Microscopic image acquisition condition reproduction and imaging method

Abstract

The present application provides a kind of microscopic image acquisition condition reproduction and imaging method, comprising: selecting objective and reading its code value, aperture stop opening degree, aperture stop opening degree, light source brightness and color temperature are calculated according to objective numerical aperture operation matching field diaphragm opening degree, acquisition reference image is encapsulated as standard imaging parameter file together with above-mentioned parameter;Call standard imaging parameter file, and drive objective conversion mechanism, field diaphragm, aperture stop, light source and observation switching module to file record working condition in parallel;By setting code type objective conversion mechanism, electric field diaphragm, electric aperture stop, adjustable color temperature light source and observation switching module and by driving unit parallel driving, the one-key linkage reset of objective magnification, double-aperture stop opening degree, light source brightness color temperature and observation station and other multiple imaging components is realized, avoid the same batch imaging parameter dispersion caused by the working condition of multiple components in existing scheme cannot be reproduced after manual adjustment, replacement operator or cross working day detection.

Description

Technical Field

[0001] This invention relates to the field of microscopic imaging, and more particularly to a method for reproducing microscopic image acquisition conditions and imaging. Background Technology

[0002] With the intelligent upgrading of the microscopic characterization industry, fully automated microscopic analysis based on artificial intelligence is widely used in scenarios such as metal grain statistics, pathological cell screening, and powder particle size detection. The model training and online inference of this type of analysis are highly dependent on the feature consistency of the input image. Once the same microstructure is imaged at different times and under different operating conditions, its brightness, depth of field, contrast, and field of view will vary, directly causing missed detections and misjudgments in artificial intelligence analysis.

[0003] An existing automated microscope control scheme improves the consistency of image brightness by memorizing and restoring the brightness parameters of the light source in the device, so that the light intensity can return to the set value after each power-on. At the image output end, post-processing algorithms such as brightness equalization and histogram normalization are used to further correct the brightness deviation of the acquired image.

[0004] However, in this scheme, the equipment only memorizes the single parameter of light source brightness. Multiple imaging components, such as objective lens magnification, field stop opening, aperture stop opening, and observation position, still rely on manual adjustment by the operator. After changing operators or testing across workdays, the working conditions of the aforementioned components cannot be accurately reproduced, and the imaging parameters of the same batch of samples become discrete. At the same time, the image post-processing algorithm can only correct the brightness deviation of the image, and cannot compensate for the physical differences in resolution and depth of field caused by changes in objective lens magnification and aperture stop opening. The normalized image is also prone to generating artifacts, which in turn interferes with the artificial intelligence's feature extraction of microstructures. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to overcome the defects of the existing technology. The present invention proposes a method for reproducing and imaging microscopic image acquisition conditions.

[0006] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: a method for reproducing and imaging microscopic image acquisition conditions, comprising: S1. Select the objective lens and read its coded value. Calculate and match the field stop opening, aperture stop opening, light source brightness and color temperature according to the objective lens numerical aperture. Acquire the reference image and package it together with the above parameters into a standard imaging parameter file. S2. Retrieve the standard imaging parameter file and drive the objective lens conversion mechanism, field stop, aperture stop, light source and observation switching module in parallel to the file recording mode. S3. Acquire real-time images and calculate four indicators: illuminance, uniformity, contrast, and resolution. Compare each indicator with the archive baseline value. If the deviation exceeds the threshold, adjust the aperture opening or the light source brightness. Repeat until the deviation values ​​of the four indicators all fall within the threshold. S4. After acquiring sample images and embedding imaging parameter metadata, the images are output to ensure that images acquired at different times maintain consistency in four indicators.

[0007] Preferably, between step S1 and step S2, the method further includes: collecting hardware parameters at each imaging moment and binding them with the corresponding sample image to store them as a traceability parameter log.

[0008] Preferably, the spot diameter corresponding to the field stop aperture is: Among them, W 相机 M is the diagonal width of the camera's sensor surface. 物镜 For objective lens magnification, k1 ranges from 1.05 to 1.20; Numerical aperture corresponding to the aperture stop: Among them, NA 物镜 K is the numerical aperture of the objective lens, and k2 ranges from 0.6 to 0.8.

