Method and device for identifying transient sources based on spatial index multi-telescope image correlation

Abstract

This application provides a transient source identification method, apparatus, device, storage medium, and program product, which can be applied to the field of astronomical information technology. The method includes: projecting an acquired observation image onto a HEALPix grid at the target level to obtain multiple tiles; performing pixel-by-pixel difference between tiles at the same location and reference tiles, identifying at least one tile from the multiple tiles whose average difference pixel value is greater than a preset average value as the difference tile, where the average difference pixel value represents the average difference between the tile and the corresponding reference tile; based on the coordinates of the difference tiles, using spatial indexing to obtain all image blocks covering the difference tiles and determining transient source candidates, wherein all image blocks come from various types of astronomical telescopes; and performing photometry on the image blocks corresponding to the transient source candidates to determine the target transient source. This method can quickly and accurately identify transient sources.

Description

Technical Field

[0001] This application relates to the field of astronomical information technology, and more specifically to a method and apparatus for identifying transient sources associated with multi-telescope images based on spatial indexing. Background Technology

[0002] Due to differences in observation bands, sampling rates, and fields of view among different astronomical telescopes, the images obtained by observing the same celestial object through different telescopes vary significantly in terms of pixel scale, image size, and projection method. These differences make it impossible to directly process the images captured by each telescope at the pixel level, thus hindering the rapid identification of transient sources.

[0003] To address the aforementioned issues in image processing prior to transient source identification, all images captured by various astronomical telescopes are typically projected onto the same equilateral coordinate system. The images are then processed within this system. However, this projection method presents several problems: image size increases linearly with the sky area, making it unsuitable for processing images covering the entire sky; furthermore, for the same celestial object, projecting images from different telescopes results in a lack of correlation between the projected images. This not only hinders rapid processing of images from various telescopes, thus affecting the efficiency of transient source identification, but also prevents multi-dimensional joint analysis of the correlation information between images from different telescopes, impacting the accuracy of transient source identification. Summary of the Invention

[0004] In view of the above problems, this application provides a method and apparatus for transient source identification based on spatial indexing of multi-telescope image association to improve the accuracy of transient source identification.

[0005] According to a first aspect of this application, a transient source identification method is provided, comprising: projecting an acquired observation image onto a HEALPix grid at a target level to obtain multiple tiles, wherein each tile corresponds to an image block, and the image blocks constitute the observation image, and the target level is determined according to the type of astronomical telescope used to capture the observation image; performing a pixel-by-pixel difference operation between tiles located at the same position and a reference tile, and determining at least one tile from the multiple tiles whose average difference pixel value is greater than a preset average value as the difference tile, wherein the average difference pixel value represents the pixel value of the tile and the corresponding reference tile. The average difference is obtained by projecting historical observation images onto the HEALPix grid. These historical observation images are obtained by the astronomical telescopes that captured the observation images at the same celestial region at the historical observation time. Based on the coordinates of the differential tiles, all image patches covering the differential tiles are obtained using spatial indexing, and transient source candidates are determined. The spatial indexing is used to obtain all image patches covering any tile based on the coordinates of any tile. All image patches come from various types of astronomical telescopes. Photometric measurements are performed on the image patches corresponding to the transient source candidates to determine the target transient source from the transient source candidates.

[0006] According to an embodiment of this application, when an observation image is acquired, the spatial index is updated based on multiple tiles so that all observation images covering the differential tiles can be obtained using the updated spatial index: parent tiles are generated for each of the multiple tiles at the target level by downsampling, wherein the level of the parent tile is lower than the target level, and the image patch corresponding to the parent tile and the image patch corresponding to the tile are obtained by two different types of astronomical telescopes photographing the same celestial region; the relationships between the multiple tiles, the parent tiles corresponding to the multiple tiles, and the parent-child tiles in the existing spatial index are associated to obtain the updated spatial index.

[0007] According to an embodiment of this application, based on the coordinates of the differential tiles, all image blocks covering the differential tiles are obtained using a spatial index, and transient source candidates are determined. This includes: recursively querying the coordinates of the differential tiles in the spatial index to obtain the first image block corresponding to the differential tile, the second image block corresponding to the parent tile of the differential tile, and the third image block corresponding to the child tile of the differential tile; extracting celestial information from the first image block, the second image block, and the third image block respectively, and using the celestial bodies represented by the celestial information as transient source candidates.

[0008] According to embodiments of this application, the transient source identification method further includes: if there is overlap between all image blocks covering the differential tiles, converting the coordinates of each tile pixel of the differential tiles into celestial coordinates; performing a first coordinate transformation on the celestial coordinates using a coordinate transformation matrix to obtain the coordinates of the image pixels corresponding to the tile pixels in the corresponding image block, wherein the coordinate transformation matrix indicates the mapping relationship between the celestial coordinates and the coordinates of the image pixels; interpolating the pixels around the image pixels to obtain the interpolated image blocks covering the differential tiles; and performing de-overlap processing on the interpolated image blocks based on an image de-overlap strategy, wherein the image de-overlap strategy includes at least one of the following: an arithmetic mean strategy for image pixels in the overlapping region, a distance-weighted strategy, and an image quality strategy. The arithmetic mean strategy refers to using the arithmetic mean of the image pixels in the overlapping region as the image pixel value of the overlapping region. The distance-weighted strategy refers to assigning weights to each image pixel based on the distance between each image pixel of the interpolated image block and the edge of the interpolated image block. The image quality strategy refers to assigning weights to each image block in the overlapping region based on the image quality of each image block in the overlapping region.

[0009] According to an embodiment of this application, converting the coordinates of each tile pixel of a differential tile into celestial coordinates includes: converting the coordinates of the tile pixel into two-dimensional coordinates using a binary bit interleaving method, wherein the two-dimensional coordinates indicate the position of the tile pixel within its respective differential tile; and converting the two-dimensional coordinates into celestial coordinates using a second coordinate transformation based on the two-dimensional coordinates and the number of the differential tile to which the tile pixel belongs, wherein the number indicates the position of the differential tile to which the tile pixel belongs among multiple tiles.

[0010] According to an embodiment of this application, determining a target transient source from transient source candidates includes: performing inverse variance weighted calculation on the photometric results of each image block corresponding to the same transient source candidate to obtain a comprehensive photometric result of each transient source candidate; and taking the transient source candidate corresponding to the comprehensive photometric result that is less than a preset threshold as the target transient source.

[0011] A second aspect of this application provides a transient source identification device, comprising: an image projection module for projecting an acquired observation image onto a HEALPix grid at a target level to obtain multiple tiles, wherein each tile corresponds to an image block, and the image blocks constitute the observation image, and the target level is determined according to the type of astronomical telescope used to capture the observation image; and a differential tile acquisition module for performing pixel-by-pixel difference operations on tiles located at the same position and reference tiles, and determining at least one tile from the multiple tiles whose average differential pixel value is greater than a preset average value as a differential tile, wherein the average differential pixel value represents the average difference between the pixel values ​​of the tile and the corresponding reference tile. The reference tile is obtained by projecting historical observation images onto the HEALPix grid. The historical observation images are obtained by the astronomical telescopes that took the observation images at the same celestial region at the historical observation time. The transient source candidate determination module is used to obtain all image blocks covering the differential tiles and determine the transient source candidates based on the coordinates of the differential tiles using spatial indexing. The spatial indexing is used to obtain all image blocks covering any tile based on the coordinates of any tile. All image blocks come from various types of astronomical telescopes. The target transient source identification module is used to perform photometry on the image blocks corresponding to the transient source candidates and determine the target transient source from the transient source candidates.

