A method for position calibration of a microscopic imaging system based on LED array illumination

Abstract

The application provides a microscopic imaging system position calibration method based on LED array illumination. Without placing a sample, an image is collected to extract a vignetting circle arc, and a radius and a corresponding circle center coordinate are obtained through circle fitting. An image coordinate system is established. An LED array illumination coordinate system is established. A deflection angle theta of the LED array illumination coordinate system relative to the image coordinate system is calculated by fitting each LED illumination corresponding image center. An image center coordinate corresponding to the center LED illumination is calculated according to the center coordinates of each image under LED illumination. The vignetting circle circumference can be regarded as an aperture angle region, and the pixel height from the LED array plane to the object plane is calculated in combination with the fitted vignetting circle pixel radius. An equation is established for each LED illumination corresponding incident angle, and the actual height h from the LED array plane to the object plane and the XY position offset difx and dify of the center LED illumination unit relative to the optical axis can be calculated. The application realizes the correction of the illumination wave vector, eliminates the influence of the system position error, and improves the Fourier stack imaging effect.

Description

Technical Field

[0001] This invention belongs to the field of computational optical microscopy imaging, specifically a method for position calibration of a microscopy imaging system based on LED array illumination. Background Technology

[0002] In 2013, Zheng et al. proposed Fourier layered imaging technology. This technology innovatively replaces the single bright field light source of traditional optical microscopes with an LED array light source that can provide multiple incident angles. Combining layered imaging, aperture synthesis, phase recovery and other technologies, it uses an iterative algorithm to alternately project and fuse the image spectrum under different incident angles in the frequency domain, and finally reconstructs a large field of view and high resolution complex amplitude spectrum. It has broad application prospects and development potential in industrial quality inspection, biomedicine and other fields.

[0003] In Fourier layered imaging, it is necessary to accurately calculate the illumination wave vector corresponding to each LED unit. This is used to determine the spectral orientation of the corresponding image during iterative updates, thereby enabling amplitude replacement to achieve aperture fusion and phase recovery. However, the calculation of the illumination wave vector is affected by the XY position offset error of the LED array, the camera deflection angle error, and the distance error between the LED array plane and the sample. These positional errors are often unavoidable during system assembly and measurement, leading to a mismatch between the calculated and actual illumination wave vectors, which severely affects the quality of the reconstructed image. Therefore, to improve the reconstruction effect of Fourier layered imaging, positional calibration of the Fourier layered microscopy system based on LED array illumination is crucial. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a position calibration method for a microscopic imaging system based on LED array illumination, so as to solve the problem that the system position error of Fourier layer imaging technology seriously affects the imaging quality.

[0005] This invention is implemented as follows:

[0006] A method for position calibration of a microscopic imaging system based on LED array illumination, the method comprising:

[0007] Step 1: Take a set of images with vignetting patterns without placing the sample;

[0008] Step 2: Preprocess and binarize each image to extract the corresponding shading arc contour;

[0009] Step 3: Perform circle fitting on the extracted hazy arc contour to obtain the effective fitting hazy circle radius and the corresponding effective hazy arc contour.

[0010] Step 4: Remove some singular values ​​from the obtained halo circle radius to obtain the effective halo circle radius;

[0011] Step 5: Perform a weighted summation of the effective halo radii to obtain the final estimated halo radius.

[0012] Step 6: Using the final estimated radius of the halos circle as the radius of the circle to be fitted, refit each effective halos circle contour to obtain the coordinates of the center of the halos circle corresponding to each image.

[0013] Step 7, obtain the image coordinate system COOR img Set the origin of the coordinate system as the center of the vignetting circle, and calculate the coordinates of the center of the vignetting circle in the image coordinate system COOR using the coordinates of the center of the vignetting circle for each image. img Below are the coordinates of the center of each image;

[0014] Step 8: Based on the central symmetry property of the illumination wave vector, average the coordinates of the image center symmetric about the central LED illumination to obtain the coordinates of the image center corresponding to the central LED illumination in the image coordinate system COOR. img The coordinates of the LED array are the coordinates of the optical axis, and the center of the halo circle is the center of the beam corresponding to the central LED illumination. The distance of the central LED illumination relative to the optical axis is calculated.

