Single-pixel microscopic depth-of-field extension method and system

Abstract

The application discloses a single-pixel microscopic depth-of-field extension method and system. The method comprises the following steps: setting a surface array detector in an imaging area to collect light intensity distribution, and determining a point spread function (PSF) spatial distribution field of a system; changing the relative position of the surface array detector and a microscopic objective lens to obtain PSF spatial distribution fields of multiple depths; collecting light intensity response signals of a sample to be measured under a modulation pattern by using a single-pixel detector; and combining the PSF spatial distribution fields of the multiple depths to perform calculation and reconstruction, thereby generating a deblurring image of the sample to be measured. The application effectively solves the problems of a shallow depth of field of a single-pixel microscopic system and easy defocus blurring of a three-dimensional sample by combining a surface array detector with a compressed sensing algorithm, and significantly extends the depth-of-field range of microscopic imaging.

Description

Technical Field

[0001] This invention relates to the field of optical imaging technology, and in particular to a method and system for extending the depth of field in single-pixel microscopy. Background Technology

[0002] In the field of computational imaging, single-pixel imaging, as an advanced imaging technology, has been widely applied in various industrial scenarios due to its wide imaging spectrum and excellent imaging performance in low-light environments. Unlike traditional imaging techniques, single-pixel imaging reconstructs the target image by projecting encoded patterns and acquiring corresponding light intensity values ​​using a single-pixel detector, combined with mathematical transformations. In the biomedical field, single-pixel microscopy has attracted much attention due to its unique advantages in applications such as observing the microscopic structure of cells.

[0003] However, the depth of field in microscopic optical imaging systems is typically very limited, especially when the object being imaged has three-dimensional characteristics. This makes it prone to defocusing, making it difficult to capture details of the object on the defocus plane. For example, for a microscope objective with a magnification of 10x and a numerical aperture of 0.25 NA, the original depth of field is only on the micrometer scale, approximately 8.8 micrometers. This shallow depth of field limits the application of single-pixel microscopy in observing three-dimensional samples. To address the imaging blur problem, accurately measuring the spatial distribution of the point spread function at the imaging plane is crucial for estimating the degree of blur in the target image and performing deblurring.

[0004] Currently, the measurement methods for the spatial distribution field of the point spread function (PSF) of a single-pixel imaging system are mainly based on fiber optic technology. This method places one end of the fiber as a light intensity receiver on the imaging plane to receive the light intensity of the modulation pattern, while the other end is coupled to a single-pixel detector for signal acquisition. An inverse transform is then used to obtain the single point spread function (PSF) at a specific location. By measuring the PSF at different locations, the spatial distribution field of the PSF of the imaging system is obtained.

[0005] However, this method has significant drawbacks when applied to single-pixel microscopy systems. First, the small field of view in single-pixel microscopy results in tiny modulation patterns, only on the millimeter scale, making alignment of the fiber optic receiver extremely difficult. It's hard to visually determine whether the modulation pattern accurately illuminates the receiver, increasing operational complexity. Second, as the magnification of the microscope objective increases, the imaging range shrinks further, exponentially increasing the difficulty of measuring the spatial distribution of the point spread function (PSF) using fiber optics. Furthermore, the fiber optic method can only acquire the PSF at one location in a single measurement. To obtain PSFs at different locations or depths, multiple mechanical movements and measurements are necessary. This is not only time-consuming but also prone to introducing measurement errors due to fiber placement variations, severely impacting subsequent depth-of-field extension.

[0006] Therefore, how to accurately obtain the spatial distribution field of the point spread function (PSF) of a single-pixel microscopic imaging system while ensuring ease of operation, and how to use the measured spatial distribution field of the point spread function (PSF) to achieve depth of field extension of the single-pixel imaging system, thereby solving the problem of easy defocusing in stereoscopic sample imaging, has become an urgent technical challenge. Summary of the Invention

[0007] The main objective of this invention is to provide a method and system for extending the depth of field in single-pixel microscopy. The aim is to accurately obtain the spatial distribution field of the point spread function (PSF) of a single-pixel microscopy imaging system while ensuring ease of operation, and to use it to extend the depth of field of the single-pixel microscopy imaging system, thereby solving the problem of easy defocusing in three-dimensional sample imaging.

