocused image of the chip marker; cutting the defocused image of the chip marker according to a cutting principle to obtain sub-images, calculating the sharpness of the sub-images using a calculation method to obtain a sharpness score; keeping the objective lens position unchanged, changing the driving current of the liquid lens to make the defocused image of the chip marker sharp, ensuring that the driving current A corresponding to the moment the image is sharp corresponds to the sharpness score Q of the original defocused image sub-image; then, resetting the liquid lens, continuing to move the objective lens one step at a time, and repeating the above steps to obtain a series of driving currents that correspond one-to-one with the image sharpness scores, thereby determining the relationship between the image sharpness of the chip marker and the driving current of the liquid lens, and completing the calibration model of the image sharpness of the chip marker and the driving current of the liquid lens.

[0008] The coarse focusing process involves continuously moving the objective lens while simultaneously performing axial scanning with a continuous exposure camera to obtain the correspondence between each image in the scanning video and the objective lens position. At the same time, a set n1 containing the marker points is determined based on the chip marker point recognition model. Each image in set n1 is segmented and analyzed to obtain the image with the highest sharpness score. The objective lens is then moved to the position corresponding to the image with the highest sharpness score, and the image with the highest sharpness score is recorded as the collimated plane image.

[0009] The fine focusing process involves keeping the objective lens position unchanged, adjusting the driving current of the liquid lens focusing assembly based on the sharpness of the image with the highest sharpness score, the sharpness of the chip marker image, and the calibration model of the liquid lens driving current, to obtain the final sharp image of the marker.

[0010] In some embodiments, the chip marker recognition model is established by: collecting a training image set with chip markers, training and learning using a neural network model, and finally obtaining the chip marker recognition model.

[0011] In some embodiments, the initial reset position of the liquid lens is determined by the following method: based on the relationship curve between the driving current and the optical power, the optical power of the liquid lens is initially set to 0 according to the curve, and this is recorded as the initial reset position of the liquid lens.

[0012] In some embodiments, the cutting principle is: the chip marker is located at the center of the sub-image, and the area of ​​the sub-image is twice that of the chip marker; the calculation method is: a no-reference image quality assessment algorithm based on the optimized BRISQUE algorithm to obtain a sharpness score of 0-1.

[0013] In some embodiments, the relationship between the image sharpness of the calibration chip marker and the driving current of the liquid lens is determined by multiple operations, and the average value is taken, or the average value is taken after removing the candidate maximum and minimum values.

[0014] In some embodiments, the coarse focusing specifically includes:

[0015] Starting from the current position of the objective lens, move the objective lens focusing assembly upward by a step length L1, denoted as position A1. Starting from position A1, set the speed of the objective lens focusing assembly to v, and continuously move the objective lens downward for a total step length of L2, where L2 = 2L1. Simultaneously, continuously expose the camera to perform axial flying scan. Take n frames of images at an average interval in the obtained flying scan video. The correspondence between the i-th image and the objective lens position Ai in these n frames is: Ai = (L2 × i) / n.

[0016] Based on the chip marker recognition model, determine whether the n frames of images contain chip markers, and obtain the set n1 containing markers.

[0017] Each image in set n1 is cut according to the cutting principle to obtain a sub-image set n1'. The sharpness of the sub-images is calculated using the calculation method to obtain a sharpness score. The image with the highest score is selected and mapped to image j in set n1 by timestamp index.

[0018] Based on the correspondence between the i-th image and the objective lens position Ai in the n frames, move the objective lens to the position Aj corresponding to image j, and denote image j as the collimated plane image.

[0019] In some embodiments, the process further includes, before coarse focusing: connecting the microfluidic chip to a continuous sample feeding device and placing it on a two-dimensional motorized scanning platform, automatically resetting the platform to align with the first sample cell, turning on the white light source and adaptively adjusting it to a suitable brightness, and resetting the liquid lens focusing assembly; and after fine focusing: moving the objective lens through the objective lens focusing assembly to perform offset compensation, obtaining the sample focal plane position corresponding to the first sample cell, and completing the automatic focusing of the sample cell.

