A topology-based image reconstruction method and system
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- HEFEI TUXUN ELECTRONICS TECH
- Filing Date
- 2025-04-29
- Publication Date
- 2026-06-23
Smart Images
Figure CN120672628B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of image reconstruction, and more specifically, to an image reconstruction method and system based on topology. Background Technology
[0002] In image processing and computer vision, image angle correction technology is a crucial step in ensuring image quality and accuracy. Traditional image angle correction techniques are mainly based on geometric transformations, which involve manually adjusting or utilizing geometric features in the image (such as lines and edges) to correct angles, for example, using Hough transform to detect lines and calculate tilt angles. However, these methods rely on manual intervention or specific scene features, resulting in poor robustness. Another approach is template matching-based correction, which uses predefined templates (such as QR codes or calibration boards) to determine the reference orientation of the target object, but this requires fixed templates and has low adaptability to complex scenes. These methods generally suffer from the technical drawback of relying on manually set rules and being unable to handle angle deviations in dynamic and unstructured environments.
[0003] In the field of dynamic microscopic imaging, when a vertically incident laser light source is projected onto a spherical or curved object, the reflection of light causes the side information of the curved surface to be missed by the camera in time. This results in the spherical or curved object having missing information during image reconstruction, with only the top surface information remaining, exhibiting a "creamy tip" appearance. To address this technical challenge, existing spherical measurements often avoid the vertical incidence method and choose triangulation, which fills in the missing information by changing the angles of the light source and camera. However, triangulation requires significant space, is highly sensitive to relative measurement angles, is highly dependent on the number of pre-acquired images, and takes a long time to reconstruct the final image, severely limiting its application in high-precision dynamic measurement systems.
[0004] Chinese patent application CN202411476815.7, published on February 11, 2025, discloses a compensation device and method for a digital holographic microscope, relating to the field of optical imaging technology. By using deformable mirror groups and real-time wavefront sensing technology, the compensation device can dynamically adjust the optical response based on real-time measured wavefront distortion data. Adaptive optics elements precisely control the light wavefront, effectively extending the imaging depth and optimizing focus control, thus enhancing resolution. Furthermore, the compensation phase algorithm, by introducing a regularization term to calculate the required compensation, not only improves the algorithm's adaptability to environmental changes but also enhances its robustness during imaging. However, this scheme lacks an automatic angle detection and correction mechanism; therefore, the reconstruction accuracy of this scheme needs further improvement. Summary of the Invention
[0005] In traditional digital holographic microscopy, when a laser is incident perpendicularly on the surface of a curved object, the sides and steeply tilted parts of the object are often not fully recorded due to the limitation of the light scattering angle, resulting in edge distortion or missing information in the reconstructed image. This application provides an image reconstruction method and system based on topology. By establishing a digital holographic microscopy model and forming a hierarchical image model, it can accurately extract the deflection angle of the object and intelligently correct this angle, effectively compensating for the loss of side information caused by perpendicular laser incidence. This achieves high-precision reconstruction of the complete surface of spherical or curved objects, solving the imaging limitations of existing technologies when processing complex curved objects.
[0006] One aspect of this application provides an image reconstruction method based on topology, comprising: establishing a digital holographic microscopy imaging model, wherein the digital holographic microscopy imaging model provides three-dimensional spatial topological data of an object by recording and reconstructing the object's light field information; converting the three-dimensional spatial topological data into a multi-layer graph structure with different focal lengths through an adaptive focusing algorithm to form a hierarchical graph model; calculating the object's deflection angle based on the topological data of the hierarchical graph model; performing angle correction on the hierarchical graph model based on the deflection angle to obtain a corrected hierarchical graph model; calculating the optimal imaging focal point position at different angles in the corrected hierarchical graph model; and performing surface topology reconstruction based on the optimal imaging focal point position to obtain a reconstructed image after angle adjustment.