[0009] Preferably, the aperture matching coefficient k2 is between 0.65 and 0.75.

[0010] Preferably, the color temperature of the light source is selected according to the sample type: 5000K for undyed samples and 3800K to 4200K for dyed samples.

[0011] Preferably, the parallel drive in step S2 is a synchronous reset of each component according to a preset timing sequence, which is completed within 3 to 5 seconds.

[0012] Preferably, the illuminance is the average gray level L of all pixels in the image; the uniformity is the ratio U of the standard deviation of the average gray level of the nine regions in the nine-grid partition of the image to the overall average; the contrast is the average absolute deviation C of the gray level of all pixels in the image from L; and the resolution is the maximum gray level gradient of the sample edge.

[0013] Preferably, the adjustment in S3 is as follows: when the illuminance deviates by more than 3%, the light source brightness is adjusted in 1% increments; when the uniformity deviates by more than 5%, the field aperture opening is adjusted in 0.5% increments; when the contrast deviates, the aperture opening is adjusted in 0.5% increments; when the resolution deviates, the objective lens code and focus position are checked.

[0014] Preferably, the maximum number of iterations in the loop is 5, and an alarm is output if the number of iterations is exceeded.

[0015] Preferably, the objective lens conversion mechanism includes a rotary objective lens mount, an objective lens mounted on the rotary objective lens mount, and an encoder mounted on the shaft end of the rotary objective lens mount, wherein the encoder has a resolution of not less than 12 bits.

[0016] Compared with the prior art, the beneficial effects of the present invention include: 1. By setting up an coded objective lens switching mechanism, an electric field diaphragm, an electric aperture diaphragm, an adjustable color temperature light source, and an observation switching module, and driving them in parallel by a drive unit, one-click linkage reset of multiple imaging components such as objective lens magnification, dual aperture opening, light source brightness and color temperature, and observation position is achieved. This avoids the dispersion of imaging parameters in the same batch caused by the reliance on manual adjustment of multiple components, changes in operators, or the inability to reproduce the working conditions after testing across working days in the existing solution.

[0017] 2. By setting up a closed-loop calibration mechanism that captures real-time images and calculates four indicators—illuminance, uniformity, contrast, and resolution—and compares them item by item with archival benchmark values, the aperture opening or light source brightness is adjusted. This achieves the reproduction of standard imaging conditions from a hardware physical level, avoiding the problem that image post-processing normalization can only correct brightness deviations, cannot compensate for physical differences in resolution and depth of field, and is prone to generating artifacts. This ensures a high degree of consistency in the four indicators for images captured at different times.

[0018] 3. By setting the method of embedding imaging parameters as metadata into the image file for output, the imaging conditions of a single image can be traced, ensuring that subsequent artificial intelligence analysis can trace and verify the source when abnormal data occurs, and the image can be directly sent to the analysis software without the need for image preprocessing and normalization. Attached Figure Description

[0019] The disclosure of this invention is illustrated with reference to the accompanying drawings. It should be understood that the drawings are for illustrative purposes only and are not intended to limit the scope of protection of this invention. In the drawings, the same reference numerals are used to refer to the same parts. Wherein: Figure 1 The schematic diagram illustrates the overall composition of a standardized microscopic imaging control system according to an embodiment of the present invention.

[0020] Figure 2 The diagram schematically shows a cross-sectional view of a coding objective lens conversion mechanism according to an embodiment of the present invention.

[0021] Figure 3 The schematic diagram shows the mounting positions of an electric field stop and an electric aperture stop according to an embodiment of the present invention.

[0022] Figure 4 The diagram illustrates the workflow of a method according to an embodiment of the present invention.

[0023] The diagram shows: 1. Host computer control terminal; 2. Drive unit; 3. Encoded objective lens conversion mechanism; 31. Objective lens; 32. Encoder 1; 4. Electric field diaphragm; 41. Stepper motor one; 42. Encoder two; 5. Electric aperture stop; 51. Stepper motor II; 52. Encoder III; 6. Adjustable color temperature LED coaxial light source; 7. Observation switching execution module; 71. Eyepiece observation position; 72. Camera imaging position; 8. Microscope camera. Detailed Implementation

[0024] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Preferred embodiments of the invention are shown in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a thorough and complete understanding of the disclosure of the invention.

