A method for separating intrinsic and extrinsic parameters of star sensors based on feature trajectory imaging

Abstract

This invention discloses a method for separating the intrinsic and extrinsic parameters of a star sensor based on feature trajectory imaging. The method includes: analyzing the imaging model of the star sensor and its calibration system; discovering and proposing the trajectory characteristics of the star sensor imaging in the calibration system, and the imaging rules of the feature trajectories; eliminating the initial alignment deviation between the star simulator and the turntable based on the imaging rules of the feature trajectories in the calibration system; separating the principal points based on the imaging rules of the feature trajectories in the calibration system; optimizing and processing the feature trajectories to further improve the accuracy of principal point separation; and, based on eliminating the initial alignment deviation between the star simulator and the turntable and separating the principal points, comprehensively solving the remaining parameters to achieve the separation of the intrinsic and extrinsic parameters of the star sensor. The invention also discloses a system, electronic equipment, and a computer-readable storage medium.

Description

Technical Field

[0001] This invention relates to the field of star sensor parameter calibration and processing technology, and in particular to a method and system for separating intrinsic and extrinsic parameters of star sensors based on feature trajectory imaging. Background Technology

[0002] Star sensors, as photoelectric measurement instruments that rely on stellar imaging to obtain high-precision attitude measurements, are currently widely used in satellites, missiles, aircraft, and many other carriers. The imaging parameters of a star sensor directly determine its attitude measurement accuracy; therefore, star sensors need to be calibrated to obtain more accurate imaging parameters, thereby ensuring their attitude measurement accuracy. Imaging parameters mainly include focal length, principal point, and distortion coefficients. Star sensor calibration research includes two categories: on-orbit calibration and ground calibration. On-orbit calibration is an online parameter correction method, and the current technical challenge lies in ground calibration methods.

[0003] The calibration method based on unified modeling of intrinsic and extrinsic parameters using star simulators and multi-axis turntables is widely used in star sensor engineering practice due to its high precision and ease of operation. Depending on the implementation method, ground-based calibration methods for star sensors are mainly divided into non-equipment calibration and equipment calibration. Non-equipment calibration, typically based on night sky observation, directly obtains calibration data through stargazing experiments, making it convenient and low-cost; however, it is susceptible to interference from atmospheric and environmental factors. Equipment calibration mainly utilizes high-precision turntables and star simulators; some methods also require the use of autocollimating theodolites for auxiliary measurements. This type of calibration method is less affected by external interference factors, has high reliability, and is widely used in engineering practice, representing the most commonly used calibration system currently. In the calibration process using a high-precision turntable and a star simulator, ideally, the optical axis of the star simulator should be perpendicular to the inner frame plane of the turntable, and the star sensor should be mounted on the inner frame of the turntable, with the star sensor coordinate system coinciding with the turntable coordinate system. However, in actual operation, initial alignment deviations between the star simulator and the turntable, as well as installation deviations of the star sensor, are inevitably introduced. To address this, Wei et al. proposed a calibration method that unifies the modeling of intrinsic and extrinsic parameters, treating focal length, principal point, and distortion coefficient as intrinsic parameters, and initial alignment deviations and installation deviations as extrinsic parameters. This method eliminates the need for installation alignment operations, improving calibration efficiency. However, it does not consider the coupling between intrinsic and extrinsic parameters. While it can obtain high-precision measurement results in scenarios using prisms, for some scenarios using only the intrinsic parameters of the star sensor, such as missile applications, the estimated values ​​of some intrinsic parameters are significantly deviated from the true values ​​due to the coupling effect of extrinsic parameters, resulting in reduced measurement accuracy. Therefore, researching methods for decoupling and separating intrinsic and extrinsic parameters based on turntables and star simulators has significant practical implications. To address the coupling problem in the unified modeling of intrinsic and extrinsic parameters in calibration systems, Xiong et al., based on the principle of intrinsic parameter invariance, simultaneously optimized and estimated the intrinsic parameters using multiple sets of different extrinsic parameters, thereby effectively weakening the coupling effect of extrinsic parameters on the estimation of intrinsic parameters. However, this method did not completely separate the intrinsic and extrinsic parameters. Fan et al., by adjusting the angles of the outer and middle frames of a three-axis turntable, made the star point positions coincide under different inner frame angles, thus separating the initial alignment deviation. They then relied on a theodolite to separate the principal point and finally calculated the installation deviation and other parameters. This method achieved complete decoupling of intrinsic and extrinsic parameters and obtained good calibration results. However, separating the principal point required manual fine-tuning of the theodolite, and sometimes it was difficult to detect weak reflected light, affecting the separation effect of the principal point.

[0004] In summary, the calibration method based on unified modeling of intrinsic and extrinsic parameters of star simulator and multi-axis turntable ignores the coupling effect between intrinsic and extrinsic parameters, which has a certain impact on the accuracy of some application scenarios that only use the intrinsic parameters of star sensor. Summary of the Invention

[0005] The purpose of this invention is to provide a method for separating intrinsic and extrinsic parameters of a star sensor based on characteristic trajectory imaging. By deeply analyzing the imaging model of the star sensor and its calibration system, and starting from the relationship between the characteristic trajectory of the star sensor's imaging in the calibration system and certain key parameters, this invention proposes a novel method for separating intrinsic and extrinsic parameters of a star sensor. First, this invention deeply analyzes the imaging model of the star sensor and its calibration system, discovering and proposing its characteristic trajectory imaging rules. Then, based on the proposed characteristic trajectory imaging rules, the initial alignment deviation between the star simulator and the turntable is eliminated, and the principal point is separated. Simultaneously, to further improve the accuracy of the principal point, this invention optimizes and processes the elliptical trajectory. Finally, the remaining parameters are comprehensively solved to achieve the separation of intrinsic and extrinsic parameters. Experiments verify the effectiveness of the proposed method.

