Image reconstruction method and device, sequencing method and device, and electronic device

Abstract

The application discloses an image reconstruction method and device, a sequencing method and device, and an electronic device. The method comprises the following steps: determining the defocus blur feature of an imaging system based on multiple sample images, wherein the multiple sample images are images generated by the imaging system and the imaging quality of the images meets preset image processing requirements; determining the motion blur feature of the imaging system based on the scanning parameters of the imaging system; fusing the defocus blur feature and the motion blur feature to obtain a target point spread function; and reconstructing an initial image generated by the imaging system based on the target point spread function to obtain a target image. Through the application, the problem that the accurate acquisition of a point spread function is difficult and the image reconstruction effect is not ideal in the related art is solved.

Description

Technical Field

[0001] This application relates to the field of image processing, and more specifically, to an image reconstruction method, apparatus, sequencing method, apparatus, and electronic device. Background Technology

[0002] Image processing technology is an important branch of modern science and technology, widely used in medical imaging, industrial inspection, remote sensing measurement, genome sequencing, and many other fields. In these applications, image quality has a significant impact on the final analysis results.

[0003] For example, in the field of genome sequencing, with the rapid development of genomics and sequencing technologies, high-precision, high-resolution sequencing images have become important tools for basic scientific research and precision medicine. However, due to limitations in system equipment, environmental influences, and other factors, sequencing images are often affected by blurring, noise, and distortion, thus reducing image quality and usability. To improve image quality, images can be deblurred. However, the accuracy of deblurring techniques depends on a precise point spread function (PSF). In complex blurring situations, accurately obtaining the PSF is difficult, making it challenging to effectively restore image quality.

[0004] There is currently no effective solution to the problem that it is difficult to accurately obtain the point spread function in related technologies, resulting in unsatisfactory image reconstruction results. Summary of the Invention

[0005] This application provides an image reconstruction method, apparatus, sequencing method, apparatus, and electronic device to solve the problem that it is difficult to accurately obtain the point spread function and the image reconstruction effect is not ideal in related technologies.

[0006] According to one aspect of this application, an image reconstruction method is provided. The method includes: determining defocus blur features of an imaging system based on multiple sample images, wherein the multiple sample images are images generated by the imaging system whose imaging quality meets preset image processing requirements; determining motion blur features of the imaging system based on scanning parameters of the imaging system; fusing the defocus blur features and motion blur features to obtain a target point spread function; and reconstructing the initial image generated by the imaging system based on the target point spread function to obtain a target image.

[0007] Optionally, determining the motion blur features of the imaging system based on the scanning parameters of the imaging system includes: calculating the motion blur length based on the scanning speed and exposure time; determining a uniformly distributed convolution kernel with a length equal to the motion blur length along the direction of motion to obtain the motion blur features.

[0008] Optionally, determining the defocus blur features of the imaging system based on multiple sample images includes: extracting frequency domain information from multiple sample images respectively; and determining the defocus blur features based on the frequency domain information of multiple sample images.

[0009] Optionally, fusing the defocused blur features and motion blur features to obtain the target point diffusion function includes: assigning a first weight to the defocused blur features and a second weight to the motion blur features; normalizing the defocused blur features and motion blur features; and performing a weighted summation of the normalized defocused blur features and normalized motion blur features to obtain the target point diffusion function.

[0010] Optionally, before performing a weighted summation of the normalized defocused blur features and the normalized motion blur features, the method further includes: setting initial weights for the defocused blur features and the motion blur features respectively; increasing the weight of the defocused blur features and decreasing the weight of the motion blur features when the degree of defocused blur in the initial image is greater than a preset blur level and the scanning speed of the imaging system is less than a preset speed; or increasing the weight of the motion blur features and decreasing the weight of the defocused blur features when the degree of defocused blur in the initial image is less than a preset blur level and the scanning speed of the imaging system is greater than a preset speed.

[0011] Optionally, reconstructing the initial image generated by the imaging system based on the target point spread function to obtain the target image includes: determining the deconvolution kernel of the target point spread function; and using a deconvolution algorithm to deblur the initial image through the deconvolution kernel to obtain the target image.

[0012] Optionally, the initial image is deblurred using a deconvolution algorithm with a deconvolution kernel to obtain the target image, including: performing a first deblurring process on the initial image based on the target point spread function and the deconvolution kernel to obtain a first reconstructed image; performing a second deblurring process on the initial image based on the target point spread function, the deconvolution kernel, and the first reconstructed image to obtain a second reconstructed image; performing a third deblurring process on the initial image based on the target point spread function, the deconvolution kernel, and the second reconstructed image to obtain a third reconstructed image, until the number of deblurring processes reaches a preset number of iterations.

[0013] Optionally, after reconstructing the initial image generated by the imaging system based on the target point spread function to obtain the target image, the method further includes: determining whether the imaging quality of the target image meets the preset image processing requirements; if the imaging quality of the target image does not meet the preset image processing requirements, if the sharpness index of the target image is less than a preset value, updating the fusion parameters and / or updating the reconstruction parameters; performing a step of fusing the defocus blur feature and motion blur feature based on the updated fusion parameters, and / or performing a step of reconstructing the initial image generated by the imaging system based on the target point spread function based on the updated reconstruction parameters.

[0014] According to another aspect of this application, an image reconstruction method is provided. The method includes: determining defocus blur features of an imaging system based on multiple sample images, wherein the multiple sample images are images generated by the imaging system whose imaging quality meets preset image processing requirements; determining a target point spread function based on the defocus blur features; and reconstructing the initial images generated by the imaging system based on the target point spread function to obtain a target image.

[0015] Optionally, determining the defocus blur features of the imaging system based on multiple sample images includes: extracting frequency domain information from multiple sample images respectively; and determining the defocus blur features based on the frequency domain information of multiple sample images.

[0016] According to another aspect of this application, a sequencing method based on the above-described image reconstruction method is provided. The method includes: determining the defocus blur features of a sequencing system based on multiple sample sequencing images, wherein the multiple sample sequencing images are images generated by the sequencing system whose imaging quality meets the requirements for gene sequencing processing; determining the motion blur features of the sequencing system based on the scanning parameters of the sequencing system; fusing the defocus blur features and the motion blur features to obtain a target point spread function; reconstructing the initial sequencing images generated by the sequencing system based on the target point spread function to obtain a target sequencing image; and performing base identification based on the target sequencing image.

[0017] According to another aspect of this application, a computer program product is provided, including a non-volatile computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described image reconstruction method or sequencing method.

[0018] According to another aspect of this application, an electronic device is also provided, comprising a processor and a memory; the memory stores computer-readable instructions, and the processor is used to execute the computer-readable instructions, wherein the computer-readable instructions, when executed, perform the aforementioned image reconstruction method or sequencing method.

[0019] This application employs the following steps: determining the defocus blur features of an imaging system based on multiple sample images, wherein the multiple sample images are images generated by the imaging system whose imaging quality meets preset image processing requirements; determining the motion blur features of the imaging system based on the scanning parameters of the imaging system; fusing the defocus blur features and motion blur features to obtain the target point spread function; and reconstructing the initial image generated by the imaging system based on the target point spread function to obtain the target image. This solves the problem in related technologies where accurate acquisition of the point spread function is difficult and the image reconstruction effect is unsatisfactory. By combining multiple clear images generated by the imaging system with the known scanning parameters generated by the imaging system, the target point spread function is accurately estimated, and the initial image is reconstructed based on the target point spread function, achieving a significant improvement in the quality of the reconstructed image and providing clear image support for subsequent image processing. Attached Figure Description

[0020] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0021] Figure 1 This is a flowchart of an image reconstruction method according to an embodiment of this application;

[0022] Figure 2 This is a flowchart of an optional image reconstruction method according to an embodiment of this application;

[0023] Figure 3 This is a flowchart of a sequencing method according to an embodiment of this application;

[0024] Figure 4 This is a schematic diagram of a sequencing system according to an embodiment of this application;

[0025] Figure 5 This is a schematic diagram of the TDI camera imaging process according to an embodiment of this application;

[0026] Figure 6 This is a schematic diagram illustrating the simultaneous existence of blurriness and clarity within a single field of view (FOV) according to an embodiment of this application.

