A high real-time daytime sky map star point extraction method

Abstract

A high real-time daytime skyglow star image star point extraction method, comprising: first, using an improved soft morphological filtering algorithm to suppress the skyglow background of the star image; using morphological open operation to remove residual single-point noise after suppressing the skyglow of the image. Secondly, extracting bright pixels from the filtered star image and using clustering method to judge four-connected domain, sorting the connected domains from high to low according to total energy, selecting the top four stars with the highest energy to perform triangular identification matching based on the bright star catalog. Thirdly, calculating the attitude based on the successfully matched star points, and combining the complete star catalog to project the star points in the field of view back to the original star image plane. Finally, according to the calculated star point image plane coordinates, the original star image is windowed, the edge pixel fitting method is used to suppress the skyglow of the windowed image, and the adaptive threshold segmentation and four-connected domain judgment algorithm are used to complete the extraction of the star points.

Description

Technical Field

[0001] This invention relates to a method for extracting star points from stray daylight star maps in high real-time, applicable to star map processing of near-ground platforms such as aircraft, drones, and ships. Background Technology

[0002] All-weather star sensors are used in environments with strong stray light, where the imaging environment is complex and variable, and easily affected by background stray light and airborne pseudo-stars. On the one hand, a large number of airborne objects enter the field of view during flight, and the bright spots formed by their reflection of sunlight will bring a complex stray light environment. At the same time, the size and shape of the accompanying airborne objects are similar to stars, forming a large number of pseudo-stars. On the other hand, the background in the atmosphere is brighter than the space environment, and the contrast of star targets is low, making them relatively more difficult to extract.

[0003] Star extraction from daytime stray star maps of all-day star sensors mainly requires addressing issues such as noise and background suppression in stray light images and the identification of false stars under daytime conditions. Current research on star extraction under daytime stray light environments, both domestically and internationally, primarily focuses on background suppression in stray star maps from airborne and other low-to-medium altitude platforms, with limited research on methods for handling false stars.

[0004] Traditional methods for suppressing background noise in stray light images mainly include multi-frame averaging, median filtering, high-pass filtering, and morphological filtering. Multi-frame averaging adds the grayscale values ​​of corresponding pixels from multiple consecutively captured star images and then calculates the average. This method can eliminate random noise, but its effectiveness is minimal under strong shot noise caused by daylight stray light. Median filtering replaces the value of the template center point with the median value of all points within the template, effectively suppressing impulse noise, but its effect on normal white noise is not significant. Furthermore, median filtering can cause pixel shifts, affecting the accuracy of centroid localization. High-pass filtering performs a Fourier transform on the star image, preserving frequency components above the cutoff frequency, which can be used to eliminate low-frequency noise. However, its noise suppression effect on stray light background images is not significant, and the processing is computationally intensive and has poor real-time performance. Morphological filtering, based on structuring elements, measures and extracts corresponding shapes in the image, effectively suppressing complex star image backgrounds. However, the filtered results still contain many residual clutter, requiring further processing.

[0005] Against complex daytime lighting conditions, floating objects in the air and residual preprocessing filters can cause a large number of false star points in star images. Traditional methods involve extracting all false stars before matching and identifying them, which requires significant storage space and computation time exceeding ten seconds, severely impacting real-time performance. To address the problem of false star identification in complex backgrounds, current methods primarily employ multi-frame comparison. This method matches star points based on whether the pixel position changes of two connected components in star images captured at different times in two frames conform to the motion patterns of star points. This approach effectively avoids interference from false star points caused by floating objects, but it is difficult to implement in scenarios where the positions remain almost constant between frames. Furthermore, multi-frame comparison with a large number of false star points involves substantial computation, affecting both real-time performance and accuracy. Summary of the Invention

[0006] The technical problem to be solved by this invention is to overcome the shortcomings of the prior art and realize the recognition and extraction of star points with high real-time accuracy.

[0007] The objective of this invention is achieved through the following technical solutions:

[0008] In a first aspect, the present invention provides a method for extracting star points from a high-real-time daytime stray light star map, comprising:

[0009] An improved soft morphological filtering algorithm is used to suppress stray light background in the star map; morphological opening operation is then applied to the image with suppressed stray light background to remove residual single-point noise, thus obtaining the filtered star map.

[0010] Bright pixels are extracted from the filtered star map and clustering is used to determine four connected components. The connected components are sorted from high to low total energy, and the four stars with the highest energy are selected to perform triangle recognition and matching based on the bright star catalog.

[0011] The pose is calculated based on the successfully matched star points, and the star points in the field of view are back-projected onto the original star image plane in combination with the complete star catalog.

[0012] The original star map is windowed based on the calculated star point image plane coordinates. The window image is then subjected to stray light suppression using an edge pixel fitting method. Finally, the star points are extracted using an adaptive threshold segmentation and four-connected-domain judgment algorithm.