[0012] A third aspect of this application provides an electronic device comprising: one or more processors; and a memory for storing one or more computer programs, wherein the one or more processors execute the one or more computer programs to implement the steps of the method described above.

[0013] A fourth aspect of this application also provides a computer-readable storage medium having a computer program or instructions stored thereon, which, when executed by a processor, implement the steps of the above-described method.

[0014] The fifth aspect of this application also provides a computer program product, including a computer program or instructions that, when executed by a processor, implement the steps of the above-described method.

[0015] The embodiments of this application have the following beneficial effects: Projecting the observation image onto the HEALPix grid can unify the data format of observation images from different astronomical telescopes, which is beneficial for subsequent direct processing of tile data using artificial intelligence models; since the spatial index associates images of the same celestial region taken by different astronomical telescopes with their corresponding tiles, images of the same celestial region taken by various types of astronomical telescopes that correspond to the differential tile can be found through the spatial index, thereby quickly obtaining all images taken by different types of astronomical telescopes that can cover the differential tile, improving the efficiency of transient source identification; from all these found images, transient source candidates are determined to avoid missing transient sources, and by performing photometry on the images corresponding to the transient source candidates, the final transient source can be identified more precisely from the transient source candidates, further improving the accuracy of transient source identification. Attached Figure Description

[0016] The above-mentioned contents, other objects, features and advantages of this application will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0017] Figure 1 The illustrations depict application scenarios of transient source identification methods, apparatuses, devices, media, and program products according to embodiments of this application.

[0018] Figure 2 A flowchart illustrating a transient source identification method according to an embodiment of this application is shown schematically.

[0019] Figure 3 This illustration schematically shows a data processing flow diagram of a central processing unit and a graphics processing unit according to an embodiment of this application;

[0020] Figure 4 This illustration schematically shows a spatial index update method according to an embodiment of this application;

[0021] Figure 5 This illustration schematically shows a process flow diagram for image de-overlapping according to an embodiment of this application;

[0022] Figure 6 This illustration schematically shows a method for obtaining all image blocks covering the celestial coordinates based on celestial coordinates according to an embodiment of this application;

[0023] Figure 7 This schematic diagram illustrates a structural block diagram of a transient source identification device according to an embodiment of this application;

[0024] Figure 8 A block diagram schematically illustrates an electronic device suitable for implementing a transient source identification method according to an embodiment of this application. Detailed Implementation

[0025] The embodiments of this application will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of this application. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of this application for ease of explanation. However, it will be apparent that one or more embodiments may be implemented without these specific details. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concepts of this application.

[0026] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0027] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0028] When using expressions such as "at least one of A, B and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (e.g., "a system having at least one of A, B and C" should include, but is not limited to, a system having A alone, a system having B alone, a system having C alone, a system having A and B, a system having A and C, a system having B and C, and / or a system having A, B and C, etc.).

[0029] This application provides a transient source identification method. It involves projecting an observation image from an astronomical telescope onto a corresponding level of HEALPix grid to obtain multiple tiles, each representing an image patch of the observation image. Then, it performs pixel-by-pixel difference calculations between tiles at the same location and a reference tile to obtain differential tiles with significant pixel deviations. A spatial index is used to recursively query these differential tiles to obtain all image patches that cover them, i.e., to acquire images taken by different types of astronomical telescopes in the corresponding celestial region of the differential tile. Based on the celestial information contained in these image patches, transient source candidates are obtained. Photometric analysis is performed on the images corresponding to the transient source candidates to obtain the final transient source. This method improves the processing speed of observation image data, thereby enabling rapid and accurate identification of transient sources.

[0030] Figure 1 The illustration shows an application scenario diagram of the transient source identification method, apparatus, device, medium, and program product according to embodiments of this application.

[0031] like Figure 1 As shown, application scenario 100 according to this embodiment may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 serves as a medium for providing a communication link between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired or wireless communication links, or fiber optic cables, etc.

[0032] Users can use the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 via the network 104 to receive or send messages, etc. The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be different types of astronomical telescopes.

[0033] Various communication client applications can be installed on the first terminal device 101, the second terminal device 102, and the third terminal device 103, such as image processing applications, data analysis applications, celestial body query applications, etc. (for example only).

[0034] The first terminal device 101, the second terminal device 102, and the third terminal device 103 can be various electronic devices with displays and support web browsing, including but not limited to smartphones, tablets, laptops, and desktop computers.

[0035] Server 105 can be a server that provides various services, such as a backend management server that supports websites browsed by users using the first terminal device 101, the second terminal device 102, and the third terminal device 103 (this is just an example). The backend management server can analyze and process data such as received user requests (e.g., process data from observed images), and feed back the processing results (e.g., obtain transient source information based on the data processing results of observed images) to the terminal devices.

[0036] It should be noted that the transient source identification method provided in this application embodiment can generally be executed by server 105. Correspondingly, the transient source identification device provided in this application embodiment can generally be located in server 105. The transient source identification method provided in this application embodiment can also be executed by a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105. Correspondingly, the transient source identification device provided in this application embodiment can also be located in a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105.

[0037] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0038] The following will be based on Figure 1 The described scene, through Figures 2-6 The transient source identification method according to the embodiments of this application will be described in detail.

[0039] Figure 2 A flowchart illustrating a transient source identification method according to an embodiment of this application is shown.

[0040] like Figure 2 As shown, the transient source identification method 200 of this embodiment includes operations S210 to S240.

[0041] In operation S210, the acquired observation image is projected onto the HEALPix grid of the target level to obtain multiple tiles, where each tile corresponds to an image block, and the image blocks constitute the observation image. The target level is determined according to the type of astronomical telescope that took the observation image.

[0042] In operation S220, a pixel-by-pixel difference operation is performed between the tiles located at the same position and the reference tile. At least one tile with a difference pixel mean greater than a preset mean is determined from multiple tiles and used as the difference tile. Here, the difference pixel mean represents the average difference between the pixel values ​​of the tile and the corresponding reference tile. The reference tile is obtained by projecting historical observation images onto the HEALPix grid. The historical observation images are obtained by the astronomical telescope that took the observation images at the same celestial region at the historical observation time.

[0043] In operation S230, based on the coordinates of the differential tiles, all image patches covering the differential tiles are obtained using spatial indexing, and transient source candidates are determined. The spatial indexing is used to obtain all image patches covering any tile based on the coordinates of any tile. All image patches come from various types of astronomical telescopes.

[0044] In operation S240, the image block corresponding to the transient source candidate is metered, and the target transient source is determined from the transient source candidate.

[0045] In some embodiments, during operation S210, hierarchical equal-area isolatitude pixelation (HEALPix) meshing can divide the entire celestial sphere (or sphere) into layers according to resolution. The entire celestial sphere can be divided into different numbers of tiles at different layers. Tiles at the same layer have equal mesh areas, and the mesh is distributed equidimensionally on the sphere. HEALPix meshing can divide observation images from different astronomical telescopes into image patches. These images are projected onto different layers, with each image patch corresponding to one tile, or one tile representing one image patch. Image patches corresponding to tiles at the same layer have equal areas, and each layer... Each tile in the observation image file has the same pixel size; for example, each tile could be 512×512 pixels. Further, by parsing the header information in the observation image file, the target layer of the observation image can be determined. The observation image is then projected onto the HEALPix grid of the target layer to obtain multiple tiles. Finally, the mapping relationship between the tiles and image blocks is obtained. By directly parsing the header information in the observation image file without loading the pixel data of the observation image, the mapping relationship between tiles and image blocks can be obtained, thus avoiding the need to decompress all the data of the observation image, improving data processing efficiency, and reducing data processing time by an order of magnitude.