[0015] Step 9, determine the coordinate system COOR for the LED array lighting. ill The origin of the coordinate system is the center of the central LED lighting. The image centers corresponding to each LED lighting are used in the image coordinate system COOR. img The coordinates are fitted to form the coordinate system COOR. ill The XY axis directions are used to calculate the LED array lighting coordinate system COOR. ill With image coordinate system COOR img The included angle θ;

[0016] Step 10: Since the circumferential region of the vignetting circle can be regarded as the incident angle being the aperture angle, and the center of the vignetting circle is the intersection of the beam center and the image plane under the LED illumination, the pixel height from the LED array plane to the sample plane is obtained by the incident angle calculation formula.

[0017] Step 11: Using the incident angle calculation formula, substitute the coordinates of the image center corresponding to each LED illumination into the formula to calculate the incident angle corresponding to each LED illumination. Based on the known actual spacing of the LED illumination, calculate the height of each LED array plane from the sample plane according to the incident angle formula. Calculate the average of the calculated heights to obtain the estimated actual height of the LED array plane from the sample plane.

[0018] Step 12: Using the pixel distance of the LED array center illumination relative to the optical axis, substitute the calculated actual height estimate into the equation to obtain the actual distance from the center LED illumination to the optical axis;

[0019] Step 13: Based on the calculated height h from the LED array plane to the sample plane, the distance from the central LED illumination to the optical axis, and the coordinate system COOR of the LED array illumination... ill Relative to the image coordinate system COOR img The deflection angle θ is used to calibrate the current optical path position information of the system;

[0020] Step 14: According to the calibrated and corrected image coordinate system COOR for each LED illumination... img The actual coordinates are used to calculate the illumination wave vector corresponding to each LED illumination after correction, eliminate the influence of system position error, and input into the Fourier stacked iterative update and reconstruction to obtain the Fourier stacked imaging effect.

[0021] Furthermore, in step 1, a low-NA objective lens and a large-aperture camera are used when acquiring images, and the ROI is centered within the full image frame. In step 2, Gaussian blur preprocessing is performed on each image, and image binarization is performed based on the Otsu algorithm. Opening and closing operations are performed on the binarized image to eliminate stray light effects. Subsequently, the vignetting arc contour C is extracted from the preprocessed binary image. j (j = 1, 2, 3, ..., n).

[0022] Furthermore, in step 4, the radius R of the halo circle calculated when removing singular values... i The mean μ and variance σ are taken as R in the range [μ-0.5*σ, μ+0.5*σ]. i For valid values, those outside the range are removed as singular values. In step 5, a weighted sum is used to calculate the estimated radius of the halo circle. Where the weight w i The normalized image vignetting arc contour length

[0023] Furthermore, the image coordinate system COOR in step 7 img The origin of the coordinate system is set as the center of the vignetting circle, and the XY axis direction is the XY direction of the image. The image center coordinates corresponding to each LED illumination are the illumination wave vectors corresponding to each LED illumination that are symmetrical about the origin of the coordinate system. In step 8, based on the central symmetry property of the illumination wave vector, the image center corresponding to the fitted central LED illumination is set as the intersection of the optical axis and the image plane. The distance between the central LED illumination and the origin is the distance of the illumination wave vector corresponding to the central LED illumination relative to the optical axis.

[0024] Furthermore, the LED array lighting coordinate system COOR in step 9 ill The origin of the coordinate system is set as the central LED lighting unit, and the XY axis is the XY direction of the LED array plate. The coordinate system COOR of the LED array lighting can be fitted using the image center corresponding to each LED lighting unit in step 7. ilThe XY axis direction.