[0027] Figure 7(a) shows the sequencing images before and after reconstruction according to the embodiments of this application. A base comparison before and after reconstruction. Figure 1 ;

[0028] Figure 7(b) shows the sequencing images before and after reconstruction according to an embodiment of this application, comparing the A bases before and after reconstruction. Figure 2 ;

[0029] Figure 7(c) shows the sequencing images before and after reconstruction according to an embodiment of this application. A base comparison before and after reconstruction. Figure 2 ;

[0030] Figure 8 This is a schematic diagram illustrating image quality evaluation according to an embodiment of this application;

[0031] Figure 9 This is a schematic diagram of an image reconstruction apparatus according to an embodiment of this application;

[0032] Figure 10 This is a schematic diagram of an optional image reconstruction apparatus according to an embodiment of this application;

[0033] Figure 11 This is a schematic diagram of a sequencing device according to an embodiment of this application;

[0034] Figure 12 This is a schematic diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0035] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0036] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0037] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application 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 for the embodiments of this application described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover 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.

[0038] It should be noted that all information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for display, data used for analysis, etc.) involved in this disclosure are information and data authorized by the user or fully authorized by all parties.

[0039] It should be noted that the information collected in this application is information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of the relevant data all comply with the relevant laws, regulations and standards of the relevant regions, necessary confidentiality measures have been taken, and they do not violate public order and good morals. Corresponding operation portals are provided for users to choose to authorize or refuse to use the data.

[0040] To address the difficulty in accurately obtaining the diffusion function for relevant technical points, which leads to unsatisfactory image reconstruction results, the following methods can be used:

[0041] Spatial domain-based deblurring methods reduce blur by analyzing local pixel relationships in an image and using techniques such as edge detection and smoothing filtering. For example, the Laplacian operator and the Sobel operator are used to enhance edge sharpness. These methods are suitable for simple out-of-focus or slightly motion-blurred scenes, but perform poorly in handling mixed blur in complex scenes.

[0042] Frequency-domain-based deblurring methods transform the blurring problem into a convolution operation in the frequency domain by performing a Fourier transform on the image. In the frequency domain, blur can be identified and reversed through spectral analysis, for example, using techniques such as Wiener filtering or inverse filtering. However, these methods are susceptible to noise, especially when noise dominates in the high-frequency region, resulting in unsatisfactory restoration results.

[0043] Deblurring methods based on statistical models: These methods use statistical models (e.g., Maximum Aposterior Estimation, MAP) or machine learning algorithms to predict the type of blur in an image and then perform deblurring. For example, complex images can be deblurred and restored using Convolutional Neural Networks (CNNs) and Generative Adversarial Networks (GANs). However, these methods rely on large amounts of training data and are costly to train, and their generalization ability is limited in some practical scenarios.

[0044] Therefore, this application aims to provide a solution that can solve the above-mentioned technical problems, the details of which will be described in subsequent embodiments.

[0045] According to embodiments of this application, an image reconstruction method is provided.

[0046] Figure 1 This is a flowchart of an image reconstruction method according to an embodiment of this application. Figure 1 As shown, the method includes the following steps:

[0047] Step S101: Determine the defocusing blur features of the imaging system based on multiple sample images, wherein the multiple sample images are images generated by the imaging system whose imaging quality meets the preset image processing requirements.

[0048] Specifically, defocus blur refers to the blurring of an image in an imaging system when an object is not accurately focused on the focal point of the imaging plane. Defocus blur manifests in various ways, including the degree and extent of blurring; for example, edge details and fine structures in the image may be "diffused" or blurred. This embodiment determines the defocus blur characteristics of an imaging system based on multiple sample images, laying the foundation for removing defocus blur in image reconstruction.

[0049] In this context, a sample image refers to an image acquired by an imaging system that meets preset image processing requirements. An imaging system is a device or combination of systems capable of optically imaging an object, such as a camera or scanner, used to acquire image data. A sample image is a clear image obtained in focus with reduced blur and noise interference. The appropriate number of sample images is selected based on system resources. While increasing the number of samples improves the deblurring effect, it also increases computational costs. The number of sample images can be set to 5 or more, for example, 20 or more, to improve the accuracy of deblurring during image reconstruction. Preset image processing requirements ensure that the sample images have a certain level of clarity or quality for analysis and extraction of out-of-focus blur features. For example, in the field of gene sequencing, meeting preset image processing requirements means that the image clearly displays sequencing sites and fluorescence signals, enabling the system to accurately identify base information and ensure the reliability and accuracy of sequencing data.

[0050] Step S102: Determine the motion blur characteristics of the imaging system based on the scanning parameters of the imaging system.

[0051] Specifically, the scanning parameters of the imaging system can be known device parameters such as scanning speed and scanning direction. Scanning speed refers to the moving speed of the imaging system, which directly affects the degree of motion blur. The faster the speed, the more severe the motion blur; speeds greater than 20 mm / s will produce motion blur. The scanning direction refers to the direction of movement of the object or imaging device, determining the direction of motion blur (e.g., horizontal or vertical). This embodiment determines the motion blur characteristics of the imaging system based on its scanning parameters, laying the foundation for removing motion blur in image reconstruction.

[0052] Step S103: The defocused blur feature and motion blur feature are fused to obtain the target point diffusion function.

[0053] The fusion method can be weighted summation, and the resulting target point diffusion function integrates the out-of-focus blur features and motion blur features, and quantifies their influence in the image, providing a basis for subsequent deblurring processing.

[0054] Step S104: Reconstruct the initial image generated by the imaging system based on the target point spread function to obtain the target image.

[0055] Specifically, the reconstruction process can involve using the joint point spread function and deconvolution algorithm to process the blurred initial image, remove the defocus blur and motion blur of the initial image, restore the edge details and structural information of the initial image, and obtain a target image whose imaging quality meets the preset image processing requirements.

[0056] The image reconstruction method provided in this application determines the defocus blur features of an imaging system based on multiple sample images, wherein the multiple sample images are images generated by the imaging system whose imaging quality meets preset image processing requirements; determines the motion blur features of the imaging system based on the scanning parameters of the imaging system; fuses the defocus blur features and the motion blur features to obtain a target point spread function; and reconstructs the initial image generated by the imaging system based on the target point spread function to obtain a target image. This solves the problem in related technologies where accurate acquisition of the point spread function is difficult and the image reconstruction effect is unsatisfactory. By combining multiple clear images generated by the imaging system with the known scanning parameters generated by the imaging system, the target point spread function is accurately estimated, and the initial image is reconstructed based on the target point spread function, achieving a significant improvement in the quality of the reconstructed image and providing clear image support for subsequent image processing.

[0057] Frequency domain information can characterize the defocusing and blurring features of an image. Optionally, in the image reconstruction method provided in this application embodiment, determining the defocusing and blurring features of the imaging system based on multiple sample images includes: extracting the frequency domain information of multiple sample images respectively; and determining the defocusing and blurring features based on the frequency domain information of multiple sample images.