[0006] The first aspect of this invention is to provide a method for separating intrinsic and extrinsic parameters of a star sensor based on feature trajectory imaging, comprising:

[0007] S1. Analyze the imaging model of the star sensor and its calibration system, discover and propose the trajectory characteristics of the star sensor in the calibration system, and the imaging rules of the characteristic trajectory.

[0008] S2, based on the imaging pattern of the characteristic trajectory in the calibration system, eliminate the initial alignment deviation between the star simulator and the turntable; wherein, the initial alignment deviation is part of the extrinsic parameters of the star sensor calibration system;

[0009] S3, Based on the imaging pattern of the characteristic trajectory in the calibration system, separate the principal points;

[0010] S4, optimize and process the feature trajectory to further improve the accuracy of the main point separation;

[0011] S5. After eliminating the initial alignment deviation between the star simulator and the turntable and separating the principal points, the remaining parameters are solved comprehensively to achieve the separation of the internal and external parameters of the star sensor.

[0012] Preferably, S1 includes:

[0013] S11. An imaging model of the star sensor and its calibration system was established for a star sensor calibration system using a high-precision three-axis turntable and star simulator as calibration equipment.

[0014] The star sensor is a high-precision attitude measurement instrument targeting a star. Its imaging system can be simplified to a pinhole imaging model, where P(u, v) represents the ideal imaging point coordinates of star S on the image sensor, O(x0, y0) is the principal point, f is the focal length, and star S can be represented by the star vector V in the star sensor coordinate system. s To measure, the specific expression is shown in (1);

[0015]

[0016] Considering distortion, the star vector V s It can be expressed in the form of equation (2), where δ x δ y These represent distortions in the X and Y directions, respectively.

[0017]

[0018] Distortion δ in both X and Y directions x δ y The expression is shown in equation (3), where k1 and k2 are the second-order radial distortion coefficients, and p1 and p2 are the second-order tangential distortions.

[0019]

[0020] Based on equations (1) to (3), since the coordinates of the star's imaging point on the image sensor are known, its star vector V in the star sensor coordinate system is... s Determined by the internal parameters (f, x0, y0, k1, k2, p1, p2);

[0021] In the star sensor calibration system that uses a high-precision three-axis turntable and a star simulator as calibration equipment, due to installation deviations, the initial star vector V0 of the star simulator is not parallel to the turntable coordinate system Z. r The initial alignment deviation between the axis, the star simulator and the turntable, and the fact that the turntable coordinate system and the star sensor coordinate system are not completely consistent;

[0022] The initial alignment deviation of the star simulator is represented by two external parameters α and β. The initial vector V0 of the star simulator can be expressed as Equation (4) in the turntable coordinate system, where α and β represent pitch and yaw angles, respectively.

[0023]

[0024] Installation deviation matrix R between the turntable coordinate system and the star sensor coordinate system m The expression is given by three external parameters, as shown in equation (5), where V r R represents the star vector in the turntable coordinate system. m This represents the rotation matrix from the turntable coordinate system to the star sensor coordinate system. These are the Euler angles of the three axes of the coordinate system around the turntable;

[0025]

[0026] Therefore, in the star sensor calibration system using a high-precision three-axis turntable and star simulator as calibration equipment, the imaging model of the star sensor and its calibration system includes intrinsic parameters (f, x0, y0, k1, k2, p1, p2) and extrinsic parameters.

[0027] S12, By analyzing the above model, we discovered and proposed the trajectory characteristics of star sensors in the calibration system, as well as the imaging rules of the characteristic trajectories.

[0028] Based on the star sensor calibration system using a high-precision three-axis turntable and a star simulator as calibration equipment, the star sensor is mounted on the inner frame of the turntable. The outer and middle frames of the turntable are kept at arbitrary angular positions, and the angle between the star vector of the star simulator and the rotation axis of the turntable remains constant. When the inner frame is rotated, the star sensor rotates with the inner frame around the rotation axis of the turntable, and the star vector forms an image on the imaging plane. According to the principle of relative motion, this process can be equivalent to the star sensor remaining stationary while the star vector rotates around the rotation axis, forming a characteristic trajectory on the imaging plane. The imaging rules of the characteristic trajectory are analyzed and determined, based on the assumption that the star sensor mounting method remains unchanged. The imaging rules of the characteristic trajectory include:

[0029] Rule 1: When the inner frame is rotated, the star vector from the star simulator rotates around the axis of rotation, and its imaging trajectory is the intersection of the imaging surface and the surface of the cone. This intersection is an ellipse.

[0030] Rule 2: When the inner frame is rotated, the target imaging position E point remains unchanged and is uniquely determined. It is on the major axis of the ellipse. The ellipses formed by star vectors with different angles to the rotation axis on the imaging surface have different eccentricities. As the angle between the star vector and the rotation axis decreases, the elliptical trajectory of the star vector on the imaging surface gradually shrinks toward the target imaging position E point.

