Single-frame astronomical image reconstruction method and device, electronic equipment and storage medium

Through blind deconvolution algorithm and iterative optimization solution based on reweighted graphs, the problem of poor fuzzy kernel estimation stability in single-frame astronomical image reconstruction is solved, and efficient and low-cost image defuzzing is achieved, improving the improvement of the improvement of the image defuzzing process. The quality of astronomical images.

CN118967513AActive Publication Date: 2024-11-15TSINGHUA UNIVERSITY
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202411139297.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-19
Publication Date
2024-11-15
Estimated Expiration
2044-08-19

AI Technical Summary

Technical Problem

Astronomical images acquired by ground optical telescopes are affected by atmospheric turbulence, resulting in a decline in image quality. The prior art relies on multi-frame images for fuzzy core estimation, which has high calculation cost and poor stability.

Method used

Using blind deconvolution algorithm based on reweighted graphs, single-frame astronomical images are reconstructed and detailed recovery are carried out. Through iterative optimization solution and light intensity correction processing, the target fuzzy kernel is determined, and non-blind deconvolution is used. The algorithm performs defuzzing processing.

Benefits of technology

High-definition and high-quality astronomical image reconstruction are achieved, reducing the computing resources and time requirements, reducing the cost of single-frame astronomical image reconstruction, and improving processing efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118967513A_ABST
    Figure CN118967513A_ABST
Patent Text Reader

Abstract

The invention relates to a single-frame astronomical image reconstruction method and device, electronic equipment and a storage medium, and the method comprises the steps: determining a blind deconvolution optimization model corresponding to an initial single-frame astronomical image through employing a blind deconvolution algorithm based on the total variation prior of a reweighted graph; re-sampling the initial single-frame astronomical image, and determining n skeleton images corresponding to the initial single-frame astronomical image; performing iterative optimization solution and light intensity correction processing according to the blind deconvolution optimization model and the n skeleton images, and determining a target blurring kernel corresponding to the initial single-frame astronomical image; and performing deblurring processing on the initial single-frame astronomical image according to the target blurring kernel, and determining a deblurred single-frame astronomical image. According to the method, the blind deconvolution algorithm based on RGTV prior can be utilized to perform iterative optimization solution and light intensity correction processing on the initial single-frame astronomical image to obtain the deblurred single-frame astronomical image with relatively high quality, required computing resources and time are relatively few, the image reconstruction cost can be reduced, and the processing efficiency can be improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates to the technical field of image processing, and in particular to a method and device for reconstructing a single-frame astronomical image, as well as electronic equipment and storage medium. Background Art

[0002] Ground-based optical telescopes are commonly used observation tools in the field of astronomy. However, when using ground-based optical telescopes to obtain astronomical images, they will be affected by factors such as atmospheric turbulence, resulting in a decrease in the image quality of astronomical images, image distortion, blurring, and astrometric jitter. Therefore, it is necessary to reconstruct the blurred astronomical images obtained using ground-based optical telescopes in order to obtain high-definition, high-quality astronomical images. Summary of the invention

[0003] In view of this, the present disclosure proposes a technical solution of a single-frame astronomical image reconstruction method, device, electronic device, and storage medium.

[0004] According to one aspect of the present disclosure, a method for reconstructing a single-frame astronomical image is provided, comprising: determining a blind deconvolution optimization model corresponding to an initial single-frame astronomical image by using a blind deconvolution algorithm based on a reweighted graph total variation prior; resampling the initial single-frame astronomical image to determine n skeleton images corresponding to the initial single-frame astronomical image, wherein n is a positive integer greater than or equal to 1; performing iterative optimization solution and light intensity correction processing according to the blind deconvolution optimization model and the n skeleton images to determine a target blur kernel corresponding to the initial single-frame astronomical image; and deblurring the initial single-frame astronomical image according to the target blur kernel to determine a deblurred single-frame astronomical image.

[0005] In a possible implementation, the iterative optimization solution and light intensity correction processing are performed according to the blind deconvolution optimization model and the n skeleton images to determine the target blur kernel corresponding to the initial single-frame astronomical image, including: for the i-th skeleton image, according to the initial blur kernel corresponding to the skeleton image and the blind deconvolution optimization model, iterative optimization solution and light intensity correction processing are performed until the convergence condition is met, and the convergent blur kernel of the skeleton image is determined, wherein i is a positive integer greater than or equal to 1 and less than or equal to n; when i is less than n, according to the convergent blur kernel corresponding to the i-th skeleton image, the initial blur kernel corresponding to the i+1-th skeleton image is determined, and the above steps are repeated according to the initial blur kernel corresponding to the i+1-th skeleton image; when i is equal to n, according to the convergent blur kernel corresponding to the i-th skeleton image, the target blur kernel is determined.

[0006] In a possible implementation, for the i-th skeleton image, according to the initial blur kernel corresponding to the skeleton image and the blind deconvolution optimization model, iterative optimization solution and light intensity correction processing are performed until the convergence condition is met to determine the converged blur kernel of the skeleton image, including: for any iterative process, light intensity correction processing is performed on the input image corresponding to the iterative process to determine the corrected input image corresponding to the iterative process, wherein the input image corresponding to the iterative process is the i-th skeleton image, or the output image corresponding to the previous iterative process; for any iterative process, according to the input blur kernel corresponding to the iterative process and the corrected input image corresponding to the iterative process, optimizing and solving the blind deconvolution optimization model to determine the output image and output blur kernel corresponding to the iterative process, wherein the input blur kernel corresponding to the iterative process is the initial blur kernel corresponding to the i-th skeleton image, or the output blur kernel corresponding to the previous iterative process; for any iterative process, when the iterative process meets the convergence condition, the convergent blur kernel corresponding to the i-th skeleton image is determined according to the output blur kernel corresponding to the iterative process.

[0007] In a possible implementation, for any iterative process, performing light intensity correction processing on the input image corresponding to the iterative process to determine the corrected input image corresponding to the iterative process includes: for any iterative process, determining the unified blur model corresponding to the iterative process based on the input image and input blur kernel corresponding to the iterative process, wherein the unified blur model is used to adjust the brightness value of each pixel in the input image corresponding to the iterative process; performing light intensity correction processing on the input image corresponding to the iterative process based on the unified blur model to determine the corrected input image corresponding to the iterative process.

[0008] In a possible implementation, the deblurring the initial single-frame astronomical image according to the target blur kernel to determine the deblurred single-frame astronomical image includes: determining a target optimization function corresponding to the initial single-frame astronomical image according to a non-blind deconvolution algorithm; and optimizing and solving the target optimization function according to the target blur kernel to determine the deblurred single-frame astronomical image.