[0046] Images from different types of astronomical telescopes are projected onto HEALPix grids of different target levels. The target level indicates the area of ​​the sky covered by a single tile. For example, in the first target level, the entire celestial sphere is divided into *a* tiles, each covering a certain area. In the second target level, the entire celestial sphere is divided into *b* tiles, where *a* and *b* are integers greater than 1, and *a* is less than *b*. Therefore, the area of ​​the sky covered by a single tile in the second target level is smaller than that in the first target level. In other words, the first target level has a larger scale, lower pixel resolution, and a lower level order than the second target level. For the same celestial region, each tile in the first target level has a coarser granularity compared to each tile in the second target level. If the level order ranges from 0 to 11, the pixel angular resolution can range from 58.6° to 0.10″.

[0047] In some embodiments, during operation S210, when the observed image is projected onto the HEALPix grid, the median deviation of the pixel values ​​of the corresponding image pixels and tile pixels is only... 87% of the pixel values ​​have a deviation of less than 0.01, thus ensuring the accuracy of the tile data.

[0048] In some embodiments, during operation S210, the HEALPix grid-based projection method can uniformly convert observation images from any astronomical telescope, any projection method, and any pixel size into tiles with the same pixel size and coordinate system, thereby eliminating heterogeneity and standardizing image data. This facilitates subsequent data processing of the observation images. For example, a large model can be used to directly parse the tile data to identify transient sources without needing to crop or scale the tile data corresponding to observation images from different astronomical telescopes, thus improving data processing speed. Furthermore, the large model can be an artificial intelligence model with reasoning capabilities suitable for the astronomical field.

[0049] In some embodiments, during operation S220, for example, various astronomical telescopes can obtain multiple historical observation images of a certain celestial region at a historical observation time. According to the above projection method, the multiple historical observation images are projected onto the HEALPix grid of the corresponding level to obtain multiple reference tiles. Each level has a reference tile. Based on the target layer corresponding to the observation image captured by the current astronomical telescope, a reference tile is obtained at that target layer. For that target layer, a pixel-by-pixel difference operation is performed between the tile at the same location and the reference tile. For example, if tile A and reference tile B are located at the same location, and the tile pixels in tile A correspond one-to-one with the tile pixels in reference tile B, the pixel value deviation between the corresponding two tile pixels is calculated. Based on the pixel value deviation of each tile pixel relative to the corresponding reference pixel value in reference tile B, the mean difference pixel value of tile A is calculated. This mean difference pixel value indicates the degree of deviation between the pixel values ​​of tile A and reference tile B. If the mean difference pixel value of tile A is greater than a preset mean, it indicates that the celestial object represented by tile A may be a transient source, and tile A is used as the difference tile. For tiles at other locations, a pixel-by-pixel difference operation is also performed on the tile and its reference tile to obtain difference tiles. Multiple difference tiles may exist at the same target layer.

[0050] In some embodiments, during operation S220, tile A and reference tile B at the same location are subjected to pixel-by-pixel difference calculation according to the following formula: , This represents the pixel value of the p-th pixel in tile A. This represents the pixel value of the p-th pixel in reference tile B, where p is an integer greater than 1. Indicates the scaling factor, via Tile A and reference tile B can be aligned to the same photometric scale to compensate for differences in zero point, depth of exposure, system response, and transparency. This represents the pixel value of the p-th tile pixel in the differential tile.

[0051] In some embodiments, during operation S220, χ² tests can be performed on the tile pixels of the tile and its reference tile at different observation times, or on the tile pixels of the tile and its reference tile at different observation bands. The χ² test is used to calculate the degree of difference between the tile pixels of the tile from different observation times or different observation bands and the tile pixels of its reference tile. Furthermore, the tile data (including the pixel values ​​and tile positions of the tile pixels), hierarchical information (including the relationship between tiles and hierarchies), observation time, and observation band information obtained based on the transient source identification method of this application can also be used as training samples for the light curve anomaly detection model or the variable source classification model. For example, historical observation images are divided into multiple groups according to the observation time, and the historical observation images of each group are projected onto the corresponding layer to obtain multiple tiles. The tiles are also divided into multiple groups according to the observation time of the images. The tiles of adjacent observation times in each group are differentially analyzed pixel by pixel to determine whether the tiles have changed, thus obtaining multiple tile pairs. Each tile pair includes two tiles of adjacent observation times. The observation image corresponding to the known transient source is mapped onto the HEALPix grid of the layer where the tile pair is located to obtain the tile corresponding to the known transient source. Each tile pair is labeled using the tile corresponding to the known transient source. The label data, tile pair, observation time, and observation band are used as training samples. Among them, the tile that has changed significantly in the tile pair and matches the tile of the known transient source can be used as a positive sample, and the tile that has not changed significantly in the tile pair can be used as a negative sample.

[0052] In some embodiments, taking observation images taken by the Einstein Probe (EP) wide-field X-ray telescope as an example, each observation image has a pixel size of 625×625 pixels, a pixel scale of 131″ / pixel, and a field of view of approximately 23°×23°. 1000 observation images total 1.5GB. The images are designated as the third level, and these 1000 observation images are projected onto a third-level HEALPix grid, resulting in 768 tiles. The mapping relationship between the tiles and image blocks is obtained through the CPU. Due to the extremely large field of view of the EP, each observation image covers a large number of tiles, with an average of 500 to 800 observation images per tile. The images exhibit high overlap, similar to overlay. Furthermore, the process of obtaining differential tiles from the 768 tiles takes only 884 milliseconds, demonstrating extremely fast data processing speed. After updating the spatial index according to the spatial index update method of this application, obtaining all images covering the region corresponding to any celestial coordinate using the spatial index takes only 8.53 seconds, an 82-fold speed improvement compared to conventional methods. In addition, the 768 acquired tiles are all standardized 512 pixel × 512 pixel tiles, and each tile can be directly used as input to the X-ray transient source detection model without further data processing, significantly improving the efficiency of transient source detection.

[0053] In some embodiments, taking the observation images captured by the Dark Energy Spectroscopic Instrument (DESI) r-band optical telescope as an example, each observation image is 3600×3600 pixels, with a pixel scale of 0.262″ / pixel and a field of view of approximately 1°. The total size of 100 observation images is 1.2GB. The target layer is designated as the 9th layer. These 100 observation images are projected onto the HEALPix grid of the 9th layer, resulting in 8891 tiles. It takes 30.21 seconds to obtain differential tiles from these tiles, which is 1.8 times faster than the general method. The pixel scale of the 8891 tiles is 0.80″ / pixel. These tiles can cover the entire sky area. The wavelength range that the r-band optical telescope can detect includes 550nm~800nm.

[0054] In some embodiments, this application can also rapidly and accurately detect discovered transient sources. For example, if a transient source is observed in a certain area of ​​the sky through an astronomical telescope, the transient source identification method of this application can be used to further determine the type of the transient source and verify whether the transient source actually exists. Specifically, a reference tile is first constructed using tens of thousands of historical observation images from DESI, which takes about an hour to complete. Then, newly acquired observation images are projected onto the target layer to obtain multiple tiles. The tiles at the target layer and the reference tiles are then differentially divided pixel by pixel to obtain differential tiles. Transient source candidates are determined based on the differential tiles. Based on the coordinates of the differential tiles, all observation images that can cover the differential tiles and their parent and child tiles are obtained through spatial indexing. Photometric measurements are performed on these observation images, and the photometric results are combined with the multi-band energy spectrum distribution to detect the final transient source from the transient source candidates. The time required to obtain a transient source candidate from a new observation image is on the order of seconds, and the time required to obtain a photometric result from a new observation image is on the order of minutes.