[0008] To achieve the above objectives, this invention proposes a single-pixel microscopic depth-of-field extension method for measuring the spatial distribution field of the point spread function (PSF) and extending the depth of field in a single-pixel microscopic imaging system. This method is applied to a single-pixel microscopic imaging system comprising a spatial light modulator, a microscope objective, and a single-pixel detector, and includes: Point spread function spatial distribution field measurement steps: Set up an area array detector in the imaging area of ​​the imaging system to receive the modulation pattern light intensity of the single pixel microscopic imaging system at the imaging plane, and perform inverse transformation on the light intensity signal received by each pixel in the area array detector to determine the point spread function spatial distribution field of the imaging system. Three-dimensional calibration steps: Change the relative position of the area array detector and the microscope objective in the optical axis direction, or change the focal plane position of the imaging system, and repeat the point spread function (PSF) spatial distribution field measurement steps to obtain the point spread function spatial distribution field corresponding to multiple different depths; Imaging acquisition steps: The sample to be tested is placed in the imaging area, and the light intensity response signal of the sample to be tested under the action of the modulation pattern projected by the spatial light modulator is acquired using the single-pixel detector; Reconstruction and expansion steps: Using the light intensity response signal and combining it with the spatial distribution fields of the point spread function at multiple different depths, a deblurred image of the sample to be tested is generated.

[0009] Preferably, the area array detector is a CCD pixel array; in the point spread function (PSF) spatial distribution field measurement step, based on the Helmholtz reciprocity principle, each pixel in the CCD pixel array is regarded as an ideal point light source, and the spatial light modulator is regarded as a light intensity signal receiving device. The point spread function spatial distribution field is determined by calculating the response of the CCD pixel array to the light intensity of the modulation pattern.

[0010] Preferably, the point spread function spatial distribution field measurement step further includes: performing response calculations on different pixel positions of the CCD pixel array within the same imaging plane, and obtaining the point spread function corresponding to different positions within the same imaging plane; in the reconstruction and expansion step, deblurring is performed using the point spread function at the corresponding position for different regions of the image of the sample to be tested.

[0011] Preferably, the imaging system further includes a screen disposed between the spatial light modulator and the microscope objective; in the point spread function spatial distribution field measurement step and the imaging acquisition step, the spatial light modulator is controlled to project the modulation pattern onto the screen, and the modulation pattern is coupled into the microscope objective through the screen.

[0012] Preferably, the three-dimensional calibration step specifically includes: when changing the relative position of the area array detector and the microscope objective in the optical axis direction, displaying the modulation pattern image acquired by the area array detector in real time using a screen connected to the area array detector, and adjusting the relative position according to the blurring degree of the modulation pattern image to determine the measured depth position.

[0013] Preferably, in the reconstruction extension step, the computational reconstruction employs a compressed sensing algorithm; the compressed sensing algorithm uses the light intensity response signal measured by the single-pixel detector as a measurement value and the defocused modulation pattern projected by the spatial light modulator on the defocus plane as a measurement matrix to recover a clear image of the sample under test.

[0014] Preferably, the modulation pattern projected by the spatial light modulator is an illumination coding pattern generated by Hadamard transform or Fourier transform; the illumination coding pattern is one of Hadamard pattern, Fourier four-step phase shift pattern, and Fourier three-step phase shift pattern.

[0015] Preferably, the reconstruction and expansion step specifically includes: convolving the measured point spread function (PSF) with the original modulation pattern projected by the spatial light modulator and shaping it into a measurement matrix; inputting the measurement matrix and the measured original signal of the sample to be tested into the compressed sensing algorithm to obtain the deblurred pattern result.

[0016] Preferably, the spatial light modulator is one of a digital micromirror device, a liquid crystal display, or a projector; and the modulation mode of the spatial light modulator in the imaging acquisition step is active.

[0017] Preferably, the physical size of each pixel in the CCD pixel array is on the micrometer scale.