[0031] Rule 3: The major axes of the elliptical trajectories formed by star vectors with different angles to the rotation axis on the imaging surface are collinear, and the line containing the major axis must pass through the principal point O of the star sensor.

[0032] Preferably, eliminating the initial alignment deviation between the star simulator and the turntable based on the imaging pattern of the characteristic trajectory in the calibration system includes: adjusting the angle between the initial star vector of the star simulator and the rotation axis according to the imaging position of the initial star vector on the imaging plane, so that the angle is 0°, the initial star vector is parallel to the rotation axis, and the corresponding imaging point position is the target imaging position E. At this time, the initial alignment deviation α = 0°, β = 90°, thereby eliminating the initial alignment deviation between the star simulator and the turntable, including:

[0033] Determining the coordinates of the target imaging position E includes: using multiple sets of elliptical trajectory imaging points (x i yi The parameters of the ellipse equation are solved using the least squares method, and the coordinates (x, y) of the target imaging position point E are determined based on these parameters. e y e );

[0034] Adjust the star simulator so that the imaging point of the initial star vector of the star simulator coincides with the imaging point E of the target.

[0035] Preferably, S3 includes:

[0036] S31, the star sensor is installed on the inner frame of the turntable using two different installation methods. Under the two installation methods, the deviation angles between the star sensor's line of sight and the rotation axis are θ1 and θ2, respectively. The outer frame of the turntable is rotated to a suitable angle and kept unchanged. The inner frame is rotated. Under the two installation methods, the star vector of the star simulator forms elliptical trajectories l1 and l2 in different directions on the imaging surface. Points E1 and E2 represent the intersection points of the rotation axis and the imaging surface under the two installation methods, respectively. The lines containing the major axes P1P2 and P3P4 of the two ellipses both pass through the principal point O. The intersection point of the major axis lines of the two ellipses is the principal point O.

[0037] S32, calculate the coordinates (x0, y0) of the principal point O, thereby separating the principal point O from the intrinsic parameters of the star sensor as a known parameter; including: calculating the slope k of the major axis line, determining the expression of the major axis line based on the slope k; and calculating the coordinates (x0, y0) of the principal point O according to the equations of the two major axis lines.

[0038] Preferably, S4 includes:

[0039] S41, with a fixed installation method, rotate the inner frame, and star vectors with different angles to the rotation axis form elliptical trajectories with different eccentricities on the imaging surface, the major axes of the elliptical trajectories being collinear; under the same installation method, change the angle between the star vector and the rotation axis to generate multiple elliptical trajectories, and use the multiple elliptical trajectories to fit the straight line of the major axis together;

[0040] S42, using multiple different installation methods to generate corresponding elliptical trajectories, and taking the average of the intersection points of multiple major axis lines of the corresponding elliptical trajectory as the final principal point coordinates, wherein: the corresponding elliptical trajectories collected under each installation method are diagonally distributed, and the deviation angle θ between the line of sight and the rotation axis of the star sensor under each installation method is between 4° and 20°.

[0041] Preferably, S5 includes:

[0042] Based on eliminating the initial alignment deviation between the star simulator and the turntable and separating the principal point (x0, y0), the optimization objective function is constructed by combining the star sensor and its calibration system model, as shown in Equation (16), where p is the parameter to be optimized, as shown in Equation (17).

[0043]

[0044] The Levenberg-Marquardt method is used to optimize the objective function. The parameter update iteration method is as shown in Equation (18), where J is the Jacobian matrix and G is the gradient.

[0045]

[0046] By using the above joint optimization method, the remaining intrinsic parameters (f, k1, k2, p1, p2) are solved in a comprehensive manner, thus achieving complete separation of intrinsic and extrinsic parameters.

[0047] A second aspect of the present invention provides a star sensor intrinsic and extrinsic parameter separation system based on feature trajectory imaging, for implementing the separation method of the first aspect, comprising:

[0048] The model building module (101) is used to analyze the imaging model of the star sensor and its calibration system, discover and propose the trajectory characteristics of the star sensor in the calibration system, and the imaging rules of the characteristic trajectory.

[0049] An initial alignment deviation elimination module (102) is used to eliminate the initial alignment deviation between the star simulator and the turntable by analyzing the imaging patterns of the characteristic trajectories in the calibration system; wherein the initial alignment deviation is part of the extrinsic parameters of the star sensor calibration system.

[0050] The principal point separation module (103) is used to separate principal points based on the imaging rules of the feature trajectories in the calibration system;

[0051] The feature trajectory optimization design module (104) is used to optimize and process the feature trajectory in order to further improve the accuracy of the main point separation.

[0052] The residual parameter solving module (105) is used to solve the residual parameters comprehensively based on eliminating the initial alignment deviation between the star simulator and the turntable and separating the principal points, so as to realize the separation of the internal and external parameters of the star sensor.

[0053] A third aspect of the present invention provides an electronic device including a processor and a memory, the memory storing a plurality of instructions, the processor being configured to read the instructions and execute the method as described in the first aspect.

[0054] A fourth aspect of the present invention provides a computer-readable storage medium storing a plurality of instructions which can be read by a processor and executed as described in the first aspect.

