A lensless fiber bundle phase imaging method

Abstract

This invention discloses a lensless fiber bundle phase imaging method. The method includes: acquiring a single-exposure image of a sample through an optical fiber under multi-wavelength illumination; obtaining light intensity data for reconstruction by multi-channel separation and correction of the image; reconstructing the diffraction image of the sample based on the obtained light intensity data by removing the fiber bundle structure; establishing a lensless fiber bundle imaging model and correlating the diffraction image with the complex amplitude distribution of the sample; constructing an objective function for optimization and solving to obtain the amplitude image and phase image of the sample. This invention eliminates the need for lenses in the fiber bundle, offering advantages such as simple structure and suitability for miniaturized systems, and enables stable quantitative phase imaging under single exposure.

Description

Technical Field

[0001] This invention belongs to the field of optical imaging, fiber bundle imaging and quantitative phase recovery technology, and in particular, a lensless fiber bundle phase imaging method. Background Technology

[0002] Fiber optic endoscopy allows for real-time observation of living tissues under minimally invasive conditions, making it particularly suitable for imaging and analyzing cellular-scale tissue structures, thus providing an important imaging tool for clinical practice. However, because biological cells typically exhibit weak absorption characteristics, traditional fiber optic endoscopy relies primarily on intensity imaging modes, often failing to effectively characterize changes in cellular refractive index, resulting in insufficient image contrast. Currently, existing technologies often employ fluorescent labeling to improve visibility; however, such labeling methods are somewhat invasive and may affect the physiological state of tissues or cells, thereby interfering with subsequent diagnosis.

[0003] In recent years, research combining lensless fiber bundle imaging with label-free quantitative phase imaging techniques has attracted widespread attention. This type of method eliminates the need for distal-end lenses in the fiber bundle, significantly reducing probe size and offering the advantage of high system integration. However, the inherent characteristics of fiber bundles, such as discrete core arrangement, spatially irregular sampling, and inter-core mode coupling, make phase retrieval based on light intensity measurement an inverse problem with ill-posed properties. Furthermore, when using color image sensors or multi-wavelength illumination for multi-channel imaging, color crosstalk easily arises in the image due to overlap between different spectra, further degrading image quality.

[0004] Existing lensless fiber bundle phase retrieval techniques largely rely on multi-frame acquisition and spatial light modulators to stabilize phase information through multi-state modulation. However, these methods cannot meet the demands of single-frame, high-speed, and miniaturized imaging, and are difficult to implement in fiber bundle systems. Current research has also attempted to reconstruct the phase using speckle patterns on the fiber bundle endface, but this typically requires stringent calibration conditions and is highly sensitive to noise and system environment.

[0005] Therefore, under the condition of single-frame light intensity measurement provided by fiber bundle, how to achieve robust phase recovery, effectively suppress the influence of color crosstalk, and improve the temporal resolution and reconstruction accuracy of imaging are technical problems that urgently need to be solved. Summary of the Invention

[0006] The purpose of this invention is to address the deficiencies or shortcomings of the existing technology by providing a lensless fiber bundle phase imaging method. This method constructs an imaging model suitable for fiber bundle structures and combines it with an optimized solution strategy to achieve quantitative reconstruction of sample amplitude and phase under single-frame light intensity measurement conditions. Furthermore, this invention proposes a multi-channel color crosstalk correction mechanism and a multi-wavelength multiplexing framework to enhance the robustness of phase recovery.

[0007] The technical solution to achieve the objective of this invention is: a lensless fiber bundle phase imaging method, the method comprising the following steps:

[0008] Step 1: Under multi-wavelength illumination, acquire a single-exposure image of the sample through an optical fiber;

[0009] Step 2: Perform color crosstalk correction and fiber core transmittance correction on the single exposure image to obtain a corrected image for reconstruction;

[0010] Step 3: Based on the corrected image, remove the fiber bundle structure and reconstruct the diffraction image of the sample;

[0011] Step 4: Establish a lensless fiber bundle imaging model to correlate the diffraction image with the complex amplitude distribution of the sample;

[0012] Step 5: Based on the lensless fiber bundle imaging model, construct an optimization objective function that includes a data consistency term and a regularization term, and obtain the amplitude and phase images of the sample by optimizing the solution.

