A false alarm suppression method, device, and medium for ground-based telescope observations

Abstract

The embodiment of the application provides a false alarm suppression method, device and medium for ground-based telescope observation, through research on the constructed false alarm data set, various false alarm characteristics can be fully obtained for different false alarm categories, based on trajectory quality characteristics, sky light identification coefficient, target identification coefficient, hot pixel identification coefficient, space-time geometric consistency, gray consistency, speed size consistency, long axis and speed size consistency, long axis and speed direction consistency, the uncertainty of the traditional experience threshold discrimination method can be got rid of, not only various scene false alarms can be effectively identified, but also false suppression of real targets can be prevented to cause missed alarms, and the false alarm suppression method is suitable for target detection after false alarm suppression in different observation environments, and the false alarm suppression effect is good.

Description

Technical Field

[0001] This application relates to the field of aerospace technology, and in particular to a method for suppressing false alarms for ground-based telescope observations, an electronic device, and a storage medium. Background Technology

[0002] In recent years, with the exponential growth in the number of space targets, small-aperture, large-field-of-view telescopes have been used more and more widely in space target observation, and space target detection technology is also showing its importance. Space target detection is of great significance to aerospace safety, aerospace scientific research, resource management and other fields. By detecting and tracking targets such as satellites, potential safety threats, such as space debris collisions, can be detected and monitored in a timely manner. At the same time, effective space target detection can help maintain the operational status of spacecraft and satellites, extend their service life and reduce resource waste.

[0003] However, false alarms are a significant problem in spatial target detection. Inappropriate detection algorithms can lead to a large number of false alarms, causing unnecessary waste of resources for early warning and security systems, as well as related manpower and materials, and significantly impacting the real-time performance of subsequent processing. Statistical analysis of false alarms detected by conventional methods reveals that complex meteorological conditions such as clouds and moonlight affect image quality, leading to misjudgments in target detection. Furthermore, various noises and interferences encountered during image acquisition and transmission, such as artifacts caused by image compression and noise inherent in the sensor itself, are often misidentified as targets, resulting in false alarms.

[0004] Since false alarm suppression can effectively reduce false alarms, improve the accuracy and precision of target detection, further ensure security, optimize resource allocation, improve work efficiency, and save costs, it is a crucial part of improving the system's perception efficiency in spatial target detection. Traditional false alarm suppression approaches focus on studying the characteristics of the raw data before outputting the target detection results, using preprocessing techniques (image denoising, trajectory correlation, etc.) to obtain better detection results and achieve the goal of suppressing false alarms.

[0005] However, regardless of the image preprocessing or trajectory association methods used, most decision-making steps in space target detection against a star map background rely on manually set thresholds based on experience. The same threshold often fails to achieve satisfactory detection results in different observation scenarios, especially with small-aperture, large-field-of-view telescopes, where the low signal-to-noise ratio makes the threshold definition even more ambiguous. Experiments with a series of detection algorithms revealed that different threshold settings lead to varying degrees of false alarms and missed alarms. Summary of the Invention

[0006] This application provides a method for suppressing false alarms in ground-based telescope observations, in order to solve the problem that space target detection cannot achieve good detection results when using the same threshold in different scenarios, thus leading to false alarms and missed alarms.

[0007] Accordingly, embodiments of this application also provide a false alarm suppression device for ground-based telescope observations, an electronic device, and a storage medium to ensure the implementation and application of the above methods.

[0008] To address the aforementioned problems, this application discloses a false alarm suppression method for ground-based telescope observations, the method comprising:

[0009] Construct a false alarm dataset and obtain multiple false alarm categories; the false alarm dataset includes multiple false alarm data, and the false alarm categories include skylight, thermal pixels, trajectory consistency, and mapping consistency between star images and motion;

[0010] Obtain the trajectory quality features corresponding to the false alarm dataset; the trajectory quality features include trajectory length, detection queue frame number, exposure time, shooting interval time, average trajectory velocity, trajectory direction, and average star image size of each trajectory point;

[0011] Based on the skylight, determine the skylight recognition coefficient and target recognition coefficient corresponding to the false alarm data;

[0012] Determine the hot pixel discrimination coefficient corresponding to the false alarm data based on the hot pixels;

[0013] Based on the trajectory consistency, determine the spatiotemporal geometric consistency, grayscale consistency, and velocity magnitude consistency of the false alarm data;

[0014] The consistency between the major axis and velocity magnitude, and the consistency between the major axis and velocity direction, are determined based on the mapping consistency between the star image and the motion.

[0015] The trajectory quality features, the skylight recognition coefficient, the target recognition coefficient, the thermal pixel discrimination coefficient, the spatiotemporal geometric consistency, the grayscale consistency, the velocity magnitude consistency, the major axis and velocity magnitude consistency, and the major axis and velocity direction consistency are used as extracted features and input into the classifier to identify target false alarms among multiple false alarm data, so as to suppress the target false alarms.

[0016] This application also discloses an electronic device, including: a processor; and a memory storing executable code thereon, which, when executed, causes the processor to perform a false alarm suppression method for ground-based telescope observations as described in any of the embodiments of this application.

[0017] This application also discloses one or more machine-readable media storing executable code, which, when executed, causes a processor to perform a false alarm suppression method for ground-based telescope observations as described in any of the embodiments of this application.

[0018] The embodiments of this application have the following advantages:

[0019] The false alarm suppression method for ground-based telescope observations provided in this application adopts a reverse approach to post-detection suppression. It studies the detection results (false alarm dataset) and can fully acquire false alarm features (trajectory quality features, skylight recognition coefficient, target recognition coefficient, thermal pixel discrimination coefficient, spatiotemporal geometric consistency, grayscale consistency, velocity magnitude consistency, major axis and velocity magnitude consistency, and major axis and velocity direction consistency) for different false alarm categories. Based on the false alarm features, it overcomes the uncertainty of traditional empirical threshold-based discrimination methods. It can not only effectively identify false alarms in various scenarios, but also prevent false suppression of real targets and missed alarms. It is applicable to post-detection false alarm suppression under different observation environments and has a good false alarm suppression effect. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating the steps of an embodiment of a false alarm suppression method for ground-based telescope observations according to this application;

[0021] Figure 2 This is a flowchart illustrating the steps of an embodiment of a false alarm suppression method for ground-based telescope observations according to this application;

[0022] Figure 3 This is a flowchart illustrating the steps of another embodiment of the false alarm suppression method for ground-based telescope observations in this application;

[0023] Figure 4 This is a flowchart illustrating the steps of another embodiment of the false alarm suppression method for ground-based telescope observations in this application;

[0024] Figure 5 This is a flowchart illustrating the steps of another embodiment of the false alarm suppression method for ground-based telescope observations in this application;

[0025] Figure 6 This is a schematic diagram of a classification confusion matrix provided in an embodiment of this application;

[0026] Figure 7 This is a schematic diagram of the structure of a device provided in an embodiment of this application. Detailed Implementation

[0027] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0028] Reference Figure 1 This is a flowchart illustrating the steps of an embodiment of a false alarm suppression method for ground-based telescope observations according to this application, including the following steps:

[0029] Step 101: Construct a false alarm dataset and obtain multiple false alarm categories; the false alarm dataset includes multiple false alarm data, and the false alarm categories include skylight, thermal pixels, trajectory consistency, and mapping consistency between star images and motion;

[0030] In this embodiment, the post-detection suppression method of the present invention adopts a reverse approach, that is, the false alarm data output by the detection algorithm is used as the research object. Specifically, the false alarm data output by the detection algorithm is used as the false alarm data. In order to distinguish false alarms in different observation scenarios, it is also necessary to clarify various false alarm categories such as sky light, thermal pixels, trajectory consistency, and mapping consistency between star images and motion, so as to lay the foundation for subsequent false alarm suppression research.