[0025] Furthermore, in step 10, the formula for calculating the incident angle of each LED lighting is as follows: Where L is the distance of the LED lighting unit relative to the optical axis, and h is the height of the LED array plane from the sample plane; based on the estimated value R of the halos radius. vig and its corresponding incident angle is the aperture angle θ. NA The formula for calculating the pixel height from the LED array plane to the sample plane is:

[0026] Further, in step 11, the illumination wave vector corresponding to each LED illumination is calculated. in L i Given the information, determine the actual height from the LED array plane to the sample plane. For h i The mean value is calculated to obtain the estimated actual height h from the LED array plane to the sample plane. Similarly, in step 12, the actual distances difx and disc from the central LED lighting unit to the optical axis are calculated.

[0027] Further, in step 13, based on the calculated actual height h from the LED array plane to the sample plane, the actual distance difx from the central LED lighting unit to the optical axis, and the LED array lighting coordinate system COOR, ill Relative to the image coordinate system COOR img The deflection angle θ, correcting the image coordinate system COOR img The following is the directional information for each LED light:

[0028] H = h

[0029]

[0030]

[0031]

[0032]

[0033] Where H is the height from the LED array plane to the sample plane. and These are the coordinates of the i-th LED relative to the center LED in the LED array lighting coordinate system COOR. ill The actual distance along the x and y axes; and The i-th LED lighting center is located in the LED array lighting coordinate system COOR. ill The actual coordinates on; and The center of the i-th LED illumination is in the image coordinate system COOR. img The coordinates below.

[0034] Furthermore, in step 14, the formula for calculating the incident angle corresponding to each LED illumination is:

[0035]

[0036]

[0037] Compared with the prior art, the beneficial effects of this invention are as follows:

[0038] This invention illuminates an LED unit and acquires an image with a vignetting pattern without placing a sample. The vignetting circle arc is extracted from the acquired image, and its center and radius are obtained through circle fitting. Subsequently, statistical analysis is performed on the center and radius obtained from the fitting of each image to obtain the final fitted vignetting circle radius and center coordinates.

[0039] Since the center of the halos can be regarded as the illumination wave vector corresponding to the central LED illumination, and the center of the camera image plane can be regarded as the intersection of the optical axis of the imaging system and the image plane, the XY position offsets difx and difx of the central LED illumination unit relative to the optical axis, the camera deflection angle θ, and the distance h from the LED array plate to the sample can be calculated. By substituting the above parameters into the wave vector calculation, the illumination wave vector after system error correction can be obtained, eliminating the influence of system position error and thus improving the Fourier stacked imaging effect. Attached Figure Description

[0040] Figure 1 This is a schematic diagram of the structure of the microscopic imaging system according to an embodiment of the present invention.

[0041] Figure 2 This is a flowchart illustrating an embodiment of the present invention;

[0042] Figure 3 shows the Fourier layered microscopy results using the USAF resolution plate as the sample under test. Figure 3(a) shows the reconstruction result without system position calibration, and Figure 3(b) shows the reconstruction result after system position calibration according to the present invention. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0044] The microscopic imaging system based on LED array illumination provided in this embodiment of the invention is as follows: Figure 1As shown, the system includes an LED flat panel array light source 1, a microscope objective 2, a tube lens 3, and a camera 4. The LED flat panel array light source is an Adafruit 32×32 LED array, with a spacing of 4mm between adjacent LED units. The microscope objective is an Olympus 4× / 0.1NA objective, and the camera is an ANDOR Zyla 5.5 with a target surface size of 16.6×14.0mm and a resolution of 2560*2160. The central LED unit of the LED array, the microscope objective, the tube lens, and the camera center are initially aligned coaxially. It is assumed that the camera's imaging center is on the optical axis. The center of the vignetting circle corresponding to each LED illumination is the region where the beam center of that LED illumination is located (i.e., the illumination wave vector). The circumference of the vignetting circle can be approximated as the region where the incident angle under that LED illumination is the aperture angle.

