An astronomical image restoration method, an electronic device, and a program product

Abstract

An embodiment of the present disclosure discloses an astronomical image restoration method, an electronic device, and a program product. The method includes: obtaining one-dimensional observed data of celestial body vectors to be recognized according to the astronomical image data to be recognized; obtaining an initial reconstructed target source vector; adding a product to the model target source vector; using the residual vector as the input value of the conjugate gradient iteration method for iteration; adding the product of the maximum value of the elements in the reconstructed target source vector and a second parameter factor to the model target source vector; obtaining a final target source vector according to the sum of the model target source vector and the iterated reconstructed target source vector; and obtaining a two-dimensional reconstructed target image based on the final target source vector. When processing the one-dimensional observed data of celestial body vectors to be recognized with a large amount of noise, the multiple celestial bodies with close distances in the finally obtained two-dimensional reconstructed target image can be accurately distinguished, improving the accuracy of distinguishing multiple celestial bodies with close distances in the astronomical image based on restoration.

Description

Technical Field

[0001] The present disclosure relates to the technical field of celestial body recognition, and particularly to an astronomical image restoration method, an electronic device, and a program product. Background Art

[0002] Radio astronomy is an important branch of astronomy. Radio astronomical equipment based on radio astronomy mainly receives and processes radio radiation from cosmic celestial bodies to obtain observation data. According to this observation data, astronomical images can be restored to facilitate the recognition of celestial bodies or astronomical phenomena in the universe based on the restored astronomical images. In related technologies, when restoring astronomical images based on observation data, the CLEAN algorithm can be used. The CLEAN algorithm belongs to a deconvolution algorithm with a relatively high signal-to-noise ratio, and it can eliminate the influence of the sidelobes of the point spread function during the process of restoring astronomical images.

[0003] In recent years, with the rapid development of information technology and manufacturing technology, human communication activities have become increasingly intensive, and the intensity of electromagnetic signals in daily life has also become greater. Although radio astronomical equipment has been continuously upgraded or replaced, the increasingly strong non-astronomical signals (such as signals from communication activities and electromagnetic signals in daily life) have had an increasingly strong impact on radio astronomical equipment, resulting in the observation data collected by radio observation equipment often being noisy. Considering that the CLEAN algorithm has poor effects when processing data with strong noise, in the above-mentioned scheme for restoring astronomical images based on observation data, when the observation data includes data of multiple celestial bodies that are relatively close to each other, the multiple celestial bodies that are relatively close to each other may not be distinguishable based on the restored astronomical images, reducing the accuracy of distinguishing multiple celestial bodies that are relatively close to each other based on the restored astronomical images. Summary of the Invention

[0004] To solve the problems in related technologies, embodiments of the present disclosure provide an astronomical image restoration method, an electronic device, and a program product.

[0005] In a first aspect, an astronomical image restoration method is provided in embodiments of the present disclosure.

[0006] Specifically, the astronomical image restoration method includes:

[0007] S1, obtaining astronomical image data to be recognized, and performing serialization processing on the astronomical image data to be recognized to obtain one-dimensional vector observation data of celestial bodies to be recognized;

[0008] S2, based on the vector observation data of celestial bodies to be recognized, obtaining an initial reconstructed target source vector through the conjugate gradient iteration method;

[0009] S3. When the maximum value of the elements in the initial reconstructed target source vector is greater than or equal to the first vector element threshold, obtain the reconstructed target source vector according to the initial reconstructed target source vector, and execute step S4. When the maximum value of the elements in the initial reconstructed target source vector is less than the first vector element threshold, obtain the reconstructed target source vector according to the initial reconstructed target source vector, and execute step S7;

[0010] S4. Add the product of the maximum value of the elements in the reconstructed target source vector and the first parameter factor to the model target source vector, and remove the convolution of the maximum value of the elements in the reconstructed target source vector and the point spread function from the observed data vector of the celestial body to be recognized. The length of the model target source vector is the same as the length of the reconstructed target source vector;

[0011] S5. Use the residual vector between the observed data vector of the celestial body to be recognized and the model observed data as the input value of the conjugate gradient iteration method for iteration to obtain the updated reconstructed target source vector. The model observed data is the convolution of the reconstructed target source vector and the point spread function;

[0012] S6. When the maximum value of the elements in the updated reconstructed target source vector is greater than or equal to the second vector element threshold, execute step S4. When the maximum value of the elements in the updated reconstructed target source vector is less than the second vector element threshold, execute step S7;

[0013] S7. Use the residual vector as the input value of the conjugate gradient iteration method for iteration, and set the values of the negative elements in the reconstructed target source vector whose values are negative to 0 during the iteration process to obtain the iterated reconstructed target source vector;

[0014] S8. Add the product of the maximum value of the elements in the iterated reconstructed target source vector and the second parameter factor to the model target source vector, and remove the convolution of the maximum value of the elements in the iterated reconstructed target source vector and the point spread function from the residual vector;

[0015] S9. When the maximum value of the elements in the iterated reconstructed target source vector is greater than or equal to the third vector element threshold, execute step S8. When the maximum value of the elements in the iterated reconstructed target source vector is less than the third vector element threshold, obtain the final target source vector according to the sum of the model target source vector and the iterated reconstructed target source vector;

[0016] S10. Obtain the two-dimensional reconstructed target image based on the final target source vector.