[0009] In a possible implementation, the method further includes: determining a global atmospheric light value and an atmospheric transmittance corresponding to the deblurred single-frame astronomical image; performing image artifact removal and enhancement processing on the deblurred single-frame astronomical image according to the global atmospheric light value and the atmospheric transmittance, and determining an enhanced image corresponding to the deblurred single-frame astronomical image.

[0010] In a possible implementation, determining the global atmospheric light value and atmospheric transmittance corresponding to the deblurred single-frame astronomical image includes: determining the global atmospheric light value based on a dark channel image corresponding to the deblurred single-frame astronomical image; and determining the atmospheric transmittance based on a dark channel prior algorithm and the global atmospheric light value.

[0011] According to another aspect of the present disclosure, a single-frame astronomical image reconstruction device is provided, including: an optimization model construction module, used to determine a blind deconvolution optimization model corresponding to an initial single-frame astronomical image by using a blind deconvolution algorithm based on a reweighted graph total variation prior; a resampling module, used to resample the initial single-frame astronomical image, and determine n skeleton images corresponding to the initial single-frame astronomical image, wherein n is a positive integer greater than or equal to 1; a blur kernel determination module, used to perform iterative optimization solution and light intensity correction processing based on the blind deconvolution optimization model and the n skeleton images, and determine a target blur kernel corresponding to the initial single-frame astronomical image; and a deblurring module, used to deblur the initial single-frame astronomical image based on the target blur kernel, and determine a deblurred single-frame astronomical image.

[0012] According to another aspect of the present disclosure, an electronic device is provided, comprising: a processor; and a memory for storing processor-executable instructions; wherein the processor is configured to implement the above method when executing the instructions stored in the memory.

[0013] According to another aspect of the present disclosure, a non-volatile computer-readable storage medium is provided, on which computer program instructions are stored, wherein the computer program instructions implement the above method when executed by a processor.

[0014] In the disclosed embodiment, a blind deconvolution algorithm based on a reweighted graph total variation prior is used to determine a blind deconvolution optimization model corresponding to an initial single-frame astronomical image. The initial single-frame astronomical image is resampled to determine n skeleton images corresponding to the initial single-frame astronomical image, so that the skeleton images of different resolution scales containing edge information of saturated regions can be used to achieve alternating iterative optimization at different scales, thereby improving the accuracy and reliability of a target blur kernel, wherein n is a positive integer greater than or equal to 1. According to the blind deconvolution optimization model and the n skeleton images, iterative optimization and light intensity correction processing can be performed to determine a target blur kernel corresponding to the initial single-frame astronomical image, and light intensity correction processing can be used to improve the reliability of edge information of saturated regions in skeleton images, and reduce the influence of overexposure of astronomical imaging on image reconstruction. According to the target blur kernel, the initial single-frame astronomical image is deblurred to determine a deblurred single-frame astronomical image with higher clarity and image quality, without the need for additional data support, and the image processing requires less computing resources and time, which can reduce the cost of single-frame astronomical image reconstruction and improve processing efficiency.

[0015] Further features and aspects of the present disclosure will become apparent from the following detailed description of exemplary embodiments with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate exemplary embodiments, features, and aspects of the disclosure and, together with the description, serve to explain the principles of the disclosure.

[0017] Figure 1 A schematic diagram showing a process of forming a blurred astronomical image according to an embodiment of the present disclosure;

[0018] Figure 2 A flowchart showing a method for reconstructing a single-frame astronomical image according to an embodiment of the present disclosure is shown;

[0019] Figure 3 A schematic diagram showing an initial single-frame astronomical image according to an embodiment of the present disclosure;

[0020] Figure 4 A schematic diagram showing a skeleton image according to an embodiment of the present disclosure;

[0021] Figure 5 A schematic diagram showing a blur kernel according to an embodiment of the present disclosure;

[0022] Figure 6 A schematic diagram showing a deblurred single-frame astronomical image according to an embodiment of the present disclosure;

[0023] Figure 7 A block diagram showing a single-frame astronomical image reconstruction device according to an embodiment of the present disclosure;

[0024] Figure 8 A block diagram of an electronic device according to an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0025] Various exemplary embodiments, features and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0026] The word “exemplary” is used exclusively herein to mean “serving as an example, example, or illustration.” Any embodiment described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments.

[0027] The term "and / or" herein is only a description of the association relationship of the associated objects, indicating that there may be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the term "at least one" herein represents any combination of at least two of any one or more of a plurality of. For example, including at least one of A, B, and C can represent including any one or more elements selected from the set consisting of A, B, and C.

[0028] In addition, in order to better illustrate the present disclosure, numerous specific details are given in the following specific embodiments. It should be understood by those skilled in the art that the present disclosure can also be implemented without certain specific details. In some examples, methods, means, components and circuits well known to those skilled in the art are not described in detail in order to highlight the main purpose of the present disclosure.

[0029] Ground-based optical telescopes are commonly used observation tools in the field of astronomy. However, due to the uneven density and temperature of the Earth's atmosphere, atmospheric turbulence is formed, which leads to uneven distribution of atmospheric refractive index, affecting the image quality of astronomical images obtained by ground-based optical telescopes.

[0030] Figure 1 A schematic diagram showing a process of forming a blurred astronomical image according to an embodiment of the present disclosure is shown. Figure 1 As shown, when observing a target object based on a ground-based optical telescope, under the influence of atmospheric turbulence, only a blurred astronomical image corresponding to the target object can be obtained.

[0031] Specifically, when the light corresponding to the target object passes through the atmosphere, it will be refracted multiple times due to the unevenness of the atmospheric refractive index, causing wavefront distortion, resulting in blurring and distortion of astronomical images acquired using ground-based optical telescopes. The spatial scattering caused by atmospheric turbulence will affect the contrast and resolution of astronomical images, resulting in spots in astronomical images; the instability of atmospheric turbulence will cause the intensity of the light corresponding to the target object to fluctuate, causing flickering in astronomical images. In addition, the aberrations of the ground-based optical telescope itself (for example, spherical aberration, coma, and astigmatism, etc.) will also affect the clarity and resolution of astronomical images, causing deformation and distortion of astronomical images.

[0032] The combined effect of the above factors leads to poor image quality of astronomical images, which can be called image degradation. Figure 1 For example, Figure 1 As shown in Figure 1, the degradation process can be expressed as adding the blur kernel and measurement noise to the target object to obtain the corresponding blurred astronomical image. The degradation process can be expressed as formula (1):

[0033]

[0034] Where y represents the blurred astronomical image obtained based on the ground-based optical telescope; x represents the clear astronomical image corresponding to the target object; k represents the blur kernel; n represents the measurement noise; Represents the convolution operator.