[0013] In a second aspect, the present invention provides a computer-readable storage medium having stored thereon computer program instructions, which, when loaded and run by a processor, cause the processor to execute the high real-time daytime stray light star map star point extraction method described in the first aspect.

[0014] Compared with the prior art, the present invention has the following advantages:

[0015] (1) This invention proposes a method to quickly match and identify the bright star catalog and the connected component sequence with the highest gray value, and then back-project to extract the remaining faint star points. This avoids the extraction and identification process of hundreds of false stars, greatly reduces the amount of computation, shortens the working time and storage space of star point identification and matching, realizes high real-time star point identification and matching, and effectively solves the problem of star point extraction in the environment of a large number of false star points caused by accompanying floating objects.

[0016] (2) In order to suppress the complex star map background formed by sunlight reflection in daytime environment and the single point noise remaining after filtering, the present invention adopts an improved soft morphological filtering and morphological opening operation method, which effectively removes the influence of stray light background and suppresses single point noise, highlights the contrast between the target area and the surrounding area, and significantly improves the probability of star point target extraction.

[0017] (3) The present invention adopts a window background fitting method to remove background stray light, which solves the accuracy problem of star point extraction under stray light environment and realizes high-precision star point extraction under complex light background environment.

[0018] (4) This invention can be extended to the extraction of faint star points in various environments. For sensors outside the atmosphere, this method can be used to detect weaker star points. It uses bright star catalogs to quickly identify and match, then back-projects the dark star catalogs and captures the window image, performs background fitting, and finally extracts the star points. It can quickly and accurately realize the detection of faint star points in space.

[0019] (5) This invention is easy to implement in hardware, requires little computation, has good real-time performance, has high application value, and can be widely used in all-day star sensors. Attached Figure Description

[0020] Figure 1 Flowchart of the algorithm for extracting star points from high real-time daytime stray light star maps;

[0021] Figure 2(a) shows the improved soft morphological dilatation filter structure element b;

[0022] Figure 2(b) shows the improved soft morphological erosion filter center structure element B1;

[0023] Figure 2(c) shows the improved soft morphological erosion filter edge structure element B2;

[0024] Figure 3 For morphological open column structure elements;

[0025] Figure 4 This is a back-projection image of star points within the detectable magnitude range. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0027] like Figure 1 As shown, a method for extracting star points from a high-real-time daytime stray light star map includes:

[0028] (1) An improved soft morphological filtering method is used to suppress stray light background in the star map, highlighting the pixel grayscale difference between the target area and the surrounding area. The specific star map processing procedure is as follows:

[0029] (1.1) Select an edge filtering template and perform soft morphological dilation processing on the original star map row by row and column by column to obtain the first-step filtering value at the center position of the template. Select a structuring element size of p×p, a structuring element center range of (p-4)×(p-4) with a pixel weight of 1 for the edge annular range. In soft morphological filtering, the structuring element is divided into a central structuring element b1 and an edge structuring element b2. Therefore, during filtering, the central structuring element is the original structuring element's center range of (p-4)×(p-4), and the edge structuring element is the p×p annular range with a blank center range of (p-4)×(p-4).

[0030] In this embodiment, an 11×11 structuring element b is selected, as shown in Figure 2(a). The center is a 7×7 area with no visible area, and the pixel weight of the edge annular range is 1. In soft morphological filtering, the central structuring element b1 is considered to be the 7×7 area at the center of structuring element b, and the edge structuring element b2 is considered to be an 11×11 annular structuring element with a 7×7 area at the center. The gray value at the center of the template is replaced by the result of dilation processing at other locations, i.e., the maximum value within the annular range of structuring element b2. The original stray star map is processed by soft morphological dilation to obtain the first-step filtered image.

[0031] (1.2) Select a filter template and perform soft morphological erosion processing on the first-step filtered image row by row and column by column to obtain the stray background value at the center of the template. The size of the structuring element is the same as in step (1.1), which is p×p. The pixel weight of the (p-4)×(p-4) range at the center of the structuring element is k (1≤k≤min{p×p / 2,(p-4)×(p-4)}), and the pixel weight of the edge annular range is 1. Use the above structuring element to perform erosion operation on the first-step filtered image. The gray value at the center of the template is represented by the kth value of the structuring element in ascending order of gray values ​​to obtain the stray background gray value of the stray star map.

[0032] The weight of the central structuring element is selected based on the image grayscale. The mean grayscale value f1 of the central structuring element and the mean grayscale value f2 of the edge structuring elements are calculated separately. The difference between f1 and f2 is calculated and compared with min{f1,f2}. When the difference is less than t times min{f1,f2}, the value of k is 1; when the difference is greater than t times min{f1,f2}, the value of k is calculated based on the current pixel grayscale. The value of t is a preset threshold based on the uniformity of ambient stray light background, 0 < t ≤ 1. The greater the background fluctuation, the higher the threshold should be set, and it can generally be set to 0.1.