[0017] In an implementation manner of the present disclosure, the method further includes:

[0018] Obtain the evaluation parameters of the final target source. The model observation evaluation parameters include at least one of the final target source residual, the chi-square value of the final target source vector, the mean value of the final target source vector, and the mean square deviation of the final target source vector. The final target source residual is the difference between the observed data of the celestial body vector to be identified and the final target source vector.

[0019] Determine the credibility of the two-dimensional reconstructed target image according to the evaluation parameters of the final target source.

[0020] In an implementation manner of the present disclosure, the value ranges of the first parameter factor and the second parameter factor are both [0, 1].

[0021] In an implementation manner of the present disclosure, before step S3, the method further includes:

[0022] Obtain a model target source vector with an initial value of 0 and a length the same as that of the reconstructed target source vector.

[0023] In an implementation manner of the present disclosure, the first vector element threshold x T1 = 3σ x1 , where σ x1 is the standard deviation of the reconstructed target source vector.

[0024] In an implementation manner of the present disclosure, the second vector element threshold x T2 = 3σ x2 , where σ x2 is the standard deviation of the residual vector.

[0025] In an embodiment of the present disclosure, the method further includes:

[0026] Display the two-dimensional reconstructed target image.

[0027] In a second aspect, an electronic device is provided in an embodiment of the present disclosure, including a memory and a processor; wherein, the memory is used to store one or more computer instructions, and one or more computer instructions are executed by the processor to implement the method steps of any one of the first aspect.

[0028] In a third aspect, a readable storage medium is provided in an embodiment of the present disclosure, on which computer instructions are stored, and when the computer instructions are executed by the processor, the method steps of any one of the first aspect are implemented.

[0029] In a fourth aspect, a computer program product is provided in an embodiment of the present disclosure, including computer instructions, and when the computer instructions are executed by the processor, the method steps of any one of the first aspect are implemented.

[0030] According to the technical solution provided by the embodiments of the present disclosure, in step S1, the astronomical image data to be recognized is obtained, and the astronomical image data to be recognized is serialized to obtain one-dimensional vector observation data of the celestial body to be recognized; in step S2, based on the vector observation data of the celestial body to be recognized, the initial reconstructed target source vector is obtained by the conjugate gradient iteration method; in step S3, when the maximum value of the elements in the initial reconstructed target source vector is greater than or equal to the first vector element threshold, the reconstructed target source vector is obtained according to the initial reconstructed target source vector, and step S4 is executed. When the maximum value of the elements in the initial reconstructed target source vector is less than the first vector element threshold, the reconstructed target source vector is obtained according to the initial reconstructed target source vector, and step S7 is executed; in step S4, the product of the maximum value of the elements in the reconstructed target source vector and the first parameter factor is added to the model target source vector, and the convolution of the maximum value of the elements in the reconstructed target source vector and the point spread function is removed from the vector observation data of the celestial body to be recognized. The length of the model target source vector is the same as the length of the reconstructed target source vector; in step S5, the residual vector between the vector observation data of the celestial body to be recognized and the model observation data is used as the input value of the conjugate gradient iteration method for iteration to obtain the updated reconstructed target source vector. The model observation data is the convolution of the reconstructed target source vector and the point spread function; in step S6, when the maximum value of the elements in the updated reconstructed target source vector is greater than or equal to the second vector element threshold, step S4 is executed. When the maximum value of the elements in the updated reconstructed target source vector is less than the second vector element threshold, step S7 is executed; in step S7, the residual vector is used as the input value of the conjugate gradient iteration method for iteration, and during the iteration process, the negative elements with negative values in the reconstructed target source vector are set to 0 to obtain the iterated reconstructed target source vector; in step S8, the product of the maximum value of the elements in the iterated reconstructed target source vector and the second parameter factor is added to the model target source vector, and the convolution of the maximum value of the elements in the iterated reconstructed target source vector and the point spread function is removed from the residual vector; in step S9, when the maximum value of the elements in the iterated reconstructed target source vector is greater than or equal to the third vector element threshold, step S8 is executed. When the maximum value of the elements in the iterated reconstructed target source vector is less than the third vector element threshold, the final target source vector is obtained by summing the model target source vector and the iterated reconstructed target source vector; in step S10, the two-dimensional reconstructed target image is obtained based on the final target source vector. The technical solution provided by the embodiments of the present disclosure can accurately distinguish multiple celestial bodies with close distances in the finally obtained two-dimensional reconstructed target image when processing the vector observation data of celestial bodies to be recognized with more noise, improving the accuracy of distinguishing multiple celestial bodies with close distances in the astronomical image based on restoration.

[0031] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present disclosure. Description of the Drawings

[0032] In conjunction with the accompanying drawings, other features, objects, and advantages of the present disclosure will become more apparent through the following detailed description of non - restrictive embodiments. In the drawings:

[0033] Figure 1 A flowchart of an astronomical image restoration method according to an embodiment of the present disclosure is shown;

[0034] Figure 2 A schematic structural block diagram of an astronomical image restoration apparatus according to an embodiment of the present disclosure is shown;

[0035] Figure 3 A structural block diagram of an electronic device according to an embodiment of the present disclosure is shown;

[0036] Figure 4 A schematic structural diagram of a computer system suitable for implementing the method according to an embodiment of the present disclosure is shown. Detailed Embodiments

[0037] Hereinafter, exemplary embodiments of the present disclosure will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily implement them. In addition, for clarity, parts irrelevant to the description of the exemplary embodiments are omitted in the drawings.