[0035] In the prior art, in order to obtain a clear image corresponding to a blurred image, a post-processing algorithm is usually used to estimate the blur kernel and clear image corresponding to the blurred image using partial information of the imaging system. For blurred astronomical images, by considering the influence of factors such as atmospheric turbulence as part of the information of the imaging system, a blind deconvolution algorithm can be used to estimate the blur kernel corresponding to the blurred astronomical image, thereby reconstructing the blurred astronomical image. However, the blind deconvolution algorithm commonly used in the prior art usually relies on multiple frames of blurred images for blur kernel estimation and image reconstruction, which requires a large amount of data support, has a high computational cost, and consumes a long image processing time. In actual application, especially in the reconstruction scenario of astronomical images, there are problems such as difficulty in data acquisition and limited computing resources, which leads to poor stability of the estimated blur kernel and is sensitive to similarity between images.

[0036] In view of this, the present disclosure provides a single-frame astronomical image reconstruction method, which can utilize a blind deconvolution algorithm based on a reweighted graph total variation (RGTV) prior to reconstruct and restore details of a blurred initial single-frame astronomical image, and perform light intensity correction processing based on the characteristics of the astronomical image to obtain a deblurred single-frame astronomical image with high clarity and image quality, without the need for additional data support, and with less computing resources and time required for image processing, which can reduce the cost of single-frame astronomical image reconstruction and improve processing efficiency. The single-frame astronomical image reconstruction method provided by the present disclosure is described in detail below.

[0037] Figure 2 A flow chart of a method for reconstructing a single-frame astronomical image according to an embodiment of the present disclosure is shown. The method for reconstructing a single-frame astronomical image can be executed by an electronic device such as a terminal device or a server. The terminal device can be a user equipment (UE), a mobile device, a user terminal, a terminal, a cellular phone, a cordless phone, a personal digital assistant (PDA), a handheld device, a computing device, a vehicle-mounted device, a wearable device, etc. The method for reconstructing a single-frame astronomical image can be implemented by a processor calling computer-readable instructions stored in a memory. Alternatively, the method for reconstructing a single-frame astronomical image can be executed by a server. Figure 2 As shown, the reconstruction method of the single-frame astronomical image includes:

[0038] In step S201, a blind deconvolution algorithm based on reweighted graph total variation prior is used to determine a blind deconvolution optimization model corresponding to an initial single-frame astronomical image.

[0039] The initial single-frame astronomical image here may represent any frame of a blurred astronomical image that needs to be reconstructed to improve clarity. The specific method of obtaining the initial single-frame astronomical image may be flexibly set according to actual usage requirements, and the present disclosure does not specifically limit this.

[0040] Figure 3 FIG. 1 is a schematic diagram showing an initial single-frame astronomical image according to an embodiment of the present disclosure. Figure 3 As shown, the initial single-frame astronomical image is an image with low definition and unable to observe the details of the astronomical object.

[0041] RGTV defines a variational regularization term on the image structure and introduces a weight adjustment mechanism to maintain the piecewise smooth form of the image, thereby effectively promoting the restoration of sharp edges in the image. Specifically, the RGTV regularization term can be expressed as formulas (2) to (3):

[0042]

[0043]

[0044] Among them, ||x|| RGTV represents the RGTV regularization term; x represents the clear image to be obtained; V represents the pixel set in the blurred image; i represents any pixel in the blurred image; W i,. (x) represents the i-th row of the weight matrix W(x) corresponding to x; represents the gradient vector corresponding to pixel i; N represents the number of pixels in the blurred image; ω i,j (x i , x j ) represents the initial edge weight; x i represents the brightness value corresponding to pixel i; x j represents the brightness value corresponding to pixel j; d = |x j -x i |, used to dynamically adjust edge weights; σ represents the scale parameter used to adjust edge weights.

[0045] The RGTV regularization term can make only two intensity distribution states exist between different pixels in the image, forming a bimodal weight distribution. Among them, one intensity distribution state is that the intensity distribution between different pixels presents a small gap to maintain the smoothness of the corresponding area in the image; the other intensity distribution state is that the intensity distribution between different pixels presents a large difference to highlight the edges of different areas in the image. Therefore, by minimizing the function containing the RGTV regularization term, the sharpness of the edges of different areas in the image can be improved, and the blurred area can be suppressed, thereby achieving deblurring of the blurred image.

[0046] Based on the above principles, a blind deconvolution algorithm based on RGTV prior can be constructed. The blind deconvolution algorithm based on RGTV prior can be expressed as formula (4):

[0047]

[0048] in, represents the image obtained after deconvolution; represents the blur kernel; x represents the ideal clear image corresponding to the astronomical object; y represents the blurred image; Represents the data fidelity term, which is used to ensure the image obtained after deconvolution Similar to the blurred image y; β represents the regularization parameter; Blur kernel The corresponding regularization term can use the l2 norm to maintain the blur kernel The smoothness of ; μ represents the regularization parameter.

[0049] According to the blind deconvolution algorithm based on RGTV prior, the blind deconvolution optimization model corresponding to the initial single-frame astronomical image can be determined to determine the target blur kernel corresponding to the initial single-frame astronomical image. The specific form of the blind deconvolution optimization model can be flexibly set according to actual usage requirements, and the present disclosure does not make specific limitations on this.

[0050] In one example, the blind deconvolution optimization model can be expressed as formulas (5) to (6):

[0051]

[0052]

[0053] in, Represents the gradient vector corresponding to the image obtained after deconvolution.

[0054] In step S202 , the initial single-frame astronomical image is resampled to determine n skeleton images corresponding to the initial single-frame astronomical image, where n is a positive integer greater than or equal to 1.

[0055] By resampling the initial single-frame astronomical image, multiple skeleton images of different resolutions corresponding to the initial single-frame astronomical image can be obtained, which can be used for subsequent optimization solutions in alternating iterations at multiple scales in combination with a blind deconvolution optimization model. Among them, the specific content of resampling can be flexibly set according to actual usage requirements, for example, it can include upsampling and downsampling, etc., and the present disclosure does not make specific restrictions on this; the specific method of implementing resampling can refer to the implementation methods in the relevant technology, such as bilinear interpolation algorithms, etc., and the present disclosure does not make specific restrictions on this; the specific value of n can be flexibly set according to actual usage requirements, and the present disclosure does not make specific restrictions on this; the specific form of the skeleton image can be flexibly set according to actual usage requirements, and the present disclosure does not make specific restrictions on this.

