A random mask based lensless microscopic imaging system and method
By integrating random masks and multi-angle illumination into a lensless microscopy system, combined with a self-calibration algorithm, the phase recovery problem in lensless microscopy is solved, achieving high-resolution and large field-of-view imaging, simplifying the system structure and improving imaging stability and speed.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- BOHAI UNIV
- Filing Date
- 2026-02-27
- Publication Date
- 2026-06-09
Smart Images
Figure CN122172435A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of lensless microscopy imaging technology, and more particularly to a lensless microscopy imaging system and method based on random masks. Background Technology
[0002] Lensless microscopy is an imaging method that directly records the diffraction pattern of a sample using an image sensor and then reconstructs a high-resolution image of the sample using reconstruction algorithms. Compared to traditional optical microscopes, this technique eliminates complex lens groups such as objectives, thus avoiding optical aberrations and overcoming the spatial bandwidth product limitations of traditional microscopes. Because it eliminates the need for lenses, the system structure is more compact, enabling a super-large field of view equal to the area of the image sensor, while also offering advantages such as low cost and high resolution. These characteristics make lensless microscopy a promising technology for applications in resource-constrained areas and portable testing.
[0003] However, the core challenge of lensless imaging technology lies in the phase recovery problem. Since image sensors can only record intensity information, the phase information of the sample is lost during imaging, making the reconstruction of the sample from a single intensity image an ill-posed inverse problem, prone to reconstruction stalls or deblurring. To address this issue, existing technologies typically employ diverse measurement strategies, such as changing the illumination angle, acquiring multiple defocus plane images, or using speckle illumination. However, these approaches often rely on sophisticated mechanical scanning devices or scattering elements that are difficult to control precisely, increasing system complexity and reducing imaging stability and speed, thus limiting the practical application of this technology. Summary of the Invention
[0004] To address the technical problems of existing lensless microscopy techniques, such as reliance on mechanical scanning, system complexity, poor stability, and poor imaging quality, this invention provides a lensless microscopy system and method based on random masks. This invention combines multi-angle illumination with an integrated random mask, resulting in a simple structure and the ability to simultaneously achieve a large field of view and high-resolution imaging. By directly integrating the random mask onto the glass protective cover of the image sensor, structural stability, a large field of view, and high-resolution imaging can be achieved without precise mechanical scanning.
[0005] The technical means employed in this invention are as follows:
[0006] A lensless microscopic imaging system based on a random mask includes an LED array, a random mask, an image sensor, and a reconstruction algorithm module, wherein: The LED array is positioned above the sample under test and is used to provide multi-angle spherical wave illumination. The random mask is a random binary pattern integrated on the glass protective cover of the image sensor, used to modulate the amplitude of the light field passing through the sample to be tested. The image sensor is positioned below the sample to be tested and is used to acquire multiple low-resolution sample intensity images formed by sequential illumination from the LED array. The reconstruction algorithm module is used for: During the first imaging, all low-resolution sample intensity images and initial random binary patterns are received. High-resolution sample images are reconstructed synchronously through an iterative phase recovery algorithm. The initial random binary pattern is self-calibrated during the first iterative imaging process to obtain the calibrated mask pattern. In non-first imaging, the calibrated mask pattern is directly invoked and the mask calibration step is skipped to reconstruct a high-resolution image of the new sample with fewer LED lighting and iterations than in the first imaging.
[0007] Furthermore, the LED array consists of multiple independently controllable LED units, providing complete illumination of the sample under test with uniform light distribution.
[0008] Furthermore, the random mask is directly fabricated on the glass protective cover of the image sensor using a chrome plating process. The position of the mask pattern is fixed during the manufacturing process, avoiding alignment deviations during use.
[0009] Furthermore, the random mask has a light-transmitting area and an opaque area, wherein the glass protective cover is an unplated area, i.e., a light-transmitting area; and the random binary pattern is a chrome-plated area, i.e., an opaque area.
[0010] Furthermore, the image sensor is a two-dimensional pixelated image sensor, including a CCD or CMOS image sensor.