[0058] It should be noted that in determining the out-of-focus blur features, the sample images are transformed from the time domain to the frequency domain using Fourier transform to obtain the frequency information of each sample image:

[0059]

[0060] Where f(x,y) represents the original spatial domain image, F(u,v) represents the corresponding frequency domain representation, M represents the number of pixels in the horizontal direction of the image, i.e., the width of the image; N represents the number of pixels in the vertical direction of the image, i.e., the height of the image; u,v represent the frequency domain coordinates used to represent points in frequency space; x,y represent the spatial domain coordinates used to represent the specific location of the image in the spatial domain; e represents the base of the complex exponent; i is the imaginary unit; and π is pi.

[0061] In the Fourier transform, each pixel position of the sample image is traversed and processed to generate a corresponding frequency domain representation, thus obtaining frequency domain information. Then, the frequency domain information of each sample image is accumulated and averaged to obtain the defocus blur PSF. defocus out-of-focus blur PSF defocus The specific mathematical representation of the defocus blur feature:

[0062]

[0063] Among them, F i (u,v) is the Fourier transform result of the i-th sharp image, N is the number of sharp images, and |F i (u,v)| is F i The absolute value of (u,v).

[0064] According to the embodiments of this application, Fourier transform is used to perform frequency domain transformation on the sample image to obtain the frequency domain information of the sample image. Frequency domain transformation can convert the details in the image into high-frequency components, while the overall structure is represented by low-frequency components. Defocus blur is represented by the attenuation of high-frequency components, so it can be more easily identified in the frequency domain. After obtaining the frequency domain information of the sample image, the defocus blur features are further determined to provide accurate parameters for subsequent image reconstruction.

[0065] Motion blur features can be simulated based on convolution kernel functions. Optionally, the motion blur features of the imaging system can be determined based on the scanning parameters of the imaging system, including: calculating the motion blur length based on the scanning speed and exposure time; determining a uniformly distributed convolution kernel with a length equal to the motion blur length along the direction of motion to obtain the motion blur features.

[0066] For example, the motion blur PSF can be obtained by simulating the linear motion blur phenomenon that may occur when an image is scanned rapidly by setting the length L and angle θ of the convolution kernel function. motion .

[0067] PSF motion The value of is uniformly distributed along the direction of motion, and the formula can be expressed as:

[0068]

[0069] Where h(x,y) represents the value of the convolution kernel at position (x,y), L is the motion blur length, θ is the angle of motion, and (x,y) are the coordinates relative to the center of the convolution kernel. The length of the motion blur L can be calculated using the relative speed of the camera or sample and the exposure time: L = v·t, where L is the length of the motion blur, v is the scanning speed, and t is the exposure time.

[0070] This embodiment simulates the blurring effect during the scanning process based on the convolution kernel function, generating corresponding motion blur features to provide accurate parameters for subsequent image reconstruction.

[0071] To accurately determine the target point diffusion function, optionally, in the image reconstruction method provided in this application embodiment, fusing the defocused blur feature and the motion blur feature to obtain the target point diffusion function includes: setting a first weight for the defocused blur feature and a second weight for the motion blur feature; normalizing the defocused blur feature and the motion blur feature; and performing a weighted summation of the normalized defocused blur feature and the normalized motion blur feature to obtain the target point diffusion function.

[0072] Specifically, the out-of-focus blur PSF defocus and motion blur PSF motion The weighted combinations are then used to form a unified joint PSF. combined , which is the target point spread function.

[0073] PSF combined =ω defocus ·PSF defocus +ω motion ·PSF motion ;

[0074] Where, ω defocus and ω motion It represents the weights of out-of-focus blur and motion blur. By setting different weight parameters, it is possible to flexibly adapt to the relative influence of out-of-focus blur and motion blur in different blurry scenarios.

[0075] The purpose of normalization is to adjust the two fuzzy features to the same scale, eliminating the difference in their numerical ranges. This allows the two fuzzy features to be weighted and summed under the same standard, resulting in a more accurate target point diffusion function. For example, when the out-of-focus fuzzy feature is in a lower numerical range while the motion fuzzy feature is in a higher numerical range, this difference may cause one feature to dominate the weighted summation process, affecting the balance of the fusion.

[0076] According to the embodiments of this application, the defocused blur feature and motion blur feature are normalized and then weighted and summed to obtain the target point spread function. By assigning appropriate weights, the relative influence of the two blur features in the target point spread function can be controlled, which can ensure that the target point spread function is used to remove the blur in the image more accurately.

[0077] To make the target point diffusion function more accurate, optionally, in the image reconstruction method provided in this application embodiment, before performing a weighted summation of the normalized defocused blur features and the normalized motion blur features, the method further includes: setting initial weights for the defocused blur features and the motion blur features respectively; increasing the weight of the defocused blur features and decreasing the weight of the motion blur features when the degree of defocused blur in the initial image is greater than a preset blur level and the scanning speed of the imaging system is less than a preset speed; or increasing the weight of the motion blur features and decreasing the weight of the defocused blur features when the degree of defocused blur in the initial image is less than a preset blur level and the scanning speed of the imaging system is greater than a preset speed.

[0078] Specifically, the initial weight ratio of defocus blur features and motion blur features can be 1:1. It should be noted that as the scanning speed increases, the degree of motion blur increases accordingly, while the impact of defocus blur is relatively small; conversely, when the scanning speed is slow and the focal point deviates from the imaging plane, the impact of motion blur decreases, while the impact of defocus blur becomes more significant. Therefore, the first weight of the defocus blur feature and the second weight of the motion blur feature can be dynamically adjusted based on the characteristics of the equipment, the actual image blur situation, and known scanning conditions. For example, if the defocus blur feature has a greater impact, the first weight should be increased; if the motion blur feature has a greater impact, the second weight should be increased.

[0079] For example, the degree of defocus blur can be judged visually. If the initial image has a higher degree of defocus blur than a preset blur level, and the scanning speed of the imaging system is lower than a preset speed (20 mm / s), it indicates that defocus blur has a greater impact on the image, while motion blur has a smaller impact. In this case, the dominant role of defocus blur is reflected by increasing the weight of the defocus blur feature and decreasing the weight of the motion blur feature. Conversely, if the initial image has a lower degree of defocus blur than a preset blur level, and the scanning speed of the imaging system is higher than a preset speed, it indicates that motion blur has a greater impact on the image, while defocus blur has a smaller impact. In this case, the weight of the motion blur feature is increased and the weight of the defocus blur feature is decreased to reflect the impact of motion blur on the image.

[0080] Based on the specific imaging conditions and image features of the embodiments of this application, the weight allocation of the target point spread function is optimized, thereby more accurately determining the target point spread function and improving the effect of subsequent image reconstruction.

[0081] Convolution algorithms can be used to reconstruct images. Optionally, in the image reconstruction method provided in this application embodiment, reconstructing the initial image generated by the imaging system based on the target point spread function to obtain the target image includes: determining the deconvolution kernel of the target point spread function; and using a deconvolution algorithm to deblur the initial image through the deconvolution kernel to obtain the target image.

[0082] Specifically, the reconstruction process uses a pre-built joint PSF. combined The initial image generated by the imaging system is processed using the Richardson-Lucy deconvolution algorithm. This algorithm iteratively reduces the blurring effects caused by defocusing and motion, restoring edge details and structural information to obtain the target image. The specific number of iterations is set according to the characteristics of the device and the actual blurring situation of the image.

[0083] The deconvolution kernel is a mathematical model that can inversely compensate for image blur caused by blur features in deconvolution processing. The process of determining the deconvolution kernel of the target point spread function is as follows: First, the target point spread function is transformed into the frequency domain, then the magnitude squared of the PSF is calculated, and then the Fourier transform of the deconvolution kernel is calculated. At the same time, a regularization parameter is added to avoid instability in the calculation, thus obtaining the representation of the deconvolution kernel in the frequency domain. The deconvolution kernel in the frequency domain is then transformed back into the spatial domain for application in subsequent deblurring processing.

