A completely blind matching calculation method for astronomical images

Abstract

The present invention provides a method for calculating the blind matching of astronomical images, including generating a theoretical star map according to a theoretical star catalog and sorting the generated theoretical star map according to features; calculating the position of each star map according to the sorted theoretical star map to obtain an index table; and calculating the center orientation and pixel scale of the astronomical image according to the theoretical star map and the index table. The present invention screens the all-sky star catalog and generates an all-sky star map grid. The star map distribution is relatively uniform and the amount of star map data is reduced. The star map matching efficiency is improved by making an index through feature sorting, so that the astronomical image can obtain the image center orientation and pixel scale without any prior information. The situation that the observation data formats of different telescopes are not unified and the shared data is difficult to use can be improved, and the utilization rate of massive astronomical images can be improved, thereby saving a lot of observation time and improving the work efficiency of astronomical researchers.

Description

Technical Field

[0001] The invention belongs to the technical field of astronomical image processing, and in particular relates to a method for calculating astronomical image full-blind matching. Background Art

[0002] Totally blind matching means that no prior information about the image is known, including but not limited to the image target surface, pixel size, field of view of the telescope used to capture the image, angular distance, and direction. Only the star information captured in the image is used, and the star catalog data is used for matching to calculate the image's center direction, pixel scale, image rotation angle and other information.

[0003] With the advent of the era of astronomical big data, the number of astronomical images will grow exponentially. Most astronomical research is conducted using astronomical images collected by a single telescope. Although a single telescope can produce correct and standard metadata, the control systems of many telescopes drift relative to the sky and can only produce approximate astrometric metadata. Some telescopes do not generate metadata, or generate metadata in a form that does not meet the standards. Even complex and highly automated measurements such as SDS can fail in the system that generates astrometric calibration information, resulting in very good astronomical image data that cannot be used due to invalid center pointing and pixel scale, resulting in these non-standard astronomical images that cannot be effectively used, resulting in a large loss of observation time and reduced work efficiency of astronomical researchers. Summary of the invention

[0004] The present invention aims at the deficiencies in the prior art and provides a method for calculating astronomical image full-blind matching.

[0005] The present invention provides a method for calculating astronomical image blind matching, comprising:

[0006] Step 1, generating a theoretical star map according to the theoretical star catalog and sorting the generated theoretical star map according to the characteristics;

[0007] Step 2, calculate the position of each star map according to the sorted theoretical star map to obtain an index table;

[0008] Step 3: Calculate the center direction and pixel scale of the astronomical image based on the theoretical star map and index table.

[0009] Furthermore, generating a theoretical star map according to the theoretical star catalog and sorting the generated theoretical star map according to the features includes:

[0010] The theoretical star catalog is divided into multiple grids according to right ascension and declination, and the star with the smallest magnitude in the theoretical star catalog data is selected in each grid to generate a polygonal theoretical star map;

[0011] The theoretical star map is sorted from 60° to 180° for the largest angle and from 0° to 60° for the second largest angle.

[0012] Furthermore, the step of calculating the center orientation and pixel scale of the astronomical image according to the theoretical star map and the index table includes:

[0013] Step 301, obtaining the centroid and grayscale information of a point target in an astronomical image; the point target is an image spot formed by a star in the astronomical image; the grayscale information includes the number of pixels of the point target, the aspect ratio of the point target, and the grayscale sum of the point target;

[0014] Step 302, sorting the stars in descending order according to the number of pixels of the point target or the grayscale of the point target to obtain a measured star map;

[0015] Step 303, performing polygon similarity matching between the measured star map and the theoretical star map according to the polygons composed of stars, to obtain multiple groups of matching star pairs;

[0016] Step 304, calculate the pixel scale of the first group of matching star pairs, and find at least one group of matching star pairs with the same scale as the first group in the remaining M-1 groups; wherein M is the number of groups of matching star pairs; if no matching star pairs with the same scale as the first group can be found in the remaining M-1 groups; calculate the pixel scale of the second group of matching star pairs, and find at least one group of matching star pairs with the same scale as the second group in the remaining M-2 groups; if no matching star pairs with the same scale as the second group can be found in the remaining M-2 groups; calculate the pixel scale of the third group of matching star pairs, and find at least one group of matching star pairs with the same scale as the third group in the remaining M-3 groups, until the pixel scale of the Mth group of matching star pairs is calculated cyclically; if there are more than or equal to two groups of matching star pairs with the same pixel scale, calculate the six-constant film model coefficients;