[0011] The present invention also provides a lensless microscopy imaging method based on the lensless microscopy imaging system, comprising: S1, Prepare the sample to be tested. Initialized as a matrix of all 1s, with a random mask pattern. Initial design values; S2. Based on the test sample and the initial values of the random mask from step S1, predict the light field on the image sensor plane. ; S3, Low-resolution sample intensity image captured by the image sensor The square root replaces the light field on the plane of the image sensor. The amplitude, The phase remains unchanged, resulting in an updated light field on the image sensor plane. :
[0012] In the above formula, To represent a complex number, Represents the imaginary unit. Indicates the phase when taking a complex number; S4. Update the light field on the image sensor plane. Backpropagation to the random mask plane yields the updated, randomly mask-modulated optical field. :
[0013] In the above formula, Indicates complex conjugation; S5. Optical field modulated based on the updated random mask The light field obtained after the updated sample modulation propagates to the random mask plane :
[0014] S6. Use a self-calibration method to calibrate the alignment deviation of the random mask to obtain the updated random mask pattern. :
[0015] S7. Perform binarization on the updated random mask pattern to preserve the binary properties of the mask:
[0016] S8, Updated Light Field The light field is backpropagated to the sample plane to obtain the updated sample-modulated light field. :
[0017] S9, Light field modulated based on updated sample Update the test samples The updated test sample was obtained. :
[0018] S10. For each low-resolution sample intensity image Execute steps S2 to S9 sequentially until... This process involves iterating through all low-resolution sample intensity images, repeating the iterations until the solution converges or the preset number of iterations is reached, ultimately outputting a high-resolution sample image. and calibrated random mask pattern .
[0019] Further, step S2 includes: S21. Different LED units in the LED array are lit sequentially. For each LED unit illuminated, the image sensor records a low-resolution sample intensity image, represented as:
[0020] In the above formula, Indicates the first Low-resolution sample intensity images recorded by the image sensor when an LED is lit. , Indicates the total number of LEDs in the array. Represents spatial location coordinates; Indicates the first The spherical waves emitted when an LED light is lit; Indicates the sample to be tested; The point spread function represents the distance from the sample to the random mask. Indicates a random mask; This represents the point spread function from a random mask to an image sensor; This represents the dot product operation. This represents the convolution operation; S22. When each LED is lit, the spherical wave emitted illuminates the sample under test. Calculate the light field modulated by the sample. :
[0021] S23, Predict the light field after sample modulation Light field propagating to the random mask plane :
[0022] S24, Light Field After random mask modulation, the modulated light field is calculated. :
[0023] S25, Predict the optical field after mask modulation Light field propagating to the image sensor plane :
[0024] S26. Calculate the final recorded light field of the image sensor. intensity : .
[0025] Compared with the prior art, the present invention has the following advantages: 1. This invention integrates a random mask onto the image sensor, avoiding mask alignment issues and improving system stability.
[0026] 2. This invention achieves multi-angle illumination by controlling an LED array, eliminating the need for precision mechanical moving parts and improving imaging speed.
[0027] 3. This invention improves the reliability of reconstruction results by automatically compensating for errors in mask processing or integration through the built-in mask self-calibration function in the reconstruction algorithm.
[0028] 4. The system structure of the present invention is simple, requiring no lenses or precision displacement stages, which is conducive to realizing portable and low-cost microscopic imaging equipment. Attached Figure Description
[0029] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0030] Figure 1 This is a schematic diagram illustrating the structural principle of the lensless microscopy imaging system of the present invention.
[0031] Figure 2 This is a schematic diagram of the LED array in the lensless microscopy imaging system of the present invention.
[0032] Figure 3 This is a schematic diagram of the random mask in the lensless microscopy imaging system of the present invention.
[0033] Figure 4 This is a schematic diagram of the CMOS image sensor for the lensless microscopy imaging system of the present invention.
[0034] In the figure: 1. LED array; 2. Sample under test; 3. Random mask; 4. Image sensor; 5. Distance between LED array and sample under test; 6. Distance between sample under test and random mask; 7. Distance between random mask and CMOS image sensor; 8. LED unit; 9. Minimum physical size of the opaque area of random mask; 10. Opaque area; 11. Transparent area; 12. Glass protective cover. Detailed Implementation
[0035] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.
[0036] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.
[0037] like Figure 1 As shown, this invention provides a lensless microscopic imaging system based on a random mask, comprising: an LED array 1, a random mask 3, an image sensor 4, and a reconstruction algorithm module, wherein: The LED array 1 is positioned above the sample 2 to provide multi-angle spherical wave illumination; The random mask 3 is a random binary pattern, integrated on the glass protective cover 12 of the image sensor 4, and is used to modulate the amplitude of the light field passing through the sample 2 to be tested. The image sensor 4 is positioned below the sample 2 to be tested and is used to acquire multiple low-resolution sample intensity images formed under sequential illumination by the LED array 1. The reconstruction algorithm module is used for: During the first imaging, all low-resolution sample intensity images and initial random binary patterns are received. High-resolution sample images are reconstructed synchronously through an iterative phase recovery algorithm. The initial random binary pattern is self-calibrated during the first iterative imaging process to obtain the calibrated mask pattern. In non-first imaging, the calibrated mask pattern is directly invoked and the mask calibration step is skipped to reconstruct a high-resolution image of the new sample with fewer LED lighting and iterations than in the first imaging.