[0084] The Richardson-Lucy deconvolution algorithm and a defined deconvolution kernel are used to deblur an initially blurred image. The deconvolution algorithm gradually restores the image's sharpness and details through iterative calculations or specific deblurring operations, thus obtaining the deblurred target image. The Richardson-Lucy deconvolution algorithm formula is as follows:

[0085]

[0086] Among them I (k) B is the estimated image at the k-th iteration, and B is the blurred image. This represents the convolution operation, where PSF is the point spread function. * It is the deconvolution kernel of PSF. After a certain number of iterations, the image reaches the preset target image standard, and the target image is obtained.

[0087] According to the embodiments of this application, a deconvolution kernel is calculated as a key parameter of the deconvolution algorithm. The initial blurred image is subjected to iterative deconvolution operations to finally restore the image's clarity and quality, thereby achieving deblurring of the blurred image.

[0088] To continuously improve image clarity in image reconstruction methods, optionally, in the image reconstruction method provided in this application embodiment, the initial image is deblurred using a deconvolution algorithm with a deconvolution kernel to obtain the target image, including: performing a first deblurring process on the initial image based on the target point spread function and the deconvolution kernel to obtain a first reconstructed image; performing a second deblurring process on the initial image based on the target point spread function, the deconvolution kernel, and the first reconstructed image to obtain a second reconstructed image; performing a second deblurring process on the initial image based on the target point spread function, the deconvolution kernel, and the second reconstructed image to obtain a third reconstructed image, until the number of deblurring processes reaches a preset number of iterations.

[0089] Specifically, multiple iterative deconvolution processes continuously improve image sharpness and quality through progressive deblurring operations. Each iteration builds upon the previous reconstruction result, gradually removing blur caused by out-of-focus and motion blur features. In each iteration, based on the current imaging quality and the convolution error of the blurred image, the fusion parameters and / or reconstruction parameters are updated, thus gradually reducing the error.

[0090] It should be noted that excessive iterations may introduce noise or artifacts. Therefore, a reasonable preset number of iterations is needed to obtain a suitable deblurring effect, ensuring that the number of iterations is neither too many nor too few. The minimum number of iterations can be 30-50, and the specific number can be determined based on the characteristics of the shooting equipment. The faster the equipment moves, the more iterations are required.

[0091] According to the embodiments of this application, the blurring process is performed through continuous iteration to repair the image affected by defocusing blur and motion blur during the imaging process, so that the edges and fine structures in the image can be gradually restored, achieving a high-precision blurring effect.

[0092] To avoid the reconstructed target image failing to meet preset image processing requirements, optionally, in the image reconstruction method provided in this application embodiment, after reconstructing the initial image generated by the imaging system based on the target point spread function to obtain the target image, the method further includes: determining whether the imaging quality of the target image meets preset image processing requirements; if the imaging quality of the target image does not meet the preset image processing requirements, if the sharpness index of the target image is less than a preset value, updating the fusion parameters, and / or updating the reconstruction parameters; performing a step of fusing defocus blur features and motion blur features based on the updated fusion parameters, and / or performing a step of reconstructing the initial image generated by the imaging system based on the target point spread function based on the updated reconstruction parameters.

[0093] To verify the effectiveness of the deblurring process, this application introduces several image sharpness metrics, including Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM), and gradient strength. PSNR, SSIM, and gradient strength are considered better when closer to 1. Preset image processing requirements are also introduced. Whether an image meets these requirements can be measured by the base recognition algorithm. This algorithm is used in sequencing to analyze base fluorescence signals. By identifying specific fluorescence signals emitted by different bases, it determines the base sequence in the gene sequence to assess signal accuracy.

[0094] In this embodiment, after the target image is reconstructed, its imaging quality is evaluated to check whether it meets preset image processing requirements. This quality check aims to ensure the deblurring effect meets the requirements. If the imaging quality of the target image does not meet the preset standard, and the sharpness index is lower than the preset value, parameter updates are triggered. The updated parameters include: fusion parameters, used to readjust the weights of defocus blur features and motion blur features to more accurately reflect the current blur situation; and reconstruction parameters, which may include the number of deconvolution iterations or other detailed parameters to further optimize the image reconstruction effect. Guided by the updated fusion parameters and / or reconstruction parameters, the fusion of defocus blur features and motion blur features is re-executed to generate a new target point spread function, and a new round of reconstruction is performed on the initial image based on this target point spread function. This iterative process ensures that the image sharpness is gradually optimized until the preset imaging quality requirements are met.

[0095] According to an embodiment of this application, an optional image reconstruction method is provided.

[0096] Figure 2 This is a flowchart of an optional image reconstruction method according to an embodiment of this application. For example... Figure 2 As shown, the method includes the following steps:

[0097] S201: Determine the defocusing blur characteristics of the imaging system based on multiple sample images, wherein the multiple sample images are images generated by the imaging system whose imaging quality meets the preset image processing requirements.

[0098] Specifically, defocus blur refers to the blurring of an image in an imaging system when an object is not accurately focused on the focal point of the imaging plane. Defocus blur manifests in various ways, including the degree and extent of the blur; for example, edge details and fine structures in the image may be "diffused" or blurred.

[0099] In this context, a sample image refers to an image acquired by an imaging system that meets preset image processing requirements. An imaging system is a device or combination of systems capable of optically imaging an object, such as a camera or scanner, used to acquire image data. A sample image is a clear image obtained in focus with reduced blur and noise interference. The appropriate number of sample images is selected based on system resources. While increasing the number of samples improves the deblurring effect, it also increases computational costs. The number of sample images can be set to 5 or more, for example, 20 or more, to improve the accuracy of deblurring during image reconstruction. Preset image processing requirements ensure that the sample images have a certain level of clarity or quality for analysis and extraction of out-of-focus blur features. For example, in the field of gene sequencing, meeting preset image processing requirements means that the image clearly displays sequencing sites and fluorescence signals, enabling the system to accurately identify base information and ensure the reliability and accuracy of sequencing data.

[0100] S202: Determine the target point diffusion function based on the defocused blur characteristics.

[0101] Specifically, the out-of-focus blur feature can be identified as the target point spread function. The target point spread function quantifies the impact of out-of-focus blur in the image, providing a basis for subsequent deblurring processing.

[0102] S203: Reconstruct the initial image generated by the imaging system based on the target point spread function to obtain the target image.

[0103] Specifically, the reconstruction process can involve processing the blurred initial image using a target point spread function and a deconvolution algorithm to remove the defocus blur and motion blur of the initial image, restore the edge details and structural information of the initial image, and obtain a target image whose imaging quality meets the preset image processing requirements.

[0104] The image reconstruction method provided in this application determines the defocus blur features of an imaging system based on multiple sample images, wherein the multiple sample images are images generated by the imaging system whose imaging quality meets preset image processing requirements; determines the target point spread function based on the defocus blur features; and reconstructs the initial image generated by the imaging system based on the target point spread function to obtain the target image. This solves the problem in related technologies where it is difficult to accurately obtain the point spread function and the image reconstruction effect is not ideal. By accurately estimating the target point spread function by combining multiple clear images generated by the imaging system, and reconstructing the initial image based on the target point spread function, the quality of the reconstructed image is significantly improved, providing clear image support for subsequent image processing.