[0038] In the present disclosure, it should be understood that terms such as "including" or "having" are intended to indicate the presence of features, numbers, steps, actions, components, parts, or combinations thereof disclosed in this specification, and are not intended to exclude the possibility of the presence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0039] In addition, it should be noted that, without conflict, the embodiments in the present disclosure and the features in the embodiments can be combined with each other. The present disclosure will be described in detail below with reference to the drawings and in conjunction with the embodiments.

[0040] Radio astronomy is an important branch of astronomy. Radio astronomy equipment based on radio astronomy mainly receives and processes radio radiation from cosmic celestial bodies to obtain observation data. Based on this observation data, astronomical images can be restored to facilitate the identification of celestial bodies or astronomical phenomena in the universe according to the restored astronomical images. In the related art, when restoring astronomical images based on observation data, the CLEAN algorithm can be used. The CLEAN algorithm belongs to a deconvolution algorithm with a relatively high signal - to - noise ratio, and it can eliminate the influence of the sidelobes of the point spread function during the process of restoring astronomical images.

[0041] In recent years, with the rapid development of information technology and manufacturing technology, human communication activities have become increasingly intensive, and the intensity of electromagnetic signals in daily life has also become greater. Although radio astronomy equipment has been continuously upgraded or replaced, the increasingly strong non-astronomical signals (such as signals from communication activities and electromagnetic signals in daily life) have had an increasingly strong impact on radio astronomy equipment, resulting in the observation data collected by radio observation equipment often being noisy. Considering that the CLEAN algorithm has poor performance in processing data with strong noise, in the above-mentioned solution for restoring astronomical images based on observation data, when the observation data includes data of multiple celestial bodies that are relatively close to each other, it may not be possible to distinguish multiple celestial bodies that are relatively close to each other based on the restored astronomical image, reducing the accuracy of distinguishing multiple celestial bodies that are relatively close to each other based on the restored astronomical image.

[0042] Considering the above defects, in the technical solution provided by the present disclosure,

[0043] In step S1, the astronomical image data to be recognized is obtained, and the astronomical image data to be recognized is serialized to obtain one-dimensional vector observation data of the celestial body to be recognized; in step S2, based on the vector observation data of the celestial body to be recognized, the initial reconstructed target source vector is obtained by the conjugate gradient iteration method; in step S3, when the maximum value of the elements in the initial reconstructed target source vector is greater than or equal to the first vector element threshold, the reconstructed target source vector is obtained according to the initial reconstructed target source vector, and step S4 is executed. When the maximum value of the elements in the initial reconstructed target source vector is less than the first vector element threshold, the reconstructed target source vector is obtained according to the initial reconstructed target source vector, and step S7 is executed; in step S4, the product of the maximum value of the elements in the reconstructed target source vector and the first parameter factor is added to the model target source vector, and the convolution of the maximum value of the elements in the reconstructed target source vector and the point spread function is subtracted from the vector observation data of the celestial body to be recognized. The length of the model target source vector is the same as the length of the reconstructed target source vector; in step S5, the residual vector between the vector observation data of the celestial body to be recognized and the model observation data is used as the input value of the conjugate gradient iteration method for iteration to obtain the updated reconstructed target source vector. The model observation data is the convolution of the reconstructed target source vector and the point spread function; in step S6, when the maximum value of the elements in the updated reconstructed target source vector is greater than or equal to the second vector element threshold, step S4 is executed. When the maximum value of the elements in the updated reconstructed target source vector is less than the second vector element threshold, step S7 is executed; in step S7, the residual vector is used as the input value of the conjugate gradient iteration method for iteration, and during the iteration process, the negative elements with negative values in the reconstructed target source vector are set to 0 to obtain the iterated reconstructed target source vector; in step S8, the product of the maximum value of the elements in the iterated reconstructed target source vector and the second parameter factor is added to the model target source vector, and the convolution of the maximum value of the elements in the iterated reconstructed target source vector and the point spread function is subtracted from the residual vector; in step S9, when the maximum value of the elements in the iterated reconstructed target source vector is greater than or equal to the third vector element threshold, step S8 is executed. When the maximum value of the elements in the iterated reconstructed target source vector is less than the third vector element threshold, the final target source vector is obtained by summing the model target source vector and the iterated reconstructed target source vector; in step S10, the two-dimensional reconstructed target image is obtained based on the final target source vector. The technical solution provided by the embodiments of the present disclosure can accurately distinguish multiple celestial bodies with close distances in the finally obtained two-dimensional reconstructed target image when processing the vector observation data of celestial bodies to be recognized with more noise, improving the accuracy of distinguishing multiple celestial bodies with close distances in the astronomical image based on restoration.

[0044] Figure 1 FIG. shows a flowchart of an astronomical image restoration method according to an embodiment of the present disclosure. As Figure 1As shown, the astronomical image restoration method includes the following steps S1 - S10:

[0045] In step S1, obtain the astronomical image data to be recognized, and perform serialization processing on the astronomical image data to be recognized to obtain one - dimensional vector observation data of the celestial body to be recognized;

[0046] In step S2, based on the vector observation data of the celestial body to be recognized, obtain the initial reconstructed target source vector through the conjugate gradient iteration method.