[0018] This application also discloses a single-pixel microscopic depth-of-field extension system for measuring and extending the spatial distribution field of the point spread function (PSF) of a single-pixel microscopic imaging system, comprising: an illumination modulation module including a spatial light modulator for projecting a modulation pattern; a microscopic imaging optical path including a microscope objective for transmitting the light field; a detection module including a single-pixel detector for acquiring light intensity response signals; a calibration module including an area array detector for being placed in the imaging region during the measurement phase of the point spread function (PSF) spatial distribution field to acquire the light intensity distribution of the modulation pattern; and a processing control module configured to perform the method described in any of the preceding claims.

[0019] The above technical solution has the following advantages: This invention effectively solves the problem of shallow depth of field in single-pixel microscopy imaging systems by using an area array detector in the imaging area to assist in measuring the spatial distribution of the point spread function (PSF) of the single-pixel microscopy imaging system, and then reconstructing it using a compressed sensing algorithm. The method of determining the PSF spatial distribution of a single-pixel microscopy imaging system by using an area array detector to collect light intensity distribution data can acquire information from multiple locations within the field of view at once, significantly improving measurement efficiency and accuracy compared to traditional fiber optic scanning methods. By changing the relative position of the area array detector and the microscope objective or adjusting the focal plane, three-dimensional PSF spatial distribution fields corresponding to different depths can be quickly established. In the subsequent imaging process of the sample under test, these precisely calibrated PSFs are used in conjunction with the compressed sensing algorithm for calculation and reconstruction. This allows for the recovery of a clear image from the light intensity response signal measured on the defocused plane, thereby extending the physical depth of field of the microscopy system several times, for example, from approximately 8.8 micrometers to approximately 40 micrometers. This achieves full-focus clear imaging of three-dimensional samples with a certain thickness. Furthermore, this method has a compact system structure, eliminates the need for complex mechanical scanning devices, and significantly reduces hardware costs and operational complexity. Attached Figure Description

[0020] The present invention will now be described in detail with reference to specific embodiments and accompanying drawings, wherein: Figure 1 The following is a schematic diagram of the system principle and process provided in the embodiments of the present invention; wherein, (a) is a schematic diagram of the single-pixel microscopic imaging system principle, (b) is a schematic diagram of the CCD array measuring the spatial distribution field of the point spread function (PSF) of the single-pixel microscopic imaging system principle, and (c) is a schematic diagram of the deblurring principle using the spatial distribution field of the point spread function (PSF).

[0021] Figure 2 The image provided in this embodiment of the invention is a comparison of the original blurred image and the images after deblurring based on PSF1 and PSF2.

[0022] Figure 3The point diffusion function patterns for different regions are provided in the embodiments of the present invention.

[0023] Figure 4 A three-dimensional histogram of PSF1 provided in an embodiment of the present invention.

[0024] Figure 5 A three-dimensional histogram of PSF2 provided in an embodiment of the present invention. Detailed Implementation

[0025] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are merely some, not all, of the embodiments of the present application. All other embodiments obtained by those skilled in the art based on the embodiments of the present application without creative effort should fall within the scope of protection of the present application.

[0026] like Figures 1 to 5 As shown, this application provides a method for measuring the spatial distribution field of the point spread function (PSF) and extending the depth of field in a single-pixel microscopic imaging system. This method is used for measuring the spatial distribution field of the PSF in a single-pixel microscopic imaging system and extending its depth of field. Existing single-pixel microscopic imaging systems have extremely shallow depths of field, typically only at the micrometer level. For example, the original depth of field of an objective lens with a magnification of 10x and a numerical aperture of 0.25 NA is approximately... This leads to the problem of blurry, out-of-focus imaging of three-dimensional samples. This application utilizes the spatial distribution field of the point spread function of the single-pixel microscopic imaging system of the area array detector-assisted measurement system, and combines it with the compressed sensing algorithm for calculation and reconstruction, thereby significantly expanding the depth range of microscopic imaging.