[0013] Furthermore, the lighting in step 1 includes dual-wavelength or multi-wavelength multiplexed lighting, and the lighting source includes one of halogen lamps, LEDs, and lasers.

[0014] Further, in step 1, a single exposure image refers to image data acquired within the same exposure time. The image data includes: a color image acquired by a single color image sensor, or a multi-channel image acquired in parallel within the same exposure time by one or more image sensors through an optical path allocation.

[0015] Furthermore, the optical path allocation includes splitting the incident light into multiple paths using a beam splitter and / or a filter element, so that each path can be acquired by a different image sensor.

[0016] Further, step 2 involves correcting the single-exposure image to obtain a corrected image for reconstruction, specifically including:

[0017] (1) In the case of using a color image sensor, channel correction processing is performed on the single-exposure color image; wherein, the channel correction processing is achieved by establishing a channel response model, which is used to describe the mapping relationship between the measured light intensity and the real light intensity;

[0018] In cases where multiple images are acquired by multiple image sensors within the same exposure time through optical path allocation, the optical signals of each channel have been physically separated during the acquisition process, eliminating the need for channel correction processing.

[0019] (2) Perform fiber transmittance correction based on background image to compensate for the transmission differences between different fibers.

[0020] Furthermore, the mapping relationship includes linear or nonlinear mapping relationships, and is represented in matrix form or an equivalent mathematical model, such as:

[0021]

[0022] In the formula, This represents a vector composed of light intensities measured from multiple channels. This represents a vector composed of the actual light intensities at corresponding wavelengths. This represents the channel response matrix.

[0023] Furthermore, step 3, reconstructing the sample diffraction image based on the corrected image, specifically includes:

[0024] Step 3-1: Determine the location of the maximum intensity of each fiber based on the calibration image, and spatially divide the calibration image;

[0025] Step 3-2: Based on the spatial division results, reconstruct the sample using interpolation to obtain the diffraction image.

[0026] Furthermore, the lensless fiber bundle imaging model described in step 4 is represented as follows:

[0027]

[0028] In the formula, The reconstructed sample diffraction image is shown below. Two-dimensional images at pixel size, This represents a nonlinear mapping operation that is equivalent to amplitude or light intensity. Indicates the interpolation operator. Let i represent the free space propagation operator at the i-th wavelength. This represents the complex amplitude information of the sample at the i-th wavelength.

[0029] Furthermore, the free space propagation operator adopts an angular spectrum propagation model or an equivalent diffraction propagation model.

[0030] Further, in step 5, the optimization objective function is expressed as:

[0031]

[0032]

[0033] In the formula, Let be the complex amplitude vector to be determined. This represents the total number of images acquired within a single exposure time, which is the same as the total number of wavelengths. The reference wavelength is selected by the user. For the i-th wavelength, and These represent the real and imaginary parts of a complex number, respectively. For data fidelity items, These are regularization or constraint terms used to introduce prior sample information or physical prior constraints. For complex amplitude information;

[0034] Complex amplitude information is obtained by optimizing and solving the objective function. ;

[0035] Based on the complex amplitude information Obtain the amplitude of the sample and phase :

[0036]

[0037] .