[0055] According to embodiments of this application, the observed image is projected onto the HEALPix grid. This method of converting observed images from any astronomical telescope into tiles unifies the data format of observed images from different astronomical telescopes, which is beneficial for subsequent direct processing of tile data using artificial intelligence models. The projected tile is then pixel-by-pixel differentiated from the corresponding reference tile to obtain the differential tile. All images covering the differential tile are then searched in the spatial index to determine transient source candidates. The spatial index allows for rapid retrieval of the image corresponding to the differential tile, as it encompasses images of the same celestial region taken by different astronomical telescopes and their corresponding tiles. Since they are linked, the spatial index can be used to find images of celestial regions that correspond to the differential tile, taken by various types of astronomical telescopes. This allows for the acquisition of all images that cover the differential tile, taken by different types of telescopes. From these images, transient source candidates are identified to avoid missing any. By performing photometry on the images corresponding to the transient source candidates, the final transient source can be identified more precisely, further improving the accuracy of transient source identification. Using the spatial index to find all images covering the differential tile enables rapid searching, thereby improving the efficiency of transient source identification.

[0056] In wide-field X-ray sky surveys, this transient source identification method has a significant advantage in detecting faint sources. Taking the example of repeatedly observing the same sky area at different times using the Einstein Probe or a wide-field X-ray telescope to obtain highly overlapping images, these images refer to multiple images with a large overlap. During observations using the Einstein Probe or a wide-field X-ray telescope, extensive exposures to the sky area create deep coverage within a limited area. Many faint sources and slowly brightening transient sources below the single-observation detection threshold often only become apparent after multi-epoch observation data superposition (i.e., superimposing observation images of the same sky area obtained at different observation times) or after differentiating the highly overlapping images with a deep reference image. The deep reference image is a baseline observation image obtained by superimposing a large amount of historical observation data. The transient source identification method of this application can project multi-epoch observation data onto the corresponding target level to obtain the tiles corresponding to each epoch observation data, thereby unifying the multi-epoch observation data into tile data. Then, the tiles and reference tiles are subjected to pixel-by-pixel difference operation to obtain difference tiles. Then, all image blocks covering this difference tile are obtained through spatial indexing. Based on all the obtained image blocks, the target transient source is further determined. Thus, with the support of spatial indexing, efficient overlay, difference and candidate backtracking measurement operations are completed, so that the effective signal of the faint target can be accumulated, while avoiding the high computational overhead caused by traditional large field-of-view data overall reprojection and full processing. By performing faint source identification on 1000 highly overlapping observation images using the transient source identification method described in this application, it was found that the time required for faint source identification is shorter than that of general faint source identification methods, reduced from 707.73 seconds to 8.56 to 10.09 seconds. The speedup ratio can reach 70.1 to 82.7 times, with a median speedup ratio of approximately 75.9 times. The speedup ratio indicates how many times the time required to identify a faint source using the transient source identification method of this application is reduced compared to the general faint source identification method; a higher speedup ratio indicates a shorter time required to identify a faint source using the transient source identification method of this application. Therefore, the transient source identification method proposed in this application can not only significantly improve the ability to identify faint sources from large field-of-view X-ray repeated observation data, but also shorten the identification time, improve identification efficiency and accuracy, and thus allow for rapid follow-up work after the faint source is identified.

[0057] Figure 3 The schematic diagram illustrates the data processing flow of the central processing unit and the graphics processing unit according to an embodiment of this application.

[0058] This application employs a hybrid execution method combining the Central Processing Unit (CPU) and the Graphics Processing Unit (GPU) for transient source identification, significantly improving data processing speed and thus enhancing transient source identification efficiency. For example... Figure 3 As shown, the central processing unit 301 executes operations S310 to S330 and operation S350, and the graphics processing unit 302 executes operations S340 and operations S360 to S3100.

[0059] In operation S310, the header information in the observed image file is parsed to determine the target level, and the observed image is projected onto the HEALPix grid of the target level to obtain multiple tiles.

[0060] In operation S320, the mapping relationship between image patches and tiles in the observed image is obtained.

[0061] During operation of S330, the above mapping relationship, the observed image, and the header information in the observed image file are transmitted to the graphics processor.

[0062] The graphics processor 302 stores the mapping relationship between image blocks and tiles 3021, the observed image 3022, and the header information 3023.

[0063] In operation S340, multiple adjacent tiles are downsampled to generate a parent tile.

[0064] In the S350 operation, associate the parent and child tiles and update the spatial index.

[0065] In operation S360, based on the coordinates of the differential tiles, all image blocks covering the differential tiles are obtained using the updated spatial index. The differential tiles are obtained by performing pixel-by-pixel difference operations on the tiles at the same location and the reference tile.

[0066] In operation S370, the coordinates of the image pixels in the corresponding image blocks are obtained based on the binary bit interleaving method and the coordinate transformation matrix in the header information, and according to the tile pixel coordinates of the differential tiles and their parent and child tiles.

[0067] In operation S380, the pixels surrounding the image pixels in each image block are interpolated to obtain multiple image blocks that cover the differential tile and its parent and child tiles after interpolation.

[0068] In operation S390, overlapping regions in multiple interpolated image blocks are de-overlapped.

[0069] In operation S3100, the light is measured on multiple image blocks after de-overlap processing to obtain the target transient source.

[0070] In some embodiments, during operation S330, the pixel data of the observed image can also be transferred to the global memory of the graphics processor in the form of a continuous floating-point array. During the process of transferring the mapping relationship, the observed image, the header information in the observed image file and the pixel data of the observed image to the graphics processor memory, if the data exceeds the graphics processor memory, a streaming caching method is used for caching. For example, cached data that has not been used for a long time is removed from the graphics processor memory, and other data to be cached is stored in the graphics processor memory.

[0071] In some embodiments, during operation S340, for example, the four adjacent tiles in each layer are downsampled to generate corresponding parent tiles for these four tiles. The level of the parent tile is lower than the level of the four adjacent tiles. Specifically, the parent tiles are generated using a 2×2 mean downsampling method. That is, the four adjacent tiles are arranged in a 2×2 pattern. The mean value of the tile pixels in each row and column of the tiles is calculated to obtain the mean value of the four tile pixels. The mean value of the four tile pixels is used as the tile pixels of the parent tile to obtain the parent tile corresponding to the four adjacent tiles. Furthermore, after generating the parent tile, based on the mapping relationship between image blocks and tiles, the central processing unit can associate each parent and child tile with the corresponding image blocks. This allows the spatial index to associate each parent and child tile with the corresponding image blocks. Therefore, based on the tile number, the spatial index can be used to find all image blocks covering the tile and its parent and child tiles. Since the image blocks covering the parent tile must cover the tile, and the image blocks covering the child tile cover a part of the tile, it can also be said that based on the tile number, the spatial index can be used to find all image blocks covering the tile.

[0072] In some embodiments, during operation S360, since the spatial index associates the tile and its corresponding image block, the parent tile and its corresponding image block, and the child tile and its corresponding image block, all image blocks covering the differential tile can be obtained using the spatial index based on the coordinates of the differential tile. These image blocks are obtained by dividing observation images from different astronomical telescopes into HEALPix grids.

[0073] In some embodiments, during operation S370, the header information includes a coordinate transformation matrix, the image pixel coordinates of the observed image, and the pixel size of the observed image.

[0074] In some embodiments, during operation S380, bilinear interpolation is used to interpolate image pixels, which can skip invalid image pixels and improve interpolation accuracy.