[0105] Frequency domain information can characterize the defocusing and blurring features of an image. Optionally, in the image reconstruction method provided in this application embodiment, determining the defocusing and blurring features of the imaging system based on multiple sample images includes: extracting the frequency domain information of multiple sample images respectively; and determining the defocusing and blurring features based on the frequency domain information of multiple sample images.

[0106] It should be noted that in determining the out-of-focus blur features, the sample images are transformed from the time domain to the frequency domain using Fourier transform to obtain the frequency information of each sample image:

[0107]

[0108] Where f(x,y) represents the original spatial domain image, F(u,v) represents the corresponding frequency domain representation, M represents the number of pixels in the horizontal direction of the image, i.e., the width of the image; N represents the number of pixels in the vertical direction of the image, i.e., the height of the image; u,v represent the frequency domain coordinates used to represent points in frequency space; x,y represent the spatial domain coordinates used to represent the specific location of the image in the spatial domain; e represents the base of the complex exponent; i is the imaginary unit; and π is pi.

[0109] In the Fourier transform, each pixel position of the sample image is traversed and processed to generate a corresponding frequency domain representation, thus obtaining frequency domain information. Then, the frequency domain information of each sample image is accumulated and averaged to obtain the defocus blur PSF. defocus out-of-focus blur PSF defocus The specific mathematical representation of the defocus blur feature:

[0110]

[0111] Among them, F i (u,v) is the Fourier transform result of the i-th sharp image, N is the number of sharp images, and |F i (u,v) is F i The absolute value of (u,v).

[0112] According to the embodiments of this application, Fourier transform is used to perform frequency domain transformation on the sample image to obtain the frequency domain information of the sample image. Frequency domain transformation can convert the details in the image into high-frequency components, while the overall structure is represented by low-frequency components. Defocus blur is represented by the attenuation of high-frequency components, so it can be more easily identified in the frequency domain. After obtaining the frequency domain information of the sample image, the defocus blur features are further determined to provide accurate parameters for subsequent image reconstruction.

[0113] According to an embodiment of this application, a sequencing method based on image reconstruction is provided.

[0114] Figure 3 This is a flowchart of a sequencing method according to an embodiment of this application. For example... Figure 3 As shown, the method includes the following steps:

[0115] S301: Determine the defocusing blur characteristics of the sequencing system based on multiple sample sequencing images, wherein the multiple sample sequencing images are images generated by the sequencing system whose imaging quality meets the requirements of gene sequencing processing.

[0116] A sequencing system is an imaging system used for genome sequencing, capable of capturing and analyzing fluorescence signals of base sequences in biological samples. Figure 4 This is a schematic diagram of a sequencing system according to an embodiment of this application, such as... Figure 4 As shown, the sequencing system includes an optical system and a motion platform. The optical system includes an autofocus module, a photoelectric conversion module, a laser, and a camera (e.g., a TDI (Time Delay Integration) camera, an area array camera, etc.).

[0117] Sample sequencing images can be clear images obtained with reduced blur and noise interference under focused conditions. The sequencing system generates sample sequencing images as follows: When the autofocus module is activated, near-infrared light is incident on the object surface. The reflected signal enters the photoelectric conversion module. The autofocus module analyzes this reflected signal, calculates the defocus amount of the sample, and drives the objective lens to move along the Z-axis until the optimal focal plane position is reached. After the Z-axis stabilizes, the fluorescence signal generated by the sample is excited by a laser and transmitted to a TDI camera for acquisition via the optical system, generating multiple sample sequencing images. Based on these sample sequencing images, defocused blur features can be accurately extracted. For example, the sequencing system uses a laser and a TDI camera to generate sequencing images. Figure 5 This is a schematic diagram of the TDI camera imaging process according to an embodiment of this application, as shown below. Figure 5 As shown: The integration direction of the TDI camera is along the laser direction. When the laser excites the sample, the camera can synchronously trigger integration. Specifically, when the laser irradiates a certain area of ​​the sample, it excites a fluorescence signal, and the corresponding TDI camera begins integration and acquisition. When the laser leaves the sample area, the camera's integration ends, completing the acquisition of the fluorescence signal. The fluorescence signal refers to the light signal emitted by fluorescent molecules in the sample under laser excitation at a specific wavelength during sequencing. Each base (such as A, T, C, G) is labeled with a different fluorescent dye, which emits a specific color of fluorescence after laser excitation, allowing the TDI camera or other imaging devices to capture these fluorescence signals. To ensure the synchronization and accuracy of the timing during acquisition, the TDI camera and the motion platform are connected via hard triggering. Connection methods include, but are not limited to, direct signal connection, synchronization controllers or trigger modules, encoder signal triggering, and digital I / O interfaces.

[0118] The sample sequencing images are microscopic images of multiple fluorescent dots that meet preset image processing requirements. Each fluorescent dot represents a base position, and the color is produced by laser excitation at different wavelengths. The sequencing system generates these sample sequencing images by laser-excited fluorescence and a TDI camera acquiring signals, which are then used for subsequent base identification and gene sequence analysis. In the field of gene sequencing, meeting preset image processing requirements means that the image can clearly display the sequencing sites and fluorescence signals, enabling the system to accurately identify base information and ensure the reliability and accuracy of the sequencing data. The number of sample images is selected appropriately based on system resources. While increasing the number of samples can improve the deblurring effect, it also increases computational costs. The number of sample images can be set to 5 or more, for example, 20 or more, to improve the accuracy of deblurring during image reconstruction. To ensure data reliability, this embodiment selects all clear images within one period, totaling 114 images.

[0119] It should be noted that defocusing blur refers to the blurring of an image in an imaging system when an object is not accurately focused on the focal point of the imaging plane. Defocusing blur manifests in both its degree and extent; for example, edge details and fine structures in the image may be "diffused" or blurred. The causes of defocusing blur in sequencing systems are as follows: samples are attached to the moving platform in various ways, such as clamping, positive and negative pressure adsorption, and embedding. These attachment methods can easily lead to unevenness on the sample surface, causing inaccurate focusing and image blurring. Furthermore, the timing of the image capture process is accurate to the millisecond level. If the autofocus module's judgment and processing are not timely enough, or if the Z-axis cannot effectively follow the movement of the sample surface, defocusing blur can also occur. This embodiment uses multiple sample images to determine the defocusing blur characteristics of the sequencing system, laying the foundation for removing defocusing blur from sequencing images during image reconstruction.

[0120] Optionally, in the image reconstruction method provided in this application embodiment, determining the defocus blur features of the sequencing system based on multiple sample sequencing images includes: extracting the frequency domain information of the multiple sample sequencing images respectively; and determining the defocus blur features based on the frequency domain information of the multiple sample sequencing images.

[0121] It should be noted that in determining the out-of-focus blur features, Fourier transform is used to convert the sample sequencing images from the time domain to the frequency domain, obtaining the frequency information of each sample sequencing image:

[0122]

[0123] Where f(x,y) represents the original spatial domain image, F(u,v) represents the corresponding frequency domain representation, M represents the number of pixels in the horizontal direction of the image, i.e., the width of the image; N represents the number of pixels in the vertical direction of the image, i.e., the height of the image; u,v represent the frequency domain coordinates used to represent points in frequency space; x,y represent the spatial domain coordinates used to represent the specific location of the image in the spatial domain; e represents the base of the complex exponent; i is the imaginary unit; and π is pi.

[0124] In the Fourier transform, each pixel position of the sample sequencing image is traversed and processed to generate the corresponding frequency domain representation, thus obtaining the frequency domain information. Then, the frequency domain information of each sample sequencing image is accumulated and averaged to obtain the defocus blur PSF. defocus out-of-focus blur PSF defocus The specific mathematical representation of the defocus blur feature:

[0125]

[0126] Among them, F i (u,v) is the Fourier transform result of the i-th sharp image, N is the number of sharp images, and |F i (u,v) is F i The absolute value of (u,v).