[0038] Compared with the prior art, the significant advantages of this invention are:

[0039] (1) This invention introduces dual-wavelength or multi-wavelength illumination within the same exposure time and simultaneously acquires light intensity information corresponding to multiple wavelengths through an optical fiber bundle under single exposure conditions. This allows for the acquisition of multi-channel measurement data for phase recovery without the need for multi-frame acquisition, mechanical scanning, or spatial light modulation, thereby achieving quantitative reconstruction of amplitude and phase information in the complex amplitude of the sample. Compared with existing phase imaging methods that rely on multi-frame acquisition or multi-state modulation, this invention effectively avoids the influence of sample movement, system jitter, and environmental disturbances on the phase reconstruction results, significantly improving imaging speed and result stability.

[0040] (2) Improve the accuracy of multi-wavelength imaging data by color crosstalk correction and fiber transmittance correction. This invention establishes a multi-channel response model to correct color crosstalk in single-exposure images to compensate for signal aliasing caused by sensor response overlap between different spectral channels; at the same time, it compensates for light intensity non-uniformity caused by differences in size, shape or transmission characteristics between different fiber cores by fiber transmittance correction based on background image.

[0041] (3) Improving imaging quality and reducing structural artifacts by removing the influence of fiber bundle structure. To address the structural artifacts caused by discrete core arrangement and spatially irregular sampling in fiber bundle imaging, this invention determines the location of maximum intensity for each fiber and performs spatial division. Then, based on the spatial division results, interpolation reconstruction is performed to convert the discretely sampled fiber bundle image into a continuous sample diffraction image. This method effectively reduces the influence of fiber bundle structure on the imaging results and reduces honeycomb artifacts and spatial distortion caused by core arrangement.

[0042] (4) This invention establishes a lensless fiber bundle imaging model, correlates the sample diffraction image after removing the fiber bundle structure with the complex amplitude distribution of the sample at different wavelengths, and introduces a free-space propagation operator into the model, thereby enabling the imaging process to realistically describe the propagation characteristics of the sample's complex amplitude information under lensless conditions. Compared with existing deep learning-based phase retrieval methods, the imaging model used in this invention has clear physical meaning, which helps to improve the stability of the phase inversion process and the reliability of the results.

[0043] (5) This invention proposes a phase retrieval method based on a lensless fiber bundle imaging model. An optimization objective function containing a data consistency term and a regularization term is constructed and solved to obtain the complex amplitude information of the sample. By introducing a regularization constraint into the optimization objective function, effective constraints can be applied to the phase retrieval process under single-frame light intensity measurement conditions, thereby alleviating the ill-posedness of the phase retrieval problem, improving the algorithm's ability to suppress noise and system errors, and making the phase reconstruction results more stable and reliable.

[0044] (6) From the perspective of the overall technical solution, the present invention addresses the problem of insufficient phase information acquisition and unstable solution under the condition of lensless fiber bundle imaging. By introducing multi-wavelength illumination under single exposure conditions, and combining it with color crosstalk correction, fiber transmittance correction, fiber bundle structure removal, lensless fiber bundle imaging modeling, and optimization solution based on the imaging model, it is possible to complete the quantitative reconstruction of sample amplitude and phase information when only a single frame of light intensity measurement data is acquired.

[0045] Through the above-described overall technical solution, this invention achieves the acquisition and stable recovery of information required for phase imaging without the need for additional imaging lenses, multi-frame acquisition, or additional optical modulation devices. While ensuring the simplification of system structure and miniaturization of size, it improves the stability and reliability of phase reconstruction, thereby enhancing the practicality and applicability of the lensless fiber bundle phase imaging method.

[0046] The present invention will now be described in further detail with reference to the accompanying drawings. Attached Figure Description

[0047] Figure 1 This is a flowchart of phase reconstruction of a lensless fiber bundle in one embodiment.

[0048] Figure 2 This is an example of an original image and a magnified view acquired by a color camera, wherein... Figure 2 (a) in the image is the original image of the sample obtained. Figure 2 (b) in the image is a magnified view of a local area in (a).