[0025] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0026] Reference Figures 1 to 4 This invention provides a method for reproducing and imaging microscopic images based on a standardized microscopic imaging control system. The signal flow of the entire control system is as follows: the host computer control terminal 1 acts as the human-computer interaction and calculation center, sending parameter commands to the main control microcontroller unit (MCU) via the communication bus; the MCU drive unit 2 calculates the commands according to the instructions and drives the coded objective lens conversion mechanism 3, the motorized field stop 4, the motorized aperture stop 5, the adjustable color temperature LED coaxial light source 6, and the observation switching execution module 7 to the target working condition; the microscopic camera 8 acquires images when the observation switching execution module 7 is in the camera imaging position 72 and sends the image data back to the host computer control terminal 1; the host computer control terminal 1 calculates four indicators—illuminance, uniformity, contrast, and resolution—of the returned images, compares them with the reference values, and then sends fine-tuning commands to the MCU drive unit 2, thus forming a closed-loop control link of command issuance, hardware execution, image acquisition, indicator verification, and parameter fine-tuning.

[0027] The host computer control terminal 1 can be an industrial control computer, an embedded industrial panel PC, or a general-purpose computer with equivalent computing power. It runs standardized imaging control software, which includes a built-in archive database, an optical matching rule module, an image index calculation module, and a closed-loop calibration and fine-tuning module. The main control MCU drive unit 2 uses a 32-bit embedded microcontroller as its computing core and is externally expanded with multiple stepper motor drive interfaces, analog-to-digital conversion interfaces, digital input / output interfaces, and communication interfaces.

[0028] Specifically, the coded objective lens conversion mechanism 3 includes a rotary objective lens mount, several objectives 31 mounted on the rotary objective lens mount, a stepper motor coaxially connected to the rotary objective lens mount, and an encoder 32 mounted on the shaft end of the rotary objective lens mount. The number of objectives 31 installed is preferably four to six, corresponding to objective lens specifications with different magnifications and numerical apertures (NAs) such as 4×, 10×, 20×, 50×, and 100×. The encoder 32 has a resolution of not less than 12 bits, that is, it divides a circle into no less than 4096 angular positions, so that each objective lens position has a unique angular coding value. In other optional embodiments, the encoder 32 can also be replaced by a high-resolution incremental encoder with a zero-position calibration signal and an electric zeroing program, which can also achieve unique identification of the objective lens position. The angular coding value is read by the main control MCU drive unit 2 and transmitted to the host computer control terminal 1 to lock the objective lens specification and the corresponding numerical aperture (NA) parameter of the current actual working objective lens, so as to avoid mismatch of calibration parameters caused by incorrect objective lens switching.

[0029] Reference Figure 3 The motorized field stop 4 and motorized aperture stop 5 have similar structures and are both mounted in the standard aperture mounting position on the optical tube. The motorized field stop 4 controls the diameter of the beam illuminating the sample surface, determining the effective imaging field of view. The motorized aperture stop 5 controls the angle of the light cone entering the objective lens, determining the numerical aperture and depth of field of the image. Each includes an aperture blade assembly, stepper motor 41 and stepper motor 51 connected to the aperture blade assembly, and encoders 42 and 52 for measuring the aperture opening. The step angle of stepper motor 41 and stepper motor 51 is no greater than 0.1. The micro stepper motor, combined with a precision gear transmission with a reduction ratio of 50:1, achieves fine adjustment of the aperture opening at the 0.5% level. The step angle and reduction ratio can balance response speed and position accuracy. In other optional embodiments, the micro stepper motor can also be replaced by a DC servo motor with a high-precision grating ruler. Encoder 2 42 and encoder 3 52 adopt 1024-line incremental encoders. Combined with the zero-position calibration signal, they achieve closed-loop position feedback within the range of 0% to 100% aperture full opening, ensuring that each aperture opening reset falls within the ±0.3% error band of the archived value.