[0127] S302: Determine the motion fuzziness characteristics of the sequencing system based on the scanning parameters of the sequencing system.

[0128] The reason for motion blur in sequencing systems is that the motion platform needs to move at high speed to improve acquisition efficiency when scanning samples. However, when the movement speed exceeds a certain threshold (20 mm / s), the platform movement will produce ghosting during image acquisition, resulting in motion blur. In particular, when the TDI camera is integrating, it triggers integration acquisition synchronously with the motion platform. If the platform movement speed is too fast or the synchronization accuracy is insufficient, the details in the image will become blurred due to the motion.

[0129] Specifically, the scanning parameters of the imaging system can be known equipment parameters such as scanning speed and scanning direction. Scanning speed refers to the moving speed of the imaging system, which directly affects the degree of motion blur. The faster the speed, the more severe the motion blur; speeds greater than 20 mm / s will produce motion blur. The scanning direction refers to the direction of movement of the object or imaging device, determining the direction of motion blur (e.g., horizontal or vertical). Optionally, determining the motion blur characteristics of the imaging system based on its scanning parameters includes: calculating the motion blur length based on the scanning speed and exposure time; and determining a uniformly distributed convolution kernel along the direction of motion to obtain the motion blur characteristics.

[0130] For example, the motion blur PSF can be obtained by simulating the linear motion blur phenomenon that may occur when an image is scanned rapidly by setting the length L and angle θ of the convolution kernel function. motion .

[0131] PSF motion The value of is uniformly distributed along the direction of motion, and the formula can be expressed as:

[0132]

[0133] Where h(x,y) represents the value of the convolution kernel at position (x,y), L is the motion blur length, θ is the angle of motion, and (x,y) are the coordinates relative to the center of the convolution kernel. The length of the motion blur L can be calculated using the relative speed of the camera or sample and the exposure time: L = v·t, where L is the length of the motion blur, v is the scanning speed, and t is the exposure time.

[0134] For example, the TDI camera's acquisition direction, the speed of the motion platform, and the angle are input. In this embodiment, the TDI camera acquires images along the Y direction, the platform speed is approximately 35 mm / s, and the angle is approximately 0°. Based on this information, motion blur features are generated.

[0135] S303: The defocused blur feature and motion blur feature are fused to obtain the target point spread function.

[0136] The fusion method can be weighted summation. The resulting target point spread function integrates the out-of-focus blur features and motion blur features, quantifying their impact on the image and providing a basis for subsequent deblurring processing. For example, the out-of-focus PSF and motion blur PSF are first resized to ensure consistent dimensions. Then, the out-of-focus PSF and motion blur PSF are weighted and combined to generate a comprehensive target PSF, used to characterize the composite blur properties. The weights can be dynamically adjusted based on device characteristics, the actual image blur situation, and known scanning conditions. For example, if the out-of-focus blur features have a greater impact, the weight of the out-of-focus PSF is increased; if the motion blur features have a greater impact, the weight of the motion blur PSF is increased. For instance, the out-of-focus PSF weight is set to 0.6, and the motion blur PSF weight is set to 0.4.

[0137] S304: Reconstruct the initial sequencing image generated by the sequencing system based on the target site diffusion function to obtain the target sequencing image, and perform base identification based on the target sequencing image.

[0138] Specifically, the reconstruction process can involve using the joint point spread function and deconvolution algorithm to process the blurred initial sequencing image, remove the defocus blur and motion blur of the initial sequencing image, restore the edge details and structural information of the initial sequencing image, and obtain a target image whose imaging quality meets the preset image processing requirements.

[0139] For example, the Richardson-Lucy deconvolution algorithm is used to progressively improve image sharpness with each iteration. The initial number of iterations is set to 30, and can be dynamically adjusted based on actual results; in this embodiment, it is ultimately set to 35. A reasonable number of iterations is crucial; too many iterations may introduce noise or artifacts, while too few may not be sufficient to recover image details.

[0140] Defocusing and motion blur significantly impact the quality of sequencing data, leading to weakened fluorescence signals, reduced signal-to-noise ratio, decreased resolution, and even base identification errors, thus affecting the accuracy and reliability of sequencing results. Fluorescence detection systems prioritize ensuring the rapid response of the autofocus module and the Z-axis to avoid image quality degradation and guarantee high system accuracy and reliability. However, even with system hardware performance at its limits, blurring may still occur in a few areas of the image.

[0141] Figure 6 This is a schematic diagram illustrating the simultaneous existence of blurriness and sharpness within a single FOV according to an embodiment of this application. A single FOV (Field of View) image is a field-of-view image captured in a sequencing system, containing multiple base signal points labeled with fluorescent dyes. These fluorescent points are distributed throughout the image, each point representing the location and type of a base, such as... Figure 6 As shown, the upper half is blurry while the lower half is clear.

[0142] Figure 7(a) shows the sequencing images before and after reconstruction according to the embodiments of this application. A base comparison before and after reconstruction. Figure 1 Figure 7(a) shows the fluorescence images before and after reconstruction. The upper half of the original image is blurred to the point that the fluorescent dots are difficult to distinguish visually. Figure 7(b) is a comparison of the sequencing images before and after reconstruction according to the embodiment of this application, showing the A bases before and after reconstruction. Figure 2 Figure 7(b) is a partial enlarged view of Figure 7(a), and Figure 7(c) is a comparison of the A bases in the sequencing images before and after reconstruction according to the embodiments of this application. Figure 3 Figure 7(c) is a magnified schematic diagram of a single fluorescent spot in Figure 7(b).

[0143] To further verify the effectiveness of the reconstruction method, several image quality evaluation metrics were introduced. Figure 8 This is a schematic diagram of image quality evaluation according to an embodiment of this application, such as... Figure 8As shown. This includes peak signal-to-noise ratio (PSNR), structural similarity analysis (SSIM), gradient strength analysis, and quality analysis of the base identification algorithm.

[0144] Analysis was performed using four base images from a single field of view (see...). Figure 8 Compared to the blurred image, the reconstructed image showed a PSNR increase of 21.3853 dB to 28.3250 dB and an SSIM increase of 0.7195 to 0.8859, both indicating significant improvements in signal-to-noise ratio and structural similarity. A PSNR value exceeding 20 dB indicates reduced noise levels and improved signal quality, while SSIM performed well in terms of brightness, contrast, and structure. The average gradient intensity reflects the sharpness of edges and details in the image; a higher gradient indicates sharper image details. In the reconstructed A-base image, the gradient intensity increased from 0.3112 to 0.638, an increase of 105.01%, significantly enhancing the sharpness of edges and details. The gradient intensity of the G-base image increased from 0.2126 to 0.5033, an increase of 136.74%, showing particularly outstanding restoration results. The gradient intensity improvements in the T and C-base images were relatively smaller, but still showed significant improvement.

[0145] Furthermore, the recognition accuracy was assessed using Base Identification Information Content (BIC) and Crosstalk Fitting Index (FIT). The BIC of the reconstructed image improved by 29.74%, and the FIT improved by 23.31%, indicating that the reconstructed image significantly improved the signal quality and discriminative ability of base recognition. The Q30 and Q40 metrics for sequencing accuracy also showed significant improvements, with Q30 increasing from 0.654 to 0.931 and Q40 from 0.578 to 0.888, representing increases of 42.35% and 53.63%, respectively, indicating a significant improvement in high-precision base recognition. The total number of recognized bases in both sets of images was consistent at 6,975,788, indicating that the reconstructed image improved recognition accuracy with the same amount of data.