[0049] Figure 3 One example shows the reconstruction result and its magnified local view, where Figure 3 (a) in the image is the reconstructed sample amplitude image; Figure 3 (b) in the image is a magnified view of a local area in (a); Figure 3 (c) in the image represents the reconstructed sample phase image; Figure 3 (d) in the image is a magnified view of a local area in (c). Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0051] It should be noted that if the embodiments of the present invention involve descriptions such as "first" and "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" and "second" may explicitly or implicitly include at least one of those features. Furthermore, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by the present invention.

[0052] In one embodiment, combined Figure 1 A lensless fiber bundle phase imaging method is provided, the method comprising the following steps:

[0053] Step 1: Under multi-wavelength illumination, acquire a single-exposure image of the sample through an optical fiber;

[0054] Here, the image is acquired through an optical fiber bundle, which consists of multiple optical fibers. Due to the discrete sampling between the optical fibers, spatial sampling irregularities are inevitable.

[0055] Here, illumination includes dual-wavelength or multi-wavelength multiplexed illumination (e.g., simultaneously illuminating the sample with red, green, and blue light-emitting diodes; the resulting light signals are transmitted via fiber optic bundles and received by a color image sensor). Because the size of a single pixel in a color image sensor is much smaller than the lateral dimension of a single optical fiber, a spatial sampling density exceeding the fiber optic bundle sampling limit can be achieved, and image information from multiple channels can be acquired simultaneously under single-exposure conditions. The original image of the sample acquired by the imaging system is shown below. Figure 2 As shown), the lighting source includes one of the following: halogen lamp, LED, laser, etc.

[0056] Step 2: Perform color crosstalk correction and fiber core transmittance correction on the single exposure image to obtain a corrected image for reconstruction;

[0057] Step 3: Based on the corrected image, remove the fiber bundle structure and reconstruct the diffraction image of the sample;

[0058] Step 4: Establish a lensless fiber bundle imaging model to correlate the diffraction image with the complex amplitude distribution of the sample;

[0059] Step 5: Based on the lensless fiber bundle imaging model, construct an optimization objective function that includes a data consistency term and a regularization term, and obtain the amplitude and phase images of the sample by optimizing the solution.

[0060] Furthermore, in one embodiment, in step 1, a single exposure image refers to image data acquired within the same exposure time. The image data includes, but is not limited to, a color image acquired by a single color image sensor (single color camera), or a multi-channel image acquired in parallel within the same exposure time by one or more image sensors through an optical path allocation.

[0061] Preferably, in some embodiments, the optical path allocation includes splitting the incident light into multiple paths by means of a beam splitter and / or a filter element, so that each path can be acquired by a different image sensor (camera).

[0062] Furthermore, in one embodiment, step 2 involves correcting the single-exposure image to obtain a corrected image for reconstruction, specifically including:

[0063] (1) For the case of using a color image sensor, channel correction processing is performed on the single-exposure color image (to correct color crosstalk in order to reduce cross interference between different spectral channels); wherein, the channel correction processing is achieved by establishing a channel response model, which is used to describe the mapping relationship between the measured light intensity and the real light intensity;

[0064] In cases where multiple images are acquired by multiple image sensors within the same exposure time through optical path allocation, the optical signals of each channel have been physically separated during the acquisition process, eliminating the need for channel correction processing.

[0065] (2) Since different optical fibers have differences in size, shape or transmission characteristics, further optical fiber transmittance correction based on background image is performed to compensate for the light intensity non-uniformity caused by differences in size, shape or transmission characteristics between different optical fibers.

[0066] Here, under fixed imaging configuration conditions, background images can be acquired without samples, and the sample images can be normalized using the background images to compensate for transmission differences between different optical fibers.

[0067] Preferably, in some embodiments, the mapping relationship between measured light intensity and true light intensity can be represented by the following exemplary linear model:

[0068]

[0069] In the formula, This represents a vector composed of light intensities measured from multiple channels. This represents a vector composed of the actual light intensities at corresponding wavelengths. This represents the channel response matrix.