[0031] In one embodiment of this application, constructing the false alarm dataset includes:

[0032] Acquire multiple frames of observation images;

[0033] Based on the centroid, trajectory tracking and detection are performed on the observed image to obtain the space target; the space target is the trajectory formed by the moving target in multiple frames of observed images, and the trajectory includes multiple trajectory points corresponding to star images;

[0034] By processing the trajectory points using the software Extractor, the background region corresponding to the star image, the star image region corresponding to the star image, the centroid coordinates of the star image in the observed image, and the connected domain of the star image are obtained.

[0035] The full width at half maximum (FWHM) of the star image in the observed image is calculated based on the background region corresponding to the star image, the star image region corresponding to the star image, and the centroid coordinates of the star image in the observed image.

[0036] The phase angle and eccentricity of the star image are calculated based on the connected domain of the star image.

[0037] The right ascension and declination of the centroid coordinates are determined by astronomical positioning, and the initial orbit is determined by the least squares method.

[0038] The initial trajectory is matched with the two rows of orbital elements to identify whether the spatial target is a false alarm data.

[0039] The false alarm dataset is constructed based on the identified multiple false alarm data.

[0040] Specifically, to construct the false alarm dataset, image data under different seasons and climates needs to be collected uniformly, including various observation images under conditions such as low visibility due to sandstorms, high humidity, clouds and fog, moonlight background, and excellent observation conditions. Data under extreme conditions is removed through manual review, and the remaining data undergoes image preprocessing such as denoising and background removal. Then, a centroid-based trajectory tracking detection method is used to obtain the detection results, with each detected trajectory representing a spatial target.

[0041] Secondly, the detected trajectory point star images are processed by the software Extractor to complete the source extraction and centroid localization of each star image, obtain the star image region corresponding to the star image, the centroid coordinates of the star image in the observed image, and the connected domain of the star image, and calculate the full width at half maximum (FWHM) size of the star image in each frame image, and calculate the phase angle and eccentricity of the star image based on the connected domain of the star image. It should be noted that in the embodiments of this application, there is no limitation on the calculation of the phase angle and eccentricity of the star image.

[0042] Then, the right ascension and declination of each centroid are obtained using astronomical positioning methods. The initial orbit is determined using the least squares method. The obtained initial orbit is matched with the publicly available two-line element set (TLE) to confirm whether the space target is a false alarm. At this point, the space targets that are false alarms can be identified as false alarm data and labeled. Finally, a one-dimensional false alarm dataset is constructed based on these labeled space targets (false alarm data) for false alarm modeling research.

[0043] Reference Figure 2 This is a schematic diagram of data format conversion provided in an embodiment of this application, which specifically illustrates the conversion process from image data (observation image) to tabular data (false alarm data table).

[0044] It should be noted that during the manual review process of constructing the false alarm dataset, in order to meet the basic requirements of most detection algorithms, it is necessary to ensure that the length of the continuous image observation sequence obtained after final screening is at least 10 frames.

[0045] Step 102: Obtain the trajectory quality features corresponding to the false alarm dataset; the trajectory quality features include trajectory length, detection queue frame count, exposure time, shooting interval time, average trajectory velocity, trajectory direction, and average star image size of each trajectory point;

[0046] In this embodiment of the application, in order to accurately distinguish the categories of false alarm data and explore the characteristic differences of different false alarms, it is necessary to obtain trajectory quality features related to the trajectory attributes, observation equipment parameters, and star image features of the false alarm data. Trajectory quality features include, but are not limited to, trajectory length, detection queue frame count, exposure time, shooting interval time, average trajectory velocity, trajectory direction, and average star image size of each trajectory point.

[0047] Specifically, the trajectory length is used to count the total number of trajectory points contained in a single false alarm trajectory, reflecting the continuous coverage of the trajectory in multiple frames of images; the detection queue frame count is used to record the number of image frames continuously captured by the observation equipment for this trajectory, i.e., the total number of observation images corresponding to the trajectory; the preset parameters of the search satellite observation equipment can be directly read, including the exposure time and the shooting interval time. The exposure time is the exposure duration of a single frame image, and the shooting interval time is the shooting time difference between two adjacent frames; the average velocity of the trajectory is calculated based on the right ascension and declination coordinates of the trajectory points in multiple frames of images, the positional difference between adjacent trajectory points is calculated, and the average velocity of the trajectory is obtained by combining the shooting interval time; the trajectory direction is determined by the coordinate change trend of continuous trajectory points, which determines the direction of the trajectory's movement in the celestial coordinate system (such as the direction of increasing right ascension, the direction of decreasing declination, etc.); the average star image size of each trajectory point is calculated by extracting the full width at half maximum (FWHM) of the star images of all trajectory points on each trajectory and calculating its average value, reflecting the average size characteristics of the star image during the observation process. It can be understood that the star image size refers to the number of pixels contained in the area of ​​the star image.

[0048] Step 103: Determine the skylight recognition coefficient and target recognition coefficient corresponding to the false alarm data based on the skylight.

[0049] In this embodiment of the application, "sky light" refers to diffuse light that covers the background area of ​​the observation image in the Earth-Moon space observation scene, formed by atmospheric scattering (such as Earth's atmospheric glow, lunar reflected light), cosmic background radiation, or stray light from the observation equipment itself. It can interfere with the clear extraction of star images and easily form background false alarms without clear trajectories. In order to quantify the degree of interference of sky light on false alarm data, distinguish between false alarms caused by sky light and real target star images, and provide a quantitative basis for subsequent determination of sky light-related false alarms based on coefficients, it is necessary to determine the corresponding sky light recognition coefficient and target recognition coefficient based on the sky light.

[0050] In one embodiment of this application, step 103, determining the skylight recognition coefficient and target recognition coefficient corresponding to the false alarm data based on the skylight, includes:

[0051] Obtain each non-zero pixel in the background area corresponding to the star image, the global pixel median of the current frame, each non-zero pixel in the star image area corresponding to the star image, and the number of track points detected on the trajectory;

[0052] Based on each non-zero pixel in the background area corresponding to the star image and the global pixel median of the current frame, calculate the first evaluation coefficient for the background area corresponding to the star image to be identified as a skylight area;

[0053] Based on each non-zero pixel in the background area corresponding to the star image and each non-zero pixel in the star image area corresponding to the star image, calculate the second evaluation coefficient for the star image area corresponding to the star image to be identified as a real target.