[0017] Step 305, calculating the central orientation and image rotation angle of the astronomical image according to the six-constant film model coefficients, and taking the pixel scale of the matching star pair with the same pixel scale as the pixel scale of the astronomical image;

[0018] Step 306, performing full-image matching of the theoretical star catalog and the measured stars according to the central orientation, image rotation angle and pixel scale of the astronomical image;

[0019] Step 307, determining whether the full image matching satisfies a preset matching rate;

[0020] Step 308, if yes, output the central direction, image rotation angle and pixel scale of the astronomical image;

[0021] If not, determine whether there are other matching star pairs with the same pixel scale greater than or equal to the two groups according to step 304;

[0022] If yes, return to step 305;

[0023] If no, determine that the center direction, image rotation angle and pixel scale of the astronomical image are wrong, and exit the calculation.

[0024] Furthermore, the theoretical star catalog is any one of the GAIA star catalog, the Hipparcos star catalog, the Tycho star catalog and the FK5 star catalog.

[0025] Furthermore, the six-constant film model is:

[0026]

[0027] Among them, (ξ,ζ) is the ideal coordinate of the star, (x,y) is the image coordinate of the star; a 1 is the constant term of the transverse equation; b 1 and c 1 are the coefficients of the first-order terms of the transverse equation; a 2 is the constant term of the longitudinal equation; b 2 and c 2 are all first-order coefficients of the longitudinal equation.

[0028] Furthermore, the step of sorting the stars in descending order according to the number of pixels of the point targets or the grayscale of the point targets to obtain the measured star map includes:

[0029] Setting star map features; the star map features include polygons, main sequence star maps and vector maps;

[0030] Setting a star map feature threshold; the star map feature threshold includes a side length range, an angle range, and the number of sides of the star map;

[0031] Select a star as the first endpoint, and select stars from the remaining stars to establish star map elements according to the star map characteristics and the star map characteristic threshold, wherein the star map elements are polygonal graphics, main sequence star maps or vector graphics; the star map elements include the point coordinates of each star in the star map, and the characteristic values ​​between each point, wherein the characteristic values ​​include side lengths and angles;

[0032] Select multiple stars as the first endpoints to establish different star map elements according to the characteristics of the star map;

[0033] Different star map elements are sorted according to the star map element characteristics to obtain a measured star map; the star map element characteristics include the lengths of each side and the angles of each corner of a polygonal figure, a main sequence star map or a vector map.

[0034] The present invention provides a method for calculating the blind matching of astronomical images, including generating a theoretical star map according to a theoretical star catalog and sorting the generated theoretical star map according to features; calculating the position of each star map according to the sorted theoretical star map to obtain an index table; and calculating the center orientation and pixel scale of the astronomical image according to the theoretical star map and the index table. The present invention screens the all-sky star catalog and generates an all-sky star map grid. The star map distribution is relatively uniform and the amount of star map data is reduced. The star map matching efficiency is improved by making an index through feature sorting, so that the astronomical image can obtain the image center orientation and pixel scale without any prior information. The situation that the observation data formats of different telescopes are not unified and the shared data is difficult to use can be improved, and the utilization rate of massive astronomical images can be improved, thereby saving a lot of observation time and improving the work efficiency of astronomical researchers. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solution of the present invention, the drawings required for use in the embodiments are briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0036] Figure 1 Astronomical images provided by embodiments of the present invention;

[0037] Figure 2 A workflow diagram of a method for calculating astronomical image blind matching provided by an embodiment of the present invention;

[0038] Figure 3 A schematic diagram of a polygon constructed using the geometric relationship of star positions provided in an embodiment of the present invention;

[0039] Figure 4 A schematic diagram of grid division provided by an embodiment of the present invention;

[0040] Figure 5 A Gaia DR2 source (star) density distribution diagram in the galactic coordinate system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0042] like Figure 1As shown in the figure, astronomical images are images of the starry sky observed through a telescope. The stars, planets, satellites and space debris in the night sky are imaged on a detector installed at the focus of the telescope, usually a CCD or CMOS camera. Stars and targets generally appear as circular spots with a certain area due to point spread. Due to reasons such as optical distortion of the telescope, there may be coma. In addition, they may be elongated into ellipses due to different observation modes.