[0007] In particular, by establishing a digital holographic microscopic imaging model, not only is the intensity information of light recorded, but also the phase information is acquired at the same time. This allows for the complete capture of the light field distribution of the object, providing complete three-dimensional spatial topological data for the subsequent all-round reconstruction of curved objects, overcoming the limitation of traditional imaging methods that can only acquire two-dimensional information.
[0008] Furthermore, a hierarchical graph model is formed, including: transforming the three-dimensional spatial topology data into a multi-layer graph structure imaged at different focal lengths using the following formula:
[0009]
[0010] Where d represents the different focal lengths of the input; For Fourier transform, It is the selected spectral level. For reference wavefront, The input is the intensity information of the hologram. For wavelength, and The coordinates are in the time domain and frequency domain, respectively; i represents a positive integer; each pixel is a node, and connections are established between adjacent pixel nodes. The edge weights are determined by phase continuity to construct the micro-layer structure; similar regions are merged by superpixel segmentation algorithm, nodes are defined as local surface patches, and connections are established between superpixel nodes by curvature similarity or spatial proximity to construct the meso-layer structure; the micro-layer structure and the meso-layer structure are integrated to obtain the hierarchical graph model.
[0011] In particular, this application achieves effective fusion of multi-scale information by constructing a two-layer structure of micro and meso layers. The micro layer captures high-frequency details at the pixel level, while the meso layer acquires low-frequency structural information at the local surface patch level. This multi-scale representation method enables the system to simultaneously process high-frequency details (such as microtextures) and low-frequency features (such as overall shape).
[0012] On the other hand, the hierarchical graph model maintains spatial topological relationships through edge weight definition, a characteristic that is crucial in the angle correction process. Traditional methods are prone to introducing topological breaks during angle transformation, while this method ensures topological consistency during the transformation process through phase continuity and curvature similarity constraints, avoiding reconstruction artifacts such as "pseudo-edges" and "faults".
[0013] Furthermore, a theoretical bridge was established from the physical properties of light waves to the digital computational model through Fourier transform and angular spectrum propagation algorithms. The formula integrates physical parameters such as reference wavefront, holographic intensity information, and wavelength, ensuring that the reconstruction process maintains a high degree of consistency with actual optical phenomena, thereby effectively reducing errors caused by model approximation.
[0014] Finally, through multi-layer information fusion and topological constraints at different focal lengths, the system can compensate for the lack of surface side information in traditional methods. Without additional hardware, it can achieve complete surface reconstruction, thus solving the "cream tip" problem in vertical incidence imaging.
[0015] Further, angle correction includes: extracting the object deflection angle θ based on the digital holographic microscopy imaging model through time and frequency domain analysis; and performing single-layer comparative analysis on the hierarchical image model based on the deflection angle θ to determine the optimal pixel position. Based on the optimal pixel position Determine the coordinates of the center point of the hierarchical graph model. , ,in, The images were acquired respectively. Number of pixels; based on center point coordinates Angle correction is performed on each pixel of the hierarchical graph model.
[0016] In particular, based on the object deflection angle extracted from time-domain and frequency-domain analysis, combined with a precise coordinate transformation formula, it is possible to accurately correct the curved surface with a large angle of inclination, effectively recovering the area that could not be fully imaged due to the limitation of the incident angle, and significantly improving the imaging quality of the side area.
[0017] Furthermore, the deflection angle θ of the extracted object is obtained using the following formula: ,in, The imaginary part of the input hologram intensity information; This represents the intensity information of the input hologram; This represents the real part of the intensity information of the input hologram.
[0018] Furthermore, based on the coordinates of the center point Angle correction is performed on each pixel of the hierarchical graph model using the following formula: ; ; ;in, Indicates the corrected pixels; This represents the holographic intensity information after angle correction; Represents holographic intensity information without angle correction; Indicates the optimal pixel position.
[0019] Furthermore, obtaining the angle-adjusted reconstructed image includes: treating each pixel as a single topological node for the microscopic layer; treating a preset local surface patch as a node for the mesoscopic layer, processing the foreground and background information of the local surface patch using Principal Component Analysis (PCA) algorithm, and using the processed foreground information as a single topological node; calculating the focus and weight of a single topological node in a hierarchical graph model; calculating the optimal imaging focus position based on the focus and weight; and performing surface topology reconstruction based on the optimal imaging focus position to obtain the angle-adjusted reconstructed image.