[0020] Another aspect of the present invention discloses an automatic focusing device for the above-mentioned microscopic imaging focusing method, comprising: an illumination optical path, a sample area to be tested, a microscope objective focusing optical path, a first image plane optical path, a 4f relay and liquid lens focusing imaging optical path; the microscope objective focusing optical path includes a microscope objective and an objective focusing assembly that drives the microscope objective to move up and down.

[0021] In some embodiments, the first image plane optical path includes a barrel lens, a first reflecting mirror, and a field lens; the barrel lens focuses the sample image onto the first image plane; the first reflecting mirror bends the optical path by 90°; and the field lens is located at the first image plane.

[0022] In some embodiments, the 4f relay and liquid lens focusing imaging optical path sequentially includes a collimating lens assembly, a second reflecting mirror, a liquid lens focusing assembly, a third reflecting mirror, a focusing mirror assembly, and a camera; the collimating lens assembly collimates the imaging signal of the first image plane; the second reflecting mirror deflects the parallel beam by 90°; the liquid lens focusing assembly changes its curvature by changing its driving current, thereby changing the image distance and obtaining a series of out-of-focus blurred images and in-focus sharp images; the third reflecting mirror deflects the parallel beam by 90°, and the third reflecting mirror and the second reflecting mirror cooperate to ensure that the liquid lens focusing assembly is horizontally fixed; the focusing mirror assembly focuses the imaging signal onto the target surface of the camera, and at the same time forms a 4f relay optical path with the collimating lens assembly; the camera receives the imaging signal.

[0023] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0024] This invention provides a microscopic imaging focusing method. First, the relationship between the image sharpness of chip marker points and the driving current of the liquid lens is calibrated. Then, the focusing range is expanded by continuous movement of the objective lens, while simultaneously using a camera to capture images quickly, achieving coarse focusing. Finally, based on the relationship between the image sharpness of the chip marker points and the driving current of the liquid lens, the driving current of the liquid lens focusing assembly is adjusted to achieve fine focusing of the sample. This method enables rapid and high-precision focusing of complex biological samples containing numerous impurities. Compared to traditional fixed-step scanning methods, the entire process time is reduced from 5-10 seconds to less than 1 second, greatly improving the efficiency of acquiring clear images and increasing focusing speed. Specifically:

[0025] 1. Capable of high-precision focusing on complex biological samples containing numerous impurities. This invention proposes a focusing method that combines coarse focusing by adjusting the object distance using an objective lens focusing assembly with fine focusing by adjusting the image distance using a liquid lens focusing assembly. Compared to traditional focusing methods, this method offers higher precision for layered samples with impurity interference.

[0026] 2. High applicability: No model adjustment is required for different biological samples, enabling full automation of the entire process. The device's built-in recognition and calibration models are based on the physical static recognition of microfluidic chip markers, rather than the dynamic recognition of sample images. This makes it universally applicable to different biological samples and does not require updates for different samples, enabling unattended operation of the entire focusing process.

[0027] 3. Fast focusing speed. This invention proposes an axial scanning recognition method based on objective lens movement. Compared with the traditional fixed step scanning method, the entire process time is shortened from 5-10 seconds to less than 1 second, which greatly improves the efficiency of acquiring clear images and increases the focusing speed. Attached Figure Description

[0028] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this invention, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. In the drawings:

[0029] Figure 1 A flowchart of a microscopic imaging focusing method provided in an embodiment of the present invention;

[0030] Figure 2 This is a schematic diagram illustrating the relationship between driving current and optical power provided in an embodiment of the present invention;

[0031] Figure 3 This is a schematic diagram of the automatic focusing device provided in an embodiment of the present invention;

[0032] Figure 4 This is a schematic diagram of a microfluidic chip provided in an embodiment of the present invention;

[0033] Figure descriptions: 101, White light source; 102, Köhler illumination assembly; 201, Microfluidic chip; 2011, Sample cell; 2012, Marker point; 202, Two-dimensional motorized scanning platform; 301, Microscope objective; 302, Objective focusing assembly; 401, Lens tube lens; 402, First reflecting mirror; 403, Field lens; 501, Collimating lens assembly; 502, Second reflecting mirror; 503, Liquid lens focusing assembly; 504, Third reflecting mirror; 505, Focusing lens assembly; 506, Camera; 6, Computer control system. Detailed Implementation

[0034] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described and illustrated below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. All other embodiments obtained by those skilled in the art based on the embodiments provided by this invention without inventive effort are within the scope of protection of this invention.