[0047] In step S3, when the maximum value of the elements in the initial reconstructed target source vector is greater than or equal to the first vector element threshold, obtain the reconstructed target source vector according to the initial reconstructed target source vector, and execute step S4. When the maximum value of the elements in the initial reconstructed target source vector is less than the first vector element threshold, obtain the reconstructed target source vector according to the initial reconstructed target source vector, and execute step S7.

[0048] Among them, when the maximum value of the elements in the initial reconstructed target source vector is less than the first vector element threshold, it indicates that the main information in the astronomical image data to be recognized has been extracted, and only non - negative - constrained iteration needs to be performed subsequently to reconstruct the fine structure in the astronomical image data to be recognized. When the maximum value of the elements in the initial reconstructed target source vector is greater than or equal to the first vector element threshold, it indicates that the main structure in the astronomical image data to be recognized has not been effectively extracted. First, the main structure in the astronomical image data to be recognized needs to be extracted through the CLEAN method without non - negative constraints, and then non - negative - constrained iteration is performed subsequently to prevent the appearance of oscillatory structures when reconstructing the fine structure in the astronomical image data to be recognized.

[0049] In step S4, add the product of the maximum value of the elements in the reconstructed target source vector and the first parameter factor to the model target source vector, and remove the convolution of the maximum value of the elements in the reconstructed target source vector and the point - source spread function from the vector observation data of the celestial body to be recognized.

[0050] Among them, the length of the model target source vector is the same as the length of the reconstructed target source vector.

[0051] By adding the product of the maximum value of the elements in the reconstructed target source vector and the first parameter factor to the model target source vector, the information in the reconstructed target source vector can be extracted into the model target source vector. Then, removing the convolution of the maximum value of the elements in the reconstructed target source vector and the point - source spread function from the vector observation data of the celestial body to be recognized can avoid the situation where the extracted information still exists in the initial reconstructed target source vector.

[0052] In step S5, the residual vector between the celestial body vector observation data to be recognized and the model observation data is used as the input value of the conjugate gradient iteration method for iteration to obtain the updated reconstructed target source vector. The model observation data is the convolution of the reconstructed target source vector and the point spread function.

[0053] In step S6, when the maximum value of the elements in the updated reconstructed target source vector is greater than or equal to the second vector element threshold, step S4 is executed. When the maximum value of the elements in the updated reconstructed target source vector is less than the second vector element threshold, step S7 is executed.

[0054] Among them, when the maximum value of the elements in the residual vector is greater than or equal to the second vector element threshold, it can be understood that at this time, more information in the astronomical image data to be recognized has not been extracted. Therefore, information extraction is continued by executing step S4. When the maximum value of the elements in the residual vector is less than the second vector element threshold, it can be understood that most of the information that can be extracted in the astronomical image data to be recognized has been extracted. Therefore, step S7 can be executed to eliminate the oscillatory structure that appears in the subsequent fine structure reconstruction process.

[0055] In step S7, the residual vector is used as the input value of the conjugate gradient iteration method for iteration. During the iteration process, the values of the negative elements in the reconstructed target source vector that are negative are set to 0 to obtain the iterated reconstructed target source vector.

[0056] In step S8, the product of the maximum value of the elements in the iterated reconstructed target source vector and the second parameter factor is added to the model target source vector, and the convolution of the maximum value of the elements in the iterated reconstructed target source vector and the point spread function is removed from the residual vector.

[0057] In step S9, when the maximum value of the elements in the iterated reconstructed target source vector is greater than or equal to the third vector element threshold, step S8 is executed. When the maximum value of the elements in the iterated reconstructed target source vector is less than the third vector element threshold, the final target source vector is obtained by summing the model target source vector and the iterated reconstructed target source vector.

[0058] Among them, when the maximum value of the elements in the reconstructed target source vector is greater than or equal to the third vector element threshold, it can be understood that more information in the celestial body vector observation data to be recognized has not been extracted. Therefore, information extraction is continued by executing step S8. When the maximum value of the elements in the reconstructed target source vector is less than the third vector element threshold, it can be understood that most of the information that can be extracted in the celestial body vector observation data to be recognized has been extracted. Therefore, the final target source vector can be obtained to facilitate obtaining the two-dimensional reconstructed target image based on the final target source vector.

[0059] In step S10, a two-dimensional reconstructed target image is obtained based on the final target source vector.