[0056] Figure 4 FIG. 1 is a schematic diagram showing a skeleton image according to an embodiment of the present disclosure. Figure 4 As shown, Figure 4 For Figure 3 The skeleton image corresponding to the initial single-frame astronomical image shown in FIG. Figure 3 The initial single-frame astronomical image shown has a higher resolution but less image detail information.

[0057] In a possible implementation, n skeleton images may form an image pyramid according to the order of resolution from small to large, so that in subsequent processing, the skeleton images are called according to the order of resolution to achieve orderly and step-by-step improvement of the accuracy of the target blur kernel.

[0058] In step S203, an iterative optimization solution and light intensity correction process are performed based on the blind deconvolution optimization model and the n skeleton images to determine the target blur kernel corresponding to the initial single-frame astronomical image.

[0059] According to the n skeleton images and the blind deconvolution optimization model, iterative optimization solutions can be alternately performed at different resolution scales to improve the accuracy of the blur kernel and the intermediate image obtained in the iterative process, so that the target blur kernel finally determined has high accuracy and reliability. The specific method of optimization solution can refer to the implementation method in the relevant technology, and this disclosure does not specifically limit this.

[0060] Furthermore, in the field of astronomical observation, astronomical images are usually acquired through long exposure, so there may be overexposed pixels in the initial single-frame astronomical image. The image areas corresponding to these pixels lack necessary detail features, resulting in errors in the edge information of the saturated areas in the image, which will affect the process of solving the target blur kernel.

[0061] In view of this characteristic of astronomical images, in order to make full use of the information of the non-saturated and saturated areas in the initial single-frame astronomical image, reduce the error influence of the edge information of the saturated area, and improve the accuracy and reliability of the target blur kernel, the embodiment of the present disclosure can synchronously perform light intensity correction processing on the skeleton image and the intermediate image obtained in each iterative process during the iterative optimization solution for any skeleton image, so as to adjust the brightness value of the skeleton image and the intermediate image obtained in each iterative process. Among them, the specific method of light intensity correction processing can refer to the implementation method in the relevant technology, and the present disclosure does not make specific limitations on this.

[0062] The specific form of the target blur kernel can be flexibly set according to actual usage requirements. For example, the target blur kernel can be expressed as a point spread function (PSF), etc., and the present disclosure does not make specific limitations on this.

[0063] Figure 5 FIG. 4 is a schematic diagram showing a blur kernel according to an embodiment of the present disclosure. Figure 5 As shown, Figure 5 For Figure 3 The blur kernel corresponding to the initial single-frame astronomical image is shown.

[0064] The following text will describe in detail the process of iteratively optimizing and solving and correcting the light intensity based on the blind deconvolution optimization model and n skeleton images to determine the target blur kernel corresponding to the initial single-frame astronomical image in combination with possible implementation methods of the present invention, which will not be elaborated here.

[0065] In step S204, the initial single-frame astronomical image is deblurred according to the target blur kernel to determine a deblurred single-frame astronomical image.

[0066] After determining the target blur kernel corresponding to the initial single-frame astronomical image, the target blur kernel can be used to perform deblurring on the initial single-frame astronomical image to obtain a deblurred single-frame astronomical image with higher definition. The specific method of deblurring can refer to the implementation method in the related art, and the present disclosure does not specifically limit this.

[0067] Figure 6 FIG. 2 is a schematic diagram showing a deblurred single-frame astronomical image according to an embodiment of the present disclosure. Figure 6 As shown, Figure 6 For Figure 3 The deblurred single-frame astronomical image corresponding to the initial single-frame astronomical image shown in FIG. Figure 6 The deblurred single-frame astronomical image shown is compared to Figure 3 The initial single-frame astronomical image shown has a high resolution and is able to capture detailed information of astronomical objects.

[0068] In the disclosed embodiment, a blind deconvolution algorithm based on a reweighted graph total variation prior is used to determine a blind deconvolution optimization model corresponding to an initial single-frame astronomical image. The initial single-frame astronomical image is resampled to determine n skeleton images corresponding to the initial single-frame astronomical image, so that the skeleton images of different resolution scales containing edge information of saturated regions can be used to achieve alternating iterative optimization at different scales, thereby improving the accuracy and reliability of a target blur kernel, wherein n is a positive integer greater than or equal to 1. According to the blind deconvolution optimization model and the n skeleton images, iterative optimization and light intensity correction processing can be performed to determine a target blur kernel corresponding to the initial single-frame astronomical image, and light intensity correction processing can be used to improve the reliability of edge information of saturated regions in skeleton images, and reduce the influence of overexposure of astronomical imaging on image reconstruction. According to the target blur kernel, the initial single-frame astronomical image is deblurred to determine a deblurred single-frame astronomical image with higher clarity and image quality, without the need for additional data support, and the image processing requires less computing resources and time, which can reduce the cost of single-frame astronomical image reconstruction and improve processing efficiency.

[0069] In a possible implementation, an iterative optimization solution and light intensity correction process are performed based on a blind deconvolution optimization model and n skeleton images to determine a target blur kernel corresponding to an initial single-frame astronomical image, including: for an i-th skeleton image, an iterative optimization solution and light intensity correction process are performed based on an initial blur kernel corresponding to the skeleton image and a blind deconvolution optimization model until a convergence condition is met, and a converged blur kernel of the skeleton image is determined, wherein i is a positive integer greater than or equal to 1 and less than or equal to n; when i is less than n, an initial blur kernel corresponding to an i+1-th skeleton image is determined based on a converged blur kernel corresponding to the i-th skeleton image, and the above steps are repeated based on the initial blur kernel corresponding to the i+1-th skeleton image; when i is equal to n, a target blur kernel is determined based on the converged blur kernel corresponding to the i-th skeleton image.

[0070] Specifically, for the i-th skeleton image, when i=1, it can be determined that the skeleton image is the first skeleton image used for iterative solution. At this time, the initial blur kernel corresponding to the skeleton image can be flexibly set according to actual usage requirements, and the present disclosure does not make specific restrictions on this. When i is greater than or equal to 1 and less than or equal to n, the convergence blur kernel corresponding to the last skeleton image that completed the iterative process can be determined as the initial blur kernel corresponding to the skeleton image, thereby achieving a resolution scale change based on the skeleton image and gradually improving the accuracy of the convergence blur kernel.

[0071] For the i-th skeleton image, after the initial blur kernel corresponding to the skeleton image is determined by the above-mentioned process, iterative optimization and light intensity correction processing can be performed according to the initial blur kernel corresponding to the skeleton image and the blind deconvolution optimization model until the convergence condition is met, thereby determining the converged blur kernel of the skeleton image. The specific form of the convergence condition can be flexibly set according to actual usage requirements, for example, it can be set to meet a preset number of iterations, etc., and the present disclosure does not make specific limitations on this.