[0035] Obviously, the accompanying drawings described below are merely some examples or embodiments of the present invention. Those skilled in the art can apply the present invention to other similar scenarios based on these drawings without any inventive effort. Furthermore, it is understood that although the efforts made in this development process may be complex and lengthy, for those skilled in the art related to the content disclosed in this invention, modifications to design, manufacturing, or production based on the technical content disclosed in this invention are merely conventional technical means and should not be construed as insufficient disclosure of the present invention.

[0036] In this invention, the reference to "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described in this invention may be combined with other embodiments without conflict.

[0037] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "a," "an," "an," "the," and similar words used in this invention do not indicate quantity limitation and may indicate singular or plural. The terms "comprising," "including," "having," and any variations thereof used in this invention are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may also include steps or units not listed, or may include other steps or units inherent to these processes, methods, products, or devices. The terms "connected," "linked," "coupled," and similar words used in this invention are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "A plurality" used in this invention refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships may exist; for example, "A and / or B" can represent: A alone, A and B simultaneously, and B alone. The character " / " generally indicates that the preceding and following objects have an "or" relationship. The terms "first," "second," and "third" used in this invention are merely to distinguish similar objects and do not represent a specific ordering of the objects.

[0038] This invention provides a method for focusing microscopic imaging, the process of which is as follows: Figure 1As shown. The process includes: preparation, coarse focusing, and fine focusing. Preparation includes calibrating the relationship between the image sharpness of chip marker 2012 and the driving current of the liquid lens. This calibration involves: positioning the initial reset position of the liquid lens, moving the objective lens one step at a time, recording the objective lens position, and obtaining the defocused image of chip marker 2012; cutting the defocused image of chip marker 2012 according to the cutting principle to obtain sub-images, calculating the sharpness of the sub-images using a calculation method, and obtaining a sharpness score; keeping the objective lens position unchanged, changing the driving current of the liquid lens to make the defocused image of chip marker 2012 sharp, ensuring that the driving current A corresponding to the moment the image is sharp corresponds to the sharpness score Q of the original defocused image's sub-image; then, resetting the liquid lens, continuing to move the objective lens one step at a time, and repeating the process. The above steps yield a series of driving currents corresponding one-to-one with the image sharpness scores, thereby determining the relationship between the image sharpness of the calibration chip markers and the driving current of the liquid lens, completing the calibration model for the image sharpness of the chip markers and the driving current of the liquid lens. Specifically, the method for making the defocused image sharp is as follows: first, the defocused image is made sharp through subjective judgment; then, the subjectively sharp image is cut using the above cutting principle; the sub-image sharpness is calculated using the above method; the driving current is fine-tuned; and the above process is repeated to obtain the driving current A1 corresponding to the highest image sharpness score. To reduce subjective judgment errors and image sharpness algorithm calculation errors, this process needs to be performed multiple times, obtaining a series of A1, A2, ..., An. The average value is taken to obtain A, or the maximum and minimum values ​​are removed from A1, A2, ..., An, and the average value is taken to obtain A. Finally, A corresponding to Q is obtained. Then, the liquid lens is reset, and the objective lens is moved step by step, repeating the above steps. A series of driving currents corresponding one-to-one with the image sharpness scores are obtained, and then the relationship between the image sharpness of the calibrated chip markers and the driving current of the liquid lens is determined, thus completing the calibration model of the image sharpness of the chip markers and the driving current of the liquid lens.

[0039] Coarse focusing involves continuously moving the objective lens while the continuous exposure camera 506 performs axial flying scans to obtain the correspondence between each image in the flying scan video and the objective lens position. At the same time, based on the recognition model of the chip marker 2012, a set n1 containing marker 2012 is determined. Each image in set n1 is segmented and analyzed to obtain the image with the highest sharpness score. The objective lens is moved to the position corresponding to the image with the highest sharpness score, and the image with the highest sharpness score is recorded as the collimated plane image.