[0054] The skylight recognition coefficient is calculated based on the trajectory and the first evaluation coefficient;

[0055] The target recognition coefficient is calculated based on the trajectory and the second evaluation coefficient;

[0056] The first evaluation coefficient is calculated using the following formula:

[0057]

[0058] in, This represents the background area corresponding to the star image. This represents each non-zero pixel within the background area corresponding to the star image. This represents the median global pixel value in the current frame. The first evaluation coefficient indicates that the background area corresponding to the star image is identified as the skylight area;

[0059] The second evaluation coefficient is calculated using the following formula:

[0060]

[0061] in, This indicates the star image region corresponding to the star image. OBJ represents each non-zero pixel within the star image region corresponding to the star image, and represents the second evaluation coefficient for the star image region corresponding to the star image to be identified as a real target.

[0062] The skylight recognition coefficient is calculated using the following formula:

[0063]

[0064] The target recognition coefficient is calculated using the following formula:

[0065]

[0066] in, For the trajectory, The skylight recognition coefficient is the coefficient used to identify the skylight. The target recognition coefficient is... This represents the first evaluation coefficient of the k-th trajectory point in the trajectory. This represents the second evaluation coefficient of the k-th trajectory point in the trajectory. This indicates the number of trajectory points detected on the trajectory.

[0067] In this embodiment of the application, the sky light recognition coefficient refers to the SKY coefficient corresponding to the entire trajectory, and the target recognition coefficient refers to the OBJ coefficient corresponding to the entire trajectory. Therefore, in order to calculate the corresponding sky light recognition coefficient and target recognition coefficient, it is necessary to calculate the SKY coefficient and OBJ coefficient corresponding to each trajectory point on the entire trajectory, namely the first evaluation coefficient and the second evaluation coefficient.

[0068] Specifically, the SKY coefficient for each trajectory point is calculated using the following formula (1):

[0069] (1)

[0070] in, This indicates the background area where the detected target is located (the background area corresponding to the star image). express Each non-zero pixel within the area (each non-zero pixel within the background area corresponding to the star image). This represents the median global pixel value in the current frame. The coefficient represents the current The region identification is the first evaluation coefficient for the skylight region.

[0071] It should be noted that the region The determination of the scope should adhere to the following two principles:

[0072] (1) Not too small: The area of ​​the star image in the region should be kept away from being too large, which would lead to The selection of the median for the region is influenced by celestial configurations;

[0073] (2) It should not be too large: When selecting When the area exceeds the actual skylight area, it results in the inclusion of a large amount of non-skylight area background, thus the selected median will not be comprehensive.

[0074] The OBJ coefficient for each trajectory point is calculated using the following formula (2):

[0075] (2)

[0076] in, Indicates the target star image region (the star image region corresponding to the star image). The OBJ coefficient represents the non-zero pixels within the star image area (each non-zero pixel in the star image area corresponding to the star image). Region identification is the second evaluation coefficient for real targets (space targets or stars).

[0077] Reference Figure 3 This is a schematic diagram illustrating the calculation of a first evaluation coefficient and a second evaluation coefficient according to an embodiment of this application. Specifically, region 1 is region... Area 2 is the area The arrows indicate the calculation direction, and the endpoints are the median values ​​of each region, specifically including the global pixel median value of the current frame.

[0078] Reference Figure 4 This diagram illustrates the calculation of the first and second evaluation coefficients for typical targets in a skylight background, as provided in an embodiment of this application. The SKY coefficient (first evaluation coefficient) and OBJ coefficient (second evaluation coefficient) designed in this embodiment can effectively identify skylight areas and separate real targets from false alarms in the background. Specifically... Figure 4 The image shows the first and second evaluation coefficients under halo and cloud cover conditions, respectively.

[0079] Ultimately, regarding the detected trajectory Its SKY coefficient can be expressed as the following formula (3):

[0080] (3)

[0081] For the detected trajectory Its OBJ coefficient can be expressed as the following formula (4):

[0082] (4)

[0083] in, and For the current trajectory The correlation coefficients of each trajectory point detected are averaged to represent the SKY and OBJ coefficients of the entire trajectory, namely the skylight recognition coefficient and the target recognition coefficient. and This represents the correlation coefficient (first evaluation coefficient and second evaluation coefficient) of the k-th trajectory point in the trajectory. Representing the trajectory The number of trajectory points detected.

[0084] Step 104: Determine the hot pixel discrimination coefficient corresponding to the false alarm data based on the hot pixels;

[0085] In this embodiment, a hot pixel refers to an isolated bright spot that appears continuously in a fixed pixel position, has abnormal brightness, and no motion trajectory in multiple frames of observation images due to the increased device temperature and circuit noise during the operation of a search satellite observation device (such as a telescope imaging sensor). It is easily misidentified as a static space target, forming a fixed-position false alarm. In order to accurately identify false alarms caused by hot pixels in the false alarm data, distinguish hot pixel bright spots from real target star images (with motion characteristics), and provide quantitative judgment indicators for subsequent targeted suppression of hot pixel false alarms, it is necessary to determine the corresponding hot pixel discrimination coefficient based on the hot pixel.

[0086] In one embodiment of this application, step 104, determining the hot pixel discrimination coefficient corresponding to the false alarm data based on the hot pixels, includes:

[0087] Obtain the star image size corresponding to the trajectory point and the pixel point corresponding to the star image; the pixel point includes a pixel peak point and the pixel peak point has a corresponding surrounding neighborhood.

[0088] Obtain the median value of the pixels in the surrounding neighborhood after removing zero-value points;

[0089] The pixels corresponding to the star image are normalized based on the pixel median to obtain the normalized pixels corresponding to the star image.

[0090] Obtain the standard deviation of the remaining pixels in the normalized surrounding neighborhood after removing zero-value points;

[0091] The third evaluation coefficient is calculated based on the star image size corresponding to the trajectory point, the pixel peak point corresponding to the normalized star image, and the standard deviation of the remaining pixels in the normalized surrounding neighborhood after removing zero points.

[0092] The thermal pixel discrimination coefficient is calculated based on the trajectory and the third evaluation coefficient;

[0093] The third evaluation coefficient is calculated using the following formula:

[0094]

[0095] in, This represents the third evaluation coefficient of the k-th trajectory point in the trajectory. This represents the size of the star image at the k-th point in the trajectory. This represents the standard deviation of the remaining pixels in the normalized surrounding neighborhood after removing zero-value points. The pixel peak points corresponding to the normalized star image;

[0096] The thermal pixel discrimination coefficient is calculated using the following formula:

[0097]

[0098] in, The thermal pixel discrimination coefficient, For the trajectory, This indicates the number of trajectory points detected on the trajectory.

[0099] In this embodiment of the application, the hot pixel discrimination coefficient refers to the HOT coefficient corresponding to the entire trajectory. Therefore, in order to calculate the corresponding hot pixel discrimination coefficient, it is necessary to calculate the HOT coefficient corresponding to each trajectory point on the entire trajectory, that is, the third evaluation coefficient.

[0100] Specifically, assuming the peak pixel points of the target detected after background removal processing are The 8-neighborhood surrounding the peak point is S (surrounding neighborhood), and the median value of the pixels in S after removing the zero-value points is... If we normalize all pixels in S with respect to the median, then we have: , The third evaluation coefficient is then obtained using the following formula (5):

[0101] (5)

[0102] in, This represents the standard deviation of the remaining pixels in S after removing zero-value points in the normalized dataset. These are the pixel peak points corresponding to the normalized star image.