[0060] According to the technical solution provided by the embodiments of the present disclosure, in step S1, astronomical image data to be recognized is obtained, and the astronomical image data to be recognized is serialized to obtain one-dimensional vector observation data of celestial bodies to be recognized; in step S2, based on the vector observation data of celestial bodies to be recognized, an initial reconstructed target source vector is obtained by the conjugate gradient iteration method; in step S3, when the maximum value of the elements in the initial reconstructed target source vector is greater than or equal to the first vector element threshold, a reconstructed target source vector is obtained according to the initial reconstructed target source vector, and step S4 is executed. When the maximum value of the elements in the initial reconstructed target source vector is less than the first vector element threshold, a reconstructed target source vector is obtained according to the initial reconstructed target source vector, and step S7 is executed; in step S4, the product of the maximum value of the elements in the reconstructed target source vector and the first parameter factor is added to the model target source vector, and the convolution of the maximum value of the elements in the reconstructed target source vector and the point spread function is removed from the vector observation data of celestial bodies to be recognized. The length of the model target source vector is the same as the length of the reconstructed target source vector; in step S5, the residual vector between the vector observation data of celestial bodies to be recognized and the model observation data is used as the input value of the conjugate gradient iteration method for iteration to obtain an updated reconstructed target source vector. The model observation data is the convolution of the reconstructed target source vector and the point spread function; in step S6, when the maximum value of the elements in the updated reconstructed target source vector is greater than or equal to the second vector element threshold, step S4 is executed. When the maximum value of the elements in the updated reconstructed target source vector is less than the second vector element threshold, step S7 is executed; in step S7, the residual vector is used as the input value of the conjugate gradient iteration method for iteration, and during the iteration process, the values of the negative elements in the reconstructed target source vector with negative values are set to 0 to obtain an iterated reconstructed target source vector; in step S8, the product of the maximum value of the elements in the iterated reconstructed target source vector and the second parameter factor is added to the model target source vector, and the convolution of the maximum value of the elements in the iterated reconstructed target source vector and the point spread function is removed from the residual vector; in step S9, when the maximum value of the elements in the iterated reconstructed target source vector is greater than or equal to the third vector element threshold, step S8 is executed. When the maximum value of the elements in the iterated reconstructed target source vector is less than the third vector element threshold, the final target source vector is obtained by summing the model target source vector and the iterated reconstructed target source vector; in step S10, a two-dimensional reconstructed target image is obtained based on the final target source vector. The technical solution provided by the embodiments of the present disclosure can accurately distinguish multiple celestial bodies with close distances in the finally obtained two-dimensional reconstructed target image when processing vector observation data of celestial bodies to be recognized with more noise, improving the accuracy of distinguishing multiple celestial bodies with close distances in the astronomical image based on restoration.

[0061] In an implementation of the present disclosure, the astronomical image restoration method further includes:

[0062] Obtain the final target source evaluation parameters, where the model observation evaluation parameters include at least one of the final target source residual, the chi-square value of the final target source vector, the mean value of the final target source vector, and the mean square deviation of the final target source vector. The final target source residual is the difference between the observed data of the celestial body vector to be identified and the final target source vector.

[0063] Determine the credibility of the two-dimensional reconstructed target image according to the final target source evaluation parameters.

[0064] Among them, the final target source residual can reflect the amount of information extracted from the original data, that is, the observed data of the celestial body vector to be identified. The chi-square value, the mean value, and the mean square deviation of the final target source vector can all reflect the distribution of the final target source vector. Since the two-dimensional reconstructed target image obtained by restoration is based on the final target source vector, the credibility of the two-dimensional reconstructed target image can be determined based on the final target source evaluation parameters, which is convenient for users to conduct corresponding research based on the two-dimensional reconstructed target image.

[0065] In an implementation of the present disclosure, the value ranges of both the first parameter factor and the second parameter factor are [0, 1].

[0066] In an implementation of the present disclosure, before step S3, the method further includes:

[0067] Obtain a model target source vector with an initial value of 0 and a length the same as that of the reconstructed target source vector.

[0068] In an implementation of the present disclosure, the first vector element threshold x T1 = 3σ x1 , where σ x1 is the standard deviation of the reconstructed target source vector.

[0069] By setting the first vector element threshold according to the standard deviation of the reconstructed target source vector in step S3, it is possible to judge the probability of the occurrence of an oscillation structure around the element with the largest value in the initial reconstructed target source vector adaptively according to the initial reconstructed target source vector in step S3.

[0070] In an implementation of the present disclosure, the second vector element threshold x T2 = 3σ x2 , where σ x2 is the standard deviation of the residual vector.

[0071] By setting the second vector element threshold according to the standard deviation of the residual vector, it is possible to adaptively determine whether there is still a large amount of information in the reconstructed target source vector in step S6 that has not been extracted based on the residual vector.

[0072] In one implementation of the present disclosure, the third vector element threshold x T3 = 3σ x3 , where σ x3 is the standard deviation of the reconstructed target source vector in step S9.

[0073] By setting the second vector element threshold according to the standard deviation of the reconstructed target source vector in step S9, it is possible to determine whether there is still a large amount of information in the reconstructed target source vector in step S9 that has not been extracted.

[0074] In one implementation of the present disclosure, the astronomical image restoration method further includes:

[0075] Displaying the two-dimensional reconstructed target image.

[0076] Among them, by displaying the two-dimensional reconstructed target image, it is convenient for the user to understand the information contained in the two-dimensional reconstructed target image.

[0077] The following is an embodiment of the device of the present disclosure, which can be used to execute the method embodiment of the present disclosure.