[0030] The adjustable color temperature LED coaxial light source 6 consists of a warm color temperature LED group, a cool color temperature LED group, a diffuser plate, a constant current drive circuit, a brightness adjustment circuit, a color temperature adjustment circuit, and a temperature compensation sensor. The color temperature of the warm color temperature LED group is 3000K, and the color temperature of the cool color temperature LED group is 7000K. By adjusting the current ratio of the two groups of LEDs, the color temperature can be continuously adjusted within the range of 3800K to 6800K. The constant current drive circuit uses pulse width modulation (PWM) to make the LED brightness approximately linearly adjustable within the range of 0% to 100%, with an adjustment resolution of 1%. The temperature compensation sensor detects the ambient temperature of the LEDs in real time. When the temperature change exceeds 5°C, the main control MCU drive unit 2 automatically compensates the drive current according to the pre-stored temperature-brightness attenuation compensation table. The temperature-brightness attenuation compensation table is obtained by performing a temperature scan calibration of the LED light source in the range of 15°C to 55°C before leaving the factory, recording a set of LED current values ​​and actual luminous flux values ​​every 5°C, fitting the temperature-drive current compensation coefficient curve accordingly, and writing it into the memory of the main control MCU drive unit 2. The temperature compensation mechanism controls the time drift of the light source brightness within ±0.5% / hour. In other optional embodiments, the light source can also be replaced by a halogen lamp with an electric color filter wheel, and the color temperature adjustment is achieved by switching different color filters.

[0031] The observation switching execution module 7 includes a beam splitter prism, a prism switching mechanism, and a switching drive motor, used for rapid switching between the eyepiece observation position 71 and the camera imaging position 72. When in the eyepiece observation position 71, the light path is deflected to the eyepiece by the beam splitter prism, making it easy for the operator to observe the sample with the naked eye. When in the camera imaging position 72, the light path directly enters the microscope camera 8, facilitating image acquisition and automated artificial intelligence analysis. The switching response time of the switching drive motor is no more than 500 milliseconds, and the switching position is accurately determined by the limit optocoupler sensor. When the prism switching mechanism reaches the target position, the limit optocoupler sensor outputs a level change signal to the main control MCU drive unit 2 to avoid image halo caused by the prism staying in the middle of the light path.

[0032] The microscope camera 8 is preferably a 10-megapixel complementary metal-oxide-semiconductor (CMOS) industrial camera with a photosensitive surface size of not less than 2 / 3 inches. It can use either a global shutter or a rolling shutter; a global shutter is preferred for fast-moving samples, while either can be used for static samples. The microscope camera 8 connects to the host computer control terminal 1 via a USB 3.0 interface or a Gigabit Ethernet interface to transmit raw RAW format standardized image data. In other optional implementations, high-speed industrial interfaces such as CameraLink or CoaXPress can also be used. The main control MCU driver unit 2 connects to the host computer control terminal 1 via an RS485 serial bus or a USB communication interface, issuing commands and providing status feedback in master-slave mode. A single complete communication cycle is no more than 10 milliseconds to ensure synchronization when multiple hardware components are driven in parallel.

[0033] A method for reproducing and imaging microscopic image acquisition conditions, comprising: S100. The operator starts the file creation program on the host computer control terminal 1, selects the standard objective lens specifications according to the detection requirements of the target sample, and the system reads the feedback signal of encoder 32 on the coded objective lens conversion mechanism 3 to confirm the magnification and numerical aperture NA of the current working objective lens. The host computer control terminal 1 automatically calculates the standard imaging parameter combination matching the current objective lens according to the built-in optical matching rules. The specific content of the optical matching rules is as follows: The standard opening of the field stop is matched to the field of view of the objective lens so that the beam diameter exactly covers the sample area corresponding to the camera's photosensitive surface, thus preventing stray light from outside the field of view from entering the optical path. It is calculated according to the formula. in, This represents the spot diameter corresponding to the standard aperture of the field of view. The width of the camera's photosensitive surface diagonal. This refers to the objective lens magnification. The field redundancy coefficient is 1.05 to 1.20, preferably 1.10, and the field aperture is preferably 80% to 95% full aperture.