[0040] Precision focusing involves keeping the objective lens position unchanged, adjusting the driving current of the liquid lens focusing assembly 503 based on the sharpness of the image with the highest sharpness score, the sharpness of the chip marker 2012 image, and the calibration model of the liquid lens driving current, to obtain the final sharp image of the marker 2012.

[0041] The aforementioned microscopic imaging focusing method first calibrates the relationship between the image sharpness of the chip marker 2012 and the driving current of the liquid lens. Then, it adjusts the objective lens position through the objective lens focusing assembly 302 for coarse focusing, and finally adjusts the image distance through the liquid lens focusing assembly 503 for fine focusing. Specifically, the continuous movement of the objective lens expands the focusing range, while the camera 506 is used for rapid image capture to locate the sample more quickly, achieving coarse focusing. Then, based on the relationship between the image sharpness of the chip marker 2012 and the driving current of the liquid lens, the driving current of the liquid lens focusing assembly 503 is adjusted to achieve fine focusing of the sample. This method achieves rapid and high-precision focusing of complex biological samples containing a large number of impurities. Compared with the traditional fixed-step scanning method, the entire process time is shortened from 5-10 seconds to less than 1 second, greatly improving the efficiency of acquiring clear images and increasing the focusing speed.

[0042] It should be noted that the aforementioned complex biological samples containing a large number of impurities can be yeast derived from a fermentation system. Therefore, the sample suspension will contain a large number of grain particles or husks from the fermentation system, such as sorghum grains, starch granules, and rice husks. In existing methods for image sharpness analysis, the impurities often dominate the image, resulting in a higher sharpness of the impurity focal plane than the sample focal plane. Consequently, the final focal plane is calculated as the impurity focal plane, leading to the failure of autofocus on the sample. The microscopic imaging focusing method of this invention can achieve high-precision focusing on complex biological samples containing a large number of impurities. It employs a focusing method combining coarse focusing based on the objective lens focusing assembly 302 adjusting the object distance and fine focusing based on the liquid lens focusing assembly 503 adjusting the image distance. For layered samples with impurity interference, this method offers higher accuracy and faster focusing speed compared to traditional focusing methods. Specifically, this invention proposes an axial scanning and recognition method based on objective lens movement. Compared to traditional fixed-step scanning methods, the entire process time is reduced from 5-10 seconds to less than 1 second, significantly improving the efficiency of acquiring clear images and increasing focusing speed.

[0043] In some embodiments, the recognition model for chip marker 2012 is established by acquiring a training image set containing chip marker 2012, training and learning using a neural network model, and finally obtaining the recognition model for chip marker 2012. Specifically, the chip is placed parallel to or at a certain angle relative to the two-dimensional electric platform, and the objective lens is moved by micro-stepping using the objective lens focusing assembly 302 to obtain a series of clear, blurred, and tilted images of marker 2012 as the training image set. The aforementioned neural network model can be YOLO, AlexNet, or ResNet, or can be freely constructed without limitation. Since marker 2012 is an easily recognizable basic image shape and is fixed on the chip glass, it is not affected by sample diversity. The model can be established after the device is assembled and adjusted, and is directly built into the control software on the computer in the device, without needing to be updated with different samples.

[0044] In some embodiments, the initial reset position of the liquid lens is determined by the following method: based on Figure 2 The curve showing the relationship between the driving current and optical power is used to initially set the optical power of the liquid lens to 0, which is recorded as the initial reset position of the liquid lens. Further, the objective lens is moved in minute steps (<0.2 μm) by the objective lens focusing assembly 302.

[0045] In some embodiments, the cutting principle is: the chip marker 2012 is located at the center of the sub-image, and the area of ​​the sub-image is twice that of the chip marker 2012; the calculation method is: a no-reference image quality assessment algorithm based on the optimized BRISQUE algorithm to obtain a sharpness score of 0-1.

[0046] To reduce subjective judgment errors and image sharpness calculation errors, the relationship between the image sharpness of chip marker 2012 and the driving current of the liquid lens is calibrated through multiple operations, and the average value is taken, or the average value is taken after removing the candidate maximum and minimum values. To further reduce errors, the above process is executed twice, sequentially from the initial position of the objective lens to moving the objective lens upwards and downwards. In the above technical solution: keeping the objective lens position unchanged, the driving current of the liquid lens is changed to make the defocused image of chip marker 2012 sharp, resulting in a sharp defocused image of chip marker 2012. The defocused image can be made sharp by subjective judgment, the image is segmented using the above-mentioned segmentation principle, and the sharpness of the sub-images is calculated using the above-mentioned calculation method.