[0078] An astronomical image restoration device according to an embodiment of the present disclosure, the astronomical image restoration device can be implemented as part or all of an electronic device through software, hardware, or a combination of both. Figure 2 Shows a schematic structural block diagram of an astronomical image restoration device according to an embodiment of the present disclosure, as Figure 2 shown, the astronomical image restoration device 200 includes:

[0079] A first vector acquisition module 201, configured to execute S1, acquire astronomical image data to be recognized, and perform serialization processing on the astronomical image data to be recognized to acquire one-dimensional astronomical object vector observation data to be recognized;

[0080] A second vector acquisition module 202, configured to execute S2, and acquire an initial reconstructed target source vector based on the astronomical object vector observation data to be recognized through a conjugate gradient iteration method;

[0081] A first judgment module 203, configured to execute S3, when the maximum value of the elements in the initial reconstructed target source vector is greater than or equal to the first vector element threshold, acquire the reconstructed target source vector according to the initial reconstructed target source vector, and execute step S4, when the maximum value of the elements in the initial reconstructed target source vector is less than the first vector element threshold, acquire the reconstructed target source vector according to the initial reconstructed target source vector, and execute step S7;

[0082] The first information extraction module 204 is configured to execute S4, add the product of the maximum value of the elements in the reconstructed target source vector and the first parameter factor to the model target source vector, and remove the convolution of the maximum value of the elements in the reconstructed target source vector and the point spread function from the celestial object vector observation data to be recognized. The length of the model target source vector is the same as that of the reconstructed target source vector.

[0083] The residual vector acquisition module 205 is configured to execute S5, take the residual vector of the celestial object vector observation data to be recognized and the model observation data as the input value of the conjugate gradient iteration method for iteration, and obtain the updated reconstructed target source vector. The model observation data is the convolution of the reconstructed target source vector and the point spread function.

[0084] The second judgment module 206 is configured to execute S6. When the maximum value of the elements in the updated reconstructed target source vector is greater than or equal to the second vector element threshold, execute step S4. When the maximum value of the elements in the updated reconstructed target source vector is less than the second vector element threshold, execute step S7.

[0085] The non - negative constraint module 207 is configured to execute S7, take the residual vector as the input value of the conjugate gradient iteration method for iteration, and set the negative elements with negative values in the reconstructed target source vector to 0 during the iteration process to obtain the iterated reconstructed target source vector.

[0086] The second information extraction module 208 is configured to execute S8, add the product of the maximum value of the elements in the iterated reconstructed target source vector and the second parameter factor to the model target source vector, and remove the convolution of the maximum value of the elements in the iterated reconstructed target source vector and the point spread function from the residual vector.

[0087] The third judgment module 209 is configured to execute S9. When the maximum value of the elements in the iterated reconstructed target source vector is greater than or equal to the third vector element threshold, execute step S8. When the maximum value of the elements in the iterated reconstructed target source vector is less than the third vector element threshold, obtain the final target source vector according to the sum of the model target source vector and the iterated reconstructed target source vector.

[0088] The restored image acquisition module 210 is configured to execute S10 to obtain a two - dimensional reconstructed target image based on the final target source vector.

[0089] According to the technical solution provided by the embodiments of the present disclosure, in step S1, the astronomical image data to be recognized is obtained, and the astronomical image data to be recognized is serialized to obtain one-dimensional vector observation data of the celestial body to be recognized; in step S2, based on the vector observation data of the celestial body to be recognized, an initial reconstructed target source vector is obtained by the conjugate gradient iteration method; in step S3, when the maximum value of the elements in the initial reconstructed target source vector is greater than or equal to the first vector element threshold, a reconstructed target source vector is obtained according to the initial reconstructed target source vector, and step S4 is executed. When the maximum value of the elements in the initial reconstructed target source vector is less than the first vector element threshold, a reconstructed target source vector is obtained according to the initial reconstructed target source vector, and step S7 is executed; in step S4, the product of the maximum value of the elements in the reconstructed target source vector and the first parameter factor is added to the model target source vector, and the convolution of the maximum value of the elements in the reconstructed target source vector and the point spread function is removed from the vector observation data of the celestial body to be recognized. The length of the model target source vector is the same as the length of the reconstructed target source vector; in step S5, the residual vector between the vector observation data of the celestial body to be recognized and the model observation data is used as the input value of the conjugate gradient iteration method for iteration to obtain an updated reconstructed target source vector. The model observation data is the convolution of the reconstructed target source vector and the point spread function; in step S6, when the maximum value of the elements in the updated reconstructed target source vector is greater than or equal to the second vector element threshold, step S4 is executed. When the maximum value of the elements in the updated reconstructed target source vector is less than the second vector element threshold, step S7 is executed; in step S7, the residual vector is used as the input value of the conjugate gradient iteration method for iteration, and during the iteration process, the negative elements in the reconstructed target source vector with negative values are set to 0 to obtain an iterated reconstructed target source vector; in step S8, the product of the maximum value of the elements in the iterated reconstructed target source vector and the second parameter factor is added to the model target source vector, and the convolution of the maximum value of the elements in the iterated reconstructed target source vector and the point spread function is removed from the residual vector; in step S9, when the maximum value of the elements in the iterated reconstructed target source vector is greater than or equal to the third vector element threshold, step S8 is executed. When the maximum value of the elements in the iterated reconstructed target source vector is less than the third vector element threshold, the final target source vector is obtained by summing the model target source vector and the iterated reconstructed target source vector; in step S10, a two-dimensional reconstructed target image is obtained based on the final target source vector. The technical solution provided by the embodiments of the present disclosure can accurately distinguish multiple celestial bodies with close distances in the finally obtained two-dimensional reconstructed target image when processing the vector observation data of celestial bodies to be recognized with more noise, improving the accuracy of distinguishing multiple celestial bodies with close distances in the astronomical image based on restoration.