[0043] like Figure 2 As shown, the embodiment of the present invention partially provides a method for calculating astronomical image blind matching, including:

[0044] Step 1: Generate a theoretical star map based on the theoretical star catalog and sort the generated theoretical star map by features.

[0045] Theoretical star catalogs are catalogs that record various parameters of celestial bodies, such as position, motion, magnitude, spectral type, etc. Compiling star catalogs through astronomical observations is one of the earliest tasks in astronomy. With the introduction of the principle of transit observation and the use of new telescopes, the accuracy of star catalogs has been increasing. Commonly used visual star catalogs in practical applications include the GAIA star catalog, the Hipparcos star catalog, the Tycho star catalog, and the FK5 star catalog.

[0046] like Figure 5 As shown in the figure, compared with the Hipparcos satellite, the Gaia satellite can improve the measurement accuracy of the three-dimensional position of celestial bodies obtained from direction and parallax information and the tangential velocity obtained from proper motion data by more than 100 times, and the equivalent angle measurement accuracy has reached 10 micro-arcseconds. In terms of observation targets, Gaia has also improved many times, with a total observation target of far more than 1 billion celestial bodies.

[0047] like Figure 4 As shown in the figure, the theoretical star catalog is divided into a target number of grids according to right ascension and declination. In each grid, the star with the smallest magnitude in the theoretical star catalog data is selected to generate a polygonal theoretical star map. A star map is a graph constructed using the geometric relationship of star positions, generally including triangles, polygons, vector star maps, etc. A triangular star map is a relatively simple and commonly used star map, such as Figure 3 As shown, it is formed by connecting the centers of any three stars. Generally, when building a star map, the number of times each star is used, the number of times each edge is used, the maximum edge length, the minimum edge length, etc. are limited to reduce the number of compositions while meeting the matching requirements. The theoretical star map is a star map constructed using a theoretical star catalog, and the measured star map is a star map constructed using stars in astronomical observation images. When the two star maps are matched, they will be converted to the same coordinate system.

[0048] The positions of stars in the star catalog are coordinates in the celestial coordinate system, with the measurement center as the center. The coordinates are (right ascension, declination) ranging from right ascension 0° to 360°, and declination -90° to 90°. Grid division is to divide the projection of stars on the celestial sphere at certain intervals to form a uniform grid. The grid density can be divided according to demand. The denser the grid, the smaller the image pixel scale that can be detected, but the generated star map data will be larger and the retrieval will be slower. In actual operation, multiple specifications of grids can be divided, and the solution can be called cyclically in the order of pixel scale from large to small. The solution will exit when the solution is successful. That is, the solution speed of images with larger pixel scales is guaranteed, while the solution capability of images with small pixel scales is taken into account, so as to ensure the efficiency and success rate of calculation.

[0049] When observing with a telescope, the angle of each pixel in the celestial coordinate system is related to the telescope focal length, field of view, detector size, etc. The angle corresponding to each pixel is the pixel scale. A completely blind person does not know the pixel scale of the telescope and can only solve it through star map matching. However, when matching and solving, different pixel scales need to be classified and sorted, and solved separately. Then, the results calculated for each pixel scale are confirmed through measured and theoretical star matching. The one that can be successfully matched is the final result.

[0050] The triangular theoretical star map is sorted in two levels: the largest angle 60°~180° sorting and the second largest angle 0°~60° sorting. The purpose of sorting is to improve the matching speed, and the sorted star map sequence is stored in a certain feature order.

[0051] Step 2: Calculate the position of each star map based on the sorted theoretical star map to obtain an index table.

[0052] Step 3: Calculate the center direction and pixel scale of the astronomical image based on the theoretical star map and index table.

[0053] Step 301, using image processing software Soure Extractor to process astronomical observation images, and obtain the centroid and grayscale information of point targets in the astronomical images; the point targets are image spots formed by stars in the astronomical images; the grayscale information includes the number of pixels of the point targets, the aspect ratio of the point targets, and the grayscale sum of the point targets.

[0054] Step 302, sorting the stars in descending order according to the number of pixels of the point targets or the grayscale of the point targets to obtain a measured star map.