[0027] Example 1 This embodiment first introduces a method for measuring the spatial distribution field of the point spread function (PSF) and extending the depth of field in a single-pixel microscopic imaging system. This method is used for measuring the spatial distribution field of the PSF in a single-pixel microscopic imaging system and extending the depth of field. For example... Figure 1 As shown, the system method mainly includes an illumination modulation module, a microscopic imaging optical path, a detection module, a calibration module, and a processing and control module.

[0028] The illumination modulation module is primarily used to provide an active, structured illumination light field. This module includes a light source and a digital micromirror device (DMD). The light source is preferably an LED capable of providing a uniform illumination light field. The DMD is used to spatially modulate the incident light to generate a specific illumination coding pattern. In this embodiment, the spatial light modulator specifically employs a DMD. Of course, in other alternative embodiments, the spatial light modulator can also be a liquid crystal display (LCD) or a projector.

[0029] The optical path for microscopic imaging is positioned along the projection direction of the spatial light modulator, with the microscope objective as its core component. The microscope objective transmits the modulation pattern projected by the spatial light modulator to the imaging area. To address the challenge of precisely coupling the tiny modulation pattern into the microscope objective, this embodiment also includes a screen between the spatial light modulator and the microscope objective. The spatial light modulator projects the modulation pattern onto the screen, which acts as a relay imaging surface, coupling the modulation pattern into the microscope objective. This design significantly reduces the difficulty of aligning the optical path.

[0030] The detection module includes a single-pixel detector (PD), such as a photodiode or photomultiplier tube. The single-pixel detector is used to acquire the total light intensity response signal of the sample under the action of the modulation pattern.

[0031] The calibration module is a key component in this embodiment for measuring the spatial distribution field of the point spread function (PSF), and it includes an area array detector. Specifically, the area array detector employs a CCD pixel array, which is placed in the imaging area only during the PSF spatial distribution field measurement phase of the system. It is used to replace the position of the sample under test to acquire the light intensity distribution transmitted through the microscope objective. The area array detector is connected to a display screen for real-time display of the acquired modulation pattern image.

[0032] The processing control module is typically a computer or a control circuit equipped with corresponding algorithms. It is communicatively connected to the spatial light modulator, single-pixel detector, and area array detector, respectively, and is used to control the projection of modulation patterns, signal acquisition, and execution of subsequent image reconstruction algorithms.

[0033] Based on the above system structure, this embodiment elaborates on a method for measuring the spatial distribution field of the point spread function (PSF) of a single-pixel microscopic imaging system and extending the depth of field. The method is used for measuring the spatial distribution field of the point spread function (PSF) of a single-pixel microscopic imaging system and extending the depth of field. The method mainly includes a point spread function (PSF) spatial distribution field measurement step, a three-dimensional calibration step, an imaging acquisition step, and a reconstruction and extension step.

[0034] First, the point spread function spatial distribution field measurement step is performed. An area array detector is positioned near the imaging plane of the microscope objective within the imaging region of the imaging system. A spatial light modulator projects a modulation pattern. This modulation pattern can be a Hadamard substrate pattern or other specific detection patterns. The area array detector receives the modulation pattern transmitted through the microscope objective. The processing and control module determines the point spread function spatial distribution field of the imaging system corresponding to the imaging plane position based on the light intensity distribution collected by the area array detector.

[0035] In this step, this embodiment creatively obtains the point spread function based on the Helmholtz reciprocity principle. Specifically, due to the reversibility of the optical path, each pixel in the CCD pixel array of the area array detector can be regarded as an ideal point light source, and the spatial light modulator can be regarded as a receiving device for light intensity signals. It is known that the physical size of each pixel in the CCD pixel array is extremely small, typically on the micrometer scale; for example, the pixel size used in this embodiment is... Specifically, a Hikvision MV-CA013-21UM camera can be used, which closely approximates the definition of an ideal point light source. By calculating the response of each pixel in the CCD pixel array to the intensity of the modulated pattern light, the point spread function at that location can be accurately derived. Compared to existing methods that use fiber optic tips for scanning measurements, this embodiment utilizes a CCD pixel array to acquire information from multiple locations within the field of view simultaneously, significantly improving measurement efficiency and accuracy.