[0090] The present disclosure also discloses an electronic device, Figure 3A structural block diagram of an electronic device according to an embodiment of the present disclosure is shown.

[0091] As Figure 3 shown, the electronic device 300 includes a memory 301 and a processor 302. Among them, the memory 301 is used to store one or more computer instructions, and the one or more computer instructions are executed by the processor 302 to implement the method according to the embodiment of the present disclosure.

[0092] Specifically, the method includes:

[0093] S1. Obtain the astronomical image data to be recognized, and perform serialization processing on the astronomical image data to be recognized to obtain one-dimensional vector observation data of the celestial body to be recognized;

[0094] S2. Based on the vector observation data of the celestial body to be recognized, obtain the initial reconstructed target source vector through the conjugate gradient iteration method;

[0095] S3. When the maximum value of the elements in the initial reconstructed target source vector is greater than or equal to the first vector element threshold, obtain the reconstructed target source vector according to the initial reconstructed target source vector, and execute step S4. When the maximum value of the elements in the initial reconstructed target source vector is less than the first vector element threshold, obtain the reconstructed target source vector according to the initial reconstructed target source vector, and execute step S7;

[0096] S4. Add the product of the maximum value of the elements in the reconstructed target source vector and the first parameter factor to the model target source vector, and remove the convolution of the maximum value of the elements in the reconstructed target source vector and the point spread function from the vector observation data of the celestial body to be recognized. The length of the model target source vector is the same as the length of the reconstructed target source vector;

[0097] S5. Use the residual vector between the vector observation data of the celestial body to be recognized and the model observation data as the input value of the conjugate gradient iteration method for iteration to obtain the updated reconstructed target source vector. The model observation data is the convolution of the reconstructed target source vector and the point spread function;

[0098] S6. When the maximum value of the elements in the updated reconstructed target source vector is greater than or equal to the second vector element threshold, execute step S4. When the maximum value of the elements in the updated reconstructed target source vector is less than the second vector element threshold, execute step S7;

[0099] S7. Use the residual vector as the input value of the conjugate gradient iteration method for iteration, and set the negative elements with negative values in the reconstructed target source vector to 0 during the iteration process to obtain the iterated reconstructed target source vector;

[0100] S8. Add the product of the maximum value of the elements in the iterated reconstructed target source vector and the second parameter factor to the model target source vector, and remove the convolution of the maximum value of the elements in the iterated reconstructed target source vector and the point spread function from the residual vector;

[0101] S9. When the maximum value of the elements in the iterated reconstructed target source vector is greater than or equal to the third vector element threshold, execute step S8. When the maximum value of the elements in the iterated reconstructed target source vector is less than the third vector element threshold, obtain the final target source vector by summing the model target source vector and the iterated reconstructed target source vector;

[0102] S10. Obtain the two-dimensional reconstructed target image based on the final target source vector.

[0103] In one implementation of the present disclosure, the method further includes:

[0104] Obtain the final target source evaluation parameter. The model observation evaluation parameter includes at least one of the final target source residual, the chi-square value of the final target source vector, the mean value of the final target source vector, and the mean square deviation of the final target source vector. The final target source residual is the difference between the vector observation data of the celestial body to be identified and the final target source vector;

[0105] Determine the credibility of the two-dimensional reconstructed target image according to the final target source evaluation parameter.

[0106] In one implementation of the present disclosure, the value ranges of the first parameter factor and the second parameter factor are both [0, 1].

[0107] In one implementation of the present disclosure, before step S3, the method further includes:

[0108] Obtain a model target source vector with all initial values being 0 and the same length as the reconstructed target source vector.

[0109] In one implementation of the present disclosure, the first vector element threshold x T1 = 3σ x1 , where σ x1 is the standard deviation of the reconstructed target source vector.

[0110] In one implementation of the present disclosure, the second vector element threshold x T2 = 3σ x2 , where σ x2 is the standard deviation of the residual vector.

[0111] In one implementation of the present disclosure, the method further includes:

[0112] Display the two-dimensional reconstructed target image.

[0113] Figure 4A schematic structural diagram of a computer system suitable for implementing the method according to an embodiment of the present disclosure is shown.

[0114] As Figure 4 shown, the computer system 400 includes a processing unit 401, which can execute various processes in the above embodiments according to a program stored in a read-only memory (ROM) 402 or a program loaded from a storage section 408 into a random access memory (RAM) 403. In the RAM 403, various programs and data required for the operation of the system 400 are also stored. The processing unit 401, the ROM 402, and the RAM 403 are connected to each other via a bus 404. An input / output (I / O) interface 405 is also connected to the bus 404.

[0115] The following components are connected to the I / O interface 405: an input section 406 including a keyboard, a mouse, etc.; an output section 407 including, for example, a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 408 including a hard disk, etc.; and a communication section 409 including a network interface card such as a LAN card, a modem, etc. The communication section 409 performs communication processing via a network such as the Internet. A drive 410 is also connected to the I / O interface 405 as required. A removable medium 411, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 410 as required so that a computer program read from it can be installed into the storage section 408 as required. Among them, the processing unit 401 can be implemented as a processing unit such as a CPU, a GPU, a TPU, an FPGA, an NPU, etc.