[0020] In particular, by combining pixel-level microstructures and superpixel-level mesostructures, the system enhances its ability to understand the overall topological structure of curved objects while ensuring the accuracy of reconstructed details. Specifically, node connections established through curvature similarity and spatial proximity enable the system to better handle transition regions on curved surfaces.
[0021] Furthermore, the focus and weight are calculated using the following formula: ; ,in, , These represent nodes at the same position in the hierarchical graph at different hierarchical levels. Corrected intensity The coefficients of the quadratic and linear terms of the fitted quadratic curve.
[0022] In particular, by calculating the focus and weight of a single topological node and combining it with the principal component analysis (PCA) algorithm to process foreground and background information, the optimal imaging position for different regions of a complex surface is accurately located, solving the problem of unclear surface reconstruction caused by a single focal plane in traditional methods.
[0023] Furthermore, the optimal imaging focal position is calculated based on the focus and weight. The following formula is used: .
[0024] Furthermore, based on the optimal imaging focal position, surface topology reconstruction is performed to obtain the angle-adjusted reconstructed image, using the following formula: In particular, surface topology reconstruction is performed based on the calculated optimal imaging focal position. Combined with the spatial relationship constraints of the topological structure, the reconstruction process not only considers single-point information but also integrates the correlation of the surrounding area. This effectively compensates for the information loss caused by uneven illumination or scattering angle limitations, and achieves complete and high-precision reconstruction of curved objects.
[0025] Another aspect of this application provides an image reconstruction system based on topology, comprising: an imaging module for establishing a digital holographic microscopic imaging model and providing three-dimensional spatial topological data of the object by recording and reconstructing the object's light field information; a hierarchical graph module for converting the three-dimensional spatial topological data into a multi-layer graph structure using an adaptive focusing algorithm, wherein the multi-layer graph structure includes a micro-layer structure and a meso-layer structure; an angle calculation module for calculating the object's deflection angle θ based on the topological data of the multi-layer graph structure; an angle correction module for performing angle correction on the multi-layer graph structure based on the object's deflection angle θ to obtain a corrected multi-layer graph structure; a focus calculation module for calculating the optimal imaging focus position at different angles based on the corrected multi-layer graph structure; and a reconstruction module for performing surface topology reconstruction based on the optimal imaging focus position to obtain a reconstructed image after angle adjustment.
[0026] Compared to existing technologies, the advantages of this application are:
[0027] On the one hand, existing technologies for solving the imaging problem of curved objects generally employ imaging methods with a single focal plane and simple angle correction algorithms. However, these methods suffer from drawbacks such as inability to handle large-angle tilted sections, low computational efficiency and limited accuracy, and difficulty in achieving multi-scale information collaborative processing. This application directly associates the phase of light waves with surface geometry through dynamic topology modeling and overcomes the imaging limitations of large-angle (greater than 10 degrees) tilted surfaces based on the physical constraints of node-edge weights in the topology, achieving physical-level precision angle self-adjustment.
[0028] On the other hand, existing technologies in real-time imaging processing generally improve speed by simplifying models or reducing resolution, but this suffers from the drawback of difficulty in achieving both accuracy and speed. This application significantly improves the algorithm's speed and efficiency through optimized implementation based on GPU parallel computing, and, combined with a fast tilt compensation method based on Zernike polynomials, overcomes the constraints and influences of angle on imaging effects in the field of high-precision imaging, achieving fast and high-precision real-time calibration imaging. Attached Figure Description
[0029] Figure 1 A schematic diagram of the digital reconstruction modeling process for this application;
[0030] Figure 2 This is an example flowchart of an image reconstruction method based on topology structure according to this application;
[0031] Figure 3 The image shows the surface reconstruction result of a wafer solder ball 3D holographic reconstruction using a traditional algorithm.