[0047] In some embodiments, coarse focusing specifically includes:

[0048] Starting from the current position of the objective lens, the objective lens focusing assembly 302 moves upward by a step length L1, denoted as position A1. Starting from position A1, the objective lens focusing assembly 302 is set to speed v, and the objective lens is continuously moved downward for a total step length of L2, where L2 = 2L1. At the same time, the camera 506 is continuously exposed to perform axial flying scan. In the obtained flying scan video, n frames of images are taken at an average interval. The correspondence between the i-th image and the objective lens position Ai in these n frames is: Ai = (L2 × i) / n.

[0049] Based on the identification model of chip marker 2012, we can determine whether the n frames of images contain chip marker 2012 and obtain the set n1 containing marker 2012.

[0050] Each image in set n1 is cut according to the cutting principle to obtain a sub-image set n1'. The sharpness of the sub-images is calculated using a calculation method to obtain a sharpness score, and the image with the highest score is selected and mapped to image j in set n1 through a timestamp index. The cutting principle is that the chip marker 2012 is located at the center of the sub-image, and the area of ​​the sub-image is twice that of the chip marker 2012. The calculation method is a no-reference image quality assessment algorithm based on the optimized BRISQUE algorithm to obtain a sharpness score (0-1).

[0051] Based on the correspondence between the i-th image and the objective lens position Ai in n frames, move the objective lens to the position Aj corresponding to image j, and denote image j as the collimated image. In this step, since the acceleration and deceleration process of the objective lens movement is ignored during the rapid scanning process, and there are issues such as uneven transfer after exposure of camera 506 when discretely taking n frames, the obtained position Aj is not the clearest position of marker point 2012, so image j is denoteed as the collimated image.

[0052] In some embodiments, the process further includes: before coarse focusing: connecting the microfluidic chip 201 to a continuous sample feeding device and placing it on a two-dimensional motorized scanning platform 202; the platform automatically resetting and aligning with the first sample cell 2011; turning on the white light source 101 and adaptively adjusting it to a suitable brightness; and resetting the liquid lens focusing assembly 503; and after fine focusing: moving the objective lens through the objective lens focusing assembly 302 to perform offset compensation, obtaining the sample focal plane position corresponding to the first sample cell 2011, and completing the automatic focusing of the sample cell 2011; repeating the above steps to complete the automatic focusing of other required sample cells 2011.

[0053] Another aspect of the present invention discloses an automatic focusing device for the above-described microscopic imaging focusing method, such as... Figure 3As shown, it includes: an illumination optical path, a sample area to be tested, a focusing optical path for the microscope objective 301, a first image plane optical path, and a 4f relay and liquid lens focusing and imaging optical path. The focusing optical path for the microscope objective 301 includes the microscope objective 301 and an objective focusing assembly 302 that moves the microscope objective 301 up and down. The objective focusing assembly 302 realizes the movement of the microscope objective 301 along the Y-axis (while the chip remains stationary), and works with the camera 506 to achieve coarse focusing.

[0054] The above-mentioned illumination path provides suitable bright field illumination for microscopic imaging, including a white light source 101, which provides a light source for bright field imaging; and a Köhler illumination component 102, which is designed based on the Köhler illumination principle and provides a light field with high uniformity.