[0146] The reconstruction method described in this embodiment is demonstrated to improve image sharpness and detail recovery using quality evaluation metrics such as PSNR, SSIM, and gradient strength. It effectively addresses the simultaneous issues of defocusing and motion blur in sequencing images. Furthermore, the significant improvement in key base recognition metrics (BIC, FIT, Q30, Q40) confirms the algorithm's high performance in data fitting and high-precision recognition.

[0147] The image reconstruction method provided in this application determines the defocus blur features of a sequencing system based on multiple sample images, wherein the multiple sample images are images generated by the sequencing system whose sequencing quality meets preset image processing requirements; determines the motion blur features of the sequencing system based on the scanning parameters of the sequencing system; fuses the defocus blur features and the motion blur features to obtain the target point spread function; and reconstructs the initial image generated by the sequencing system based on the target point spread function to obtain the target image. This solves the problem of difficulty in accurately obtaining the point spread function and unsatisfactory image reconstruction results in related technologies. By combining multiple clear images generated by the imaging system with the known scanning parameters generated by the imaging system, the target point spread function is accurately estimated, and the initial image is reconstructed based on the target point spread function, achieving a significant improvement in the quality of the reconstructed image and providing clear image support for subsequent image processing.

[0148] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0149] This application also provides an image reconstruction apparatus. It should be noted that the image reconstruction apparatus of this application can be used to execute the image reconstruction method provided in this application. The image reconstruction apparatus provided in this application is described below.

[0150] Figure 9 This is a schematic diagram of an image reconstruction apparatus according to an embodiment of this application. Figure 9 As shown, the device includes: a first determining unit 901, a second determining unit 902, a first fusion unit 903, and a first reconstruction unit 904.

[0151] The first determining unit 901 determines the defocusing blur characteristics of the imaging system based on multiple sample images, wherein the multiple sample images are images generated by the imaging system whose imaging quality meets preset image processing requirements.

[0152] The second determining unit 902 determines the motion blur features of the imaging system based on the scanning parameters of the imaging system.

[0153] The first fusion unit 903 fuses the defocused blur feature and the motion blur feature to obtain the target point diffusion function.

[0154] The first reconstruction unit 904 reconstructs the initial image generated by the imaging system based on the target point diffusion function to obtain the target image.

[0155] The image reconstruction apparatus provided in this application solves the problem that it is difficult to accurately obtain the point spread function in related technologies, resulting in unsatisfactory image reconstruction effects. By combining multiple clear images generated by the imaging system with known scanning parameters generated by the imaging system, the target point spread function is accurately estimated. Based on the target point spread function, the initial image is reconstructed, which significantly improves the quality of the reconstructed image and provides clear image support for subsequent image processing.

[0156] Optionally, in the image reconstruction apparatus provided in this application embodiment, the first determining unit 901 includes: a first extraction module, used to extract frequency domain information of multiple sample images respectively; and a first determining module, used to determine defocus blur features based on the frequency domain information of multiple sample images.

[0157] Optionally, in the image reconstruction apparatus provided in this application embodiment, the first fusion unit 903 includes: a first setting module, used to set a first weight for the out-of-focus blur feature; a normalization module, used to set a second weight for the motion blur feature; normalizing the out-of-focus blur feature and the motion blur feature; and a weighted summation module, used to perform a weighted summation on the normalized out-of-focus blur feature and the normalized motion blur feature to obtain the target point diffusion function.

[0158] Optionally, in the image reconstruction apparatus provided in this application embodiment, the apparatus further includes: a second setting module, configured to set initial weights for the defocused blur features and motion blur features respectively before performing a weighted summation of the normalized defocused blur features and the normalized motion blur features; a first adjustment module, configured to increase the weight of the defocused blur features and decrease the weight of the motion blur features when the degree of defocused blur in the initial image is greater than a preset blur level and the scanning speed of the imaging system is less than a preset speed; or a second adjustment module, configured to increase the weight of the motion blur features and decrease the weight of the defocused blur features when the degree of defocused blur in the initial image is less than a preset blur level and the scanning speed of the imaging system is greater than a preset speed.

[0159] Optionally, in the image reconstruction apparatus provided in this application embodiment, the first reconstruction unit 904 includes: a second determining module, used to determine the deconvolution kernel of the target point spread function; and a deblurring processing module, used to perform deblurring processing on the initial image using a deconvolution algorithm through the deconvolution kernel to obtain the target image.

[0160] Optionally, in the image reconstruction apparatus provided in this application embodiment, the deblurring processing module includes: a first deblurring processing submodule, used to perform a first deblurring processing on the initial image based on the target point spread function and the deconvolution kernel to obtain a first reconstructed image; a second deblurring processing submodule, used to perform a second deblurring processing on the initial image based on the target point spread function, the deconvolution kernel, and the first reconstructed image to obtain a second reconstructed image; and a third deblurring processing submodule, used to perform a second deblurring processing on the initial image based on the target point spread function, the deconvolution kernel, and the second reconstructed image to obtain a third reconstructed image, until the number of deblurring processes reaches a preset number of iterations.

[0161] Optionally, in the image reconstruction apparatus provided in this application embodiment, the apparatus further includes: a judgment unit, configured to determine whether the imaging quality of the target image meets preset image processing requirements after reconstructing the initial image generated by the imaging system based on the target point spread function to obtain a target image; an update unit, configured to update the fusion parameters and / or update the reconstruction parameters if the sharpness index of the target image is less than a preset value when the imaging quality of the target image does not meet the preset image processing requirements; and an execution unit, configured to perform the step of fusing the defocus blur feature and motion blur feature based on the updated fusion parameters, and / or perform the step of reconstructing the initial image generated by the imaging system based on the target point spread function based on the updated reconstruction parameters.

[0162] The image reconstruction device includes a processor and a memory. The first determining unit 901, the second determining unit 902, the first fusing unit 903, and the first reconstruction unit 904 are all stored in the memory as program units. The processor executes the program units stored in the memory to realize the corresponding functions.

[0163] Figure 10 This is a schematic diagram of an optional image reconstruction apparatus according to an embodiment of this application. For example... Figure 10 As shown, the device includes: a third determining unit 1001, a fourth unit 1002, and a second reconstruction unit 1003.

[0164] The third determining unit 1001 determines the defocusing blur characteristics of the imaging system based on multiple sample images, wherein the multiple sample images are images generated by the imaging system whose imaging quality meets preset image processing requirements.

[0165] Unit 4, Question 1002: Determining the target point diffusion function based on defocused blur characteristics.

[0166] The second reconstruction unit 1003 reconstructs the initial image generated by the imaging system based on the target point diffusion function to obtain the target image.

[0167] The optional image reconstruction apparatus provided in this application embodiment solves the problem that it is difficult to accurately obtain the point spread function in related technologies and the image reconstruction effect is not ideal. By combining multiple clear images generated by the imaging system, the target point spread function is accurately estimated, and the initial image is reconstructed based on the target point spread function, which significantly improves the quality of the reconstructed image and provides clear image support for subsequent image processing.

[0168] Optionally, in the image reconstruction apparatus provided in this application embodiment, the third determining unit 1001 includes: a second extraction module, used to extract frequency domain information of multiple sample images respectively; and a third determining module, used to determine defocus blur features based on the frequency domain information of the multiple sample images.

[0169] The image reconstruction device described above includes a processor and a memory. The third determining unit 1001, the fourth unit 1002, and the second reconstruction unit 1003 are all stored in the memory as program units. The processor executes the program units stored in the memory to realize the corresponding functions.