[0072] When i is less than n, after completing the iterative process for the i-th skeleton image, the convergent blur kernel corresponding to the i-th skeleton image can be determined as the initial blur kernel corresponding to the i+1-th skeleton image, and the above steps are repeated until the iterative process for all skeleton images is completed. When i is equal to n, the convergent blur kernel corresponding to the i-th skeleton image can be determined as the target blur kernel.

[0073] In a possible implementation, for the i-th skeleton image, according to the initial blur kernel and the blind deconvolution optimization model corresponding to the skeleton image, iterative optimization solution and light intensity correction processing are performed until the convergence condition is met to determine the converged blur kernel of the skeleton image, including: for any iterative process, light intensity correction processing is performed on the input image corresponding to the iterative process to determine the corrected input image corresponding to the iterative process, wherein the input image corresponding to the iterative process is the i-th skeleton image, or the output image corresponding to the previous iterative process; for any iterative process, according to the input blur kernel corresponding to the iterative process and the corrected input image corresponding to the iterative process, the blind deconvolution optimization model is optimized and solved to determine the output image and output blur kernel corresponding to the iterative process, wherein the input blur kernel corresponding to the iterative process is the initial blur kernel corresponding to the i-th skeleton image, or the output blur kernel corresponding to the previous iterative process; for any iterative process, when the iterative process meets the convergence condition, the convergent blur kernel corresponding to the i-th skeleton image is determined according to the output blur kernel corresponding to the iterative process.

[0074] Specifically, for any iterative process of the i-th skeleton image, if the iterative process is the first iterative process corresponding to the skeleton image, the skeleton image may be subjected to light intensity correction processing to determine the corrected input image corresponding to the first iterative process. If the iterative process is not the first iterative process corresponding to the skeleton image, the input image corresponding to the iterative process is the output image obtained by the previous iterative process. At this time, the output image corresponding to the previous iterative process may be subjected to light intensity correction processing to determine the corrected input image corresponding to the iterative process.

[0075] For any iterative process, after completing the light intensity correction process, the input blur kernel corresponding to the iterative process and the corrected input image corresponding to the iterative process can be input into the blind deconvolution optimization model for optimization and solution to obtain the output image and output blur kernel corresponding to the iterative process. If the iterative process does not meet the convergence conditions, the output image and output blur kernel corresponding to the iterative process can be determined as the input image and input blur kernel corresponding to the next iterative process, and the above steps are repeated. If the iterative process meets the convergence conditions, the output blur kernel corresponding to the iterative process can be used to determine the converged blur kernel corresponding to the i-th skeleton image, and the iterative process for the i-th skeleton image is terminated.

[0076] In one possible implementation, for any iterative process, a light intensity correction process is performed on the input image corresponding to the iterative process to determine the corrected input image corresponding to the iterative process, including: for any iterative process, based on the input image and input blur kernel corresponding to the iterative process, determining a unified blur model corresponding to the iterative process, wherein the unified blur model is used to adjust the brightness value of each pixel in the input image corresponding to the iterative process; based on the unified blur model, a light intensity correction process is performed on the input image corresponding to the iterative process to determine the corrected input image corresponding to the iterative process.

[0077] Specifically, for any iterative process, the unified fuzzy model corresponding to the iterative process can be determined according to the input image and input fuzzy kernel corresponding to the iterative process, so as to adjust the brightness value of each pixel in the input image corresponding to the iterative process, so that the brightness values ​​of all pixels in the input image corresponding to the iterative process are within the preset range. The specific form of the unified fuzzy model can refer to the implementation in the relevant technology, and the present disclosure does not make specific restrictions on this; the specific value of the preset range can be flexibly set according to the actual use requirements, and the present disclosure does not make specific restrictions on this.

[0078] In one example, the unified fuzzy model can be expressed as formulas (7) to (8):

[0079]

[0080]

[0081] Where y represents the corrected input image; m represents the light intensity correction operation; represents the Hadamard product; x represents the input image, and k represents the input blur kernel.

[0082] By using the unified blur model, the input image corresponding to this iterative process can be subjected to light intensity correction processing to determine the corrected input image corresponding to this iterative process, thereby making full use of the information of the non-saturated and saturated areas in the initial single-frame astronomical image and reducing the error influence of the edge information in the saturated area, so as to improve the accuracy and reliability of the converged blur kernel.

[0083] In a possible implementation, an initial single-frame astronomical image is deblurred according to a target blur kernel to determine a deblurred single-frame astronomical image, including: determining a target optimization function corresponding to the initial single-frame astronomical image according to a non-blind deconvolution algorithm; and optimizing and solving the target optimization function according to the target blur kernel to determine the deblurred single-frame astronomical image.

[0084] After determining the target blur kernel, the initial single-frame astronomical image can be deblurred using a non-blind deconvolution algorithm to determine a deblurred single-frame astronomical image. The specific form of the non-blind deconvolution algorithm here can refer to the implementation in the relevant technology, for example, a non-blind deconvolution algorithm based on a super-Laplacian prior, a Gaussian deblurring algorithm, etc., which is not specifically limited in the present disclosure.

[0085] In one example, a non-blind deconvolution algorithm based on a super-Laplacian prior may be used to deblur an initial single-frame astronomical image.

[0086] The super Laplace prior combines the smoothness of the Gaussian distribution and the sparseness of the Laplace distribution, models the heavy-tailed characteristics of the gradient in natural scenes, and can facilitate the restoration of smooth regions and edge information of blurred images. The super Laplace prior distribution can be expressed in formula (9):

[0087]

[0088] Wherein, x represents a blurred image; k represents a parameter describing the image scale corresponding to the blurred image; α represents a preset parameter for processing the gradient distribution corresponding to the blurred image, and its specific value can be flexibly set according to actual usage requirements. For example, it can be set to 0.5≤α≤0.8, etc., and the present disclosure does not make specific limitations on this.

[0089] On this basis, the target optimization function can be determined according to the non-blind deconvolution algorithm based on super-Laplacian prior. The target optimization function can be expressed as formula (10):

[0090]

[0091] in, represents the deblurred single-frame astronomical image; x represents the ideal clear image corresponding to the astronomical object; k represents the target blur kernel; y represents the initial single-frame astronomical image; β represents the regularization parameter.

[0092] By optimizing and solving the target optimization function and minimizing the target optimization function, a deblurred single-frame astronomical image can be obtained. The specific method of optimizing and solving the target optimization function can refer to the implementation methods in the relevant technology, for example, it can include a semi-quadratic penalty method and an alternating minimization method, etc., which is not specifically limited in the present disclosure.