[0103] Meanwhile, based on the single-point mutation of hot pixels, the area of ​​the star image detected by template convolution will not be too large, usually not exceeding 5 pixels. In order to further optimize the quality of sample data, prior information is added to the feature calculation. Formula (5) can be updated to the following formula (6), that is, the third evaluation coefficient in the embodiment of this application is obtained by the following formula (6):

[0104] (6)

[0105] in, This represents the HOT coefficient (third evaluation coefficient) of the k-th trajectory point in the trajectory. This represents the size of the star image at the k-th trajectory point. For trajectory points larger than 5 pixels, it can be assumed that they are not hot pixels due to prior knowledge and therefore do not participate in the HOT coefficient evaluation. Thus, their HOT coefficient is directly set to 1, indicating that there are no peak pixels within their surrounding neighborhood S. Finally, for the detected trajectory... Its HOT coefficient (hot pixel discrimination coefficient) The result can be obtained using the following formula (7):

[0106] (7)

[0107] in, The HOT coefficient (hot pixel discrimination coefficient) is the coefficient for the entire trajectory, obtained by analyzing the current trajectory. The HOT coefficients of each trajectory point detected are averaged to represent the HOT coefficient of the entire trajectory. This represents the HOT coefficient (third evaluation coefficient) of the k-th trajectory point in the trajectory. Representing the trajectory The number of trajectory points detected.

[0108] Step 105: Determine the spatiotemporal geometric consistency, grayscale consistency, and velocity magnitude consistency corresponding to the false alarm data based on the trajectory consistency.

[0109] In this embodiment, trajectory consistency refers to the degree of continuous matching of the spatial position change, star image grayscale features, and motion speed of the trajectory of false alarm data in multiple frames of observed images. That is, the real target trajectory usually has the characteristics of continuous position without jumps, stable grayscale without sudden changes, and uniform speed without sudden changes. However, false alarm trajectories often show abnormal breaks or fluctuations in spacetime, grayscale, or speed. In order to quantify the continuity and rationality of the trajectory from the three core dimensions of spacetime, grayscale, and speed, and to distinguish between real target trajectories with stable consistency and false alarm trajectories with consistency anomalies, it is necessary to determine the corresponding spacetime geometric consistency, grayscale consistency, and speed magnitude consistency based on the trajectory consistency.

[0110] In one embodiment of this application, step 105, determining the spatiotemporal geometric consistency, grayscale consistency, and velocity magnitude consistency corresponding to the false alarm data based on the trajectory consistency, includes the following steps:

[0111] In one embodiment of this application, determining the spatiotemporal geometric consistency corresponding to the false alarm data based on the trajectory consistency includes:

[0112] Obtain the position of the trajectory point;

[0113] The trajectory is fitted using the least squares method to obtain a straight line model;

[0114] The spatiotemporal geometric consistency is calculated based on the position of the trajectory points, the straight line model, and the number of trajectory points detected on the trajectory.

[0115] The spatiotemporal geometric consistency is calculated using the following formula:

[0116]

[0117] in, This indicates the spatiotemporal geometric consistency. Let A, B, and C be the positions of the trajectory points, and let A, B, and C be the model parameters of the straight line model. This indicates the number of trajectory points detected on the trajectory.

[0118] In one embodiment of this application, determining the grayscale consistency corresponding to the false alarm data based on the trajectory consistency includes:

[0119] Obtain the grayscale value of each trajectory point, and integrate the grayscale values ​​of each trajectory point on the trajectory into a grayscale set;

[0120] Based on the gray values ​​of each trajectory point, determine the mean gray value and the standard deviation of the gray value corresponding to the gray value set;

[0121] The grayscale consistency is calculated based on the mean grayscale value and the standard deviation of grayscale values ​​corresponding to the grayscale set.

[0122] The grayscale consistency is calculated using the following formula:

[0123]

[0124] in, This indicates the grayscale consistency. This represents the mean gray level corresponding to the gray level set. This represents the standard deviation of the grayscale values ​​corresponding to the grayscale set.

[0125] In one embodiment of this application, determining the speed magnitude consistency corresponding to the false alarm data based on the trajectory consistency includes:

[0126] Obtain the velocity magnitude of each trajectory point and integrate the velocity magnitudes of each trajectory point on the trajectory into a trajectory point velocity set;

[0127] Calculate the mean velocity and standard deviation of the velocity set corresponding to the trajectory points based on the velocity magnitude of each trajectory point;

[0128] The consistency of velocity magnitude is calculated based on the mean velocity and the standard deviation of the velocity corresponding to the set of trajectory points.

[0129] The consistency of the speed magnitude is calculated using the following formula:

[0130]

[0131] in, This indicates that the speed magnitudes are consistent. This represents the average velocity of the set of velocities of the trajectory points. This represents the standard deviation of the velocity set of the trajectory points.

[0132] In this embodiment of the application, it is assumed that the trajectory output by the detection algorithm is The linear model fitted by least squares is as follows: Then A, B, and C are the model parameters of the linear model. In this case, the spatiotemporal geometric consistency is... It can be obtained by the following formula (8):

[0133] (8)

[0134] in, Let A, B, and C be the positions of the trajectory points, and let A, B, and C be the model parameters of the straight line model. This represents the average spatial distance between each trajectory point and the fitted trajectory, i.e., spatiotemporal geometric consistency. Representing the trajectory The number of trajectory points detected.

[0135] However, for synchronously orbiting targets moving at extremely slow speeds in images, the centroid positioning error can significantly impact the trajectory fitting results, leading to a large deviation between the fitted trajectory and the true trajectory. Therefore, the fitted trajectory is unsuitable as a decision-making basis. To address this, this application's embodiments establish a slow-motion protection zone. Specifically, for targets in the detection queue with a longest movement distance of less than 10 pixels, it is assumed that they satisfy spatiotemporal geometric consistency, and... .

[0136] In this embodiment of the application, the grayscale of each trajectory point is That is, a grayscale set, where grayscale uniformity is achieved. It can be obtained by the following formula (9):

[0137] (9)

[0138] in, Represents a set The mean gray level of (a grayscale set). Represents a set The standard deviation of gray levels. Specifically, The smaller the value, the stronger the grayscale consistency.

[0139] In this embodiment of the application, the position of each trajectory point in the pixel coordinate system is: At this point, the magnitude of the velocity of the kth trajectory point is obtained by the following formula (10):

[0140] (10)

[0141] in, The unit is pixel / frame. The difference in frame number between the k-th trajectory point and the (k-1)-th trajectory point yields the velocity set of the trajectory points. Then the speed is consistent It can be obtained through the following formula (11):

[0142] (11)

[0143] in, Represents a set The average velocity of (the set of velocities of trajectory points). Represents a set The speed standard deviation. Specifically, The smaller the value, the stronger the consistency of its speed.

[0144] Similar to the assessment of spatiotemporal geometric consistency, this also faces the problem of large measurement errors in the velocity of slowly moving targets. Therefore, by setting the same low-speed motion protection zone, targets with an average motion velocity of less than 0.5 pixels / frame are assumed to satisfy velocity magnitude consistency, and... .

[0145] Step 106: Determine the consistency between the major axis and velocity magnitude, and the consistency between the major axis and velocity direction, corresponding to the false alarm data, based on the mapping consistency between the star image and motion.