[0070] It should be noted that, not limited to the linear form mentioned above, it can also be implemented using an equivalent nonlinear mapping method, and represented in matrix form or an equivalent mathematical model.

[0071] Furthermore, in one embodiment, step 3, reconstructing the sample diffraction image based on the corrected image, specifically includes:

[0072] Step 3-1: Determine the location of the maximum intensity of each fiber based on the calibration image, and spatially divide the calibration image;

[0073] Step 3-2: Based on the spatial division results, reconstruct the sample using interpolation to obtain the diffraction image.

[0074] Preferably, the spatial partitioning adopts, but is not limited to, triangulation, and the interpolation method adopts, but is not limited to, centroid interpolation or equivalent interpolation methods.

[0075] Furthermore, in one embodiment, the lensless fiber bundle imaging model described in step 4 is represented as follows:

[0076]

[0077] In the formula, The reconstructed sample diffraction image is shown below. Two-dimensional images at pixel size, This represents a nonlinear mapping operation that is equivalent to amplitude or light intensity. Indicates the interpolation operator. Let i represent the free space propagation operator at the i-th wavelength. This represents the complex amplitude information of the sample at the i-th wavelength.

[0078] Preferably, the free space propagation operator adopts, but is not limited to, an angular spectrum propagation model or an equivalent diffraction propagation model.

[0079] Specifically:

[0080] The image output by the fiber optic bundle can be represented as:

[0081]

[0082] Among them, the image output by the fiber bundle for A two-dimensional image at pixel size, where x and y represent spatial coordinates. The complex amplitude of the sample at the i-th wavelength is represented by the symbol. This indicates a modulo operation. This describes the modulation characteristics of a lensless fiber bundle. This point-by-point modulation is equivalent to a diagonal matrix. . The propagation operator is defined by the following formula:

[0083]

[0084] Discretization and quantization can be written as:

[0085]

[0086] in, For the propagation matrix, Corresponding to the sample complex amplitude function The vectorized form, This corresponds to the vectorized form of images acquired using fiber optic bundles. For example... Figure 1 In step 3, further... Trigonometric interpolation is performed to eliminate the influence of the fiber sampling structure, resulting in:

[0087]

[0088] in, The interpolation matrix can be used in various forms such as triangular interpolation and bilinear interpolation.

[0089] Furthermore, in one embodiment, in step 5, the optimization objective function is expressed as:

[0090]

[0091]

[0092] In the formula, Let be the complex amplitude vector to be determined. This represents the total number of images acquired within a single exposure time, which is the same as the total number of wavelengths. The reference wavelength parameter is selected by the user. For the i-th wavelength parameter, and These represent the real and imaginary parts of a complex number, respectively. For data fidelity items, These are regularization or constraint terms used to introduce prior sample information or physical prior constraints. For complex amplitude information;

[0093] Complex amplitude information is obtained by optimizing and solving the objective function. ;

[0094] Based on the complex amplitude information Obtain the amplitude of the sample and phase :

[0095]

[0096] .

[0097] By way of example, the method of the present invention, Figure 2 The reconstruction results are as follows Figure 3 As shown.

[0098] Preferably, the regularization term may include one or more of total variational regularization, sparse prior, or L1 norm prior. The optimization solution may employ alternating iteration, gradient descent, conjugate gradient, quasi-Newton method, or optimization algorithms based on iterative shrinkage / threshold strategies, including Fast Iterative Shrinkage Thresholding Algorithm (FISTA) and Two-Step Iterative Shrinkage Thresholding Algorithm (TwIST). This invention is not limited to any specific method.

[0099] For example, the objective function can be expressed as:

[0100]

[0101]

[0102] In the formula, This represents the total number of images acquired within a single exposure time. and These represent obtaining the real and imaginary parts of a complex number, respectively. For data fidelity items, It is a total variational (TV) regularization, which can be replaced by L1 regularization, sparse prior, etc. For non-negative absorption constraints, the corresponding set is:

[0103] .