[0055] The beneficial effects of the method and system of the present invention are as follows:

[0056] By deeply analyzing the imaging model of the star sensor and its calibration system, this invention proposes a method for separating the intrinsic and extrinsic parameters of the star sensor based on feature trajectory imaging. First, this invention analyzes the imaging model of the star sensor and its calibration system, discovering and proposing its feature trajectory imaging rules. Then, based on the proposed feature trajectory imaging rules, the initial alignment deviation between the star simulator and the turntable is eliminated, and the principal point is separated. Furthermore, to further improve the accuracy of the principal point, this invention optimizes and processes the feature trajectory. Finally, the remaining parameters are solved comprehensively to achieve the separation of intrinsic and extrinsic parameters. Experiments verify the effectiveness of the proposed method. Experimental results show that the method achieves high accuracy in separating the intrinsic parameters. During the elimination of the initial alignment deviation, the deviation of the target imaging point E in both the X and Y directions is approximately on the order of e-03 pixels, and the principal point coordinate separation accuracy can reach 2 pixels. This method has significant technical advantages and application potential in engineering applications. Attached Figure Description

[0057] To more clearly illustrate the technical solutions in the specific embodiments or related technologies of the present invention, the drawings used in the description of the specific embodiments or related technologies will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0058] Figure 1 This is a flowchart of a star sensor intrinsic and extrinsic parameter separation method based on feature trajectory imaging according to an embodiment of the present invention;

[0059] Figure 2 This is a schematic diagram of the star sensor structure and imaging model provided in an embodiment of the present invention;

[0060] Figure 2 (a) is a schematic diagram of the star sensor structure provided in an embodiment of the present invention;

[0061] Figure 2 (b) is a schematic diagram of the star sensor pinhole imaging model provided in the embodiment of the present invention;

[0062] Figure 3 This is a schematic diagram of the structural principle of the star sensor calibration system provided in an embodiment of the present invention;

[0063] Figure 4This is a diagram illustrating the characteristic trajectory imaging model of a star sensor in a calibration system, as provided in an embodiment of the present invention.

[0064] Figure 5 A schematic diagram illustrating the separation principle of the initial alignment parameters α and β provided in an embodiment of the present invention;

[0065] Figure 6 This is a schematic diagram of the principal point (x0, y0) separation principle provided in an embodiment of the present invention;

[0066] Figure 7 This is a schematic diagram illustrating the principle of the influence of the deviation angle θ on the proportion of the elliptical trajectory provided in an embodiment of the present invention;

[0067] Figure 8 The trajectory distribution diagram provided in the embodiments of the present invention;

[0068] Figure 8 (a) A trajectory distribution diagram of the first installation method provided in the embodiment of the present invention;

[0069] Figure 8 (b) is a trajectory distribution diagram of the second installation method provided in the embodiment of the present invention;

[0070] Figure 8 (c) is a trajectory distribution diagram of the third installation method provided in the embodiment of the present invention;

[0071] Figure 8 (d) is a trajectory distribution diagram of the fourth installation method provided in the embodiment of the present invention;

[0072] Figure 9 The experimental trajectory distribution diagram provided in the embodiments of the present invention;

[0073] Figure 10 This is a schematic diagram of the internal and external parameter separation system of a star sensor based on feature trajectory imaging, provided in an embodiment of the present invention.

[0074] Figure 11 This is a structural diagram of an electronic device provided according to an embodiment of the present invention. Detailed Implementation

[0075] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0076] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0077] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0078] like Figure 1 As shown, this embodiment provides a method for separating intrinsic and extrinsic parameters of a star sensor based on feature trajectory imaging, including:

[0079] S1. Analyze the imaging model of the star sensor and its calibration system, discover and propose the trajectory characteristics of the star sensor in the calibration system, and the imaging rules of the characteristic trajectory.

[0080] In a preferred embodiment, S1 includes:

[0081] S11. An imaging model of the star sensor and its calibration system was established for a star sensor calibration system using a high-precision three-axis turntable and star simulator as calibration equipment.

[0082] The star sensor is a high-precision attitude measurement instrument targeting a star, and its structural diagram is shown below. Figure 2 As shown in (a), since starlight is parallel light at infinity, the star sensor imaging system can be simplified into a pinhole imaging model, such as... Figure 2 As shown in (b). Where P(u, v) are the ideal imaging point coordinates of star S on the image sensor, O(x0, y0) is the principal point, f is the focal length, and star S can be represented by the star vector V in the star sensor coordinate system. s To measure this, the specific expression is shown in (1).

[0083]

[0084] Considering distortion, the star vector V s It can be expressed in the form of equation (2), where δx δ y These represent distortions in the X and Y directions, respectively.

[0085]

[0086] Distortion δ in both X and Y directions x δ y The expression is shown in equation (3), where k1 and k2 are the second-order radial distortion coefficients, and p1 and p2 are the second-order tangential distortions.