[0045] See Figure 2 As shown, a method for position calibration of a microscopic imaging system based on LED array illumination includes:

[0046] Step 1: Using the central LED lighting unit as the center, illuminate a 5×5 LED array, lighting up each LED unit row by row from the top left to the bottom right. Take a 2048×2048 pixel ROI centered within the camera's image area and capture a corresponding set of images (Img) with a vignetting pattern. j (j = 1, 2, 3, ..., n)(n = 25), where Img j This represents the image captured corresponding to the illumination of the j-th LED unit.

[0047] Step 2: Gaussian blur preprocessing is performed on each image. Image binarization is then performed based on the Otsu algorithm. Opening and closing operations are then performed on the binarized images to eliminate stray light effects. Finally, the vignetting arc contour C is extracted from the preprocessed binary image. j (j = 1, 2, 3, ..., n);

[0048] Step 3, for the shading arc contour C j Perform circle fitting on (j = 1, 2, 3, ..., n) to obtain the radius R of the halos circle obtained from the effective fitting. i (i = 1, 2, 3, ..., m), and the corresponding effective shading arc contours. Where m≤n;

[0049] Step 4, adjust the radius R of the obtained halo circle. i (i = 1, 2, 3, ..., m) Calculate the mean μ and variance σ, and take a radius value R in the range of [μ - 0.5*σ, μ + 0.5*σ]. i For valid values, radii outside the range are treated as singular values ​​and removed to obtain valid halos. Where k≤m;

[0050] Step 5, for the effective shading circle half-stroke Perform a weighted summation, with weights w i The length of the vignetting arc contour in the corresponding image after normalization. Thus, the final estimated value of the radius of the halo circle is obtained.

[0051] Step 6, using the final estimated radius R of the halos. vig Using the radius of the circle to be fitted, the effective shading arc contours are then re-evaluated. Perform circle fitting to obtain the coordinates of the center of the corresponding halo circle.

[0052] Step 7, obtain the image coordinate system COOR img Set the origin of the coordinate system as the center of the shading circle, and use the coordinates of the center of the shading circle corresponding to each image. The COOR in the image coordinate system can be calculated. img Below, the coordinates of the center of each image

[0053] Step 8: Based on the central symmetry property of the illumination wave vector, in the images that can be effectively circle-fitted, match the image symmetric about the central LED illumination for each pair of LED illuminations, and then calculate the center coordinates of all matched images. Calculate the average value to obtain the image center coordinate X corresponding to the central LED illumination. axis Y axis Calculate the center LED illumination in the image coordinate system COOR img The pixel distance in the XY directions relative to the optical axis, dx = X axis dy = X axis ;

[0054] Step 9, determine the coordinate system COOR for the LED array lighting. ill and for the x in the lighting coordinate system ill Each column in the direction, y ill Linear fitting is performed on the effective image center points of each row in the direction. in The slope of the straight line obtained by linear fitting the center point of the i-th column / row along the dir axis is used to calculate the COOR of the LED array lighting coordinate system. ill x ill y ill Orientation and image coordinate system COOR img x img y img Directional deflection angle The average value of the deflection angle is used to obtain the coordinate system COOR for the LED array lighting.ill Relative to the image coordinate system COOR img deflection angle

[0055] Step 10: Use the circumference of the halo circle as the boundary between the bright and dark fields, i.e., the corresponding incident angle is the aperture angle θ. NA sinθ NA =NA, and the approximate relationship between the center of the vignetting circle and the intersection of the optical axis and the image plane, can be calculated using the incident angle formula. Where θ is the incident angle, L is the distance from the object point to the optical axis, and h is the distance from the LED array plane to the sample plane, the pixel height from the LED array plane to the sample plane is calculated.