[0075] In some embodiments, each thread in the GPU is responsible for acquiring the coordinates of one image pixel. That is, the data processing within each differential tile is executed in parallel. For example, each differential tile has 512×512 tile pixels, and there are 512 thread blocks. Each thread block includes 512 thread units, and one thread unit is responsible for processing one tile pixel. One thread is responsible for acquiring an interpolated image block. In other words, the data processing between each differential tile is executed in parallel. For example, assuming there are F differential tiles, a total of F×512 thread blocks are started to process the data of these differential tiles, where F is an integer greater than 1.

[0076] In some embodiments, the method of using a combination of a central processing unit and a graphics processing unit to perform the above data processing operations to identify transient sources, especially for datasets with high image overlap, can significantly improve data processing speed. The GPU can complete the projection operation of 1,000 highly overlapping images in 1 second, and the acquisition operation of differential tiles in the S360 operation can also be completed in seconds. This allows for the rapid identification of transient sources such as gamma burst afterglow and tidal collapse, which are transient sources whose brightness or energy changes drastically in a very short time and then rapidly weakens or disappears, thus improving the timeliness of transient source detection.

[0077] Figure 4 A schematic diagram illustrating a spatial index update method according to an embodiment of this application is shown.

[0078] In some embodiments, when an observation image is acquired, the spatial index is updated based on multiple tiles to obtain all observation images covering the differential tiles using the updated spatial index: parent tiles are generated for each of the multiple tiles at the target level by downsampling, wherein the level of the parent tile is lower than the target level, and the image patch corresponding to the parent tile and the image patch corresponding to the tile are obtained by two different types of astronomical telescopes photographing the same celestial region; the relationships between the multiple tiles, the parent tiles corresponding to the multiple tiles, and the parent-child tiles in the existing spatial index are associated to obtain the updated spatial index.

[0079] like Figure 4As shown, assuming the target level is in the second level, the parent tiles (i.e., the first parent tile 401) are generated for the first child tile 402 to the fourth child tile 403 in the second level using a 2×2 mean downsampling method. Similarly, the parent tiles (i.e., the Mth parent tile 404) are generated for the Nth child tile 405 to the (N+3)th child tile 406 in the second level using the same 2×2 mean downsampling method. M and N are integers greater than 1. The first parent tile 401 and the Mth parent tile 404 are in the first level. The level order of the first level is less than that of the second level. The higher the level order... "Low" indicates that the granularity of the tiles at this level is coarser (or the resolution of this level is lower, and the tiles at this level can display less detail information); each child tile in the second level is a parent tile of some tiles in the third level. For example, the first child tile 402 is the parent tile of the Qth child tile 407 to the (Q+3)th child tile 408 in the third level, and the (N+3)th child tile 406 is the parent tile of the Yth child tile 409 to the (Y+3)th child tile 410 in the third level. The level order of the second level is lower than that of the third level, and Q and Y are integers greater than 1.

[0080] Since the existing spatial index 415 already associates some parent-child tiles and their corresponding image patches—for example, the parent tile of the R-th child tile 411 to the (R+3)-th child tile 412 in the i-th level is a tile in the (i-1)-th level, and the parent tile of the U-th child tile 413 to the (U+3)-th child tile 414 in the i-th level is another tile in the (i-1)-th level, where i, R, and U are integers greater than 1, and the level order of the i-th level is greater than that of the third level—the parent-child relationship between the tiles in the existing spatial index and the tiles in the target level can be obtained through 2×2 mean downsampling. That is, the tiles in the target level are associated with the tiles in the existing spatial index. Since each tile and its corresponding observed image... There are mapping relationships between image blocks, so we can further associate tiles in the target level, tiles in the existing spatial index, and the image blocks corresponding to each of these tiles to obtain an updated spatial index. In other words, based on the updated spatial index, we can quickly find the parent and child tiles of any tile, as well as the image blocks corresponding to that tile and its parent and child tiles. These image blocks are obtained from images of the same celestial region taken by different astronomical telescopes. That is, the image block corresponding to the tile was taken by one astronomical telescope, the image block corresponding to the parent tile was taken by another astronomical telescope, and the image block corresponding to the child tile was taken by yet another astronomical telescope. Through the spatial index, we can quickly obtain images of the same celestial region taken by different astronomical telescopes.

[0081] In some embodiments, for newly acquired observation images, the spatial index is updated using an incremental update mode. That is, only the tiles corresponding to the newly acquired observation image and the relationships between tiles associated with that tile are updated in the spatial index. In this way, second-level updates can be achieved in new observation scenarios.

[0082] According to the embodiments of this application, by dynamically generating low-level parent tiles and establishing associations with high-level child tiles, spatial index updates of image data of the same celestial region under different astronomical telescopes and different resolutions are realized. This enables rapid retrieval of multi-source astronomical observation data, which is beneficial to improving the data processing speed in the field of astronomical data processing. It also enables data traceability and improves the accuracy of data processing.

[0083] In some embodiments, based on the coordinates of the differential tiles, all image blocks covering the differential tiles are obtained using a spatial index and transient source candidates are determined, including: recursively querying the coordinates of the differential tiles in the spatial index to obtain the first image block corresponding to the differential tile, the second image block corresponding to the parent tile of the differential tile, and the third image block corresponding to the child tile of the differential tile; extracting celestial information from the first image block, the second image block, and the third image block respectively, and using the celestial bodies represented by the celestial information as transient source candidates.

[0084] For example, the recursive query process is as follows: Based on the coordinates of the differential tile, its tile number can be determined. Then, based on the tile number, the tile numbers of its parent and child tiles can be determined. Further querying upwards based on the tile number yields the subdirectory where the differential tile is located. The subdirectory determines the layer of the differential tile, and the corresponding image block within that layer is obtained. Based on the parent tile number, the subdirectory of the parent tile is determined, and the layer of the parent tile is obtained, yielding the corresponding image block within that layer. Finally, based on the child tile number, the subdirectory of the child tile is determined, and the layer of the child tile is obtained, yielding the corresponding image block within that layer. Furthermore, each 10,000 tiles can be divided into a subdirectory. The storage structure of tile information in the spatial index can be: Norder{k} / Dir{d} / Npix{p} / , where p represents the tile number, d represents the subdirectory number, k represents the level order, Npix represents the index for retrieving the tile number, Dir represents the index for retrieving the subdirectory number, and Norder represents the index for retrieving the level order. During the recursive query process, only tiles with data are indexed, and all tiles are no longer traversed, which can greatly reduce invalid indexes. For sparse sky areas, this can reduce invalid searches by more than 99%.

[0085] Furthermore, due to the significant differences between the differential tiles and the reference tiles, the celestial objects in the differential tiles and their parent-child tiles can be considered transient sources. After obtaining the first, second, and third image patches through spatial indexing, since these three image patches were obtained from observations of the same celestial region by different astronomical telescopes at different times, the brightness, shape, color, and position information of the celestial objects in these image patches can be extracted. By comprehensively analyzing the changes in brightness, shape, and position of the celestial objects in these three image patches, transient source candidates can be identified. For example, if the brightness of the celestial objects in these three image patches differs greatly, with a brightness variation exceeding 10 magnitudes, and the color shows a trend of changing from blue to red, the transient sources in these three image patches can be considered to be Type Ia supernovae. Alternatively, if the celestial object extracted from these three image patches is relatively dim, with a peak apparent magnitude of approximately 2.9, and a red color, and has three halos around it, the celestial object can be considered to be a Type II supernova.