[0104] In one embodiment, a lensless fiber bundle phase imaging system is provided, the system comprising:

[0105] The first module is used to acquire single-exposure images of samples via optical fiber under multi-wavelength illumination.

[0106] The second module is used to perform color crosstalk correction and fiber core transmittance correction on the single exposure image to obtain a corrected image for reconstruction.

[0107] The third module is used to remove the fiber bundle structure and reconstruct the diffraction image of the sample based on the corrected image;

[0108] The fourth module is used to establish a lensless fiber bundle imaging model to correlate the diffraction image with the complex amplitude distribution of the sample.

[0109] The fifth module is used to construct an optimization objective function containing a data consistency term and a regularization term based on the lensless fiber bundle imaging model, and obtain the amplitude and phase images of the sample by optimizing the solution.

[0110] Specific limitations regarding the lensless fiber bundle phase imaging system can be found in the limitations of the lensless fiber bundle phase imaging method described above, and will not be repeated here. Each module in the aforementioned lensless fiber bundle phase imaging system can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0111] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements:

[0112] Step 1: Under multi-wavelength illumination, acquire a single-exposure image of the sample through an optical fiber;

[0113] Step 2: Perform color crosstalk correction and fiber core transmittance correction on the single exposure image to obtain a corrected image for reconstruction;

[0114] Step 3: Based on the corrected image, remove the fiber bundle structure and reconstruct the diffraction image of the sample;

[0115] Step 4: Establish a lensless fiber bundle imaging model to correlate the diffraction image with the complex amplitude distribution of the sample;

[0116] Step 5: Based on the lensless fiber bundle imaging model, construct an optimization objective function that includes a data consistency term and a regularization term, and obtain the amplitude and phase images of the sample by optimizing the solution.

[0117] For specific limitations on each step, please refer to the limitations on the lensless fiber bundle phase imaging method mentioned above, which will not be repeated here.

[0118] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program being implemented when executed by a processor:

[0119] Step 1: Under multi-wavelength illumination, acquire a single-exposure image of the sample through an optical fiber;

[0120] Step 2: Perform color crosstalk correction and fiber core transmittance correction on the single exposure image to obtain a corrected image for reconstruction;

[0121] Step 3: Based on the corrected image, remove the fiber bundle structure and reconstruct the diffraction image of the sample;

[0122] Step 4: Establish a lensless fiber bundle imaging model to correlate the diffraction image with the complex amplitude distribution of the sample;

[0123] Step 5: Based on the lensless fiber bundle imaging model, construct an optimization objective function that includes a data consistency term and a regularization term, and obtain the amplitude and phase images of the sample by optimizing the solution.

[0124] For specific limitations on each step, please refer to the limitations on the lensless fiber bundle phase imaging method mentioned above, which will not be repeated here.

[0125] In summary, this invention eliminates the need for lenses in the fiber bundle, offering advantages such as simple structure and suitability for miniaturized systems. Furthermore, it enables stable quantitative phase imaging under single exposure.

[0126] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention without departing from its spirit and scope should be included within the protection scope of the present invention.