[0036] Furthermore, considering that the aberrations at the edge and center of the microscope's field of view may differ, the measurement step of the point spread function (PSF) spatial distribution field also includes calculating the response of the CCD pixel array at different pixel positions within the same imaging plane. The processing control module acquires the point spread function corresponding to different positions within the same imaging plane (e.g., Figure 4 and Figure 5 (As shown in different forms), construct a spatial distribution field of point spread function that varies with spatial location.

[0037] Next, a three-dimensional calibration step is performed to obtain the spatial distribution of the point spread function (PSF) at different depths. The relative positions of the area array detector and the microscope objective along the optical axis (Z-axis) are changed using a precision displacement stage. Alternatively, the area array detector can be kept stationary, and the focal plane position can be changed by adjusting the focusing knob of the microscope imaging system. At each new depth location, the above point spread function (PSF) spatial distribution measurement steps are repeated to obtain the spatial distribution at multiple different depths. Point spread function .

[0038] To improve operational efficiency and ensure measurement validity during the 3D calibration process, this embodiment utilizes a screen connected to the area array detector to display the acquired modulation pattern image in real time. Operators can roughly adjust the relative position based on the degree of blurriness of the image displayed on the screen, avoiding blind scanning and significantly reducing operational complexity.

[0039] After calibration, the area array detector is removed, and the imaging acquisition step begins. The sample to be tested is placed on the stage in the imaging area. A series of illumination-coded patterns are projected by the spatial light modulator. In this embodiment, the illumination-coded patterns are generated by Hadamard transform. Of course, Fourier four-step phase-shift patterns or Fourier three-step phase-shift patterns can also be used. These patterns are projected onto the sample to be tested after being coupled into the microscope objective via a screen. The light intensity response signal of the sample to be tested under each illumination-coded pattern is acquired using a single-pixel detector.

[0040] Finally, the reconstruction extension step is performed. The processing control module uses the light intensity response signal collected by the single-pixel detector and combines it with the previously calibrated point spread functions at multiple different depths to perform calculations and reconstruction, generating a deblurred image of the sample under test.

[0041] The specific computational reconstruction process employs either compressed sensing or inverse transform algorithms.

[0042] First, a mathematical model of the imaging process is constructed. This model includes all defocus modulation patterns. The measurement matrix formed The distribution function of the image of the sample under test Its vectorized form is I, and the measurement value B of the single-pixel detector. Therefore, the measurement value of the single-pixel detector can be described as... .

[0043] The degree of defocus in an imaging system can be mathematically described as the point spread function (PSF). The PSF is typically described as a Gaussian filter, and its distribution function is: It is given by the following formula (1) σ is the standard deviation of the Gaussian distribution, and (x,y) are the spatial coordinates. Defocus modulation pattern. Mathematically, it can be described as the original modulation pattern. In relation to the point spread function (PSF) of the imaging system (i.e.) The convolution of ) is (2) When the sample to be tested is on the defocused plane, N defocused modulation patterns Then the nth defocus modulation pattern Corresponding single-pixel detector measurement value for: (3) When using compressed sensing reconstruction, equation (3) is usually described in vector form: (4) B and I are and The vectorized form, It is all defocus modulation patterns The measurement matrix formed The nth row is the nth defocus pattern. The vectorized form of I. Formula (4) is solved using the compressed sensing algorithm. That is, I is reconstructed from B. For a three-dimensional sample with a certain thickness, this embodiment utilizes multiple point spread functions at different depths obtained during the three-dimensional calibration step. For different regions or depth layers of the sample image under test, point spread functions (PSFs) at corresponding locations or depths are used for deblurring. For example, the image can be divided into blocks, and different PSFs can be used to restore each block. Based on a gradient evaluation function, such as Tenengrad gradient sharpness or Laplacian sharpness, the result with the best restoration effect is selected as the final image for that region. Alternatively, a multi-PSF joint deconvolution algorithm can be used to directly restore the full-focus image.