[0038] In a specific implementation, as a preferred embodiment of the present invention, the LED array 1 consists of multiple independently controllable LED units 8, providing complete illumination of the sample under test with uniform illumination. For example... Figure 2 The LED array shown consists of 9×9 LED units 8 with a center wavelength of 530nm, a spacing of 4mm between adjacent LEDs, and a distance 5 of 270mm between the LED array 1 and the sample 2 under test.
[0039] In specific implementation, as a preferred embodiment of the present invention, such as Figure 3 As shown, the random mask 3 is directly fabricated on the glass protective cover 12 of the image sensor 4 using a chrome plating process. The position of the mask pattern is fixed during manufacturing to avoid alignment deviations during use. The random mask 3 has a light-transmitting area 11 and an opaque area 10. The glass protective cover 12 is an unchrome-plated area, i.e., the light-transmitting area 11; the random binary pattern is a chrome-plated area, i.e., the opaque area 10. The minimum physical size 11 of the opaque area 10 is 20 μm, the overall light transmittance of the mask pattern is 50%, and the distance 7 between the random mask 3 and the image sensor 4 is 1 mm.
[0040] In a specific implementation, as a preferred embodiment of the present invention, the image sensor 4 is a two-dimensional pixelated image sensor, including a CCD or CMOS image sensor. For example... Figure 4 The image sensor 4 shown is a CMOS image sensor, placed below the sample, with a pixel size of 3.45μm and a resolution of 2448×2048.
[0041] The present invention also provides a lensless microscopy imaging method based on the above-mentioned lensless microscopy imaging system, comprising: S1, Prepare the sample to be tested. Initialized as a matrix of all 1s, with a random mask pattern. Initial design values; S2. Based on the test sample and the initial values of the random mask from step S1, predict the light field on the image sensor plane. ; S3, Low-resolution sample intensity image captured by the image sensor The square root replaces the light field on the plane of the image sensor. The amplitude, The phase remains unchanged, resulting in an updated light field on the image sensor plane. :
[0042] In the above formula, To represent a complex number, Represents the imaginary unit. Indicates the phase when taking a complex number; S4. Update the light field on the image sensor plane. Backpropagation to the random mask plane yields the updated, randomly mask-modulated optical field. :
[0043] In the above formula, Indicates complex conjugation; S5. Optical field modulated based on the updated random mask The light field obtained after the updated sample modulation propagates to the random mask plane :
[0044] S6. Use a self-calibration method to calibrate the alignment deviation of the random mask to obtain the updated random mask pattern. :
[0045] S7. Perform binarization on the updated random mask pattern to preserve the binary properties of the mask:
[0046] S8, Updated Light Field The light field is backpropagated to the sample plane to obtain the updated sample-modulated light field. :
[0047] S9, Light field modulated based on updated sample Update the test samples The updated test sample was obtained. :
[0048] S10. For each low-resolution sample intensity image Execute steps S2 to S9 sequentially until... This process involves iterating through all low-resolution sample intensity images, repeating the iterations until the solution converges or the preset number of iterations is reached, ultimately outputting a high-resolution sample image. and calibrated random mask pattern .
[0049] In a specific implementation, as a preferred embodiment of the present invention, step S2 includes: S21. Different LED units in the LED array are lit sequentially. For each LED unit illuminated, the image sensor records a low-resolution sample intensity image, represented as:
[0050] In the above formula, Indicates the first Low-resolution sample intensity images recorded by the image sensor when an LED is lit. , Indicates the total number of LEDs in the array. Represents spatial location coordinates; Indicates the first The spherical waves emitted when an LED light is lit; Indicates the sample to be tested; The point spread function represents the distance from the sample to the random mask. Indicates a random mask; This represents the point spread function from a random mask to an image sensor; This represents the dot product operation. This represents the convolution operation; in this embodiment, a total of 81 low-resolution sample intensity images were recorded.
[0051] S22. When each LED is lit, the spherical wave emitted illuminates the sample under test. Calculate the light field modulated by the sample. :
[0052] S23, Predict the light field after sample modulation Light field propagating to the random mask plane :
[0053] S24, Light Field After random mask modulation, the modulated light field is calculated. :
[0054] S25, Predict the optical field after mask modulation Light field propagating to the image sensor plane :
[0055] S26. Calculate the final recorded light field of the image sensor. intensity : .
[0056] In this embodiment, after the initial reconstruction and mask calibration, a precise mask pattern is obtained. When observing new samples, it is not necessary to re-illuminate all LED array 1; only a portion of LED units 8 (e.g., 9 units) need to be illuminated sequentially to record the corresponding number of low-resolution sample intensity images. All low-resolution sample intensity images and the calibrated random mask pattern are input into the reconstruction algorithm, initializing the samples as an all-one matrix and skipping the mask pattern update step. This significantly reduces the number of iterations while rapidly obtaining high-quality reconstructed sample images. This method, after the initial mask calibration, can drastically reduce data acquisition and accelerate reconstruction, making it highly suitable for rapid imaging applications.