[0086] According to the embodiments of this application, recursive queries using spatial indexes can find all image blocks covering differential tiles in a very short time, thereby efficiently locating all relevant images covering differential tiles, avoiding the need to search the entire image set, and improving efficiency. By extracting astronomical information from the found image blocks and determining transient source candidates, transient source candidates can be quickly and accurately screened from massive amounts of observation data, improving the timeliness and accuracy of transient source identification.

[0087] Figure 5 The illustration shows a schematic diagram of the image de-overlap processing flow according to an embodiment of the present application.

[0088] If there is overlap between all image blocks covering the differential tiles, then operations S510 to S540 are performed.

[0089] In operation S510, the coordinates of each tile pixel of the differential tile are converted to celestial coordinates.

[0090] In operation S520, the first coordinate transformation of the celestial coordinates is performed using the coordinate transformation matrix to obtain the coordinates of the image pixels corresponding to the tile pixels in the corresponding image block. The coordinate transformation matrix indicates the mapping relationship between the celestial coordinates and the coordinates of the image pixels.

[0091] In operation S530, the pixels surrounding the image pixels are interpolated to obtain the interpolated image block covering the differential tiles.

[0092] In operation S540, based on the image de-overlap strategy, the interpolated image blocks are de-overlapped. The image de-overlap strategy includes at least one of the following: an arithmetic mean strategy of the image pixels in the overlapping region, a distance-weighted strategy, and an image quality strategy. The arithmetic mean strategy refers to using the arithmetic mean of the image pixels in the overlapping region as the image pixel value of the overlapping region. The distance-weighted strategy refers to assigning weights to each image pixel based on the distance between each image pixel in the interpolated image block and the edge of the interpolated image block. The image quality strategy refers to assigning weights to each image block in the overlapping region based on the image quality of each image block in the overlapping region.

[0093] In some embodiments, during the de-overlap processing of interpolated image blocks, atomic floating-point addition accumulation can be used to merge pixel values ​​from the same location in different images into a single pixel value. That is, each thread in the GPU is responsible for processing one image pixel in an image block. When multiple threads perform floating-point addition accumulation operations on the pixel values ​​of multiple image pixels, using atomic floating-point addition accumulation ensures that each addition operation is not split, avoiding data loss caused by simultaneous read / write operations. Furthermore, the image de-overlap strategy aims to merge pixel values ​​from the same location in different images into a single pixel value, thereby achieving image de-overlap.

[0094] According to the embodiments of this application, the image pixel coordinates of the image block corresponding to the differential tile are obtained based on the tile pixel coordinates of the differential tile. Then, the image block that can cover the differential tile is reconstructed based on the neighboring pixels of the image pixel. Finally, the image block obtained by the interpolation is processed to remove the overlap by combining the image de-overlap strategy. This can achieve accurate alignment and smooth transition of pixels in the overlapping area, effectively eliminate the problems of local image distortion caused by coordinate offset and overlap redundancy when stitching overlapping image blocks, and retain the detailed information of the image corresponding to the differential tile, thereby improving the de-overlap effect of the image block.

[0095] In some embodiments, converting the coordinates of each tile pixel of a differential tile to celestial coordinates includes: converting the coordinates of the tile pixel to two-dimensional coordinates using a binary bit interleaving method, wherein the two-dimensional coordinates indicate the position of the tile pixel within its respective differential tile; and converting the two-dimensional coordinates to celestial coordinates using a second coordinate transformation based on the two-dimensional coordinates and the number of the differential tile to which the tile pixel belongs, wherein the number indicates the position of the differential tile to which the tile pixel belongs among multiple tiles.

[0096] According to the embodiments of this application, using the binary bit interleaving method for coordinate transformation not only ensures that the relative positional relationship between two adjacent coordinates remains unchanged before and after the transformation, but also eliminates the need for multiplication, division, or floating-point operations during the coordinate transformation process, which can greatly improve the transformation speed and facilitate parallel processing; based on the two-dimensional coordinates and tile numbers, the two-dimensional coordinates are converted into celestial coordinates. This method is simple to implement and does not require floating-point operations, which can avoid the accumulation of nonlinear distortion caused by floating-point operations and improve the accuracy of coordinate transformation.

[0097] In some embodiments, determining the target transient source from transient source candidates includes: performing an inverse variance weighted operation on the photometric results of each image block corresponding to the same transient source candidate to obtain the comprehensive photometric result of each transient source candidate; and taking the transient source candidate corresponding to the comprehensive photometric result that is less than a preset threshold as the target transient source.

[0098] For example, suppose there are multiple differential tiles, each corresponding to a different celestial sphere region. From the astronomical information contained in the image blocks corresponding to each differential tile and its parent and child tiles, at least one transient source candidate can be obtained. That is, at least one transient source candidate can be obtained based on a differential tile. This transient source candidate may exist in multiple image blocks. For example, this transient source candidate exists in the image blocks corresponding to a certain differential tile and its parent tile. Then, the inverse variance weighted operation is performed on the photometric results of the image blocks corresponding to the differential tile and its parent tile to obtain the comprehensive photometric result of this transient source candidate. If the comprehensive photometric result is less than a preset threshold, then this transient source candidate can be used as the final transient source (i.e., the target transient source). If multiple transient source candidates are obtained based on a differential tile, then the inverse variance weighted operation is performed on the photometric results of the image blocks corresponding to the transient source candidates of the same type to obtain the comprehensive photometric results of each type of transient source candidate. The transient source candidate whose comprehensive photometric result is less than a preset threshold is taken as the final determined transient source.

[0099] Furthermore, the metering method can be aperture metering or point spread function (PSF) metering.

[0100] Furthermore, taking the generation of a transient source candidate based on a differential tile as an example, the inverse variance weighted calculation is performed according to the following formula based on the photometric results of multiple image patches corresponding to this transient source candidate, where the multiple image patches corresponding to the transient source candidate come from different astronomical telescopes or different wavelengths:

[0101] ;

[0102] In the formula, This indicates the combined flux of the transient source candidate. This represents the flux measurement value of the j-th image patch. This represents the variance of the uncertainty in the flux measurement of the j-th image patch. It is an integer greater than 1. The photometric result of each image patch can be represented by a flux measurement value, which can characterize the radiation intensity of celestial bodies. The overall photometric result can be represented by the combined flux. After inverse variance weighting, transient sources can be identified more accurately, eliminating identification results with large photometric errors and improving the accuracy of transient source identification.

[0103] According to the embodiments of this application, inverse variance weighting is performed on each image block representing the same transient source candidate. Weights can be assigned to transient source candidates based on the comprehensive photometric results. For example, the smaller the comprehensive photometric result, the smaller the photometric error of the image block corresponding to the transient source candidate, and the higher the recognition accuracy of the transient source candidate. This reduces the false alarm rate caused by photometric errors and improves the accuracy and robustness of transient source recognition. It is very suitable for rapid and accurate screening of transient sources under different observation conditions.

[0104] Figure 6 The illustration shows a schematic diagram of a method for obtaining all image blocks covering the celestial coordinates based on celestial coordinates according to an embodiment of this application.

[0105] like Figure 6 As shown, the method for obtaining all observation images covering this celestial coordinate at different observation scales using spatial indexing based on any given celestial coordinate includes operations S610 to S620.

[0106] In operation S610, the celestial coordinates are converted to tile pixel coordinates and the tile number is determined in O(1) time.

[0107] When operating S620, based on tile numbering, all images covering the region corresponding to the celestial coordinates at different observation scales are obtained through spatial indexing. Among them, the images at different observation scales include observation images taken by different astronomical telescopes, observation images acquired at different observation bands and different observation times.