[0034] The standard aperture opening is calculated using the formula. in, This is the numerical aperture corresponding to the aperture stop. The numerical aperture of the objective lens. The aperture matching coefficient has a value ranging from 0.6 to 0.8, and is more preferably from 0.65 to 0.75. This coefficient value ensures that the aperture stop is neither too small, which would cause diffraction limitation and resolution degradation, nor too large, which would cause shallow depth of field and increased edge chromatic aberration.

[0035] The standard brightness value of the LED is determined by matching the objective lens magnification with the sample type and referring to the objective lens magnification-brightness comparison table. For each increase in magnification level, the brightness increases by about 5% to 10%. Alternatively, during the calibration stage, the average gray level of the image can be used to reach the target reference gray level. For example, if the target gray level is 128, the system will automatically adjust the LED brightness in reverse to the brightness level corresponding to the target gray level.

[0036] The standard color temperature value of LEDs should be selected by referring to the color temperature reference table according to the sample type. For unstained samples such as metallographic and inorganic powders, a standard daylight color temperature of 5000K should be selected. For biological stained samples, a warm light color temperature of 3800K to 4200K should be selected to enhance the staining contrast.

[0037] The default workstation setting for the observation module is camera imaging position 72.

[0038] The system automatically drives each component to the matching condition according to the parameter combination calculated by the above optical matching rules. The microscope camera 8 acquires a reference image, and the host computer control terminal 1 calculates four reference index values ​​of the reference image: average illuminance, illuminance uniformity, image contrast, and effective resolution. Finally, the objective lens code, field stop opening, aperture stop opening, LED brightness, LED color temperature, observation position, reference image, and the four reference index values ​​are packaged and encapsulated, assigned a unique file number, and stored in the file database of the host computer control terminal 1.

[0039] S200. During the formal sample testing process, the main control MCU drive unit 2 collects the hardware parameters at each image exposure moment in real time, including objective lens code, real-time opening of the field stop, real-time opening of the aperture stop, real-time LED brightness, real-time LED color temperature, and observation position. The above parameters are bound and stored with the corresponding sample image to form a traceability parameter log for a single image, which facilitates the traceability and verification of the original imaging conditions when the subsequent AI analysis results are abnormal.

[0040] S300. When repeated imaging of a certain standard test item is required, the operator retrieves the target standard imaging parameter file number on the host computer control terminal 1. The host computer control terminal 1 sends a synchronous execution command to the main control MCU drive unit 2. The main control MCU drive unit 2 drives each hardware component in parallel according to a preset timing sequence: drives the coded objective lens conversion mechanism 3 to operate to the coded position recorded in the file, continuously reads the feedback signal of encoder 1 32 during operation, and stops when the reading is consistent with the file code; drives the stepper motor 1 41 of the electric field aperture 4 to rotate, and encoder 2 42 provides real-time feedback on the aperture opening, locking when the opening reaches the recorded value; drives the stepper motor 2 51 of the electric aperture 5 to adjust to the recorded value in the same closed-loop manner; drives the constant current drive circuit of the adjustable color temperature LED coaxial light source 6 to switch the current ratio of the two sets of LED beads according to the recorded brightness and color temperature value, while the temperature compensation sensor corrects the temperature drift in real time; drives the observation switching execution module 7 to switch to the file recording position. All hardware components execute in parallel, and the entire one-click restore process completes the hardware reset of all parameters within 3 to 5 seconds.

[0041] After the initial hardware reset is completed (S400), the microscope camera 8 acquires a real-time frame and transmits it back to the host computer control terminal 1. The host computer control terminal 1 calculates the four index values ​​of the real-time frame according to the following method.

[0042] The average illuminance of the screen is calculated using the formula; Where L is the average illumination of the image, N is the total number of pixels in the image, and G(i,j) is the gray value of the pixel at coordinate (i,j), with a gray value range of 0 to 255. The summation is applied to all pixels in the image, and the resulting L value reflects the imaging brightness level.