[0033]

[0034] In the formula, k(i,j) represents the weight value at pixel (i,j), and f(i,j) represents the image grayscale value at pixel (i,j). The structuring elements B are sorted in ascending order according to the pixel weights of the central structuring element. The k-th value is selected as the grayscale value of the center of the structuring element after the erosion operation. When |f1-f2|>tmin{f1,f2}, the average value of k(i,j) is taken to obtain the k value.

[0035] In this embodiment, an 11×11 structuring element B is selected. The central structuring element B1 is a 7×7 area at the center of the original structuring element B, with a pixel weight of k (1≤k≤49), as shown in Figure 2(b). The edge structuring element B2 is an 11×11 ring-shaped structuring element with a 7×7 blank center, and a pixel weight of 1, as shown in Figure 2(c). The above operations are performed row by row and column by column on the first-step filtered image to obtain the second-step filtered image.

[0036] (1.3) Subtract the gray value of the original stray star map from the gray value of the filtered image in the second step, and calculate the difference at the same pixel position. If the gray value of the original star map is less than the filtered value, the gray value is set to 0, and the star image after suppressing the stray background is obtained.

[0037] (2) Perform morphological opening operation on the filtered star map to remove single-point noise in the image and smooth the target edge.

[0038] A 2×1 column structuring element is selected to perform a morphological opening operation on the filtered image. The structuring element is as follows: Figure 3 As shown, abnormal pixels in an image are eliminated. First, the column structuring element is moved on the filtered image, with the center value of the structuring element being the minimum gray value of all pixels in the template, resulting in the minimum value image. Next, the column structuring element is moved on the minimum value image, with the center value of the structuring element being the maximum gray value of all pixels in the template, resulting in a filtered star map with noise removed.

[0039] (3) Combine the star sensor star table to extract, identify and match real star points in the image, and back-project the star points in the star table onto the image plane according to the current attitude.

[0040] The processed star map is traversed, and the M pixels with the highest grayscale values ​​are extracted (M is preset according to the actual flight environment, generally 500≤M≤5000). Clustering is used to determine the four-connected components of the extracted bright pixels, identifying all connected components with an area greater than 4. Considering the imaging of the stars on the image plane, connected components that do not conform to the shape and size of the stars are removed. The remaining connected components are then sorted according to their total energy grayscale value.

[0041] Select the four connected components with the highest gray values ​​and combine them with the bright star list of the star sensor (generally a star list with a magnitude lower than 4.5) to perform triangle recognition. If the recognition is successful, stop matching. If it is unsuccessful, add the fifth connected component in the sorted order for recognition, and so on, until four stars are successfully recognized.

[0042] Based on the four identified stars, a coarse attitude calculation is performed. Then, according to the current attitude, the stars in the complete star catalog (based on the star sensor's detection sensitivity, such as a magnitude 7 catalog) within the star sensor's field of view are back-projected onto the image, obtaining the coordinates of the stars on the image plane. The back-projection results of stars within the detectable magnitude range are as follows: Figure 4 As shown, the positions of the stars in the image and the instrument magnitude are marked in the figure.

[0043] (4) Window the original star map according to the star point image plane coordinates, and use the edge pixel fitting method to suppress stray light in the window map to realize star point extraction and attitude calculation. The specific steps are as follows:

[0044] (4.1) Based on the star point image plane coordinates obtained in step (3), open a window in the original star map to obtain a window image of size Npix×Npix (usually set to 24×24).

[0045] (4.2) Fit the edge pixels of each window image to obtain the stray light background, and subtract them to achieve stray light suppression in the star-shaped window image. The window image background fitting can be performed using edge pixel mean fitting method, information entropy fitting method, etc. In this invention, the least squares method is used to fit the window edge pixels.

[0046] Specifically, the window image stray light fitting method based on the least squares method is as follows:

[0047] Suppose that the edge portion of the window image consists of n pixels, where the i-th edge pixel is located at the y-th edge. i line x i The column has a pixel grayscale value of E. i The fitted stray light background surface is as follows: Where a i Let be the fitting coefficients for the stray light background surface, and let Then X = [a1 a2 a3 a4 a5 a6] T This is the fitting result, thus having

[0048]

[0049] From X=A -1 b can be used to obtain the fitting coefficients for the stray background, and the stray background map of the entire window can be obtained based on the fitting results. The gray level of the window image is subtracted from the stray background map at the same position. If the gray level of the window is less than the gray level of the stray background, it is 0, thus obtaining the window star map with the stray background removed.