[0032] Figure 4 This is a rendering of the 3D holographic reconstruction result of the wafer tin ball surface using this algorithm. Detailed Implementation
[0033] The present application will now be described in detail with reference to the accompanying drawings and specific embodiments.
[0034] like Figure 1 and Figure 2 As shown, a digital holographic microscopy model is established. This model provides three-dimensional spatial topological data of the object by recording and reconstructing its light field information. Unlike traditional imaging, which only records light intensity, digital holographic microscopy, based on the wave characteristics of light, simultaneously records the amplitude and phase information of the light wave. When a laser illuminates a curved object, the object light wave scattered by the object and the reference light wave form an interference pattern on the interference surface, containing complete three-dimensional spatial topological information.
[0035] The 3D spatial topology data is transformed into a multi-layered graph structure with different focal lengths using an adaptive focusing algorithm, forming a hierarchical graph model. For curved objects, the optimal imaging plane differs for different parts. The adaptive focusing algorithm calculates the intensity variance at different d-values to find the optimal focus position for each region, resulting in a multi-layered graph structure containing information about different focal lengths.
[0036] The object's deflection angle is calculated based on the topological data of the hierarchical graph model. The hierarchical graph model is then angle-corrected based on the deflection angle to obtain the corrected hierarchical graph model. The optimal imaging focal point position for different angles in the corrected hierarchical graph model is calculated, and surface topology reconstruction is performed based on the optimal imaging focal point position to obtain the reconstructed image after angle adjustment. This application achieves complete reconstruction of curved objects under perpendicularly incident laser light through this multi-level fusion method, avoiding the "cream tip" phenomenon in traditional methods while maintaining the system's simplicity and efficiency.
[0037] Step 1: Build the model.
[0038] First, a digital holographic microscopy (DHM) model incorporating image sharpness evaluation metrics is established. This model can assess the quality of imaging based on image sharpness, and simultaneously utilizes holographic structures to describe the spatial relationships of objects. By recording and reconstructing the complete light field information (including amplitude and phase) of the objects, it can provide high-resolution 3D spatial topological data. The detailed modeling process can be found in [link to detailed modeling process]. Figure 1 In detail, the core physical principle of digital holographic microscopy (DHM) is based on the wave nature of light and interference phenomena. When coherent light shines on the surface of an object, the reflected light wave (object wave O) encounters the unmodulated reference light wave R, forming an interference pattern. This pattern is recorded by a digital sensor as a hologram I. The hologram records not only the intensity information of the light but also complete phase information.
[0039] Step 2: Hierarchical graph modeling. Hierarchical graph modeling transforms holographic information into a multi-layered structure. Its physical essence is to numerically reconstruct the hologram at different reconstruction depths through the angular spectrum propagation algorithm.
[0040] In detail, through an adaptive focus algorithm, within the data range of the focus... , in the set The 3D information reconstructed from the hologram is transformed into a multi-layered image structure imaged at different focal lengths. The specific formula for the intensity information of each layer is as follows:
[0041] (1)
[0042] in For Fourier transform, It is the selected spectral level. For reference wavefront, The input is the intensity information of the hologram. For wavelength, and These are the coordinates corresponding to the time domain and frequency domain, respectively.
[0043] Number of layers in hierarchical graph modeling At the microscopic level of image data, each pixel serves as a node, carrying attributes such as phase, gradient, and curvature. Adjacent pixel nodes are connected, with edge weights determined by phase continuity. At the mesoscopic level, similar regions are merged through superpixel segmentation, and nodes represent local surface patches. Superpixel nodes are connected through curvature similarity or spatial proximity. The integration of these two layers forms a hierarchical graph model, enabling the system to retain high-precision pixel-level details while simultaneously understanding the overall structural features of objects at the mesoscopic level, providing a solid mathematical foundation for subsequent angle detection and correction. This fusion of physical optics and computational topology allows the system to effectively address the problem of missing side information in curved objects under perpendicular incidence.