[0146] In this embodiment, the consistency of the mapping between star image and motion refers to the inherent correlation between the morphological characteristics of the real target star image (especially the direction and length of the major axis) and the motion state (velocity magnitude and velocity direction) in Earth-Moon space observation. Because the real moving target is in motion relative to the observation equipment, the star image is likely to be an ellipse stretched along the direction of motion, and the direction of the major axis is usually consistent with the velocity direction, and the length of the major axis will increase accordingly with the increase of velocity magnitude. However, false alarm star images (such as sky light interference and hot pixels) often do not have this matching pattern. In order to quantify the rationality of false alarm data from the correlation dimension of star image morphology-motion state, distinguish between real targets with matching star image and motion mapping and false alarm targets with disordered mapping, and provide a two-dimensional (size and direction) judgment index for subsequent accurate identification of false alarms with mismatched star image and motion, it is necessary to determine the consistency between the corresponding major axis and velocity magnitude, as well as the consistency between the major axis and velocity direction, based on the consistency of the mapping between star image and motion.

[0147] In one embodiment of this application, step 106, determining the consistency between the major axis and velocity magnitude, and the consistency between the major axis and velocity direction, corresponding to the false alarm data based on the mapping consistency between the star image and motion, includes:

[0148] Obtain the set of motion velocities at the center of mass of the star and the velocities of the trajectory points;

[0149] The displacement length of the trajectory point during the exposure time is calculated based on the motion velocity at the centroid of the star and the velocity set of the trajectory point.

[0150] The major axis and minor axis of the star image are calculated based on the velocity of the star image at its centroid and the displacement length of the trajectory point within the exposure time.

[0151] Obtain the phase angle of the major axis and the phase angle of the motion direction for each trajectory point;

[0152] A fourth evaluation coefficient is calculated based on the major axis of the star image, the minor axis of the star image, the exposure time, and the velocity of motion at the center of mass of the star image.

[0153] The fifth evaluation coefficient is calculated based on the phase angle of the major axis direction and the phase angle of the motion direction;

[0154] The consistency between the major axis and the velocity magnitude is calculated based on the fourth evaluation coefficient and the number of trajectory points detected on the trajectory.

[0155] The consistency between the major axis and the velocity direction is calculated based on the fifth evaluation coefficient and the number of trajectory points detected on the trajectory;

[0156] The fourth evaluation coefficient is calculated using the following formula:

[0157]

[0158] in, This represents the fourth evaluation coefficient for the k-th trajectory point. The major axis of the star image at the k-th trajectory point is represented. The minor axis of the star image at the k-th trajectory point is represented. This indicates the exposure time. This indicates the shooting interval time. This represents the velocity at the centroid of the star in the k-th frame;

[0159] The fifth evaluation coefficient is calculated using the following formula:

[0160]

[0161] in, This represents the fifth evaluation coefficient for the k-th trajectory point. This represents the phase angle along the major axis of the k-th trajectory point. This represents the phase angle of the motion direction of the k-th trajectory point;

[0162] The consistency between the major axis and the magnitude of the velocity is calculated using the following formula:

[0163]

[0164] The alignment of the major axis with the velocity direction is calculated using the following formula:

[0165]

[0166] in, This indicates that the major axis is consistent with the magnitude of the velocity. This indicates that the major axis is aligned with the velocity direction. This indicates the number of trajectory points detected on the trajectory.

[0167] In this embodiment of the application, it is assumed that the camera exposure time is The shooting interval is , , , Let the positions of the centroids of the (k-1), k, and k+1 trajectory points be respectively. Then, the velocity at the centroid of the star image in the k-th frame can be calculated using the following formula (12):

[0168] (12)

[0169] In this application scenario, calculating the centroid velocity is more accurate than frame-by-frame calculation, but the above calculation requires k-1, k, and k+1 to be three consecutive frames. For cases where the last frame or the preceding or following frame has gaps, frame-by-frame calculation is still used.

[0170] Assuming the set of velocities of the trajectory points is obtained The speed unit is pixel / frame. The displacement length of the k-th trajectory point during the exposure time is calculated by the following formula (13):

[0171] (13)

[0172] like Figure 5 The diagram shown illustrates the relationship between a star image and its displacement according to an embodiment of this application. Based on the camera integration principle, it can be known that the target... The displacement within the target is equal to the distance between the target's centroids at the start and end of the camera exposure. Due to the camera's PSF (Point Spread Function) and atmospheric effects, a stationary target still occupies a certain circular area in the image; let its radius be... , combined Figure 5 It can be seen that, for the kth trajectory point, the major axis of the star image and the displacement during the exposure time satisfy the following formula (14):

[0173] (14)

[0174] in, The major axis of the star image at the k-th trajectory point is represented. This represents the original radius of the k-th trajectory point in a stationary state.

[0175] During the camera's integration process, the star image of a target moving in a straight line within the field of view expands along the direction of motion, becoming the major axis of an ellipse, while the size of the star image perpendicular to the direction of motion remains unchanged, ultimately resulting in the minor axis of the ellipse. Therefore, the following equation holds true.

[0176] (15)

[0177] in, Let represent the minor axis of the star image at the k-th trajectory point. Therefore, formula (15) can be transformed into the following formula (16).

[0178] (16)

[0179] Therefore, the relationship between the major axis and minor axis of the star image and the magnitude of the velocity and the shooting system time is obtained as follows (17):

[0180] (17)

[0181] in, The unit is pixel / frame. Furthermore, based on the above analysis, it can be known that the major axis of the target is parallel to the direction of motion. Assuming the phase angle of the major axis at the k-th trajectory point is... The phase angle of its direction of motion is Then the following formula (18) holds true:

[0182] (18)

[0183] The phase angle of the direction of motion is calculated using the following formula (19):

[0184] (19)

[0185] In this embodiment of the application, a function is defined. The value of satisfies the conditions restricted by the following formula (20):

[0186] (20)

[0187] in, , However, the specific direction of the major axis can be either positive or negative, making further distinction impossible. Therefore, when comparing it with the direction of motion velocity, It should be converted to The comparison is made within a certain range, only determining whether they are parallel, without distinguishing between positive and negative directions.

[0188] Due to factors such as atmospheric turbulence, large field-of-view distortion, and inaccuracies in measuring faint star images, formulas (17) and (18) are not strictly valid, but have different degrees of error in different scenarios. This application aims to use this error as an indicator for false alarm detection. Under normal circumstances, the star image of a space target is in an unsaturated state. Therefore, under the premise of a certain exposure time, the length of its star image's major axis should be proportional to the magnitude of the motion velocity. Using this relationship, a consistency index between the target's major axis and the magnitude of the velocity can be established. In this application embodiment, the consistency between the major axis and the magnitude of the velocity refers to the LEN coefficient corresponding to the trajectory. Therefore, it needs to be obtained by calculating the LEN coefficient (fourth evaluation coefficient) of each trajectory point on the trajectory. Similarly, the consistency between the major axis and the velocity direction refers to the DIR coefficient corresponding to the trajectory. Therefore, it needs to be obtained by calculating the DIR coefficient (fifth evaluation coefficient) of each trajectory point on the trajectory.