[0055] In this step, the star map features are set; the star map features include polygons, main sequence star maps and vector maps;

[0056] Setting a star map feature threshold; the star map feature threshold includes a side length range, an angle range, and the number of sides of the star map;

[0057] Select a star as the first endpoint, and select stars from the remaining stars to establish star map elements according to the star map characteristics and the star map characteristic threshold, wherein the star map elements are polygonal graphics, main sequence star maps or vector graphics; the star map elements include the point coordinates of each star in the star map, and the characteristic values ​​between each point, wherein the characteristic values ​​include side lengths and angles;

[0058] Select multiple stars as the first endpoints to establish different star map elements according to the characteristics of the star map;

[0059] Different star map elements are sorted according to the star map element characteristics to obtain a measured star map; the star map element characteristics include the lengths of each side and the angles of each corner of a polygonal figure, a main sequence star map or a vector map.

[0060] Step 303: perform polygon similarity matching between the measured star map and the theoretical star map according to the polygons composed of stars to obtain multiple groups of matching star pairs.

[0061] Step 304, calculate the pixel scale of the first group of matching star pairs, and find at least one group of matching star pairs with the same scale as the first group in the remaining M-1 groups; wherein M is the number of groups of matching star pairs; if no matching star pairs with the same scale as the first group can be found in the remaining M-1 groups; calculate the pixel scale of the second group of matching star pairs, and find at least one group of matching star pairs with the same scale as the second group in the remaining M-2 groups; if no matching star pairs with the same scale as the second group can be found in the remaining M-2 groups; calculate the pixel scale of the third group of matching star pairs, and find at least one group of matching star pairs with the same scale as the third group in the remaining M-3 groups, until the pixel scale of the Mth group of matching star pairs is calculated in a loop; if there are more than or equal to two groups of matching star pairs with the same pixel scale, calculate the six-constant film model coefficients.

[0062] Step 305, calculating the central orientation and image rotation angle of the astronomical image according to the six-constant film model coefficients, and taking the pixel scale of the matching star pair with the same pixel scale as the pixel scale of the astronomical image.

[0063] The six-constant film model is:

[0064]

[0065] Among them, (ξ,ζ) is the ideal coordinate of the star, (x,y) is the image coordinate of the star; a 1 is the constant term of the transverse equation; b 1 and c 1 are the coefficients of the first-order terms of the transverse equation; a 2 is the constant term of the longitudinal equation; b 2 and c 2 are all first-order coefficients of the longitudinal equation.

[0066] Step 306, performing full-image matching of the theoretical star catalog and the measured stars according to the central orientation, image rotation angle and pixel scale of the astronomical image;

[0067] Step 307, determining whether the full image matching satisfies a preset matching rate;

[0068] Step 308, if yes, output the central direction, image rotation angle and pixel scale of the astronomical image;

[0069] If not, determine whether there are other matching star pairs with the same pixel scale greater than or equal to the two groups according to step 304;

[0070] If yes, return to step 305;

[0071] If no, determine that the center direction, image rotation angle and pixel scale of the astronomical image are wrong, and exit the calculation.

[0072] The center point is the right ascension and declination value of the image center coordinates in the celestial coordinate system. Assuming the image is p×q, the coordinates That is, the center of the image, and its corresponding right ascension and declination value in the celestial coordinate system is (alpha, delta) angle. Through this value, i.e., the pixel scale, we can clearly know which stars should be in the image and where they should be located in the image. We can then obtain a star position map projected from the theoretical star catalog to the observed image. This map and the actual observed map can be completely aligned by rotation. The same stars are in the same position in the observed map, but due to the telescope's detection capability and the measured star map, they cannot completely contain the stars in the theoretical star catalog, because some stars are too far away, too dim, and the telescope aperture is too small to see. Matching does not require matching all stars, only a part of them. If the matching requirements are not met, this match is wrong, otherwise it is correct.

[0073] The present invention has been described in detail above in conjunction with specific implementations and exemplary examples, but these descriptions cannot be understood as limiting the present invention. Those skilled in the art understand that, without departing from the spirit and scope of the present invention, a variety of equivalent substitutions, modifications or improvements may be made to the technical solution of the present invention and its implementation methods, all of which fall within the scope of the present invention. The scope of protection of the present invention shall be subject to the attached claims.