Claims

1. A lensless fiber bundle phase imaging method, characterized in that, The method includes the following steps: Step 1: Under multi-wavelength illumination, acquire a single-exposure image of the sample through an optical fiber; Step 2: Perform color crosstalk correction and fiber core transmittance correction on the single exposure image to obtain a corrected image for reconstruction; Step 3: Based on the corrected image, remove the fiber bundle structure and reconstruct the diffraction image of the sample; Step 4: Establish a lensless fiber bundle imaging model to correlate the diffraction image with the complex amplitude distribution of the sample; Step 5: Based on the lensless fiber bundle imaging model, construct an optimization objective function that includes a data consistency term and a regularization term, and obtain the amplitude and phase images of the sample by optimizing the solution. Step 2 involves correcting the single-exposure image to obtain a corrected image for reconstruction, specifically including: (1) In the case of using a color image sensor, channel correction processing is performed on the single-exposure color image; wherein, the channel correction processing is achieved by establishing a channel response model, which is used to describe the mapping relationship between the measured light intensity and the real light intensity; In cases where multiple images are acquired by multiple image sensors within the same exposure time through optical path allocation, the optical signals of each channel have been physically separated during the acquisition process, eliminating the need for channel correction processing. (2) Perform fiber transmittance correction based on background image to compensate for the transmission differences between different fibers; Step 3, which involves reconstructing the sample diffraction image based on the corrected image, specifically includes: Step 3-1: Determine the location of the maximum intensity of each fiber based on the calibration image, and spatially divide the calibration image; Step 3-2: Based on the spatial division results, reconstruct the sample's diffraction image using interpolation. The lensless fiber bundle imaging model described in step 4 is represented as follows: ； In the formula, The reconstructed sample diffraction image is shown below. Two-dimensional images at pixel size, This represents a nonlinear mapping operation that is equivalent to amplitude or light intensity. Indicates the interpolation operator. Let i represent the free space propagation operator at the i-th wavelength. This represents the complex amplitude information of the sample at the i-th wavelength; In step 5, the optimization objective function is expressed as: ； ； In the formula, Let be the complex amplitude vector to be determined. This represents the total number of images acquired within a single exposure time, which is the same as the total number of wavelengths. The reference wavelength is selected by the user. For the i-th wavelength, and These represent the real and imaginary parts of a complex number, respectively. For data fidelity items, These are regularization or constraint terms used to introduce prior sample information or physical prior constraints. For complex amplitude information; Complex amplitude information is obtained by optimizing and solving the objective function. ; Based on the complex amplitude information Obtain the amplitude of the sample and phase : ； 。 2. The lensless fiber bundle phase imaging method according to claim 1, characterized in that, The lighting in step 1 includes dual-wavelength or multi-wavelength multiplexed lighting, and the lighting source includes one of halogen lamps, LEDs, and lasers.

3. The lensless fiber bundle phase imaging method according to claim 1, characterized in that, In step 1, a single exposure image refers to image data acquired within the same exposure time. The image data includes: a color image acquired by a single color image sensor, or a multi-channel image acquired in parallel within the same exposure time by one or more image sensors through an optical path allocation.

4. The lensless fiber bundle phase imaging method according to claim 3, characterized in that, The optical path allocation includes splitting the incident light into multiple paths using a beam splitter and / or a filter element, so that each path can be acquired by a different image sensor.

5. The lensless fiber bundle phase imaging method according to claim 1, characterized in that, The mapping relationship includes linear or nonlinear mapping relationships, and is represented in matrix form or an equivalent mathematical model, such as: ； In the formula, This represents a vector composed of light intensities measured from multiple channels. This represents a vector composed of the actual light intensities at corresponding wavelengths. This represents the channel response matrix.

6. The lensless fiber bundle phase imaging method according to claim 1, characterized in that, The free space propagation operator adopts the angular spectrum propagation model or its equivalent diffraction propagation model.

7. A lensless fiber bundle phase imaging system based on the method of any one of claims 1 to 6, characterized in that, The system includes: The first module is used to acquire single-exposure images of samples via optical fiber under multi-wavelength illumination. The second module is used to perform color crosstalk correction and fiber core transmittance correction on the single exposure image to obtain a corrected image for reconstruction. The third module is used to remove the fiber bundle structure and reconstruct the diffraction image of the sample based on the corrected image; The fourth module is used to establish a lensless fiber bundle imaging model to correlate the diffraction image with the complex amplitude distribution of the sample. The fifth module is used to construct an optimization objective function containing a data consistency term and a regularization term based on the lensless fiber bundle imaging model, and obtain the amplitude and phase images of the sample by optimizing the solution.

8. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 6.

Images