[0087]

[0088] Based on equations (1) to (3), since the coordinates of the star's imaging point on the image sensor are known, its star vector V in the star sensor coordinate system is... s Determined by the internal parameters (f, x0, y0, k1, k2, p1, p2);

[0089] The star sensor calibration system, which uses a high-precision three-axis rotary table and a star simulator as calibration equipment, is as follows: Figure 2 As shown, it consists of a high-precision three-axis turntable, a star simulator, and a star sensor. The star sensor is mounted on the inner frame of the turntable. The star simulator works in conjunction with the high-precision three-axis turntable to simulate star vectors with different incident directions, where OX... r Y r Z r Denotes the coordinate system of the turntable, OX s Y s Z s This represents the coordinate system of the star sensor.

[0090] Due to installation deviations, the initial star vector V0 of the star simulator is not parallel to the turntable coordinate system Z. r The axis represents the initial alignment deviation of the star simulator. At the same time, the coordinate system of the turntable is not completely consistent with the coordinate system of the star sensor.

[0091] The initial alignment deviation of the star simulator can be represented by two external parameters α and β. Specifically, the initial vector V0 of the star simulator in the turntable coordinate system can be expressed as Equation (4), where α and β represent pitch and yaw angles, respectively.

[0092]

[0093] The installation deviation matrix (or rotation matrix) between the turntable coordinate system and the star sensor coordinate system R m It can be expressed in the form of equation (5) with three external parameters, where V r R represents the star vector in the turntable coordinate system. m This represents the rotation matrix from the turntable coordinate system to the star sensor coordinate system. These are the Euler angles of the three axes of the coordinate system around the turntable.

[0094]

[0095] Therefore, in the star sensor calibration system using a high-precision three-axis turntable and star simulator as calibration equipment, the imaging model of the star sensor and its calibration system includes intrinsic parameters (f, x0, y0, k1, k2, p1, p2) and extrinsic parameters.

[0096] S12, By analyzing the above model, we discovered and proposed the trajectory characteristics of star sensors in the calibration system, as well as the imaging rules of the characteristic trajectories.

[0097] In this embodiment, the specific imaging point coordinates of the star vector on the star sensor image sensor are the result of the combined effect of intrinsic and extrinsic parameters in the calibration system. However, existing parameter solving methods are basically based on mathematical analytical expressions and use generalized optimization strategies, without fully exploring the close relationship between intrinsic and extrinsic parameters and imaging. Furthermore, the trajectory analysis of traditional parameter separation methods mainly focuses on the static characteristics of trajectory points, emphasizing the uniform distribution of sampling points within the field of view. However, this trajectory ignores the dynamic connections between trajectory points.

[0098] In view of this, the embodiments of the present invention start from the dynamic characteristics of the trajectory, regard each trajectory point as an interconnected whole, and through in-depth analysis of the imaging model of the star sensor and its calibration system, discover and propose the trajectory characteristics of the star sensor in the calibration system, as well as the imaging rules of the characteristic trajectory.

[0099] The specific content includes:

[0100] (1) Based on the calibration system in the star sensor elliptical trajectory imaging and calibration system model, the star sensor is installed on the inner frame of the turntable, keeping the outer frame and middle frame of the turntable at any angle, and keeping the star vector of the star simulator and the Z-axis of the turntable coordinate system constant. r With the included angle of the axes remaining constant, rotating the inner frame causes the star sensor to rotate around the Z-axis, following the rotation of the inner frame, and the star vector forms an image on the imaging plane. According to the principle of relative motion, this process can be equivalent to the star sensor remaining stationary while the star vector rotates around the Z-axis. r The characteristic trajectory model of the star vector imaged on the imaging plane is as follows: Figure 4 As shown, where θ is the star sensor's line of sight Z. s With axis Z r The deviation angle between them, δ i(i＝1，2，3) For the star vector and rotation axis Z of the star simulator r The included angle, O sPoint O is the optical center of the optical lens, point E is the principal point of the star sensor, the intersection of the optical axis of the optical lens and the imaging plane, f is the focal length of the star sensor, and point E is the rotation axis Z. r With imaging plane X c OY c The intersection point is also the point where the elevation line intersects with the imaging plane X. c OY c The intersection point.

[0101] (2) Figure 4 Further analysis of the characteristic trajectory model revealed and proposed imaging rules for the characteristic trajectories. These rules are determined based on the assumption that the star sensor mounting method remains unchanged. The characteristic trajectory imaging rules include:

[0102] Rule 1: Rotate the inner frame; the star vector from the star simulator rotates around the turntable's Z-axis. r Rotation, its imaging trajectory is the imaging plane X c OY c The line of intersection with the surface of the cone is an ellipse. For example... Figure 4 As shown in the diagram, l1, l2, and l3 represent the axes of rotation Z, respectively. r The motion trajectories of the star vectors with included angles of δ1, δ2, and δ3 correspond to the imaging trajectories l1′, l2′, and l3′ on the imaging plane, respectively.

[0103] Rule 2: When the inner frame is rotated, point E remains unchanged and is uniquely determined. It lies on the major axis of the ellipse and is parallel to the axis of rotation Z. r The trajectories of star vectors with different included angles on the imaging plane are ellipses with different eccentricities, and these trajectories change with the rotation axis Z. r As the angle decreases, the elliptical trajectory of the star vector on the imaging plane gradually shrinks towards point E.

[0104] Rule 3, with the Z-axis r The major axes of the elliptical trajectories formed by star vectors with different included angles on the imaging plane are collinear, and the straight line containing them must pass through the principal point O of the star sensor, such as... Figure 4 The solid line in the lower middle section is shown.