[0043] Illumination uniformity is calculated using the nine-grid image partitioning method. The image is divided into three equal parts along the horizontal and vertical directions, forming nine equal-area regions of 3×3. The gray mean value Lk (k=1, 2, …, 9) of each region is calculated according to the above formula, and then the illuminance uniformity U is calculated according to the formula. Where σ is the standard deviation of the gray-scale mean Lk of the nine regions, and Lavg is the total mean of the gray-scale mean Lk of the nine regions. The smaller the obtained U value, the more uniform the illuminance distribution.

[0044] Image contrast is calculated using a formula. Where N is the total number of pixels in the image, G(i,j) is the gray value of the pixel at coordinate (i,j), and L is the average illumination of the image.

[0045] The C value is the average of the absolute deviations between the gray values ​​of all pixels in the image and the average illuminance L of the image. The larger the C value, the more significant the contrast between light and dark.

[0046] Effective resolution is characterized by the maximum gray-level gradient of the sample edge in the image, specifically the maximum absolute value of the gray-level difference between adjacent pixels in a preset region of interest in the image. In other optional embodiments, the spatial frequency corresponding to the modulation transfer function (MTF) dropping to 50% can also be used as the effective resolution index. The modulation transfer function is MTF, and this index is marked as MTF50.

[0047] The four indicator values ​​of the real-time image are compared one by one with the four benchmark indicator values ​​of the pre-stored benchmark image in the archive. When any indicator deviates from the benchmark value by more than a preset threshold, a fine-tuning correction subroutine is activated: when the average illuminance of the image deviates by more than 3%, the brightness of the LED light source is adjusted slightly in steps of 1% in the direction of deviation; when the illuminance uniformity deviates by more than 5%, the field aperture opening is adjusted slightly in steps of 0.5% in the direction of deviation; when the image contrast deviates by more than a preset threshold, the aperture opening is adjusted slightly in steps of 0.5% in the direction of deviation; when the effective resolution deviates, the objective lens code and focus position are checked first, and then the aperture opening is used as an auxiliary adjustment means. The error threshold values ​​of 3% and 5% can take into account the consistency requirements of AI recognition and the achievable accuracy of the imaging system under normal environmental disturbances. Those skilled in the art can make adaptive adjustments to the sensitivity of the input image according to the specific AI model.

[0048] After each fine-tuning, images are reacquired and the four index values ​​are recalculated. The acquisition and comparison are repeated until all four index deviations fall within the corresponding threshold range, at which point the closed-loop calibration is complete. To prevent the closed loop from getting stuck in an infinite loop under extreme conditions, the maximum number of fine-tuning iterations is preset to 5. This value is based on experience and 5 iterations can cover most normal environmental disturbance scenarios. If the number of iterations is exceeded, the system will output an alarm to prompt the operator to intervene and check for foreign objects in the optical path or abnormalities in the sample, thus avoiding unnecessary iterations that waste time.

[0049] After S500 and closed-loop calibration are completed, the microscope camera 8 acquires imaging images of the formal samples. The host computer control terminal 1 embeds all imaging parameters corresponding to this image as metadata into the exchangeable image file format (i.e., Exif) field or additional text field of the image file. The standardized image with metadata is directly sent to the AI ​​analysis software. Without the need for image post-processing algorithms such as brightness equalization, histogram normalization, and gamma correction, the feature consistency of the entire batch of sample images can be guaranteed, and automatic and unified data analysis can be achieved.

[0050] The practical application of the above method will be further illustrated below with three specific embodiments.

[0051] Example 1 This embodiment uses an upright metallographic fully automated microscope as the execution device, aiming to perform batch AI statistical analysis of grain size of steel samples. The operator selects a 50× plan achromatic objective as the standard objective on the host computer control terminal 1, and the encoder-32 feeds back the numerical aperture (NA) of the objective as 0.65. The host computer control terminal 1 automatically calculates the standard imaging parameter combination according to the optical matching rules: the field redundancy coefficient k1 is 1.10, corresponding to a field stop opening of 95%; the aperture matching coefficient k2 is 0.7, corresponding to a numerical aperture (NA) stop of 0.455 and an aperture mechanism opening of 68%. The LED color temperature is fixed at 5000K to simulate standard daylight illumination, and the LED brightness is adjusted to 72% in reverse according to the target reference grayscale of 128. The observation switching execution module 7 switches to the camera imaging position 72. The system resets the hardware according to the above parameters, and the microscope camera 8 acquires the reference image. The average illuminance of the image is calculated to be 148 grayscale value, the illuminance uniformity is 3.2%, the average grayscale deviation of the image contrast is 26.5, and the MTF50 spatial frequency corresponding to the effective resolution is 68 line pairs / mm. The above parameters and reference indicators are packaged and encapsulated, assigned the file number JJ-50-01, and stored in the file database.