[0108] In some embodiments, during operation S610, O(1) represents constant time complexity, indicating the time or resource consumption of the algorithm execution. Regardless of the size of the data to be processed, the running time or resource consumption of the algorithm within O(1) time is a fixed value.

[0109] In some embodiments, spatial indexing operations can be completed in an extremely short time. Tests have shown that a single spatial indexing operation can be completed in 0.019 milliseconds, demonstrating extremely fast indexing speed. Spatial indexing can also be used for source tracing, which is beneficial for verifying transient sources.

[0110] Based on the above transient source identification method, this application also provides a transient source identification device. The following will be combined with... Figure 7 The device is described in detail.

[0111] Figure 7 A schematic block diagram of a transient source identification device according to an embodiment of this application is shown.

[0112] like Figure 7 As shown, the transient source identification device 700 of this embodiment includes an image projection module 701, a differential tile acquisition module 702, a transient source candidate determination module 703, and a target transient source identification module 704.

[0113] The image projection module 701 is used to project the acquired observation image onto the HEALPix grid at the target level, resulting in multiple tiles, where each tile corresponds to an image block, and the image blocks constitute the observation image. The target level is determined according to the type of astronomical telescope used to capture the observation image. In one embodiment, the image projection module 701 can be used to perform the operation S210 described above, which will not be repeated here.

[0114] The differential tile acquisition module 702 is used to perform pixel-by-pixel difference operations between tiles located at the same position and reference tiles, and to determine at least one tile whose average difference pixel value is greater than a preset average value from a plurality of tiles as the differential tile. The average difference pixel value represents the average difference between the pixel values ​​of the tile and the corresponding reference tile. The reference tile is obtained by projecting historical observation images onto a HEALPix grid. The historical observation images are obtained by the astronomical telescope that captured the observation images at the same celestial region at the historical observation time. In one embodiment, the differential tile acquisition module 702 can be used to perform the operation S220 described above, which will not be repeated here.

[0115] The transient source candidate determination module 703 is used to obtain all image blocks covering the differential tiles based on the coordinates of the differential tiles using a spatial index and to determine transient source candidates. The spatial index is used to obtain all image blocks covering any given tile based on its coordinates, and all image blocks come from various types of astronomical telescopes. In one embodiment, the transient source candidate determination module 703 can be used to perform the operation S230 described above, which will not be repeated here.

[0116] The target transient source identification module 704 is used to meter the image block corresponding to the transient source candidate and determine the target transient source from the transient source candidate. In one embodiment, the target transient source identification module 704 can be used to perform the operation S240 described above, which will not be repeated here.

[0117] According to the embodiments of this application, the transient source identification device 700 can quickly and accurately identify transient sources, improve the accuracy and efficiency of transient source identification, and can also convert the image data of observation images observed by different astronomical telescopes into a unified format to eliminate heterogeneity, so as to directly process the image data of observation images from different astronomical telescopes through a large model, reduce the image data processing process, and improve the data processing speed.

[0118] In some embodiments, the transient source identification device 700 is further configured to: when an observation image is acquired, update the spatial index based on multiple tiles so as to acquire all observation images covering the differential tiles using the updated spatial index; generate corresponding parent tiles for each of the multiple tiles at the target level by downsampling, wherein the level of the parent tile is lower than the target level, and the image patch corresponding to the parent tile and the image patch corresponding to the tile are obtained by two different types of astronomical telescopes photographing the same celestial region; associate the relationships between the multiple tiles, the parent tiles corresponding to the multiple tiles, and the parent-child tiles in the existing spatial index to obtain the updated spatial index.

[0119] In some embodiments, the transient source candidate determination module 703 is specifically used to: recursively query the coordinates of the differential tile in the spatial index to obtain the first image block corresponding to the differential tile, the second image block corresponding to the parent tile of the differential tile, and the third image block corresponding to the child tile of the differential tile; extract celestial information from the first image block, the second image block, and the third image block respectively, and use the celestial body represented by the celestial information as a transient source candidate.

[0120] In some embodiments, the temporary source candidate determination module 703 is further configured to: if there is overlap between all image blocks covering the differential tiles, convert the coordinates of each tile pixel of the differential tiles to celestial coordinates; perform a first coordinate transformation on the celestial coordinates using a coordinate transformation matrix to obtain the coordinates of the image pixels corresponding to the tile pixels in the corresponding image block, wherein the coordinate transformation matrix indicates the mapping relationship between the celestial coordinates and the coordinates of the image pixels; interpolate the pixels around the image pixels to obtain the interpolated image blocks covering the differential tiles; and perform de-overlap processing on the interpolated image blocks based on an image de-overlap strategy, wherein the image de-overlap strategy includes at least one of the following: an arithmetic mean strategy for image pixels in the overlapping region, a distance-weighted strategy, and an image quality strategy. The arithmetic mean strategy refers to using the arithmetic mean of the image pixels in the overlapping region as the image pixel value of the overlapping region. The distance-weighted strategy refers to assigning weights to each image pixel based on the distance between each image pixel of the interpolated image block and the edge of the interpolated image block. The image quality strategy refers to assigning weights to each image block in the overlapping region based on the image quality of each image block in the overlapping region.

[0121] In some embodiments, the transient source candidate determination module 703 is further configured to: convert the coordinates of the tile pixel into two-dimensional coordinates according to the binary bit interleaving method, wherein the two-dimensional coordinates indicate the position of the tile pixel within its respective differential tile; and convert the two-dimensional coordinates into celestial coordinates through a second coordinate transformation based on the two-dimensional coordinates and the number of the differential tile to which the tile pixel belongs, wherein the number indicates the position of the differential tile to which the tile pixel belongs among multiple tiles.

[0122] In some embodiments, the target transient source identification module 704 is specifically used to: perform inverse variance weighted calculation on the photometric results of each image block corresponding to the same transient source candidate to obtain the comprehensive photometric result of each transient source candidate; and take the transient source candidate corresponding to the comprehensive photometric result that is less than a preset threshold as the target transient source.

[0123] According to embodiments of this application, any multiple modules among the image projection module 701, differential tile acquisition module 702, transient source candidate determination module 703, and target transient source identification module 704 can be combined into one module, or any one of these modules can be split into multiple modules. Alternatively, at least part of the functionality of one or more of these modules can be combined with at least part of the functionality of other modules and implemented in one module. According to embodiments of this application, at least one of the image projection module 701, differential tile acquisition module 702, transient source candidate determination module 703, and target transient source identification module 704 can be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system-on-a-chip, a system-on-a-substrate, a system-on-package, an application-specific integrated circuit (ASIC), or any other reasonable means of integrating or packaging the circuitry, or implemented in software, hardware, or firmware, or in any suitable combination of any of these three implementation methods. Alternatively, at least one of the image projection module 701, the differential tile acquisition module 702, the transient source candidate determination module 703, and the target transient source identification module 704 can be at least partially implemented as a computer program module, which can perform corresponding functions when the computer program module is run.

[0124] Figure 8 A block diagram schematically illustrates an electronic device suitable for implementing a transient source identification method according to an embodiment of this application.

[0125] like Figure 8As shown, an electronic device 800 according to an embodiment of this application includes a processor 801, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 802 or a program loaded from a storage portion 808 into a random access memory (RAM) 803. The processor 801 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 801 may also include onboard memory for caching purposes. The processor 801 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of this application.

[0126] RAM 803 stores various programs and data required for the operation of electronic device 800. Processor 801, ROM 802, and RAM 803 are interconnected via bus 804. Processor 801 executes various operations of the method flow according to embodiments of this application by executing programs in ROM 802 and / or RAM 803. It should be noted that the programs may also be stored in one or more memories other than ROM 802 and RAM 803. Processor 801 may also execute various operations of the method flow according to embodiments of this application by executing programs stored in said one or more memories.