[0044] Step 3: Holographic topology data is used for angle calculation. In dynamic measurements, high-speed digital holography is used to record transient phase changes, and the deflection angle is extracted through correlation analysis in the time and frequency domains.
[0045] (2)
[0046] in, The imaginary part of the input hologram intensity information; This represents the intensity information of the input hologram; This represents the real part of the intensity information of the input hologram.
[0047] Specifically, digital holography records the complex amplitude information of light waves, including real and imaginary parts. When the surface of an object is tilted, the phase distribution of the reflected light changes. This phase change is reflected in the complex representation of the hologram, and the tilt angle can be calculated using the ratio of the real to the imaginary parts. When the surface of an object is tilted, the wavefront of the reflected light wave also tilts accordingly. This tilt leads to a change in spatial frequency, which is then manifested as a change in phase distribution in the hologram. By analyzing the complex representation of the hologram, this tilt angle can be extracted. Compared with traditional angle detection methods, this application utilizes phase information rather than just intensity information, improving the accuracy of angle detection. Furthermore, it is equally effective for unstructured surfaces, requiring no feature points or reference markers. Finally, it can handle minute angle changes (theoretically accurate to 0.01 degrees), making it suitable for high-precision microscopic imaging.
[0048] Step 4: Angle correction calibration hierarchy diagram.
[0049] By comparing the hierarchical image layer by layer, the optimal pixel position is calibrated and the angle is corrected to obtain the corrected hierarchical image.
[0050] Based on the same pixel point in different level maps Calculate the average gradient of the pixel and its four adjacent pixels (up, down, left, and right). (Excluding edge points).
[0051] (3)
[0052] in, The number of layers corresponding to the minimum value position This represents the optimal pixel position layer for that pixel. Specifically, on the optimal focusing layer of the object's surface, the intensity gradient between adjacent pixels should be minimal. This is because the image is sharpest and the intensity changes between pixels are smoothest on the optimal focusing plane.
[0053] Specifically, first, move the origin of the coordinate system to the center of the image. Then, a rotation transformation with angle θ is performed; finally, the origin of the coordinate system is moved back to its original position; the corrected pixel value I'(x,y) is equal to the original pixel value I(x',y') at the corresponding position after the rotation transformation; this process is equivalent to performing a geometric transformation on the light field information recorded in the hologram to compensate for the phase distribution deformation caused by the tilt of the object's surface. Through this transformation, the curved side areas that could not be fully imaged due to the limitation of the incident angle can be effectively recovered.
[0054] Based on the above logic, iterate through all pixels to find the optimal pixel position for each pixel. Among them, single-layer correction strength The specific steps are as follows:
[0055] First, determine the coordinates of the center point of the hierarchical diagram. ,in, The images were acquired respectively. Number of pixels. The number of pixels in the corrected image. The corresponding coordinate changes
[0056] (4)
[0057] (5)
[0058] (6)
[0059] Specifically, this application utilizes a matrix transformation with the image center as the origin of rotation at an angle θ, effectively redirecting the local tangent plane of the curved surface to be parallel to the imaging plane, thereby maximizing imaging quality. This application preserves the topological constraints between nodes in the hierarchical graph model, ensuring no artifacts or distortions are introduced during the transformation. Simultaneously, it employs bilinear interpolation to handle non-integer coordinate mapping, guaranteeing the smoothness of the reconstructed image. This angle correction method, combined with topological constraints, effectively solves the problem of missing side information of curved objects under perpendicularly incident laser light, eliminating the need for additional light sources or cameras, and significantly improving imaging quality while maintaining system simplicity.
[0060] Step 5: Calculate the optimal imaging focal point position at different angles.
[0061] For the micro-layer, per pixel Nodes, calculating the focus and weight of a single topological node in the hierarchical graph.
[0062] (7)
[0063] (8)
[0064] in, , These represent nodes at the same position in the hierarchical graph at different hierarchical levels. Corrected intensity The coefficients of the quadratic and linear terms of the fitted quadratic curve.