[0093] In a possible implementation, the method also includes: determining a global atmospheric light value and an atmospheric transmittance corresponding to the deblurred single-frame astronomical image; performing image artifact removal and enhancement processing on the deblurred single-frame astronomical image according to the global atmospheric light value and the atmospheric transmittance, and determining an enhanced image corresponding to the deblurred single-frame astronomical image.

[0094] The deblurred single-frame astronomical image obtained based on the above process may contain artifacts and other factors that affect image clarity and image quality. Therefore, in order to further improve the quality of the astronomical image, the deblurred single-frame astronomical image may be subjected to image artifact removal and enhancement processing.

[0095] Specifically, according to the characteristics of astronomical imaging, the artifacts in the deblurred single-frame astronomical image can be regarded as the "fog" caused by the influence of atmospheric scattering. The imaging equation of the atmospheric scattering model can be expressed as formula (11):

[0096] I(k)=x(k)t(k)+A(1-t(k)) (11)

[0097] Among them, I(k) represents the foggy astronomical image; x(k) represents the fog-free astronomical image; t(k) represents the atmospheric transmittance; A represents the global atmospheric light; k represents any pixel in the image.

[0098] As can be seen from the above, by determining the global atmospheric light value and atmospheric transmittance corresponding to the deblurred single-frame astronomical image, the global atmospheric light value and atmospheric transmittance can be used to perform image artifact enhancement processing on the deblurred single-frame astronomical image, reduce the artifact influence in the deblurred single-frame astronomical image, and obtain an enhanced image corresponding to the deblurred single-frame astronomical image. Among them, the specific method of determining the global atmospheric light value and atmospheric transmittance corresponding to the deblurred single-frame astronomical image can refer to the implementation method in the relevant technology, and the present disclosure does not make specific limitations on this.

[0099] In a possible implementation, a global atmospheric light value and an atmospheric transmittance corresponding to a deblurred single-frame astronomical image are determined, including: determining the global atmospheric light value based on a dark channel image corresponding to the deblurred single-frame astronomical image; and determining the atmospheric transmittance based on a dark channel prior algorithm and the global atmospheric light value.

[0100] The dark channel prior can be expressed as that in a fog-free astronomical image, for some pixels in the non-sky area, their corresponding brightness values ​​are close to zero in at least one channel. In other words, for a fog-free astronomical image, its corresponding dark channel image can satisfy formula (12):

[0101]

[0102] Among them, x dark (k) represents the dark channel image; represents any one of the R, G, and B channels of the image; Ω(k) represents the neighborhood centered on any pixel k.

[0103] By performing simultaneous transformation and minimization operations on formula (11) and formula (12), we can obtain formula (13):

[0104]

[0105] When the global atmospheric light is known and constant, the atmospheric transmittance can be determined according to formula (13). Specifically, the atmospheric transmittance can be expressed as formula (14):

[0106]

[0107] In actual astronomical observation scenes, even in clear weather, there will be some particles in the atmosphere, which makes the observation of astronomical objects hazy and the depth of the scene. Therefore, when performing image artifact enhancement processing on the deblurred astronomical image, some "fog" can be retained to improve the realism and depth of the obtained enhanced image. In this case, the atmospheric projection rate can be corrected, and the corrected atmospheric projection rate can be expressed as formula (15):

[0108]

[0109] Herein, ω represents a correction coefficient, and its specific value can be flexibly set according to actual usage requirements, and it only needs to satisfy 0≤ω≤1. For example, ω=0.95 can be set, and the present disclosure does not make any specific limitation on this.

[0110] When the global atmospheric light is unknown, the global atmospheric light can be determined by deblurring the dark channel image corresponding to a single-frame astronomical image. Specifically, the pixels whose brightness values ​​in the dark channel image meet the preset requirements can be determined based on the brightness values ​​corresponding to each pixel, and based on these pixels, the values ​​corresponding to the maximum brightness points in the original image are used as the values ​​corresponding to the global atmospheric light. The preset requirements here can be flexibly set according to actual usage requirements. For example, the brightness value can be set to be in the first 0.1% of the brightness values ​​corresponding to all pixels, etc. This disclosure does not make specific limitations on this.

[0111] After determining the global atmospheric light and atmospheric transmittance through the above process, the deblurred single-frame astronomical image can be subjected to image artifact enhancement processing to reduce the artifact influence in the deblurred single-frame astronomical image and obtain an enhanced image corresponding to the deblurred single-frame astronomical image. The enhanced image can be expressed as formula (16):

[0112]

[0113] Here, t′ is a constant used to prevent the denominator from taking the value of 0. Its specific value can be flexibly set according to actual usage requirements. For example, t′=0.05 can be set, and the present disclosure does not make specific limitations on this.

[0114] In the disclosed embodiment, a blind deconvolution algorithm based on a reweighted graph total variation prior is used to determine a blind deconvolution optimization model corresponding to an initial single-frame astronomical image. The initial single-frame astronomical image is resampled to determine n skeleton images corresponding to the initial single-frame astronomical image, so that the skeleton images of different resolution scales containing edge information of saturated regions can be used to achieve alternating iterative optimization at different scales, thereby improving the accuracy and reliability of a target blur kernel, wherein n is a positive integer greater than or equal to 1. According to the blind deconvolution optimization model and the n skeleton images, iterative optimization and light intensity correction processing can be performed to determine a target blur kernel corresponding to the initial single-frame astronomical image, and light intensity correction processing can be used to improve the reliability of edge information of saturated regions in skeleton images, and reduce the influence of overexposure of astronomical imaging on image reconstruction. According to the target blur kernel, the initial single-frame astronomical image is deblurred to determine a deblurred single-frame astronomical image with higher clarity and image quality, without the need for additional data support, and the image processing requires less computing resources and time, which can reduce the cost of single-frame astronomical image reconstruction and improve processing efficiency. Furthermore, by using the dark channel prior algorithm, the deblurred single-frame astronomical image can be subjected to image artifact enhancement processing while ensuring image authenticity and scene depth, thereby reducing the impact of artifacts and obtaining an enhanced image corresponding to the deblurred single-frame astronomical image.

[0115] It can be understood that the above-mentioned various method embodiments mentioned in the present disclosure can be combined with each other to form a combined embodiment without violating the principle logic. Due to space limitations, the present disclosure will not repeat them. It can be understood by those skilled in the art that in the above-mentioned method of the specific implementation method, the specific execution order of each step should be determined according to its function and possible internal logic.

[0116] In addition, the present disclosure also provides a single-frame astronomical image reconstruction device, an electronic device and a computer-readable storage medium, all of which can be used to implement any single-frame astronomical image reconstruction method provided by the present disclosure. The corresponding technical solutions and descriptions are referred to the corresponding records in the method part and will not be repeated here.