[0105] S2, based on the imaging rules of the characteristic trajectories in the calibration system, eliminate the initial alignment deviation between the star simulator and the turntable;

[0106] In a preferred embodiment, eliminating the initial alignment deviation between the star simulator and the turntable based on the relationship between the characteristic trajectory and certain key parameters includes: adjusting the initial star vector V0 of the star simulator and the rotation axis Z according to the imaging position of the initial star vector V0 of the star simulator on the imaging plane. r The angle δ between them is made so that the angle δ is 0, and the initial star vector V0 is perpendicular to the rotation axis Z. rWhen parallel, the corresponding imaging point position is the target imaging position. At this time, the initial alignment deviation α = 0° and β = 90° is eliminated, thereby eliminating the initial alignment deviation between the star simulator and the turntable.

[0107] In this embodiment, when the star sensor installation method remains unchanged, rotating the inner frame causes the star vector to align with the rotation axis Z. r As the angle δ between them decreases, the elliptical trajectory formed by the star vector on the imaging plane gradually shrinks towards point E until the imaging trajectory completely degenerates into point E. At this point, the star vector and the rotation axis Z... r The angle between the star vector and the rotation axis Z is 0°. r Completely parallel.

[0108] Based on the above principles, this invention proposes a novel method for separating the initial alignment parameters α and β. A schematic diagram of this method is shown below. Figure 5 As shown, the specific implementation method is as follows:

[0109] ① Determine the coordinates of the target imaging position E. Install the star sensor normally on the inner frame of the turntable, keeping the outer frame of the turntable at zero position. Due to the initial alignment deviation between the star simulator and the turntable, the initial star vector V0 of the star simulator and the rotation axis Z... r The included angle δ is not 0°. Rotating the inner frame causes the initial star vector V0 of the star simulator to generate an elliptical trajectory l on the imaging plane. The rotation axis Z is then calculated based on this trajectory. r The coordinates of point E, the intersection with the imaging plane. Since the star sensor is normally mounted on the inner frame of the turntable, the star sensor's line of sight Z... s With axis Z r Since the included angle θ is relatively small, point E can be considered as the center point of the elliptical trajectory.

[0110] ② Adjust the star simulator so that the imaging point of the initial star vector V0 coincides with the target point E. The imaging point P of the initial star vector V0 on the imaging plane is a random point distributed around point E. Adjust the position adjustment knob of the star simulator so that the imaging point P of the initial star vector V0 gradually approaches point E until it coincides with point E. At this time, the initial star vector V0 of the star simulator is parallel to the turntable Z. r The axis, which can be represented in the rotary table coordinate system, is as follows: This eliminates the initial vector alignment deviation of the starlight simulator, thereby separating the parameters α and β.

[0111] The specific method for solving the coordinates of the target imaging position E is as follows: Assume the general equation of the elliptical trajectory on the imaging surface is as shown in equation (6). This general equation of the ellipse contains 5 unknowns. Using multiple sets of elliptical trajectory imaging points (x... i y iThe parameters of the ellipse equation are solved using the least squares method. The coordinates (x, y) of point E are then obtained using the ellipse equation parameters obtained in equation (6). e y e ), which is Equation (7).

[0112] x 2 +Axy+By 2 +Cx+Dy+E=0 (6);

[0113]

[0114] S3, based on the imaging pattern of characteristic trajectories in the calibration system, separate the principal points;

[0115] As a preferred embodiment, separating principal points based on the imaging patterns of characteristic trajectories in the calibration system includes:

[0116] S31, under the same installation method, is related to the Z-axis. r Star vectors with different included angles form elliptical trajectories on the imaging plane whose major axes are collinear, and this straight line must pass through the principal point O of the star sensor. If the star sensor is mounted on a turntable using one mounting method, and the outer frame of the turntable is rotated to a certain angle while the inner frame is rotated, an elliptical trajectory is formed, and the line containing the major axis of this elliptical trajectory passes through the principal point O. Similarly, changing the mounting method can generate another elliptical trajectory, whose major axis also passes through the principal point O. Based on this characteristic, this invention proposes a new method for separating the principal point (x0, y0), the principle of which is as follows: Figure 6 As shown, the specific process is as follows:

[0117] The star sensor was mounted on the inner frame of the turntable using two different mounting methods. Under both methods, the star sensor's line of sight Z... s With axis Z r (Z′ r The deviation angles between θ1 and θ2 are respectively. The outer frame of the turntable is rotated to a suitable angle and kept constant. The inner frame is rotated, and under both installation methods, the star vector of the star simulator forms elliptical trajectories l1 and l2 in different directions on the imaging plane. Points E1 and E2 represent the rotation axis Z, respectively. r (Z′ r In both installation methods, the intersection point with the imaging plane, the lines containing the major axes P1P2 and P3P4 of the two ellipses both pass through the principal point O. Therefore, the intersection point of the major axes of the two ellipses is the principal point O.

[0118] S32, Based on the principal point separation method described above, calculate the principal point coordinates (x0, y0), so that the principal point can be separated from the intrinsic parameters of the star sensor as a known parameter; including:

[0119] The slope k of the major axis line is expressed as equation (11), where the values ​​of A, B, and E can be substituted into equation (6) to obtain the ellipse equation parameters, (x e y e The result can be substituted into equation (7) to calculate θ. r The direction of the ellipse.