Claims

1. A method for calculating astronomical image blind matching, characterized in that: include: Step 1, generating a theoretical star map according to the theoretical star catalog and sorting the generated theoretical star map according to the characteristics; Step 2, calculate the position of each star map according to the sorted theoretical star map to obtain an index table; Step 3, calculating the center direction and pixel scale of the astronomical image according to the theoretical star map and index table; The step of generating a theoretical star map according to the theoretical star catalog and sorting the generated theoretical star map according to the features comprises: The theoretical star catalog is divided into multiple grids according to right ascension and declination, and the star with the smallest magnitude in the theoretical star catalog data is selected in each grid to generate a polygonal theoretical star map; The theoretical star map is sorted by the largest angle of 60° to 180° and the second largest angle of 0° to 60°; The method of calculating the center direction and pixel scale of the astronomical image according to the theoretical star map and the index table includes: Step 301, obtaining the centroid and grayscale information of a point target in an astronomical image; the point target is an image spot formed by a star in the astronomical image; the grayscale information includes the number of pixels of the point target, the aspect ratio of the point target, and the grayscale sum of the point target; Step 302, sorting the stars in descending order according to the number of pixels of the point target or the grayscale of the point target to obtain a measured star map; Step 303, performing polygon similarity matching between the measured star map and the theoretical star map according to the polygons composed of stars, to obtain multiple groups of matching star pairs; Step 304, calculate the pixel scale of the first group of matching star pairs, and find at least one group of matching star pairs with the same scale as the first group in the remaining M-1 groups; wherein M is the number of groups of matching star pairs; if no matching star pairs with the same scale as the first group can be found in the remaining M-1 groups; calculate the pixel scale of the second group of matching star pairs, and find at least one group of matching star pairs with the same scale as the second group in the remaining M-2 groups; if no matching star pairs with the same scale as the second group can be found in the remaining M-2 groups; calculate the pixel scale of the third group of matching star pairs, and find at least one group of matching star pairs with the same scale as the third group in the remaining M-3 groups, until the pixel scale of the Mth group of matching star pairs is calculated cyclically; if there are more than or equal to two groups of matching star pairs with the same pixel scale, calculate the six-constant film model coefficients; Step 305, calculating the central orientation and image rotation angle of the astronomical image according to the six-constant film model coefficients, and taking the pixel scale of the matching star pair with the same pixel scale as the pixel scale of the astronomical image; Step 306, performing full-image matching of the theoretical star catalog and the measured stars according to the central orientation, image rotation angle and pixel scale of the astronomical image; Step 307, determining whether the full image matching satisfies a preset matching rate; Step 308, if yes, output the central direction, image rotation angle and pixel scale of the astronomical image; If not, determine in step 304 whether there are other matching star pairs with the same pixel scale greater than or equal to the two groups; If yes, return to step 305; If no, determine that the center direction, image rotation angle and pixel scale of the astronomical image are wrong, and exit the calculation.

2. The astronomical image blind matching calculation method according to claim 1, characterized in that: The theoretical star catalog is any one of the GAIA star catalog, the Hipparcos star catalog, the Tycho star catalog and the FK5 star catalog.

3. The astronomical image blind matching calculation method according to claim 1, characterized in that: The six-constant film model is: Among them, (ξ,ζ) are the ideal coordinates of the star, (x,y) are the image coordinates of the star; a1 is the constant term of the transverse equation; b1 and c1 are the linear coefficients of the transverse equation; a2 is the constant term of the longitudinal equation; b2 and c2 are the linear coefficients of the longitudinal equation.

4. The astronomical image blind matching calculation method according to claim 1, characterized in that: The method of obtaining a measured star map by sorting the stars in descending order according to the number of pixels of the point target or the grayscale of the point target includes: Setting star map features; the star map features include polygons, main sequence star maps and vector maps; Setting a star map feature threshold; the star map feature threshold includes a side length range, an angle range, and the number of sides of the star map; Select a star as the first endpoint, and select stars from the remaining stars to establish star map elements according to the star map characteristics and the star map characteristic threshold, wherein the star map elements are polygonal graphics, main sequence star maps or vector graphics; the star map elements include the point coordinates of each star in the star map, and the characteristic values ​​between each point, wherein the characteristic values ​​include side lengths and angles; Select multiple stars as the first endpoints to establish different star map elements according to the characteristics of the star map; Different star map elements are sorted according to the star map element characteristics to obtain a measured star map; the star map element characteristics include the lengths of each side and the angles of each corner of a polygonal figure, a main sequence star map or a vector map.

Images