[0052] Subsequently, 300 steel samples were continuously batch-tested. The operator retrieved file number JJ-50-01 on the host computer control terminal 1, and the system automatically completed hardware reset and closed-loop calibration. Minor luminous flux drift caused by changes in ambient temperature and accumulated LED operating time, as well as image illuminance fluctuations due to differences in reflectivity between samples, were all eliminated during the closed-loop calibration process through LED brightness fine-tuning. Statistical analysis showed that the average fluctuation of image illuminance for the 300 samples was controlled within ±1.5%, the deviation between AI grain size statistics and manual annotation results was less than 1.5%, and the total time for image reset and calibration of a single sample did not exceed 6 seconds.

[0053] Example 2 This embodiment uses a biological upright microscope as the execution device, aiming to automatically identify cell types in pathological sections treated with hematoxylin and eosin staining (i.e., HE staining). The operator selects a 20× biological objective as the standard objective, and the encoder-32 reports that the numerical aperture (NA) of this objective is 0.40. The host computer control terminal 1 automatically calculates the standard imaging parameter combination: the field redundancy coefficient k1 is set to 1.10, corresponding to an aperture diaphragm opening of 85%; the aperture matching coefficient k2 is set to 0.7, corresponding to the numerical aperture of the aperture diaphragm. The value is 0.28, and the aperture mechanism opening is 62%. For the red and purple staining characteristics of HE-stained samples, the LED color temperature is selected as 4000K warm light to enhance staining contrast, and the brightness is 65%. The observation is switched to the camera imaging position 72. After completing the acquisition of the baseline image and recording of the four baseline indicators, the file number SH-20-01 is assigned.

[0054] The hospital's pathology department conducts three batches of tests daily: morning, noon, and evening. After each test, the operator retrieves the files (SH-20-01) and restores all parameters with a single click. Within a day, the luminous flux of the LED light source decreases by approximately 2% to 3% due to continuous operation, and the color temperature drift caused by changes in the laboratory environment is approximately 50K to 100K. The closed-loop calibration process automatically activates the LED brightness fine-tuning compensation mechanism by monitoring the deviation of the illuminance index, stabilizing the image parameters of each batch to near the benchmark value. Long-term statistics show that the accuracy rate of AI cell type recognition has remained stable at over 96%.

[0055] Example 3 This embodiment uses an oil immersion metallurgical microscope as the execution device, aiming to perform AI image analysis on the particle size distribution of graphite powder in lithium-ion battery negative electrodes. The operator selects a 100× flat oil immersion microscope as the standard objective lens, and the encoder-32 reports that the numerical aperture (NA) of this objective lens is 1.25. The host computer control terminal 1 automatically calculates the standard imaging parameter combination: the field redundancy coefficient k1 is set to 1.10, corresponding to a field stop opening of 90%; the aperture matching coefficient k2 is set to 0.7, corresponding to the numerical aperture of the aperture stop. The value is 0.875, the aperture mechanism opening is 72%; the LED color temperature is 5000K, and the brightness is 78%. After completing the baseline image acquisition, the file number LD-100-01 is assigned.

[0056] Due to the long R&D cycle of lithium battery materials, samples from the same batch need to be retested and compared one month apart. During the retest, the operator retrieves the file number LD-100-01, restores all hardware conditions with one click, and performs closed-loop calibration. During the monthly interval, the luminous flux attenuation of the LED light source due to long-term aging is about 3% to 5%, and the slight optical path shift caused by seasonal changes in laboratory temperature and humidity is compensated in the closed-loop calibration process. There is no visible difference between the retested image and the sample image from one month ago in terms of brightness, field of view, and contrast. The repeatability deviation of the AI ​​granularity analysis results is less than 2%.