[0120]

[0121] Based on the calculation results above, the expression for the major axis line is determined as shown in equation (12):

[0122] y = kx + (y e -kx e (12);

[0123] Based on the equations of the two major axes, the coordinates of the principal point O(x0, y0) are calculated as shown in equation (13), where k1 and k2 are the slopes of the two lines, and b1 and b2 are the intercepts of the two lines.

[0124]

[0125] S4, optimize and process the elliptical trajectory to further improve the accuracy of the main point separation;

[0126] In this embodiment, in practical applications, centroid positioning errors and limitations of the star sensor's field of view can cause errors in the fitted elliptical trajectory, major axis, and major axis intersection points obtained using the above method. Therefore, to improve the accuracy of principal point calculation, the trajectory and data processing methods in principal point separation will be optimized below. Due to the limitations of the field of view, the star vector rotates around the Z-axis. r Rotation means that the imaging trajectory on the imaging sensor may only be a part of an ellipse. The ratio of the arc length of the elliptical trajectory imaged by the imaging sensor to the circumference of the entire ellipse is called the elliptical trajectory percentage. Generally, the higher the elliptical trajectory percentage, the better the ellipse fitting effect, which is more conducive to principal point separation.

[0127] However, the star sensor's field of view is fixed. To increase the proportion of the elliptical trajectory, the star sensor's Z-axis needs to be adjusted. s With axis Z r The deviation angle θ between the two points affects the proportion of the elliptical trajectory. The larger the field of view of the star sensor, the higher the proportion of the elliptical trajectory. However, the field of view of the star sensor is pre-designed and fixed. To increase the proportion of the elliptical trajectory, the star sensor's line of sight Z needs to be adjusted. s With axis Z r The deviation angle θ between them. The deviation angle θ can be adjusted by changing the installation deviation parameter. The value is used to achieve the deviation parameter. Star sensor line of sight Z s With axis Zr The relationship between the deviation angles θ is shown in equation (14):

[0128]

[0129] In the field of view and star vector and rotation axis Z r Given a fixed included angle δ, the effect of the deviation angle θ on the proportion of the elliptical trajectory is illustrated as follows: Figure 7 As shown, the smaller the deviation angle θ, the larger the proportion of the elliptical trajectory, until the entire elliptical trajectory is presented on the imaging sensor. While considering the proportion of the elliptical trajectory and reducing the deviation angle θ, it is also necessary to ensure the elliptical features. The more obvious the elliptical features, the more beneficial it is to fit the elliptical trajectory. The elliptical features are generally measured by the eccentricity e. The eccentricity e of the ellipse formed by the oblique cutting of the cone can be expressed as Equation (15), where γ is equal to the angle between the oblique cutting plane and the height line of the cone, and the star sensor's line of sight Z s With axis Z r The complementary angle θ between the two, δ is the angle between the star vector and the rotation axis Z. r As shown in formula (15), with δ fixed, the larger θ is, the smaller γ is, the larger the eccentricity e of the ellipse, and the more obvious the elliptical characteristics. Based on the above analysis, for the star sensor used in this invention, the deviation angle θ should preferably be between 4° and 20°.

[0130]

[0131] Choosing an appropriate angle θ can improve the fitting accuracy of a single ellipse. Furthermore, this invention also requires fitting the major axis line of the ellipse and calculating the intersection points and principal points of the major axis lines corresponding to different installation methods. To improve the accuracy of principal point coordinate calculation, this invention further determines the following strategy:

[0132] ① Fix one installation method, rotate the inner frame, and align it with the Z-axis. r Star vectors with different included angles δ form elliptical trajectories with different eccentricities but in the same direction on the imaging plane, and their major axes are collinear. To improve the fitting accuracy of the elliptical major axis straight line, under the same installation method, the relationship between the star vector and the rotation axis Z can be changed. r The included angle δ between the points generates multiple elliptical trajectories, which are then used to fit the straight line of their major axis. This invention collects three elliptical trajectories in the same direction but with different eccentricities at equal intervals for each installation method.

[0133] ② The intersection of the major axis lines of the elliptical trajectories corresponding to different installation methods is the principal point. In principle, calculating the intersection of two major axis lines is sufficient to determine the principal point coordinates. However, to improve the fitting accuracy of the principal point, this invention uses the average of the intersection points of the four major axis lines corresponding to the elliptical trajectories generated under the four installation methods as the final principal point coordinates. The elliptical trajectories acquired under each installation method are diagonally distributed, and the Z-axis of the star sensor under each installation method is set to... s With axis Z r The deviation angle θ between them is between 4° and 20°.

[0134] Based on the above strategy, the trajectory is designed, and the trajectory distribution is as follows: Figure 8 As shown, where O c Let be the geometric center of the image sensor. (a) to (d) represent the elliptical trajectories with collinear major axes acquired at equal intervals under four different mounting methods.

[0135] use Figure 8 The principal point (x0, y0) of the 12 elliptical trajectories designed in the model is found through the following process:

[0136] ① Take the average slope of the major axis of the three elliptical trajectories corresponding to each installation method and the average coordinates of the center point as the slope k of the major axis of the elliptical trajectories. i(i＝1，2，3，4) With the center point coordinates M i(i＝1，2，3，4) The specific analytical expressions for the four major axes are obtained, as shown in equation (16).