[0055] The sample area to be tested includes a microfluidic chip 201 containing the sample and a two-dimensional motorized scanning platform 202. The microfluidic chip 201 is fixed to the two-dimensional motorized scanning platform 202 via an adapter plate or directly, ensuring that there is no relative displacement between the chip and the scanning platform during movement. By controlling the two-dimensional motorized scanning platform 202, the microfluidic chip 201 can move in the XY directions to identify samples in different sample cells 2011. The microfluidic chip 201 in the device of this invention is as follows... Figure 4 As shown, the chip contains eight sample cells 2011 channels. It is formed by bonding three layers: the top and bottom layers are glass of different thicknesses, and the middle layer is a PSA layer. The sample cells 2011 channels are created by hollowing out the PSA layer. Each sample cell 2011 has a marker point 2012 next to it. The marker point 2012 can be a basic, easily identifiable image shape such as a circle, triangle, or crosshair, and it is etched onto the upper surface of the lower glass layer using a laser. During the pretreatment stage, the sample is deposited on the upper surface of the lower glass layer of the chip through steps such as centrifugation. Therefore, by identifying the position of the marker point 2012 and focusing, the depth information of the marker point 2012 can be compensated to obtain the focal plane position of the sample. The depth information of the marker point 2012 is a fixed bias. The etching depth of the marker point 2012 can be precisely controlled by controlling the laser exposure power. To avoid accidental errors in compensation, the depth can be measured multiple times and the average value taken. Furthermore, to avoid large or small deviations in the measured values, the maximum and minimum values ​​can be removed before averaging. The measurement method can be to measure directly under a microscope after actual sample injection, or to measure by projection magnification. The detection of this depth offset is also an important part of chip process maturation and quality inspection.

[0056] The focusing optical path of the microscope objective 301 achieves coarse focusing by adjusting the objective position to change the object distance. The microscope objective 301, used for microscopic magnification and imaging, is characterized by high magnification (≥50X), high NA (≥0.8), super-flat field, and apochromatic. The objective focusing assembly 302 can be electrically controlled to move the objective up and down with a stroke of ±5mm, a minimum step value of 0.1μm, and a repeatability accuracy of 1μm during movement.

[0057] In some embodiments, the first image plane optical path includes a barrel lens 401, a first reflecting mirror 402, and a field lens 403. The barrel lens 401 focuses the sample image onto the first image plane; the first reflecting mirror 402 bends the optical path by 90°; and the field lens 403 is located at the first image plane. Specifically, the first image plane optical path achieves infinity-corrected microscopic imaging. It includes the barrel lens 401, which focuses the sample image onto the first image plane, characterized by wide-field imaging, apochromatic aberration, and diffraction-limited performance across the entire field of view. The first reflecting mirror 402 bends the optical path by 90° to compress its volume. The field lens 403, located at the first image plane, does not contribute to the optical power of the entire optical system and therefore does not affect its performance. Its function is to compress the field of view of the off-axis imaging beam to ensure that the aperture of optical devices in subsequent optical paths is not excessively large.

[0058] In some embodiments, the 4f relay and liquid lens focusing imaging optical path sequentially includes a collimating lens assembly 501, a second reflecting mirror 502, a liquid lens focusing assembly 503, a third reflecting mirror 504, a focusing lens assembly 505, and a camera 506; the collimating lens assembly 501 collimates the imaging signal of the first image plane; the second reflecting mirror 502 bends the parallel beam by 90°; the liquid lens focusing assembly 503 changes its curvature by changing its driving current, thereby changing the image distance and obtaining a series of out-of-focus blurred images and in-focus clear images; the third reflecting mirror 504 bends the parallel beam by 90°, and the third reflecting mirror 504 and the second reflecting mirror 502 cooperate to ensure that the liquid lens focusing assembly 503 is horizontally fixed; the focusing lens assembly 505 focuses the imaging signal onto the target surface of the camera 506, and at the same time forms a 4f relay optical path with the collimating lens assembly 501; the camera 506 receives the imaging signal. This technical solution limits the liquid lens focusing assembly 503 to be located between the collimating lens assembly 501 and the focusing lens assembly 505, ensuring that the liquid lens focusing assembly 503 can perform its function while effectively avoiding the impact of the addition of the liquid lens focusing assembly 503 on the optical path.