[0044] Through the above method, this embodiment can reduce the original quantity to approximately The physical depth of field is extended to approximately It achieves approximately 4 times the depth of field improvement, effectively solving the problem of defocusing of three-dimensional samples in single-pixel microscopy, and requires no complex mechanical scanning structure, making the system compact and cost-controllable.

[0045] Example 2 This embodiment, based on Embodiment 1, further elaborates on the specific algorithmic implementation logic of the reconstruction and expansion steps in the above system, as well as the method for generating the illumination modulation pattern. This embodiment focuses on solving how to accurately recover a clear image of the sample under test using a specific illumination coding pattern and computational model, given the point spread function. The overall algorithm principle and flow are as follows: Figure 1 As shown in (c).

[0046] In the imaging acquisition step, the modulation pattern projected by the spatial light modulator preferably adopts the illumination-coded pattern generated by the Hadamard transform. Compared with random speckle patterns, the Hadamard substrate pattern has higher sampling efficiency and reconstruction quality. In actual operation, the single-pixel detector sequentially acquires the light intensity response signals under the illumination of the positive and negative Hadamard substrate patterns. Through differential calculation, background noise can be effectively removed and the spectral coefficients of the image of the object under test can be extracted.

[0047] In the reconstruction and extension steps, the core lies in constructing an accurate measurement model and solving the inverse problem. The processing control module constructs a measurement matrix containing all defocused modulation patterns. This measurement matrix is ​​based on the convolution of the measured point spread function (PSF) with all the original undefocused modulation patterns projected by the spatial light modulator. Accurately obtaining the degree of defocus in the image of the sample under test is crucial for deblurring the defocused image. This measurement model is based on the measurement matrix obtained by convolving all the original undefocused modulation patterns projected by the spatial light modulator with the measured point spread function (PSF). The distribution function of the original image of the sample to be tested This is established based on the relationship between the light intensity response signal and the single-pixel detector measurement value B. In the physical process of single-pixel imaging, this relationship is crucial. It can be represented as: I is the distribution function of the original image of the sample to be tested. The vectorized form of .

[0048] Since directly inverting the convolution matrix is ​​often ill-posed and computationally intensive, this embodiment employs a compressed sensing algorithm for computational reconstruction. The compressed sensing algorithm uses the light intensity response signal as the measurement value. The measured point spread function (PSF) and the convolution result of all the original undefocused modulation patterns projected by the spatial light modulator are used as the measurement matrix and input together into the compressed sensing algorithm for image reconstruction. Specifically, the minimum Lagrange multiplication method is used to reconstruct the image. (5) in express Norm, and It is a Lagrange multiplier. yes In the The discrete gradient vector at point β and μ are the penalty parameters of the model. In this way, the system can directly recover the clear deconvolutioned image from compressed single-pixel measurement data without first reconstructing the blurred image and then performing post-processing to deblur, thereby reducing information loss and noise amplification.

[0049] Furthermore, this embodiment further explains the calculation details based on the Helmholtz reciprocity principle for obtaining the point spread function (PSF). In the spatial distribution field measurement step of the PSF, the area array detector is a CCD pixel array. Based on reciprocity, each pixel in the CCD pixel array is regarded as an ideal point light source, and the spatial light modulator is regarded as a light intensity signal receiving device. When the spatial light modulator projects a set of orthogonal and complete substrate patterns, such as a Hadamard substrate or a Fourier substrate, each pixel on the CCD pixel array will record a corresponding set of light intensity values. This set of light intensity values ​​is actually the coefficient of the PSF at the pixel location in the transform domain, i.e., the frequency domain or the Hadamard domain. By performing the corresponding inverse transform on these coefficients, the PSF at the pixel location can be directly obtained in the spatial domain. This method avoids the positional errors caused by mechanical scanning of point light sources, and can acquire PSF information for all positions within the field of view in parallel using a single measurement sequence.

[0050] Example 3 This embodiment supplements the foregoing embodiments, focusing on the specific processing strategies for targets with large depth of field or multiple depths, as well as the specific experimental parameter configurations.