[0116] Specifically, according to an embodiment of the present disclosure, the method described above can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes computer instructions that, when executed by a processor, implement the method steps described above. In such an embodiment, the computer program product can be downloaded and installed from a network through the communication section 409, and / or installed from the removable medium 411.

[0117] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a segment of a program, or a part of code that contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, as well as combinations of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system that performs the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0118] The units or modules described in the embodiments of the present disclosure can be implemented in software or in programmable hardware. The described units or modules can also be provided in a processor, and the names of these units or modules do not, in some cases, constitute a limitation on the units or modules themselves.

[0119] As another aspect, the present disclosure also provides a computer-readable storage medium, which can be the computer-readable storage medium included in the electronic device or computer system in the above embodiments; or it can exist separately and be a computer-readable storage medium not assembled into the device. The computer-readable storage medium stores one or more programs, and the one or more programs are used by one or more processors to execute the methods described in the present disclosure.

[0120] The above description is only a preferred embodiment of the present disclosure and an explanation of the applied technical principles. Those skilled in the art should understand that the scope of the invention involved in the present disclosure is not limited to the technical solutions formed by the specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the inventive concept. For example, technical solutions formed by mutually replacing the above features with (but not limited to) technical features having similar functions disclosed in the present disclosure.

Claims

1. An astronomical image restoration method, characterized in that, The method includes the following steps: S1. Obtain the astronomical image data to be recognized, and perform serialization processing on the astronomical image data to be recognized to obtain one-dimensional vector observation data of the celestial body to be recognized; S2. Based on the vector observation data of the celestial body to be recognized, obtain the initial reconstructed target source vector through the conjugate gradient iteration method; S3. When the maximum value of the elements in the initial reconstructed target source vector is greater than or equal to the first vector element threshold, obtain the reconstructed target source vector according to the initial reconstructed target source vector, and execute step S4. When the maximum value of the elements in the initial reconstructed target source vector is less than the first vector element threshold, obtain the reconstructed target source vector according to the initial reconstructed target source vector, and execute step S7; S4. Add the product of the maximum value of the elements in the reconstructed target source vector and the first parameter factor to the model target source vector, and remove the convolution of the maximum value of the elements in the reconstructed target source vector and the point spread function from the vector observation data of the celestial body to be recognized. The length of the model target source vector is the same as the length of the reconstructed target source vector; S5. Use the residual vector between the vector observation data of the celestial body to be recognized and the model observation data as the input value of the conjugate gradient iteration method for iteration to obtain the updated reconstructed target source vector. The model observation data is the convolution of the reconstructed target source vector and the point spread function; S6. When the maximum value of the elements in the updated reconstructed target source vector is greater than or equal to the second vector element threshold, execute step S4. When the maximum value of the elements in the updated reconstructed target source vector is less than the second vector element threshold, execute step S7; S7. Use the residual vector as the input value of the conjugate gradient iteration method for iteration, and set the values of the negative elements in the reconstructed target source vector that are negative to 0 during the iteration process to obtain the iterated reconstructed target source vector; S8. Add the product of the maximum value of the elements in the iterated reconstructed target source vector and the second parameter factor to the model target source vector, and remove the convolution of the maximum value of the elements in the iterated reconstructed target source vector and the point spread function from the residual vector; S9. When the maximum value of the elements in the iterated reconstructed target source vector is greater than or equal to the third vector element threshold, execute step S8. When the maximum value of the elements in the iterated reconstructed target source vector is less than the third vector element threshold, sum the model target source vector and the iterated reconstructed target source vector to obtain the final target source vector; S10. Obtain the two-dimensional reconstructed target image based on the final target source vector.

2. The astronomical image restoration method according to claim 1, characterized in that, The method further includes: Obtain the final target source evaluation parameter, where the final target source evaluation parameter includes at least one of the final target source residual, the chi-square value of the final target source vector, the mean value of the final target source vector, and the mean square deviation of the final target source vector. The final target source residual is the difference between the vector observation data of the celestial body to be recognized and the final target source vector; Determine the credibility of the two-dimensional reconstructed target image according to the final target source evaluation parameter.

3. The astronomical image restoration method according to claim 1, characterized in that, The value ranges of the first parameter factor and the second parameter factor are both [0, 1].

4. The astronomical image restoration method according to claim 1, characterized in that, Before the step S3, the method further includes: Obtaining a model target source vector with initial values all being 0 and a length the same as that of the reconstructed target source vector.

5. The astronomical image restoration method according to claim 1, characterized in that, The first vector element threshold , where is the standard deviation of the reconstructed target source vector.

6. The astronomical image restoration method according to claim 1, characterized in that, The second vector element threshold , where is the standard deviation of the residual vector.

7. The astronomical image restoration method according to claim 1, characterized in that, The method further includes: Displaying the two-dimensional reconstructed target image.

8. An electronic device, characterized in that, Comprising a memory and a processor; wherein, the memory is used for storing one or more computer instructions, and the one or more computer instructions are executed by the processor to implement the method steps described in any one of claims 1-7.

9. A readable storage medium, characterized in that, Stored thereon are computer instructions which, when executed by a processor, implement the method steps described in any one of claims 1-7.

10. A computer program product, characterized in that, Comprising computer instructions which, when executed by a processor, implement the method steps described in any one of claims 1-7.

Images