[0189] Specifically, the fourth evaluation coefficient (LEN coefficient of each trajectory point on the trajectory) is calculated using the following formula (21):

[0190] (twenty one)

[0191] when The closer the coefficient (fourth evaluation coefficient) is to 0, the more accurate the measurement. Similarly, the direction of the major axis of the actual measured target should also have an error angle with the direction of its motion velocity. Therefore, the fifth evaluation coefficient (DIR coefficient of each trajectory point on the trajectory) is calculated by the following formula (22):

[0192] (twenty two)

[0193] The operation is defined as follows (23):

[0194] (twenty three)

[0195] in, .when The closer the coefficient is to 0, the more accurate the measurement. Ultimately, the detected trajectory... ,That coefficients and sum The coefficient can be expressed as follows (24)

[0196] (twenty four)

[0197] in, and These are, respectively, the consistency between the major axis and the magnitude of the velocity, and the consistency between the major axis and the direction of the velocity, i.e., through the current trajectory. The trajectory points detected in coefficients and sum The coefficients are averaged to represent the entire trajectory. coefficients and sum coefficient, and This represents the k-th trajectory point in the trajectory. coefficients and sum coefficient, Representing the trajectory The number of trajectory points detected.

[0198] Through steps 101 to 106 above, multiple false alarm features can be obtained. Specifically, the false alarm features are classified according to each false alarm category. The summary of each false alarm feature and its corresponding function is shown in Table 1 below:

[0199] Table 1 Characteristics of False Alarms

[0200]

[0201] Step 107: Input the extracted features, namely the trajectory quality features, the skylight recognition coefficient, the target recognition coefficient, the thermal pixel discrimination coefficient, the spatiotemporal geometric consistency, the grayscale consistency, the velocity magnitude consistency, the major axis and velocity magnitude consistency, and the major axis and velocity direction consistency, into the classifier to identify target false alarms among multiple false alarm data, so as to suppress the target false alarms.

[0202] In this embodiment of the application, the calculation results corresponding to each false alarm feature in Table 1 above are used as the input of the vector machine classifier. By performing calculation and classification through the vector machine, it can be determined whether the false alarm data belongs to the real target or to the false alarm. Then, the false alarm data that belongs to the false alarm is identified as the target false alarm, so as to suppress the target false alarm.

[0203] To enable those skilled in the art to better understand the false alarm suppression method for ground-based telescope observations provided in the embodiments of this application, the following examples are used for illustration:

[0204] Using the research group's telescope as a research platform, this application analyzes and classifies the main sources of false alarms in different observation scenarios under staring mode, describes each type of false alarm as a corresponding theoretical problem, and designs corresponding false alarm indices. Based on this, a general method for suppressing false alarms after detection is proposed. This method detects various false alarm indices for unknown targets, uses these as features, and employs a support vector machine for binary classification to achieve the purpose of separating false alarms.

[0205] Two hundred thousand observation images of the telescope in different scenarios under staring mode were collected to create a corresponding false alarm dataset, as shown in Table 2 below. Based on the application of traditional algorithms on this dataset, it can be seen that the data volume, total number of detected targets, and number of false alarms correspond to different data sources.

[0206] Table 2 Application of traditional algorithms on datasets

[0207]

[0208] To verify the robustness of the false alarm suppression method provided in this application embodiment and to ensure its applicability to different observation environments, the following cross-experiment was designed in this application embodiment:

[0209] (1) Training set: Select 50% of the target trajectories detected in each time period (data source) in Table 2 as the training set, and ensure that at least 30% of false alarms and 30% of spatial targets are selected in each time period; then mix the training data extracted from different time periods for training.

[0210] (2) Validation set: The trained model is applied to the data in each time period (data source) in Table 2 to analyze the classification results.

[0211] This cross-experiment verifies the false alarm suppression method for ground-based telescope observations shown in the embodiments of this application, and the results are as follows: Figure 6 The results are shown.

[0212] like Figure 6 The diagram shown is a schematic of a classification confusion matrix provided in an embodiment of this application. The classification confusion matrix represents the comprehensive test results on sample data from various observation time periods. The horizontal axis represents the true category, and the vertical axis represents the predicted category of the false alarm suppression method in this embodiment. Different grayscale blocks are used to distinguish different data. Two blocks 1 represent the number and total percentage of correctly identified spatial targets and false alarms, respectively; two blocks 2 represent the number and total percentage of incorrectly identified samples; four blocks 3 represent the percentage of correctly identified targets in that category; and one block 4 represents the total number and percentage of correctly identified targets. This shows that the recognition accuracy reaches 91.0%, further demonstrating the robustness of the false alarm suppression method in this embodiment.

[0213] The false alarm suppression method for ground-based telescope observations provided in this application adopts a reverse approach to post-detection suppression. It studies the detection results (false alarm dataset) and can fully acquire false alarm features (trajectory quality features, skylight recognition coefficient, target recognition coefficient, thermal pixel discrimination coefficient, spatiotemporal geometric consistency, grayscale consistency, velocity magnitude consistency, major axis and velocity magnitude consistency, and major axis and velocity direction consistency) for different false alarm categories. Based on the false alarm features, it overcomes the uncertainty of traditional empirical threshold-based discrimination methods. It can not only effectively identify false alarms in various scenarios, but also prevent false suppression of real targets and missed alarms. It is applicable to post-detection false alarm suppression under different observation environments, with good false alarm suppression effect and a false alarm recognition accuracy of up to 91.0%, and has strong robustness.

[0214] It should be noted that, for the sake of simplicity, the method embodiments are all described as a series of actions. However, those skilled in the art should understand that the embodiments of this application are not limited to the described order of actions, because according to the embodiments of this application, some steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also understand that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily required by the embodiments of this application.

[0215] This application also provides a non-volatile readable storage medium storing one or more modules (programs). When these modules are applied to a device, they enable the device to execute the instructions for the method steps in this application.

[0216] This application provides one or more machine-readable media storing instructions that, when executed by one or more processors, cause an electronic device to perform one or more of the methods described in the above embodiments. In this application, the electronic device includes various types of devices such as terminal devices and servers (clusters).

[0217] The embodiments of this disclosure can be implemented as an apparatus configured as desired using any suitable hardware, firmware, software, or any combination thereof, including electronic devices such as terminal devices, servers (clusters), etc. Figure 7 An exemplary apparatus 700 is schematically shown that can be used to implement the various embodiments described in this application.

[0218] In one embodiment, Figure 7An exemplary device 700 is shown, which includes one or more processors 702, a control module (chipset) 704 coupled to at least one of the processors 702, a memory 706 coupled to the control module 704, a non-volatile memory (NVM) / storage device 708 coupled to the control module 704, one or more input / output devices 710 coupled to the control module 704, and a network interface 712 coupled to the control module 704.

[0219] Processor 702 may include one or more single-core or multi-core processors, and processor 702 may include any combination of general-purpose processors or special-purpose processors (e.g., graphics processors, application processors, baseband processors, etc.). In some embodiments, device 700 can serve as a terminal device, server (cluster), or other device as described in the embodiments of this application.

[0220] In some embodiments, apparatus 700 may include one or more computer-readable media (e.g., memory 706 or NVM / storage device 708) having instructions 714 and one or more processors 702 that are combined with the one or more computer-readable media and configured to execute the instructions 714 to implement the module and thus perform the actions described in this disclosure.