[0065] The optimal imaging focal position for this node is calculated based on the calculated focal point and weight.
[0066] (9)
[0067] For the meso-level, using local surface patches as nodes, Principal Component Analysis (PCA) is used to preprocess the foreground and background information of the local surface patches, distinguishing between foreground information (representing valid signals) and background information (representing noise). After retaining the projection data corresponding to the main singular values, the system applies the same quadratic fitting method (Equation 7-9) to the dimensionality-reduced data as in the micro-level, but focuses more on the overall optical response characteristics of the surface patch rather than the local changes of individual pixels in a physical sense.
[0068] Step 6: Surface topology reconstruction.
[0069] The calculated optimal imaging focal position Substituting back into formula (10) to perform surface topology reconstruction, we obtain a reconstructed image after large-angle self-adjustment. .
[0070] (10)
[0071] Specifically, from the perspective of light wave propagation theory, the reconstruction method of this application is equivalent to extracting the amplitude and phase information of the wave field at different depths, and then reconstructing the complete three-dimensional structure according to the optimal focusing conditions. This process solves the "cream tip" problem in the traditional perpendicular incidence method; in detail, in the traditional method, due to the law of light reflection, perpendicularly incident light waves cannot effectively capture the side information of the curved surface; this method achieves all-round reconstruction of the curved surface through angle correction and multi-level topology analysis; and the topological continuity in the reconstruction process is guaranteed by the physical constraints of node-edge weights in the hierarchical graph model.
[0072] In a specific embodiment of this application, a spherical or curved object lacks some information during image reconstruction, possessing only top surface information, exhibiting a "creamy tip" appearance. A wafer tin ball is selected as the measurement sample. Without using the algorithm of this application, the result of simply reconstructing the surface morphology is as follows: Figure 3 As shown in the figure, it can be clearly seen from the colors in the legend that most of the solder ball top surface reconstructions have abruptly colored, high-pointed, creamy tips. Figure 4 Using this application to reconstruct the surface of the wafer solder balls after angle self-adjustment, it can be observed that the creamy tip condition is significantly improved, and the spheres are clearly and fully displayed. Test results of the wafer solder balls ( Figure 3 and Figure 4 (Comparison) This visually demonstrates the algorithm's effectiveness: under the traditional method, a "creamy tip" appears on the top of the solder ball, while after applying this algorithm, the ball's structure is complete and clear, verifying the effectiveness of this method in practical applications.
[0073] The foregoing illustrative description of the invention and its embodiments is not restrictive and can be implemented in other specific forms without departing from the spirit or essential characteristics of the invention. The accompanying drawings are only one embodiment of the invention, and the actual structure is not limited thereto. No reference numerals in the claims should limit the scope of the claims. Therefore, if a person skilled in the art, inspired by this description, designs a similar structure and embodiment without departing from the spirit of the invention, such design should fall within the scope of protection of this patent. Furthermore, the word "comprising" does not exclude other elements or steps, and the word "a" preceding an element does not exclude the inclusion of "a plurality" of that element. Multiple elements stated in the product claims can also be implemented by a single element through software or hardware. The terms "first," "second," etc., are used to indicate names and do not indicate any specific order.