[0059] Regarding the aforementioned 4f relay and liquid lens focusing imaging optical path, specifically, the functions of the 4f relay and liquid lens focusing imaging optical path are: forming a telecentric optical path to avoid magnification changes caused by focusing; adjusting the image distance through the liquid lens focusing assembly 503 for fine focusing; and receiving the imaging signal using the camera 506. Specifically, it includes: a collimating lens assembly 501 to collimate the imaging signal of the first image plane; a second reflecting mirror 502 to bend the parallel beam by 90° to compress the optical path volume; the liquid lens focusing assembly 503, which changes its curvature through changes in its driving current, thereby changing the image distance and obtaining a series of out-of-focus blurred images and in-focus sharp images; and a third reflecting mirror 504 to bend the parallel beam by 90° to compress the optical path volume, while simultaneously cooperating with the second reflecting mirror 502 to ensure the liquid lens focusing assembly 503 is horizontally fixed, avoiding the issues that occur when vertically fixed. The liquid gravity within the components affects the imaging quality of the lens; the focusing lens assembly 505 focuses the imaging signal onto the target surface of the camera 506, and simultaneously forms a 4f relay optical path with the collimating lens assembly 501, meaning that the focal lengths of the collimating lens assembly 501, the focusing lens assembly 505, and the barrel lens 401 are equal, and the distance between the two components is twice the focal length of the collimating lens assembly 501; the camera 506 is used to receive the imaging signal, and its characteristics are for sample imaging, with features such as a large surface area (≥1 inch), high resolution (>12 million pixels), and high frame rate (>100fps). In this optical path, the reflecting mirror can be a flat mirror, a reflecting prism, or other reflective optical components with the same function; the field lens 403 can be a plano-convex lens, a biconvex lens, or a cemented lens, etc.; the collimating lens assembly 501 and the focusing lens assembly 505 can be plano-convex lenses, cemented lenses, or combined lens groups, etc.

[0060] Furthermore, the aforementioned autofocus device also includes a computer control system 6. The computer control system 6 serves as the control and display center for the entire device, issuing control commands and processes via host computer software. Specifically, it controls the switching and brightness of the white light source 101; controls the speed, direction, step value, and distance parameters of the two-dimensional motorized scanning platform 202; controls the speed, direction, step value, and distance parameters of the objective lens focusing assembly 302; controls the driving current of the liquid lens; and controls the exposure time and white balance parameters of the camera 506. Additionally, this system is also responsible for displaying the image plane.

[0061] The aforementioned automatic focusing device and the microscopic imaging focusing method based on it are highly applicable, requiring no model adjustment for different biological samples and enabling full automation of the entire process. The device's built-in recognition and calibration models are based on the physical static recognition of the microfluidic chip 201 marker points 2012, rather than the dynamic recognition of the sample image. This makes it universally applicable to different biological samples, eliminating the need for updates based on sample type, and enabling unattended operation throughout the focusing process.

[0062] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0063] The above embodiments merely illustrate several implementation methods of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A method for focusing microscopic imaging, characterized in that, In order, they include: Preparation work, including calibrating the relationship between the image sharpness of the chip marker points and the driving current of the liquid lens, includes: Position the liquid lens at its initial reset position, move the objective lens one step at a time, record the objective lens position, and obtain the defocused image of the chip marker. Cut the defocused image of the chip marker according to the cutting principle to obtain sub-images. Calculate the sharpness of the sub-images using a calculation method to obtain a sharpness score. Keeping the objective lens position unchanged, change the driving current of the liquid lens to make the defocused image of the chip marker clear. Make the driving current A corresponding to the moment the image is clear correspond to the sharpness score Q of the original defocused image's sub-image. Then, reset the liquid lens, continue to move the objective lens one step at a time, and repeat the above steps to obtain a series of driving currents that correspond one-to-one with the image sharpness scores. This determines the relationship between the sharpness of the chip marker image and the driving current of the liquid lens, completing the calibration model of the chip marker image sharpness and the driving current of the liquid lens. Coarse focusing includes continuously moving the objective lens while the continuous exposure camera performs axial scanning to obtain the correspondence between each image in the scanning video and the objective lens position. At the same time, a set n1 containing the marker points is determined according to the chip marker point recognition model. Each image in set n1 is segmented and analyzed to obtain the image with the highest sharpness score. The objective lens is moved to the position corresponding to the image with the highest sharpness score, and the image with the highest sharpness score is recorded as the collimated plane image. Fine focusing includes keeping the objective lens position unchanged, adjusting the driving current of the liquid lens focusing assembly based on the sharpness of the image with the highest sharpness score, the sharpness of the chip marker image, and the calibration model of the liquid lens driving current, to obtain the final sharp image of the marker.