[0051] In actual microscopic imaging scenarios, the sample under test often has a certain thickness, which may cause different areas in the same image to be in different out-of-focus states. Figure 3 The two-dimensional morphologies of two typical point spread functions, PSF1 and PSF2, obtained in the experiment are shown. Figure 4 and Figure 5 The three-dimensional energy distribution histograms corresponding to these two point spread functions are shown respectively. To address this issue, the point spread function (PSF) spatial distribution field measurement step in this embodiment further includes calculating the response of different pixel positions of the CCD pixel array within the same imaging plane, obtaining the point spread function corresponding to different positions within the same imaging plane. Simultaneously, multiple depth values ​​obtained from the three-dimensional calibration step are also considered. The point spread function library.

[0052] In the reconstruction and expansion step, the processing control module adopts a block processing and fusion strategy. For different regions of the sample image under test, point spread functions at corresponding locations are used for deblurring. Figure 2 The comparison of specific deblurring effects is shown. Among them, Figure 2 (a) is a blurry original image that was directly captured, and the details are difficult to discern; Figure 2 (b)(c) To deblur using the multi-depth, multi-location PSF (combining PSF1 and PSF2) proposed in this application. Specifically, the processing control module first uses calibrated point spread functions at different depths. Reconstruction calculations were performed on the test data to obtain a set of intermediate image sequences focused at different depths. .

[0053] Subsequently, the processing control module fuses this set of intermediate image sequences. To determine the optimal depth of focus for each pixel or region in the image, this embodiment introduces a sharpness evaluation metric. For example, the Tenengrad gradient sharpness or Laplacian sharpness of the image is calculated. For each local window in the image, its image at different depths is compared. The image patch with the highest sharpness evaluation value is selected as the final output for that region. Experimental results show that the higher the index value, the better the deblurring effect. Through this pixel-level or patch-level fusion, a sharp extended depth-of-field image covering the entire field of view can be synthesized.

[0054] In one specific configuration of this embodiment, the microscope objective has a magnification of 10x and a numerical aperture of 0.25 NA. Before employing the method of this application, the original physical depth of field of this microscope objective, calculated according to the Rayleigh criterion, is approximately [missing information]. If the undulations of the sample under test along the optical axis exceed this range, conventional imaging will result in severe blurring. Using the method of this embodiment, by moving the area array detector or adjusting the focal plane along the optical axis, for example... or PSF calibration is performed using the step size, covering approximately the optical axis direction. The range. The physical size of each pixel in the area array detector is on the micrometer scale, specifically selected... The final reconstructed image is approximately... It remains sharp throughout the depth range, achieving approximately 4 times the depth of field extension.

[0055] Furthermore, the spatial light modulator mentioned in this embodiment is not limited to a digital micromirror device (DMD). In other optional embodiments, the spatial light modulator can also be a liquid crystal display (LCD) or a projector. For the scanning device in this system, in addition to a conventional displacement stage, a high-speed rotating device or a multi-faceted reflective device can be used. For the active spatial light modulator in this system, a passive single-pixel imaging method can also be used instead. Correspondingly, the illumination light field in the imaging acquisition step is an active structured illumination light field, and its pattern generation and projection methods need to be adapted and adjusted according to the modulator's refresh rate and grayscale characteristics. For example, when using an LCD, its relatively slow response speed needs to be considered, and the phase shift step number or integration time should be adjusted appropriately.

[0056] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the claims.

Claims

1. A single-pixel microscopic depth-of-field extension method, used for measuring the spatial distribution field of the point spread function (PSF) and extending the depth of field in a single-pixel microscopic imaging system, applied to a single-pixel microscopic imaging system including a spatial light modulator, a microscope objective, and a single-pixel detector, characterized in that... include: Point spread function spatial distribution field measurement steps: Set up an area array detector in the imaging area of ​​the imaging system to receive the modulation pattern light intensity of the single pixel microscopic imaging system at the imaging plane, and perform inverse transformation on the light intensity signal received by each pixel in the area array detector to determine the point spread function spatial distribution field of the imaging system. Three-dimensional calibration steps: Change the relative position of the area array detector and the microscope objective in the optical axis direction, or change the focal plane position of the imaging system, and repeat the point spread function (PSF) spatial distribution field measurement steps to obtain the point spread function spatial distribution field corresponding to multiple different depths; Imaging acquisition steps: The sample to be tested is placed in the imaging area, and the light intensity response signal of the sample to be tested under the action of the modulation pattern projected by the spatial light modulator is acquired using the single-pixel detector; Reconstruction and expansion steps: Using the light intensity response signal and combining it with the spatial distribution fields of the point spread function at multiple different depths, a deblurred image of the sample to be tested is generated.