[0056] Step 11: Calculate the illumination wave vector corresponding to each LED illumination. make but The incident angle (illumination wave vector) corresponding to the center of each LED lighting beam is calculated, based on the known actual distance L between each LED lighting unit and the center LED lighting unit. i Calculate the actual height from the LED array plane to the sample plane. For h i Calculate the mean to obtain an estimate of the actual height from the LED array plane to the sample plane.

[0057] Step 12, by Where dx and dy are the wave vectors of the central LED illumination in the image coordinate system COOR. img The pixel distance from the optical axis in the XY direction, difx, difx are the actual distances from the center LED lighting unit to the optical axis, and h is the pixel distance from the center LED lighting unit to the optical axis. p Calculate the pixel height from the LED array plane to the sample plane (where h is the actual height from the LED array plane to the sample plane).

[0058] Step 13: Estimate h based on the calculated actual height h from the LED array plane to the sample plane, the actual distance difx and difx from the central LED lighting unit to the optical axis, and the LED array lighting coordinate system COOR. ill Relative to the image coordinate system COOR img The deflection angle θ is used to calibrate the current system optical path position information, which is used to correct the optical path position in the image coordinate system COOR. img The location information of each LED light is shown below.

[0059] H = h

[0060]

[0061]

[0062]

[0063]

[0064] Where H is the height from the LED array plane to the sample plane. and These are the coordinates of the i-th LED relative to the center LED in the LED array lighting coordinate system COOR. ill The actual distance along the x and y axes; and The i-th LED lighting center is located in the LED array lighting coordinate system COOR. ill The actual coordinates on; and The center of the i-th LED illumination is in the image coordinate system COOR. img The coordinates below.

[0065] Step 14, based on the i-th LED lighting center in the image coordinate system COOR img Using the coordinates below, the corrected incident angle of the illumination wave vector can be calculated, thereby eliminating the influence of system position errors and obtaining better Fourier stacked reconstruction results. and Let the incident light under the illumination of the i-th LED be in the image coordinate system COOR. img The angle between the optical axis and the optical axis in the x and y directions.

[0066]

[0067]

[0068] Figure 3 shows the Fourier layered microscopy imaging results using a USAF resolution plate as the test sample. Figure 3(a) is the reconstruction result without system position calibration; Figure 3(b) is the reconstruction result after system position calibration according to the present invention. It can be seen that the method of the present invention achieves the correction of the illumination wave vector, eliminates the influence of system position error, and thus improves the Fourier layered imaging effect.