[0057] The above three embodiments involve three typical application scenarios: metallography, biological pathology, and inorganic powders, respectively. They cover three magnification levels (low, medium, and high) and two imaging methods (immersion and non-immersion). This invention can achieve fully automatic reset and standardized output of imaging conditions under different detection objects, different magnifications, and different time spans.

[0058] 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.

[0059] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, 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 reproducing and imaging microscopic image acquisition conditions, characterized in that, include: S1. Select the objective lens and read its coded value. Calculate and match the field stop opening, aperture stop opening, light source brightness and color temperature according to the objective lens numerical aperture. Acquire the reference image and package it together with the above parameters into a standard imaging parameter file. S2. Retrieve standard imaging parameter files and drive the objective lens conversion mechanism (3), field stop (4), aperture stop (5), light source (6) and observation switching module (7) to file recording mode in parallel. S3. Acquire real-time images and calculate four indicators: illuminance, uniformity, contrast, and resolution. Compare each indicator with the archive baseline value. If the deviation exceeds the threshold, adjust the aperture opening or the light source brightness. Repeat until the deviation values ​​of the four indicators all fall within the threshold. S4. After acquiring sample images and embedding imaging parameter metadata, the images are output to ensure that images acquired at different times maintain consistency in four indicators.

2. The method for reproducing and imaging microscopic image acquisition conditions according to claim 1, characterized in that, Between steps S1 and S2, the following step is also included: collecting hardware parameters at each imaging moment and binding them with the corresponding sample image to store them as a traceability parameter log.

3. The method for reproducing and imaging microscopic image acquisition conditions according to claim 2, characterized in that, The spot diameter corresponding to the field stop opening: Where W is the camera's diagonal width of the camera's photosensitive surface, M is the objective lens's magnification, and k1 ranges from 1.05 to 1.

20. Numerical aperture corresponding to the aperture stop: Wherein, NA objective is the numerical aperture of the objective, and k2 is 0.6 to 0.

8.

4. The method for reproducing and imaging microscopic image acquisition conditions according to claim 3, characterized in that, The aperture matching coefficient k2 is between 0.65 and 0.

75.

5. The method for reproducing and imaging microscopic image acquisition conditions according to claim 1, characterized in that, The color temperature of the light source is selected according to the sample type: 5000K for undyed samples and 3800K to 4200K for dyed samples.

6. The method for reproducing and imaging microscopic image acquisition conditions according to claim 1, characterized in that, The parallel drive described in step S2 is the synchronous reset of each component according to a preset timing sequence, which is completed within 3 to 5 seconds.

7. The method for reproducing and imaging microscopic image acquisition conditions according to claim 1, characterized in that, The illuminance is the average gray level L of all pixels in the image; the uniformity is the ratio U of the standard deviation of the average gray level of the nine regions in the nine-grid partition of the image to the overall average; the contrast is the average absolute deviation C of the gray level of all pixels in the image from L. The resolution is the maximum grayscale gradient at the edge of the sample.

8. The method for reproducing and imaging microscopic image acquisition conditions according to claim 1, characterized in that, The adjustment described in S3 is as follows: when the illuminance deviates by more than 3%, adjust the brightness of the light source (6) by 1% step; when the uniformity deviates by more than 5%, adjust the opening of the field stop (4) by 0.5% step. When the contrast deviates, adjust the aperture stop (5) opening in 0.5% increments; when the resolution deviates, check the objective lens code and focus position.

9. The method for reproducing and imaging microscopic image acquisition conditions according to claim 1, characterized in that, The maximum number of iterations in the loop is 5; an alarm will be output if this number is exceeded.

10. The method for reproducing and imaging microscopic image acquisition conditions according to claim 1, characterized in that, The objective lens conversion mechanism (3) includes a rotary objective lens mount, an objective lens (31) mounted on the rotary objective lens mount, and an encoder (32) mounted on the shaft end of the rotary objective lens mount. The encoder (32) has a resolution of not less than 12 bits.

Images