[0117] Figure 7 FIG. 1 is a block diagram of a single-frame astronomical image reconstruction device according to an embodiment of the present disclosure. Figure 7 As shown, the apparatus 700 includes:

[0118] The optimization model building module 701 is used to determine the blind deconvolution optimization model corresponding to the initial single-frame astronomical image by using a blind deconvolution algorithm based on a reweighted graph total variation prior;

[0119] A resampling module 702 is used to resample the initial single-frame astronomical image to determine n skeleton images corresponding to the initial single-frame astronomical image, where n is a positive integer greater than or equal to 1;

[0120] The blur kernel determination module 703 is used to perform iterative optimization and light intensity correction processing according to the blind deconvolution optimization model and the n skeleton images to determine the target blur kernel corresponding to the initial single-frame astronomical image;

[0121] The deblurring module 704 is used to perform deblurring processing on the initial single-frame astronomical image according to the target blur kernel to determine a deblurred single-frame astronomical image.

[0122] In one possible implementation, the blur kernel determination module 703 is specifically used to: for the i-th skeleton image, perform iterative optimization solution and light intensity correction processing according to the initial blur kernel and blind deconvolution optimization model corresponding to the skeleton image until the convergence condition is met, and determine the convergent blur kernel of the skeleton image, wherein i is a positive integer greater than or equal to 1 and less than or equal to n; when i is less than n, determine the initial blur kernel corresponding to the i+1-th skeleton image according to the convergent blur kernel corresponding to the i-th skeleton image, and repeat the above steps according to the initial blur kernel corresponding to the i+1-th skeleton image; when i is equal to n, determine the target blur kernel according to the convergent blur kernel corresponding to the i-th skeleton image.

[0123] In one possible implementation, the blur kernel determination module 703 is specifically used to: for any iterative process, perform light intensity correction processing on the input image corresponding to the iterative process, and determine the corrected input image corresponding to the iterative process, wherein the input image corresponding to the iterative process is the i-th skeleton image, or the output image corresponding to the previous iterative process; for any iterative process, optimize and solve the blind deconvolution optimization model according to the input blur kernel corresponding to the iterative process and the corrected input image corresponding to the iterative process, and determine the output image and output blur kernel corresponding to the iterative process, wherein the input blur kernel corresponding to the iterative process is the initial blur kernel corresponding to the i-th skeleton image, or the output blur kernel corresponding to the previous iterative process; for any iterative process, when the iterative process meets the convergence condition, determine the convergence blur kernel corresponding to the i-th skeleton image according to the output blur kernel corresponding to the iterative process.

[0124] In one possible implementation, the blur kernel determination module 703 is specifically used to: for any iterative process, determine the unified blur model corresponding to the iterative process based on the input image and input blur kernel corresponding to the iterative process, wherein the unified blur model is used to adjust the brightness value of each pixel in the input image corresponding to the iterative process; perform light intensity correction processing on the input image corresponding to the iterative process according to the unified blur model to determine the corrected input image corresponding to the iterative process.

[0125] In a possible implementation, the deblurring module 704 is specifically used to: determine the target optimization function corresponding to the initial single-frame astronomical image according to a non-blind deconvolution algorithm; optimize and solve the target optimization function according to the target blur kernel to determine the deblurred single-frame astronomical image.

[0126] In one possible implementation, the device 700 also includes an image enhancement module, which is used to: determine the global atmospheric light value and atmospheric transmittance corresponding to the deblurred single-frame astronomical image; perform image artifact removal and enhancement processing on the deblurred single-frame astronomical image according to the global atmospheric light value and the atmospheric transmittance, and determine an enhanced image corresponding to the deblurred single-frame astronomical image.

[0127] In a possible implementation, the image enhancement module is specifically used to: determine the global atmospheric light value based on the dark channel image corresponding to the deblurred single-frame astronomical image; and determine the atmospheric transmittance based on the dark channel prior algorithm and the global atmospheric light value.

[0128] In some embodiments, the functions or modules included in the device provided by the embodiments of the present disclosure can be used to execute the method described in the above method embodiments. The specific implementation can refer to the description of the above method embodiments, and for the sake of brevity, it will not be repeated here.

[0129] The embodiment of the present disclosure also provides a computer-readable storage medium on which computer program instructions are stored, and the computer program instructions implement the above method when executed by a processor. The computer-readable storage medium can be a volatile or non-volatile computer-readable storage medium.

[0130] An embodiment of the present disclosure further proposes an electronic device, comprising: a processor; and a memory for storing instructions executable by the processor; wherein the processor is configured to implement the above method when executing the instructions stored in the memory.

[0131] Figure 8 1900 is a block diagram of an electronic device according to an embodiment of the present disclosure. For example, the apparatus 1900 may be provided as a server or a terminal device. Figure 8 , the apparatus 1900 includes a processing component 1922, which further includes one or more processors, and a memory resource represented by a memory 1932 for storing instructions, such as an application, that can be executed by the processing component 1922. The application stored in the memory 1932 may include one or more modules, each corresponding to a set of instructions. In addition, the processing component 1922 is configured to execute instructions to perform the above method.

[0132] The device 1900 may also include a power supply component 1926 configured to perform power management of the device 1900, a wired or wireless network interface 1950 configured to connect the device 1900 to a network, and an input / output interface 1958 (I / O interface). The device 1900 may operate based on an operating system stored in the memory 1932, such as Windows Server 2000. TM , MacOS X TM , Unix TM ,Linux TM , FreeBSD TM or similar.

[0133] In an exemplary embodiment, a non-volatile computer-readable storage medium is also provided, such as a memory 1932 including computer program instructions, which can be executed by the processing component 1922 of the device 1900 to perform the above method.

[0134] The present disclosure may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0135] A computer-readable storage medium may be a tangible device that can hold and store instructions used by an instruction execution device. A computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples of computer-readable storage media (a non-exhaustive list) include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, such as a punch card or a raised structure in a groove on which instructions are stored, and any suitable combination of the foregoing. As used herein, a computer-readable storage medium is not to be interpreted as a transient signal per se, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., a light pulse through a fiber optic cable), or an electrical signal transmitted through a wire.

[0136] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing / processing device.

[0137] The computer program instructions for performing the operation of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages, such as Smalltalk, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. Computer-readable program instructions may be executed completely on a user's computer, partially on a user's computer, as an independent software package, partially on a user's computer, partially on a remote computer, or completely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect via the Internet). In some embodiments, an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), may be customized by utilizing the state information of the computer-readable program instructions, and the electronic circuit may execute the computer-readable program instructions, thereby realizing various aspects of the present disclosure.