2. The method according to claim 1, characterized in that, The area array detector is a CCD pixel array; in the point spread function (PSF) spatial distribution field measurement step, based on the Helmholtz reciprocity principle, each pixel in the CCD pixel array is regarded as an ideal point light source, and the spatial light modulator is regarded as a light intensity signal receiving device. The point spread function spatial distribution field is determined by calculating the response of the CCD pixel array to the light intensity of the modulation pattern.

3. The method according to claim 2, characterized in that, The point spread function spatial distribution field measurement step further includes: performing response calculations on different pixel positions of the CCD pixel array within the same imaging plane, and obtaining the point spread function corresponding to different positions within the same imaging plane; in the reconstruction and expansion step, deblurring is performed using the point spread function at the corresponding position for different regions of the image of the sample to be tested.

4. The method according to claim 1, characterized in that, The imaging system further includes a screen disposed between the spatial light modulator and the microscope objective; in the point spread function spatial distribution field measurement step and the imaging acquisition step, the spatial light modulator is controlled to project the modulation pattern onto the screen, and the modulation pattern is coupled into the microscope objective through the screen.

5. The method according to claim 1, characterized in that, The three-dimensional calibration step specifically includes: when changing the relative position of the area array detector and the microscope objective in the optical axis direction, displaying the modulation pattern image acquired by the area array detector in real time using a screen connected to the area array detector, and adjusting the relative position according to the blurring degree of the modulation pattern image to determine the measured depth position.

6. The method according to claim 1, characterized in that, In the reconstruction extension step, the computational reconstruction adopts a compressed sensing algorithm; the compressed sensing algorithm uses the light intensity response signal measured by the single-pixel detector as the measurement value and the defocused modulation pattern projected by the spatial light modulator on the defocus plane as the measurement matrix to recover a clear image of the sample under test.

7. The method according to claim 6, characterized in that, The modulation pattern projected by the spatial light modulator is an illumination coding pattern generated by Hadamard transform or Fourier transform; the illumination coding pattern is one of Hadamard pattern, Fourier four-step phase shift pattern, and Fourier three-step phase shift pattern.

8. The method according to claim 1, characterized in that, The reconstruction and expansion steps specifically include: convolving the measured point spread function (PSF) with the original modulation pattern projected by the spatial light modulator and shaping it into a measurement matrix; inputting the measurement matrix and the original signal of the measured sample into the compressed sensing algorithm to obtain the deblurred pattern result.

9. The method according to claim 1, characterized in that, The spatial light modulator is one of a digital micromirror device, a liquid crystal display, or a projector; the modulation mode of the spatial light modulator in the imaging acquisition step is active.

10. The method according to claim 2, characterized in that, The physical size of each pixel in the CCD pixel array is on the micrometer scale.

11. A single-pixel microscopic depth-of-field extension system, used for measuring the spatial distribution field of the point spread function (PSF) and extending the depth of field in a single-pixel microscopic imaging system, characterized in that... include: An illumination modulation module, including a spatial light modulator, is used to project a modulation pattern; The microscopic imaging optical path includes a microscope objective for transmitting the light field; and a detection module including a single-pixel detector for acquiring light intensity response signals. The calibration module includes an area array detector, which is placed in the imaging region during the measurement phase of the point spread function (PSF) spatial distribution field to acquire the intensity distribution of the modulated pattern light. And a processing control module, configured to perform the method as described in any one of claims 1 to 10.

Images