[0069] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for position calibration of a microscopic imaging system based on LED array illumination, characterized in that, The method includes: Step 1. Take a set of images with vignetting pattern without placing the sample wherein represent the first LED unit illuminates the corresponding captured image; Step 2: Preprocess and binarize each image to extract the corresponding shading arc contour. ; Step 3: Perform circle fitting on the extracted hazy arc contour to obtain the radius of the effectively fitted hazy circle. And the corresponding effective gradient arc contour. ,in ; Step 4, adjust the radius of the obtained halo circle. Remove some singular values ​​to obtain an effective vignetting circle radius. ,in ; Step 5, determine the effective gradation radius. By performing a weighted summation, we obtain the final estimated value of the radius of the halo circle. Weight The normalized length of the corresponding image's vignetting arc contour; Step 6: Use the final estimated radius of the halos. Using the radius of the circle to be fitted, the effective gradient arc contours are re-evaluated. Perform circle fitting to obtain the coordinates of the center of the halos circle corresponding to each image. ; Step 7, Establish the image coordinate system Set the origin of the coordinate system as the center of the shading circle, and use the coordinates of the center of the shading circle corresponding to each image. Calculation in image coordinate system Below, the center coordinates of each image ; Step 8: Based on the central symmetry property of the illumination wave vector, in the images that can be effectively circle-fitted, match the image symmetric about the central LED illumination for each pair of LED illuminations, and then calculate the center coordinates of all matched images. Calculate the average value to obtain the image center coordinates corresponding to the central LED illumination. Calculate the center LED illumination in the image coordinate system of Pixel distance relative to the optical axis: ; Step 9, determine the coordinate system of the LED array lighting. and for the lighting coordinate system Each column in the direction Linear fitting is performed on the effective image center points of each row in the direction. ,in refer to axial direction, the first The slope of the line obtained by linear fitting the center point of the column / row is used to calculate the coordinate system of the LED array lighting. of Orientation and Image Coordinate System of Directional deflection angle The LED array lighting coordinate system is obtained by averaging the deflection angles. Relative to image coordinate system deflection angle ; Step 10, based on the estimated radius of the halos. and its corresponding incident angle is the aperture angle. Calculate the pixel height from the LED array plane to the sample plane. The formula is: ; Step 11: Calculate the illumination wave vector corresponding to each LED illumination. ,make ,but The illumination wave vector corresponding to the center of each LED illumination beam is calculated, based on the known actual distance between each LED illumination beam and the center LED illumination unit. Calculate the actual height from the LED array plane to the sample plane. ,right Calculate the mean to obtain an estimate of the actual height from the LED array plane to the sample plane. ; Step 12: Substitute the calculated actual height estimate from the pixel distance of the LED array center illumination relative to the optical axis into the formula. This yields the actual distance from the center LED illumination to the optical axis. The actual distance from the central LED lighting unit to the optical axis; Step 13: Estimate the actual height from the LED array plane to the sample plane based on the calculated height. The actual distance from the central LED lighting unit to the optical axis LED array lighting coordinate system Relative to image coordinate system deflection angle Correcting the image coordinate system The following is the directional information for each LED light: ， ， ， ， ， in The height from the LED array plane to the sample plane. and They are the first Each LED lighting unit relative to the center LED lighting unit in the LED array lighting coordinate system of The actual distance along the axis; and It is the first The center of each LED lighting unit is located in the LED array lighting coordinate system. The actual coordinates on; and It is the first The center of each LED lighting unit is in the image coordinate system The coordinates below; Step 14: According to the calibrated and corrected image coordinate system of each LED illumination... The actual coordinates are used to calculate the illumination wave vector corresponding to each LED illumination after correction, thereby eliminating the influence of system position error. This vector is then used in the Fourier stacked layer iterative update and reconstruction to obtain the Fourier stacked layer imaging effect.

2. The position calibration method for a microscopic imaging system based on LED array illumination according to claim 1, characterized in that... In step 1, a low-NA objective lens and a large-aperture camera are used to acquire images, with the ROI centered within the full image frame. In step 2, Gaussian blur preprocessing is performed on each image, followed by binarization based on the Otsu algorithm. Opening and closing operations are then performed on the binarized images to eliminate stray light effects. Finally, the vignetting arc contour is extracted from the preprocessed binary image. .

3. The position calibration method for a microscopic imaging system based on LED array illumination according to claim 1, characterized in that... In step 4, the radius of the halo circle is calculated when removing singular values. mean and variance The range is within of Values ​​outside the valid range are removed as singular values. In step 5, the weights are... The normalized image vignetting arc contour length .

4. The position calibration method for a microscopic imaging system based on LED array illumination according to claim 1, characterized in that... The image coordinate system in step 7 The origin of the coordinate system is set as the center of the halo circle, and the XY axis direction is the XY direction of the image. The image center coordinates corresponding to each LED lighting are the lighting wave vectors corresponding to each LED lighting that are symmetrical about the origin of the coordinate system.

5. The position calibration method for a microscopic imaging system based on LED array illumination according to claim 1, characterized in that... In step 14, the formula for calculating the illumination wave vector corresponding to each LED illumination is: ， ， in, For the first Under LED illumination, the incident light in the image coordinate system of The angle between the direction and the optical axis, For the first Under LED illumination, the incident light in the image coordinate system of The angle between the direction and the optical axis.

Images