[0127] According to embodiments of this application, the electronic device 800 may further include an input / output (I / O) interface 805, which is also connected to a bus 804. The electronic device 800 may also include one or more of the following components connected to the input / output (I / O) interface 805: an input section 806 including a keyboard, mouse, etc.; an output section 807 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 808 including a hard disk, etc.; and a communication section 809 including a network interface card such as a LAN card, modem, etc. The communication section 809 performs communication processing via a network such as the Internet. A drive 810 is also connected to the input / output (I / O) interface 805 as needed. A removable medium 811, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 810 as needed so that computer programs read from it can be installed into the storage section 808 as needed.

[0128] This application also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of this application.

[0129] According to embodiments of this application, the computer-readable storage medium can be a non-volatile computer-readable storage medium, such as including but not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, according to embodiments of this application, the computer-readable storage medium may include ROM 802 and / or RAM 803 and / or one or more memories other than ROM 802 and RAM 803 described above.

[0130] Embodiments of this application also include a computer program product comprising a computer program containing program code for performing the methods shown in the flowchart. When the computer program product is run on a computer system, the program code enables the computer system to implement the transient source identification method provided in the embodiments of this application.

[0131] When the computer program is executed by the processor 801, it performs the functions defined in the system / apparatus of this application embodiment. According to the embodiments of this application, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0132] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and may be downloaded and installed via the communication section 809, and / or installed from a removable medium 811. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0133] In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 809, and / or installed from the removable medium 811. When the computer program is executed by the processor 801, it performs the functions defined in the system of this application embodiment. According to the embodiments of this application, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0134] According to embodiments of this application, program code for executing the computer programs provided in the embodiments of this application can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C", or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0135] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0136] Those skilled in the art will understand that the features described in the various embodiments of this application can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in this application. In particular, the features described in the various embodiments of this application can be combined and / or combined in various ways without departing from the spirit and teachings of this application. All such combinations and / or combinations fall within the scope of this application.

Claims

1. A transient source identification method, characterized in that, The method includes: The acquired observation image is projected onto the HEALPix grid of the target level to obtain multiple tiles, where each tile corresponds to an image block, and the image blocks constitute the observation image. The target level is determined according to the type of astronomical telescope that captured the observation image. A pixel-by-pixel difference operation is performed between tiles located at the same position and a reference tile. At least one tile with a difference pixel mean greater than a preset mean is selected from the plurality of tiles and used as the difference tile. The difference pixel mean represents the average difference between the pixel values ​​of the tile and the corresponding reference tile. The reference tile is obtained by projecting a historical observation image onto the HEALPix grid. The historical observation image is obtained by the astronomical telescope that took the observation image at the same celestial region at the historical observation time. Based on the coordinates of the differential tiles, all image patches covering the differential tiles are obtained using a spatial index, and transient source candidates are determined. The spatial index is used to obtain all image patches covering any given tile based on the coordinates of any given tile. All image patches are from various types of astronomical telescopes. Photometrically measure the image block corresponding to the transient source candidate, and determine the target transient source from the transient source candidate.

2. The method according to claim 1, characterized in that, Once the observed image is acquired, the spatial index is updated based on the multiple tiles so that all observed images covering the differential tiles can be obtained using the updated spatial index: Parent tiles are generated for each of the multiple tiles at the target level by downsampling. The level of the parent tile is lower than the target level. The image patch corresponding to the parent tile and the image patch corresponding to the tile are obtained by two different types of astronomical telescopes photographing the same celestial region. The spatial index is updated by associating the relationships between the multiple tiles, their corresponding parent tiles, and the parent-child tiles in the existing spatial index.

3. The method according to claim 2, characterized in that, The step of obtaining all image blocks covering the differential tiles and determining transient source candidates based on the coordinates of the differential tiles using spatial indexing includes: The coordinates of the differential tile are recursively queried in the spatial index to obtain the first image block corresponding to the differential tile, the second image block corresponding to the parent tile of the differential tile, and the third image block corresponding to the child tile of the differential tile. Celestial information is extracted from the first image block, the second image block, and the third image block, respectively, and the celestial body represented by the celestial information is used as the candidate transient source.

4. The method according to claim 1, characterized in that, The method further includes: If there is overlap between all image blocks covering the differential tiles, the coordinates of each tile pixel of the differential tiles are converted to celestial coordinates; The celestial coordinates are transformed using a coordinate transformation matrix to obtain the coordinates of the image pixel corresponding to the tile pixel in the corresponding image block. The coordinate transformation matrix indicates the mapping relationship between the celestial coordinates and the coordinates of the image pixel. Interpolate the pixels surrounding the image pixels to obtain an interpolated image block covering the differential tiles; Based on the image de-overlap strategy, the interpolated image blocks are subjected to de-overlap processing. The image de-overlap strategy includes at least one of the following: an arithmetic mean strategy of image pixels in the overlapping region, a distance-weighted strategy, and an image quality strategy. The arithmetic mean strategy refers to using the arithmetic mean of the image pixels in the overlapping region as the image pixel value of the overlapping region. The distance-weighted strategy refers to assigning weights to each image pixel based on the distance between each image pixel of the interpolated image block and the edge of the interpolated image block. The image quality strategy refers to assigning weights to each image block in the overlapping region based on the image quality of each image block in the overlapping region.

5. The method according to claim 4, characterized in that, The step of converting the coordinates of each tile pixel of the differential tile to celestial coordinates includes: The coordinates of the tile pixels are converted into two-dimensional coordinates using a binary bit interleaving method. The two-dimensional coordinates indicate the position of the tile pixel within its respective differential tile. Based on the two-dimensional coordinates and the differential tile number to which the tile pixel belongs, the two-dimensional coordinates are converted into celestial coordinates through a second coordinate transformation, wherein the number indicates the position of the differential tile to which the tile pixel belongs among the plurality of tiles.

6. The method according to claim 1, characterized in that, The step of determining the target transient source from the transient source candidates includes: The photometric results of each image patch corresponding to the same transient source candidate are subjected to inverse variance weighting to obtain the comprehensive photometric results of each transient source candidate. The transient source candidate corresponding to the comprehensive photometric result that is less than a preset threshold is taken as the target transient source.

7. A transient source identification device, characterized in that, The device includes: The image projection module is used to project the acquired observation image onto the HEALPix grid of the target level to obtain multiple tiles, wherein each tile corresponds to an image block, and the image blocks constitute the observation image. The target level is determined according to the type of astronomical telescope that captured the observation image. The differential tile acquisition module is used to perform pixel-by-pixel difference operations between tiles located at the same position and reference tiles, and to determine at least one tile with a differential pixel mean greater than a preset mean from the plurality of tiles as the differential tile. The differential pixel mean represents the average difference between the pixel values ​​of the tile and the corresponding reference tile. The reference tile is obtained by projecting a historical observation image onto the HEALPix grid. The historical observation image is obtained by the astronomical telescope that took the observation image at the same celestial region at the historical observation time. The transient source candidate determination module is used to obtain all image blocks covering the differential tiles and determine transient source candidates based on the coordinates of the differential tiles using a spatial index. The spatial index is used to obtain all image blocks covering any tile based on the coordinates of any tile. All image blocks come from various types of astronomical telescopes. The target transient source identification module is used to perform photometry on the image block corresponding to the transient source candidate and determine the target transient source from the transient source candidate.

8. An electronic device, comprising: One or more processors; Memory, used to store one or more computer programs. The characteristic feature is that the one or more processors execute the one or more computer programs to implement the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 6.

Images