[0221] In one embodiment, the control module 704 may include any suitable interface controller to provide any suitable interface to at least one of the processors 702 and / or any suitable device or component communicating with the control module 704.

[0222] The control module 704 may include a memory controller module to provide an interface to the memory 706. The memory controller module may be a hardware module, a software module, and / or a firmware module.

[0223] Memory 706 may be used, for example, to load and store data and / or instructions 714 for device 700. In one embodiment, memory 706 may include any suitable volatile memory, such as suitable DRAM. In some embodiments, memory 706 may include double data rate type quad synchronous dynamic random access memory (DDR4 SDRAM).

[0224] In one embodiment, the control module 704 may include one or more input / output controllers to provide an interface to the NVM / storage device 708 and (one or more) input / output devices 710.

[0225] For example, NVM / storage device 708 may be used to store data and / or instructions 714. NVM / storage device 708 may include any suitable non-volatile memory (e.g., flash memory) and / or may include any suitable (one or more) non-volatile storage devices (e.g., one or more hard disk drive (HDD), one or more optical disc (CD) drives, and / or one or more digital universal optical disc (DVD) drives).

[0226] NVM / storage device 708 may include storage resources that are physically part of a device on which device 700 is mounted, or that are accessible to the device but do not necessarily have to be part of the device. For example, NVM / storage device 708 may be accessed via a network via one or more input / output devices 710.

[0227] One or more input / output devices 710 may provide an interface for device 700 to communicate with any other suitable device. Input / output devices 710 may include communication components, audio components, sensor components, etc. A network interface 712 may provide an interface for device 700 to communicate via one or more networks. Device 700 may wirelessly communicate with one or more components of a wireless network according to any of one or more wireless network standards and / or protocols, such as accessing a wireless network based on a communication standard, such as WiFi, 2G, 3G, 4G, 5G, etc., or a combination thereof.

[0228] In one embodiment, at least one of the processors 702 may be logically packaged with one or more controllers (e.g., memory controller modules) of the control module 704. In one embodiment, at least one of the processors 702 may be logically packaged with one or more controllers of the control module 704 to form a system-in-package (SiP). In one embodiment, at least one of the processors 702 may be integrated with the logic of one or more controllers of the control module 704 on the same die. In one embodiment, at least one of the processors 702 may be integrated with the logic of one or more controllers of the control module 704 on the same die to form a system-on-a-chip (SoC).

[0229] In various embodiments, device 700 may be, but is not limited to, a server, desktop computing device, or mobile computing device (e.g., laptop computing device, handheld computing device, tablet computer, netbook, etc.). In various embodiments, device 700 may have more or fewer components and / or different architectures. For example, in some embodiments, device 700 includes one or more cameras, a keyboard, a liquid crystal display (LCD) screen (including a touchscreen display), a non-volatile memory port, multiple antennas, a graphics chip, an application-specific integrated circuit (ASIC), and a speaker.

[0230] The detection device can use a main control chip as a processor or control module, and sensor data, position information, etc. can be stored in a memory or NVM / storage device. The sensor group can be used as an input / output device, and the communication interface can include a network interface.

[0231] As the device embodiment is basically similar to the method embodiment, the description is relatively simple, and relevant parts can be found in the description of the method embodiment.

[0232] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0233] This application describes embodiments with reference to flowchart illustrations and / or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable false alarm suppression terminal device for ground-based telescope observations to produce a machine, such that the instructions, which execute via the computer or other programmable false alarm suppression terminal device for ground-based telescope observations, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0234] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable false alarm suppression terminal device for ground-based telescope observations to operate in a specific manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including an instruction means, which is implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0235] These computer program instructions can also be loaded onto a computer or other programmable false alarm suppression terminal for ground-based telescope observations, causing a series of operational steps to be executed on the computer or other programmable terminal to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable terminal for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0236] Although preferred embodiments of the present application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present application.

[0237] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0238] The foregoing has provided a detailed description of a false alarm suppression method and apparatus for ground-based telescope observations, an electronic device, and a storage medium provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the method and its core ideas. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for suppressing false alarms in ground-based telescope observations, characterized in that, The method includes: Construct a false alarm dataset and obtain multiple false alarm categories; the false alarm dataset includes multiple false alarm data, and the false alarm categories include skylight, thermal pixels, trajectory consistency, and mapping consistency between star images and motion; Obtain the trajectory quality features corresponding to the false alarm dataset; the trajectory quality features include trajectory length, detection queue frame count, exposure time, shooting interval time, average trajectory velocity, trajectory direction, and average star image size of each trajectory point; Based on the skylight, determine the skylight recognition coefficient and target recognition coefficient corresponding to the false alarm data; Determine the hot pixel discrimination coefficient corresponding to the false alarm data based on the hot pixels; Based on the trajectory consistency, determine the spatiotemporal geometric consistency, grayscale consistency, and velocity magnitude consistency of the false alarm data; The consistency between the major axis and velocity magnitude, and the consistency between the major axis and velocity direction, are determined based on the mapping consistency between the star image and the motion. The trajectory quality features, the skylight recognition coefficient, the target recognition coefficient, the thermal pixel discrimination coefficient, the spatiotemporal geometric consistency, the grayscale consistency, the velocity magnitude consistency, the major axis and velocity magnitude consistency, and the major axis and velocity direction consistency are used as extracted features and input into the classifier to identify target false alarms among multiple false alarm data, so as to suppress the target false alarms.

2. The method according to claim 1, characterized in that, The construction of the false alarm dataset includes: Acquire multiple frames of observation images; Based on the centroid, trajectory tracking and detection are performed on the observed image to obtain the space target; the space target is the trajectory formed by the moving target in multiple frames of observed images, and the trajectory includes multiple trajectory points corresponding to star images; By processing the trajectory points using the software Extractor, the background region corresponding to the star image, the star image region corresponding to the star image, the centroid coordinates of the star image in the observed image, and the connected domain of the star image are obtained. The full width at half maximum (FWHM) of the star image in the observed image is calculated based on the background region corresponding to the star image, the star image region corresponding to the star image, and the centroid coordinates of the star image in the observed image. The phase angle and eccentricity of the star image are calculated based on the connected domain of the star image. The right ascension and declination of the centroid coordinates are determined by astronomical positioning, and the initial orbit is determined by the least squares method. The initial trajectory is matched with the two rows of orbital elements to identify whether the spatial target is a false alarm data. The false alarm dataset is constructed based on the identified multiple false alarm data.