[0050] (4.3) For each window image after removing stray light background, the target star point region is selected by adaptive threshold segmentation, and the distribution of star points is determined according to the four connected components. The centroid position of the star points is calculated to achieve high-precision star point recognition and extraction under stray light background.

[0051] The contents not described in detail in this specification are common knowledge to those skilled in the art.

[0052] Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make possible changes and modifications to the technical solutions of the present invention by utilizing the methods and techniques disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solutions of the present invention shall fall within the protection scope of the technical solutions of the present invention.

Claims

1. A method for extracting star points from daytime stray light star maps in high real-time, characterized in that, include: An improved soft morphological filtering algorithm is used to suppress stray light background in star charts; Morphological opening operations are used to remove residual single-point noise from the image with suppressed stray light background, resulting in a filtered star map; Bright pixels are extracted from the filtered star map and clustering is used to determine four connected components. The connected components are sorted from high to low total energy, and the four stars with the highest energy are selected to perform triangle recognition and matching based on the bright star catalog. The pose is calculated based on the successfully matched star points, and the star points in the field of view are back-projected onto the original star image plane in combination with the complete star catalog. The original star map is windowed based on the calculated star point image plane coordinates. The window image is then subjected to stray light suppression using an edge pixel fitting method. The star points are extracted using an adaptive threshold segmentation and four-connected domain judgment algorithm. The improved soft morphological filtering algorithm for suppressing stray light background in star maps includes: The original star map is subjected to soft morphological dilation processing row by row and column by column by selecting the edge filtering template to obtain the first-step filtered image at the center of the template. The filter template is selected and soft morphological erosion is performed on the first-step filtered image row by row and column by column to obtain the stray light background value at the center of the template. Using the original star map and stray background values, a star image with stray background suppressed is obtained; Soft morphological dilation processes include: The structuring element is selected to be p×p, with the center area of ​​the structuring element (p-4)×(p-4) set to 0, and the pixel weight of the edge annular area set to 1; in soft morphological filtering, the structuring element is divided into a central structuring element. and edge structural elements During filtering, the central structuring element is the original structuring element's center within a (p-4)×(p-4) range, while the edge structuring element is a p×p annular range with a center of (p-4)×(p-4) blank space. The grayscale value at the center of the edge filtering template is replaced by the result of dilation processing at other locations, i.e., the structuring element... The maximum value within the annular region; Soft morphological etching treatment includes: The pixel weights within the (p-4)×(p-4) range of the structuring element center are: ,1≤ ≤ The pixel weight within the edge ring range is 1; the above structuring element is used to perform an erosion operation on the first-step filtered image, and the gray level at the center of the edge filtering template is sorted in ascending order by the gray level of the structuring element. Each value represents the stray light background value of the stray light star map; The weights of the structural element centers are selected based on the image grayscale, and the mean grayscale value of each structural element center is calculated. The average gray value of edge structuring elements ,calculate and The difference, and with Compare, when the difference is less than times hour, The value is 1; when the difference is greater than times hour, The value is calculated based on the current pixel's grayscale. The value is a preset threshold based on the uniformity of ambient stray light background; In the formula, express Weight value at pixel express Image grayscale value at a pixel; structuring elements based on the pixel weight of the central structuring element. Sort in ascending order and select the first one. This value is used as the gray value of the center of the structural element after the erosion operation. At that time, Take the average to get value.

2. The method for extracting star points from high-real-time daytime stray light star maps according to claim 1, characterized in that, Subtract the grayscale value of the original star image from the stray background value to calculate the difference at the same pixel position. If the grayscale value of the original star image is less than the difference between the grayscale values ​​of the stray background and the original star image, the grayscale value is set to 0, thus obtaining the star image after suppressing the stray background.

3. The method for extracting star points from high-real-time daytime stray light star maps according to claim 1, characterized in that, Morphological opening operations are used to remove residual single-point noise from images with suppressed stray light background, including: A 2×1 column structuring element is selected to perform a morphological opening operation on the image with suppressed stray light background. First, the column structuring element is moved on the image with suppressed stray light background, and the center value of the column structuring element is the minimum gray value of all pixels in the template to obtain the minimum value image. Then, the column structuring element is moved on the minimum value image, and the center value of the column structuring element is the maximum gray value of all pixels in the template to obtain the filtered star map with noise removal.

4. The method for extracting star points from high-real-time daytime stray light star maps according to claim 1, characterized in that, When using clustering to determine four-connected components, all connected components with an area greater than 4 are identified, and connected components that do not conform to the shape and size of the star points are removed.

5. The method for extracting star points from high-real-time daytime stray light star maps according to claim 1, characterized in that, Window image background fitting methods include edge pixel mean fitting, information entropy fitting, or least squares fitting.

6. A computer-readable storage medium having stored thereon computer program instructions, which, when loaded and run by a processor, cause the processor to perform the method as described in any one of claims 1 to 5.

Images