[0057] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.
Claims
1. A lensless microscopic imaging system based on random masks, characterized in that, include: The system comprises an LED array, a random mask, an image sensor, and a reconstruction algorithm module, among which: The LED array is positioned above the sample under test and is used to provide multi-angle spherical wave illumination. The random mask is a random binary pattern integrated on the glass protective cover of the image sensor, used to modulate the amplitude of the light field passing through the sample to be tested. The image sensor is positioned below the sample to be tested and is used to acquire multiple low-resolution sample intensity images formed by sequential illumination from the LED array. The reconstruction algorithm module is used for: During the first imaging, all low-resolution sample intensity images and initial random binary patterns are received. High-resolution sample images are reconstructed synchronously through an iterative phase recovery algorithm. The initial random binary pattern is self-calibrated during the first iterative imaging process to obtain the calibrated mask pattern. In non-first imaging, the calibrated mask pattern is directly invoked and the mask calibration step is skipped to reconstruct a high-resolution image of the new sample with fewer LED lighting and iterations than in the first imaging.
2. The lensless microscopic imaging system based on a random mask according to claim 1, characterized in that, The LED array consists of multiple independently controllable LED units, providing complete illumination of the sample under test with uniform light distribution.
3. The lensless microscopic imaging system based on a random mask according to claim 1, characterized in that, The random mask is directly fabricated on the glass protective cover of the image sensor using a chrome plating process. The position of the mask pattern is fixed during the manufacturing process to avoid alignment deviations during use.
4. The lensless microscopic imaging system based on a random mask according to claim 3, characterized in that, The random mask has a light-transmitting area and an opaque area. The glass protective cover is an unplated area, i.e., a light-transmitting area; the random binary pattern is a chrome-plated area, i.e., an opaque area.
5. A lensless microscopic imaging system based on a random mask according to claim 1, characterized in that, The image sensor is a two-dimensional pixelated image sensor, including CCD or CMOS image sensors.
6. A lensless microscopy method based on the lensless microscopy imaging system according to any one of claims 1-5, characterized in that, include: S1, Prepare the sample to be tested. Initialized as a matrix of all 1s, with a random mask pattern. Initial design values; S2. Based on the test sample and the initial values of the random mask from step S1, predict the light field on the image sensor plane. ; S3, Low-resolution sample intensity image captured by the image sensor The square root replaces the light field on the plane of the image sensor. The amplitude, The phase remains unchanged, resulting in an updated light field on the image sensor plane. : In the above formula, To represent a complex number, Represents the imaginary unit. Indicates the phase when taking a complex number; S4. Update the light field on the image sensor plane. Backpropagation to the random mask plane yields the updated, randomly mask-modulated optical field. : In the above formula, Indicates complex conjugation; S5. Optical field modulated based on the updated random mask The light field obtained after the updated sample modulation propagates to the random mask plane : S6. Use a self-calibration method to calibrate the alignment deviation of the random mask to obtain the updated random mask pattern. : S7. Perform binarization on the updated random mask pattern to preserve the binary properties of the mask: S8, Updated Light Field The light field is backpropagated to the sample plane to obtain the updated sample-modulated light field. : S9, Light field modulated based on updated sample Update the test samples The updated test sample was obtained. : S10. For each low-resolution sample intensity image Execute steps S2 to S9 sequentially until... This process involves iterating through all low-resolution sample intensity images, repeating the iterations until the solution converges or the preset number of iterations is reached, ultimately outputting a high-resolution sample image. and calibrated random mask pattern .
7. The lensless microscopic imaging method according to claim 6, characterized in that, Step S2 includes: S21. Different LED units in the LED array are lit sequentially. For each LED unit illuminated, the image sensor records a low-resolution sample intensity image, represented as: In the above formula, Indicates the first Low-resolution sample intensity images recorded by the image sensor when an LED is lit. , This indicates the total number of LEDs in the array. Represents spatial location coordinates; Indicates the first The spherical waves emitted when an LED light is lit; Indicates the sample to be tested; The point spread function represents the distance from the sample to the random mask. Indicates a random mask; This represents the point spread function from a random mask to an image sensor; This represents the dot product operation. This represents the convolution operation; S22. When each LED is lit, the spherical wave emitted illuminates the sample under test. Calculate the light field modulated by the sample. : S23, Predict the light field after sample modulation Light field propagating to the random mask plane : S24, Light Field After random mask modulation, the modulated light field is calculated. : S25, Predict the optical field after mask modulation Light field propagating to the image sensor plane : S26. Calculate the final recorded light field of the image sensor. intensity : 。