3. The method according to claim 2, characterized in that, The step of determining the skylight recognition coefficient and target recognition coefficient corresponding to the false alarm data based on the skylight includes: Obtain each non-zero pixel in the background area corresponding to the star image, the global pixel median of the current frame, each non-zero pixel in the star image area corresponding to the star image, and the number of track points detected on the trajectory; Based on each non-zero pixel in the background area corresponding to the star image and the global pixel median of the current frame, calculate the first evaluation coefficient for the background area corresponding to the star image to be identified as a skylight area; Based on each non-zero pixel in the background area corresponding to the star image and each non-zero pixel in the star image area corresponding to the star image, calculate the second evaluation coefficient for the star image area corresponding to the star image to be identified as a real target. The skylight recognition coefficient is calculated based on the trajectory and the first evaluation coefficient; The target recognition coefficient is calculated based on the trajectory and the second evaluation coefficient; The first evaluation coefficient is calculated using the following formula: in, This represents the background area corresponding to the star image. This represents each non-zero pixel within the background area corresponding to the star image. This represents the median global pixel value of the current frame. The first evaluation coefficient indicates that the background area corresponding to the star image is identified as the skylight area; The second evaluation coefficient is calculated using the following formula: in, This indicates the star image region corresponding to the star image. OBJ represents each non-zero pixel within the star image region corresponding to the star image, and represents the second evaluation coefficient for the star image region corresponding to the star image to be identified as a real target. The skylight recognition coefficient is calculated using the following formula: The target recognition coefficient is calculated using the following formula: in, For the trajectory, The skylight recognition coefficient is the coefficient of the skylight. The target recognition coefficient is... This represents the first evaluation coefficient of the k-th trajectory point in the trajectory. This represents the second evaluation coefficient of the k-th trajectory point in the trajectory. This indicates the number of trajectory points detected on the trajectory.

4. The method according to claim 2, characterized in that, The step of determining the hot pixel discrimination coefficient corresponding to the false alarm data based on the hot pixels includes: Obtain the star image size corresponding to the trajectory point and the pixel point corresponding to the star image; the pixel point includes a pixel peak point and the pixel peak point has a corresponding surrounding neighborhood. Obtain the median value of the pixels in the surrounding neighborhood after removing zero-value points; The pixels corresponding to the star image are normalized based on the pixel median to obtain the normalized pixels corresponding to the star image. Obtain the standard deviation of the remaining pixels in the normalized surrounding neighborhood after removing zero-value points; The third evaluation coefficient is calculated based on the star image size corresponding to the trajectory point, the pixel peak point corresponding to the normalized star image, and the standard deviation of the remaining pixels in the normalized surrounding neighborhood after removing zero-value points. The thermal pixel discrimination coefficient is calculated based on the trajectory and the third evaluation coefficient; The third evaluation coefficient is calculated using the following formula: in, This represents the third evaluation coefficient of the k-th trajectory point in the trajectory. This represents the size of the star image at the k-th trajectory point in the trajectory. This represents the standard deviation of the remaining pixels in the normalized surrounding neighborhood after removing zero-value points. The pixel peak points corresponding to the normalized star image; The thermal pixel discrimination coefficient is calculated using the following formula: in, The thermal pixel discrimination coefficient, For the trajectory, This indicates the number of trajectory points detected on the trajectory.

5. The method according to claim 2, characterized in that, The step of determining the spatiotemporal geometric consistency corresponding to the false alarm data based on the trajectory consistency includes: Obtain the position of the trajectory point; The trajectory is fitted using the least squares method to obtain a straight line model; The spatiotemporal geometric consistency is calculated based on the position of the trajectory points, the straight line model, and the number of trajectory points detected on the trajectory. The spatiotemporal geometric consistency is calculated using the following formula: in, This indicates the spatiotemporal geometric consistency. Let A, B, and C be the positions of the trajectory points, and let A, B, and C be the model parameters of the straight line model. This indicates the number of trajectory points detected on the trajectory.

6. The method according to claim 2, characterized in that, The step of determining the grayscale consistency corresponding to the false alarm data based on the trajectory consistency includes: Obtain the grayscale value of each trajectory point, and integrate the grayscale values ​​of each trajectory point on the trajectory into a grayscale set; Based on the gray values ​​of each trajectory point, determine the mean gray value and the standard deviation of the gray value corresponding to the gray value set; The grayscale consistency is calculated based on the mean grayscale value and the standard deviation of grayscale values ​​corresponding to the grayscale set. The grayscale consistency is calculated using the following formula: in, This indicates the grayscale consistency. This represents the mean gray level corresponding to the gray level set. This represents the standard deviation of the grayscale values ​​corresponding to the grayscale set.

7. The method according to claim 2, characterized in that, The step of determining the speed magnitude consistency corresponding to the false alarm data based on the trajectory consistency includes: Obtain the velocity magnitude of each trajectory point and integrate the velocity magnitudes of each trajectory point on the trajectory into a trajectory point velocity set; Calculate the mean velocity and standard deviation of the velocity set corresponding to the trajectory points based on the velocity magnitude of each trajectory point; The consistency of velocity magnitude is calculated based on the mean velocity and the standard deviation of the velocity corresponding to the set of trajectory points. The consistency of the speed magnitude is calculated using the following formula: in, This indicates that the speed magnitudes are consistent. This represents the average velocity of the set of velocities of the trajectory points. This represents the standard deviation of the velocity set of the trajectory points.

8. The method according to claim 2, characterized in that, The step of determining the consistency between the major axis and velocity magnitude, and the consistency between the major axis and velocity direction, corresponding to the false alarm data based on the mapping consistency between the star image and motion includes: Obtain the set of motion velocities at the center of mass of the star and the velocities of the trajectory points; The displacement length of the trajectory point during the exposure time is calculated based on the motion velocity at the centroid of the star and the velocity set of the trajectory point. The major axis and minor axis of the star image are calculated based on the velocity of the star image at its centroid and the displacement length of the trajectory point within the exposure time. Obtain the phase angle of the major axis and the phase angle of the motion direction for each trajectory point; A fourth evaluation coefficient is calculated based on the major axis of the star image, the minor axis of the star image, the exposure time, and the velocity of motion at the center of mass of the star image. The fifth evaluation coefficient is calculated based on the phase angle of the major axis direction and the phase angle of the motion direction; The consistency between the major axis and the velocity magnitude is calculated based on the fourth evaluation coefficient and the number of trajectory points detected on the trajectory. The consistency between the major axis and the velocity direction is calculated based on the fifth evaluation coefficient and the number of trajectory points detected on the trajectory; The fourth evaluation coefficient is calculated using the following formula: in, This represents the fourth evaluation coefficient for the k-th trajectory point. The major axis of the star image at the k-th trajectory point is represented. The minor axis of the star image at the k-th trajectory point is represented. This indicates the exposure time. This indicates the shooting interval time. This represents the velocity at the centroid of the star in the k-th frame; The fifth evaluation coefficient is calculated using the following formula: in, This represents the fifth evaluation coefficient for the k-th trajectory point. This represents the phase angle along the major axis of the k-th trajectory point. This represents the phase angle of the motion direction of the k-th trajectory point. This indicates that the phase angle of the major axis of the k-th trajectory point and the phase angle of the motion direction of the k-th trajectory point are respectively converted to... Internal; definition The calculation is as follows: in, ; The consistency between the major axis and the magnitude of the velocity is calculated using the following formula: The alignment of the major axis with the velocity direction is calculated using the following formula: in, This indicates that the major axis is consistent with the magnitude of the velocity. This indicates that the major axis is aligned with the velocity direction. This indicates the number of trajectory points detected on the trajectory.

9. An electronic device, characterized in that, include: processor; The processor contains a memory on which executable code is stored, which, when executed, causes the processor to perform the false alarm suppression method for ground-based telescope observations as described in any one of claims 1-8.

10. A machine-readable medium having executable code stored thereon, which, when executed, causes a processor to perform the false alarm suppression method for ground-based telescope observations as described in any one of claims 1-8.

Images