Claims
1. An image reconstruction method based on topology, characterized in that, include: A digital holographic microscopy model is established, which provides three-dimensional spatial topological data of the object by recording and reconstructing the object's light field information; The three-dimensional spatial topology data is transformed into a multi-layer graph structure with different focal lengths through an adaptive focusing algorithm, forming a hierarchical graph model. Calculate the object's deflection angle based on the topological data of the hierarchical graph model; The hierarchical graph model is angle-corrected based on the deflection angle to obtain the corrected hierarchical graph model. Calculate the optimal imaging focal point position at different angles in the corrected hierarchical graph model, and perform surface topology reconstruction based on the optimal imaging focal point position to obtain the reconstructed image after angle adjustment. To form a hierarchical graph model, including: The following formula can be used to transform 3D spatial topology data into a multi-layer graph structure imaged at different focal lengths: Where d represents the different focal lengths of the input; For Fourier transform, It is the selected spectral level. For reference wavefront, The input is the intensity information of the hologram. For wavelength, and These are the coordinates in the time domain and frequency domain, respectively; i represents a positive integer. Represents holographic intensity information without angle correction; Using each pixel as a node and establishing connections between adjacent pixel nodes, edge weights are determined through phase continuity to construct a micro-layer structure; By merging similar regions through superpixel segmentation algorithm, nodes are defined as local surface patches, and connections between superpixel nodes are established through curvature similarity or spatial proximity to construct the mesoscopic layer structure. By integrating the microscopic and mesoscopic structures, a hierarchical graph model is obtained; Angle correction, including: Based on the digital holographic microscopy model, the object deflection angle θ is extracted through time-domain and frequency-domain analysis. Based on the deflection angle θ, a single-layer comparative analysis is performed on the hierarchical graph model to determine the optimal pixel position. ; Based on the optimal pixel position Determine the coordinates of the center point of the hierarchical graph model. , ,in, The images were acquired respectively. Number of pixels; Based on the coordinates of the center point Angle correction is performed on each pixel of the hierarchical graph model; The deflection angle θ of the extracted object is obtained using the following formula: in, The imaginary part of the input hologram intensity information; This represents the intensity information of the input hologram; This represents the real part of the intensity information of the input hologram.
2. The image reconstruction method based on topology structure according to claim 1, characterized in that: Based on the coordinates of the center point Angle correction is performed on each pixel of the hierarchical graph model using the following formula: in, Indicates the corrected pixel coordinates; This represents the holographic intensity information after angle correction; Indicates the optimal pixel position.
3. The image reconstruction method based on topology according to claim 1 or 2, characterized in that: The reconstructed image after angle adjustment is obtained, including: For the microscopic layer, each pixel is treated as a single topological node; For the mesoscopic layer, a preset local surface patch is used as a node. The foreground and background information of the local surface patch is processed by the principal component analysis algorithm (PCA), and the processed foreground information is used as a single topological node. Calculate the focus and weight of a single topological node in a hierarchical graph model; Calculate the optimal imaging focus position based on the focus and weight. Based on the optimal imaging focal position, surface topology reconstruction is performed to obtain the reconstructed image after angle adjustment.
4. The image reconstruction method based on topology structure according to claim 3, characterized in that: The focus and weight are calculated using the following formula: ; ,in, , These represent nodes at the same position in the hierarchical graph at different hierarchical levels. Corrected strength The coefficients of the quadratic and linear terms of the fitted quadratic curve.
5. The image reconstruction method based on topology structure according to claim 4, characterized in that: Calculate the optimal imaging focal position based on the focus and weight. The following formula is used: .
6. The image reconstruction method based on topology structure according to claim 5, characterized in that: Based on the optimal imaging focal position, surface topology reconstruction is performed to obtain the angle-adjusted reconstructed image, using the following formula: ,in, For Fourier transform, It is the selected spectral level; For reference wavefront; Wavelength; The coordinates are in the frequency domain; i represents a positive integer. This represents the holographic intensity information after angle correction.
7. A topology-based image reconstruction system for implementing the method according to any one of claims 1 to 6, characterized in that, include: The imaging module establishes a digital holographic microscopic imaging model and provides three-dimensional spatial topological data of the object by recording and reconstructing the object's light field information. The hierarchical graph module transforms three-dimensional spatial topology data into a multi-layer graph structure through an adaptive focusing algorithm. The multi-layer graph structure includes a micro-layer structure and a meso-layer structure. The angle calculation module calculates the object's deflection angle θ based on the topological data of the multi-layer graph structure. The angle correction module corrects the multi-layer graph structure based on the object's deflection angle θ, and obtains the corrected multi-layer graph structure. The focus calculation module calculates the optimal imaging focus position at different angles based on the corrected multi-layer image structure. The reconstruction module performs surface topology reconstruction based on the optimal imaging focal position to obtain a reconstructed image with adjusted angle.