[0138] Various aspects of the present disclosure are described herein with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present disclosure. It should be understood that each box in the flowchart and / or block diagram and the combination of each box in the flowchart and / or block diagram can be implemented by computer-readable program instructions.

[0139] These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, thereby producing a machine, so that when these instructions are executed by the processor of the computer or other programmable data processing device, a device that implements the functions / actions specified in one or more boxes in the flowchart and / or block diagram is generated. These computer-readable program instructions can also be stored in a computer-readable storage medium, and these instructions cause the computer, programmable data processing device, and / or other equipment to work in a specific manner, so that the computer-readable medium storing the instructions includes a manufactured product, which includes instructions for implementing various aspects of the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0140] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device so that a series of operating steps are performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to implement the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0141] The flow chart and block diagram in the accompanying drawings show the possible architecture, function and operation of the system, method and computer program product according to multiple embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a part of a module, program segment or instruction, and a part of the module, program segment or instruction includes one or more executable instructions for realizing the specified logical function. In some alternative implementations, the function marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two continuous square boxes can actually be executed substantially in parallel, and they can sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs the specified function or action, or can be implemented with a combination of special hardware and computer instructions.

[0142] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or technical improvements in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A method for reconstructing a single-frame astronomical image, characterized in that: include: Using the blind deconvolution algorithm based on reweighted graph total variation prior, the blind deconvolution optimization model corresponding to the initial single-frame astronomical image is determined; Resampling the initial single-frame astronomical image to determine n skeleton images corresponding to the initial single-frame astronomical image, where n is a positive integer greater than or equal to 1; According to the blind deconvolution optimization model and the n skeleton images, iterative optimization solution and light intensity correction processing are performed to determine the target blur kernel corresponding to the initial single-frame astronomical image; The initial single-frame astronomical image is deblurred according to the target blur kernel to determine a deblurred single-frame astronomical image.

2. The method according to claim 1, characterized in that The step of performing iterative optimization and light intensity correction processing according to the blind deconvolution optimization model and the n skeleton images to determine the target blur kernel corresponding to the initial single-frame astronomical image includes: For the i-th skeleton image, according to the initial blur kernel corresponding to the skeleton image and the blind deconvolution optimization model, iterative optimization solution and light intensity correction processing are performed until the convergence condition is met, and the converged blur kernel of the skeleton image is determined, where i is a positive integer greater than or equal to 1 and less than or equal to n; When i is less than n, determine the initial blur kernel corresponding to the i+1th skeleton image according to the convergent blur kernel corresponding to the i-th skeleton image, and repeat the above steps according to the initial blur kernel corresponding to the i+1th skeleton image; When i is equal to n, the target blur kernel is determined according to the convergent blur kernel corresponding to the i-th skeleton image.

3. The method according to claim 2, characterized in that The method of performing iterative optimization and light intensity correction processing on the i-th skeleton image according to the initial blur kernel corresponding to the skeleton image and the blind deconvolution optimization model until a convergence condition is met to determine a converged blur kernel of the skeleton image includes: For any iterative process, performing light intensity correction processing on the input image corresponding to the iterative process to determine the corrected input image corresponding to the iterative process, wherein the input image corresponding to the iterative process is the i-th skeleton image, or the output image corresponding to the previous iterative process; For any iterative process, according to the input blur kernel corresponding to the iterative process and the corrected input image corresponding to the iterative process, the blind deconvolution optimization model is optimized and solved to determine the output image and output blur kernel corresponding to the iterative process, wherein the input blur kernel corresponding to the iterative process is the initial blur kernel corresponding to the i-th skeleton image, or the output blur kernel corresponding to the previous iterative process; For any iterative process, when the iterative process satisfies the convergence condition, the convergence blur kernel corresponding to the i-th skeleton image is determined according to the output blur kernel corresponding to the iterative process.

4. The method according to claim 3, characterized in that The step of performing light intensity correction processing on the input image corresponding to any iterative process to determine the corrected input image corresponding to the iterative process includes: For any iterative process, a unified fuzzy model corresponding to the iterative process is determined according to the input image and input fuzzy kernel corresponding to the iterative process, wherein the unified fuzzy model is used to adjust the brightness value of each pixel in the input image corresponding to the iterative process; According to the unified blur model, the input image corresponding to the iteration process is subjected to light intensity correction processing to determine the corrected input image corresponding to the iteration process.

5. The method according to any one of claims 1 to 3, characterized in that: The step of performing a deblurring process on the initial single-frame astronomical image according to the target blur kernel to determine the deblurred single-frame astronomical image comprises: Determining a target optimization function corresponding to the initial single-frame astronomical image according to a non-blind deconvolution algorithm; The target optimization function is optimized and solved according to the target blur kernel to determine the deblurred single-frame astronomical image.

6. The method according to any one of claims 1 to 3, characterized in that: The method further comprises: Determine the global atmospheric light value and atmospheric transmittance corresponding to the deblurred single-frame astronomical image; According to the global atmospheric light value and the atmospheric transmittance, image artifact removal and enhancement processing is performed on the deblurred single-frame astronomical image to determine an enhanced image corresponding to the deblurred single-frame astronomical image.

7. The method according to claim 6, characterized in that The determining of the global atmospheric light value and the atmospheric transmittance corresponding to the deblurred single-frame astronomical image comprises: Determining the global atmospheric light value according to the dark channel image corresponding to the deblurred single-frame astronomical image; The atmospheric transmittance is determined according to a dark channel priori algorithm and the global atmospheric light value.

8. A single-frame astronomical image reconstruction device, characterized in that: include: An optimization model building module is used to determine the blind deconvolution optimization model corresponding to the initial single-frame astronomical image by using a blind deconvolution algorithm based on a reweighted graph total variation prior; A resampling module, used to resample the initial single-frame astronomical image to determine n skeleton images corresponding to the initial single-frame astronomical image, where n is a positive integer greater than or equal to 1; A blur kernel determination module, used to perform iterative optimization solution and light intensity correction processing according to the blind deconvolution optimization model and the n skeleton images, to determine the target blur kernel corresponding to the initial single-frame astronomical image; The deblurring module is used to perform deblurring processing on the initial single-frame astronomical image according to the target blur kernel to determine a deblurred single-frame astronomical image.

9. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to implement the method described in any one of claims 1 to 7 when executing the instructions stored in the memory.

10. A non-volatile computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Citation Information

Patent Citations

  • Image blind deblurring method based on edge self-adaption

    CN103761710A

  • Blind convolutional motion fuzzy image restoration method

    CN106530251A

  • Radio astronomical image reconstruction method based on activation function

    CN112712486A

  • De-Noising Image Content Using Directional Filters for Image De-Blurring

    US20140133775A1