2. The microscopic imaging focusing method according to claim 1, characterized in that, The chip marker recognition model is established by the following method: collecting a training image set with chip markers, using a neural network model for training and learning, and finally obtaining the chip marker recognition model.

3. The microscopic imaging focusing method according to claim 1, characterized in that, The initial reset position of the liquid lens is determined by the following method: based on the relationship curve between the driving current and the optical power, the optical power of the liquid lens is initially set to 0 according to the curve, and this is recorded as the initial reset position of the liquid lens.

4. The microscopic imaging focusing method according to claim 1, characterized in that, The cutting principle is as follows: the chip marker is located at the center of the sub-image, and the area of ​​the sub-image is twice that of the chip marker; the calculation method is: a no-reference image quality assessment algorithm based on the optimized BRISQUE algorithm, to obtain a sharpness score of 0-1.

5. The microscopic imaging focusing method according to claim 1, characterized in that, The relationship between the image clarity of the calibration chip marker and the driving current of the liquid lens is determined by performing multiple operations and taking the average value, or by removing the candidate maximum and minimum values ​​and then taking the average value.

6. The microscopic imaging focusing method according to claim 1, characterized in that, The coarse focusing specifically includes: Starting from the current position of the objective lens, move the objective lens focusing assembly upward by a step length L1, denoted as position A1. Starting from position A1, set the speed of the objective lens focusing assembly to v, and continuously move the objective lens downward for a total step length of L2, where L2 = 2L1. Simultaneously, continuously expose the camera to perform axial flying scan. Take n frames of images at an average interval in the obtained flying scan video. The correspondence between the i-th image and the objective lens position Ai in these n frames is: Ai = (L2 × i) / n. Based on the chip marker recognition model, determine whether the n frames of images contain chip markers, and obtain the set n1 containing markers. Each image in set n1 is cut according to the cutting principle to obtain a sub-image set n1'. The sharpness of the sub-images is calculated using the calculation method to obtain a sharpness score. The image with the highest score is selected and mapped to image j in set n1 by timestamp index. Based on the correspondence between the i-th image and the objective lens position Ai in the n frames, move the objective lens to the position Aj corresponding to the image j, and denote the image j as the collimated plane image.

7. The microscopic imaging focusing method according to claim 1, characterized in that, It also includes, before the coarse focusing: connecting the microfluidic chip to the continuous sample feeding device and placing it on the two-dimensional electric scanning platform, the platform automatically resetting and aligning with the first sample cell, turning on the white light source and adaptively adjusting it to a suitable brightness, and resetting the liquid lens focusing assembly; as well as After fine focusing: the objective lens is moved by the objective lens focusing assembly to perform offset compensation, and the sample focal plane position corresponding to the first sample cell is obtained, thus completing the automatic focusing of the sample cell.

8. An automatic focusing device for the microscopic imaging focusing method according to any one of claims 1-7, characterized in that, include: Illumination optical path, sample area to be tested, microscope objective focusing optical path, first image plane optical path, 4f relay and liquid lens focusing imaging optical path; The focusing optical path of the microscope objective includes the microscope objective and a focusing assembly that moves the microscope objective up and down.

9. The automatic focusing device according to claim 8, characterized in that, The first image plane optical path includes: A telescope lens that focuses a sample image onto a first image plane; The first reflecting mirror bends the light path by 90°. A field lens, which is located at the first image plane.

10. The automatic focusing device according to claim 9, characterized in that, The 4f relay and liquid lens focusing imaging optical path includes, in sequence: A collimating lens assembly, wherein the collimating lens assembly collimates the imaging signal of the first image plane; The second reflecting mirror deflects the parallel light beam at a 90° angle. A liquid lens focusing assembly, wherein the curvature of the liquid lens focusing assembly is changed by the change of its driving current, thereby changing the image distance and obtaining a series of out-of-focus blurred images and in-focus clear images; The third reflecting mirror bends the parallel light beam by 90°, and the third reflecting mirror and the second reflecting mirror cooperate to ensure that the liquid lens focusing assembly is horizontally fixed. A focusing lens assembly that focuses the imaging signal onto the target surface of the camera, and together with the collimating lens assembly, forms a 4f relay optical path; A camera that receives imaging signals.

Images