[0170] Figure 11 This is a schematic diagram of a sequencing device according to an embodiment of this application. Figure 11 As shown, the device includes: a fifth determining unit 1101, a sixth determining unit 1102, a second fusion unit 1103, and a third reconstruction unit 1104.

[0171] The fifth determining unit 1101 determines the defocusing and blurring features of the sequencing system based on multiple sample images, wherein the multiple sample images are images generated by the sequencing system whose sequencing quality meets preset image processing requirements.

[0172] The sixth determining unit 1102 determines the motion blur features of the sequencing system based on the scanning parameters of the sequencing system; the second fusion unit 1103 fuses the defocus blur features and the motion blur features to obtain the target point diffusion function.

[0173] The third reconstruction unit 1104 reconstructs the initial image generated by the sequencing system based on the target point diffusion function to obtain the target image.

[0174] The optional image reconstruction apparatus provided in this application embodiment solves the problem that it is difficult to accurately obtain the point spread function in related technologies, resulting in unsatisfactory image reconstruction effects. By combining multiple clear images generated by the imaging system with known scanning parameters generated by the imaging system, the target point spread function is accurately estimated. Based on the target point spread function, the initial image is reconstructed, achieving a significant improvement in the quality of the reconstructed image and providing clear image support for subsequent image processing.

[0175] The image reconstruction device described above includes a processor and a memory. The fifth determining unit 1101, the sixth determining unit 1102, the second fusion unit 1103, and the third reconstruction unit 1104 are all stored in the memory as program units. The processor executes the program units stored in the memory to realize the corresponding functions.

[0176] The processor contains a kernel, which retrieves the corresponding program units from memory. One or more kernels can be configured; adjusting kernel parameters can significantly improve sequencing image quality, providing clear and accurate image support for subsequent data analysis.

[0177] The memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0178] This application also provides a computer storage medium for storing a program, wherein the program, when running, controls the device where the non-volatile storage medium is located to execute an image reconstruction method.

[0179] This application also provides an electronic device. Figure 12 This is a schematic diagram of an electronic device according to an embodiment of this application. The electronic device 120 includes a processor and a memory; the memory stores computer-readable instructions, and the processor is used to execute the computer-readable instructions, wherein the computer-readable instructions execute an image reconstruction method. The electronic device in this document may be a server, PC, PAD, mobile phone, etc.

[0180] This application also provides a computer program product, including a non-volatile computer-readable storage medium storing a computer program, which implements an image reconstruction method when executed by a processor.

[0181] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0182] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0183] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0184] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0185] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0186] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, like read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0187] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0188] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0189] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method of image reconstruction, characterized by, include: The defocusing and blurring features of the imaging system are determined based on multiple sample images, wherein the multiple sample images are images generated by the imaging system whose imaging quality meets preset image processing requirements; The motion blur characteristics of the imaging system are determined based on the scanning parameters of the imaging system. The defocus blur feature and the motion blur feature are fused to obtain the target point spread function; The target image is obtained by reconstructing the initial image generated by the imaging system based on the target point spread function.

2. The image reconstruction method of claim 1, wherein, The defocusing and blurring features of an imaging system determined based on multiple sample images include: Extract the frequency domain information from each of the multiple sample images; The defocus blur feature is determined based on the frequency domain information of the multiple sample images.

3. The image reconstruction method of claim 1, wherein, Determining the motion blur characteristics of the imaging system based on its scanning parameters includes: Calculate the motion blur length based on the scanning speed and exposure time; The motion blur feature is obtained by determining a uniformly distributed convolution kernel with a length equal to the motion blur length along the direction of motion.

4. The image reconstruction method of claim 1, wherein, The target point spread function is obtained by fusing the defocused blur feature and the motion blur feature. A first weight is assigned to the defocused blur feature, and a second weight is assigned to the motion blur feature; Normalize the defocus blur feature and the motion blur feature; The target point diffusion function is obtained by weighted summation of the normalized defocused blur feature and the normalized motion blur feature.

5. The image reconstruction method of claim 4, wherein, Before performing a weighted summation of the normalized defocused blur features and the normalized motion blur features, the method further includes: Initial weights are assigned to the defocused blur feature and the motion blur feature, respectively; If the defocus blur level of the initial image is greater than a preset blur level, and the scanning speed of the imaging system is less than a preset speed, then the weight of the defocus blur feature is increased, and the weight of the motion blur feature is decreased; or If the degree of defocus blur in the initial image is less than the preset blur degree, and the scanning speed of the imaging system is greater than the preset speed, the weight of the motion blur feature is increased, and the weight of the defocus blur feature is decreased.

6. The image reconstruction method of claim 1, wherein, The target image is reconstructed based on the target point spread function of the initial image generated by the imaging system, resulting in the following: Determine the deconvolution kernel of the spread function at the target point; The initial image is deblurred using a deconvolution algorithm with a deconvolution kernel to obtain the target image.

7. The image reconstruction method of claim 6, wherein, The initial image is deblurred using a deconvolution algorithm with the deconvolution kernel to obtain the target image, including: Based on the target point diffusion function and the deconvolution kernel, the initial image is subjected to a first deblurring process to obtain a first reconstructed image; Based on the target point diffusion function, the deconvolution kernel, and the first reconstructed image, the initial image is subjected to a second deblurring process to obtain the second reconstructed image; Based on the target point diffusion function, the deconvolution kernel, and the second reconstructed image, the initial image is subjected to a second deblurring process to obtain a third reconstructed image, until the number of deblurring processes reaches a preset number of iterations.

8. The image reconstruction method of claim 1, wherein, After reconstructing the initial image generated by the imaging system based on the target point diffusion function to obtain the target image, the method further includes: Determine whether the imaging quality of the target image meets the preset image processing requirements; If the imaging quality of the target image does not meet the preset image processing requirements, and if the sharpness index of the target image is less than the preset value, update the fusion parameters and / or update the reconstruction parameters. The steps of fusing the defocused blur features and the motion blur features are performed based on the updated fusion parameters, and / or the steps of reconstructing the initial image generated by the imaging system based on the target point diffusion function are performed based on the updated reconstruction parameters.

9. An image reconstruction method, characterized by, include: The defocusing and blurring features of the imaging system are determined based on multiple sample images, wherein the multiple sample images are images generated by the imaging system whose imaging quality meets preset image processing requirements; The target point diffusion function is determined based on the aforementioned defocusing blur characteristics; The target image is obtained by reconstructing the initial image generated by the imaging system based on the target point spread function.

10. The image reconstruction method of claim 9, wherein, The defocusing and blurring features of an imaging system determined based on multiple sample images include: Extract the frequency domain information from each of the multiple sample images; The defocus blur feature is determined based on the frequency domain information of the multiple sample images.

11. A sequencing method based on the image reconstruction method according to any one of claims 1 to 10, characterized in that, include: The defocusing and blurring features of the sequencing system are determined based on multiple sample sequencing images, wherein the multiple sample sequencing images are images generated by the sequencing system whose imaging quality meets the requirements of gene sequencing processing; The motion ambiguity features of the sequencing system are determined based on the scanning parameters of the sequencing system. The defocus blur feature and the motion blur feature are fused to obtain the target point spread function; The initial sequencing image generated by the sequencing system is reconstructed based on the target point diffusion function to obtain the target sequencing image, and base identification is performed based on the target sequencing image.

12. A computer program product, characterised in that, The method includes a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, implements the image reconstruction method of any one of claims 1 to 10 or the sequencing method of claim 11.

13. An electronic device comprising a memory and a processor, characterized in that The memory stores a computer program, and the processor is configured to execute the image reconstruction method of any one of claims 1 to 10 or the sequencing method of claim 11 through the computer program.

Images