[0137]

[0138] ② The major axis lines of the elliptical loci at non-diagonal positions intersect each other pairwise, and the corresponding four principal points O are calculated. i(i＝1，2，3，4) The average of its coordinate values ​​is taken as the final coordinate value of the principal point O(x0, y0), as shown in equation (17).

[0139]

[0140] Figure 9 The trajectory design scheme shown, along with the processing method of jointly calculating the principal point coordinates using elliptical trajectories under various installation methods, improves the accuracy of principal point coordinate calculation.

[0141] S5. After eliminating the initial alignment deviation between the star simulator and the turntable and separating the principal points, the remaining parameters are solved comprehensively to achieve the separation of the internal and external parameters of the star sensor.

[0142] The initial alignment parameters α and β have been separated, making α = 0° and β = 90°. Simultaneously, the principal point (x0, y0) has also obtained accurate values, which can be used as constants. Therefore, the calibration system now only has intrinsic parameters (f, k1, k2, p1, p2) and extrinsic parameters. The remaining internal and external parameters are not yet solved, and their coupling is negligible. The remaining parameters are solved by comprehensively solving the designed trajectory. Combining the star sensor elliptical trajectory imaging and calibration system model, a comprehensive model is constructed to optimize the objective function, as shown in equation (18), where (x ij y ij ) i＝1，2，3，4 For theoretical imaging coordinates, The actual observed imaging coordinates are given, and p is the parameter to be optimized, as shown in equation (19).

[0143]

[0144] The above problem is optimized using the Levenberg-Marquardt method. The parameter update iteration method is as shown in Equation (20), where J is the Jacobian matrix and G is the gradient.

[0145]

[0146] Using the above joint optimization method, the remaining intrinsic parameters (f, k1, k2, p1, p2) can be solved comprehensively, achieving complete separation of intrinsic and extrinsic parameters.

[0147] Verification Example:

[0148] The present invention conducted experiments to verify the effectiveness of the proposed method. The field of view of the star sensor used in the experiment was 10°×10°, the resolution of the CMOS image sensor was 2048 pixels×2048 pixels, and the pixel size in both the X and Y directions was 5.5 micrometers.

[0149] The intrinsic parameters of the star sensor are shown in Table 1. Considering that the random centroid positioning error of the star sensor used in this invention is at the 0.05 pixel level, during the calibration process, the coordinate value of the same point is generally taken as the average of 30 consecutive sampling results. This reduces the random centroid positioning error to 0.01 pixels. Therefore, Gaussian noise with a standard deviation of 0.01 pixels is added to simulate the random centroid positioning error.

[0150] Table 1

[0151]

[0152] (1) Initial alignment deviation parameters α, β separation results

[0153] This invention demonstrates the effectiveness of the method for eliminating initial alignment biases α and β by generating three different elliptical trajectories and calculating the deviation between the fitted value and the true value of the coordinates of point E corresponding to each elliptical trajectory. The fitted values ​​of the coordinates of point E are shown in Table 2, and the results of the deviation calculation are shown in Table 3.

[0154] Table 2

[0155]

[0156] Table 3

[0157]

[0158] The experimental results above show that the fitted coordinates of point E deviate very little from the true value, and the deviations in both the X and Y directions are on the order of e-03 pixels, which verifies the effectiveness of the extrinsic parameter α and β separation method proposed in this invention.

[0159] (2) Separation result of principal point (x0, y0)

[0160] This invention demonstrates the effectiveness of the principal point separation method by calculating the deviation between the separated principal point coordinates and the true values. It is known that the direction of the ellipse depends only on the installation angle. Regarding this, in conjunction with the trajectory optimization scheme, 12 sets of elliptical trajectory points were collected for 4 different combinations of installation angles, such as... Figure 9 As shown in Table 4, the results obtained from the processing are shown in Table 4.

[0161] Table 4

[0162]

[0163] The separation results show that the coordinates of the separated principal points deviate by less than 2 pixels in both the X and Y directions, and the separation accuracy of the principal points reaches the level of 2 pixels.

[0164] (3) Solving for other parameters

[0165] After eliminating the initial alignment deviations α and β and separating the principal point (x0, y0), the remaining parameters are solved. The results of the intrinsic parameter separation are shown in Table 5. The results show that the deviation between the intrinsic parameter separation value and the true value is very small, and the method proposed in this invention has significant effects.

[0166] Table 6

[0167]

[0168] Example 2

[0169] See Figure 10 This embodiment provides a star sensor intrinsic and extrinsic parameter separation system based on feature trajectory imaging, used to implement the separation method of Embodiment 1, including:

[0170] The model building module 101 is used to analyze the imaging model of the star sensor and its calibration system, discover and propose the trajectory characteristics of the star sensor in the calibration system, and the imaging rules of the characteristic trajectory.

[0171] The initial alignment deviation elimination module 102 is used to eliminate the initial alignment deviation between the star simulator and the turntable by analyzing the imaging pattern of the feature trajectory in the calibration system; wherein, the initial alignment deviation is part of the extrinsic parameters of the star sensor calibration system.

[0172] The principal point separation module 103 is used to separate principal points based on the imaging rules of the feature trajectories in the calibration system.

[0173] The feature trajectory optimization design module 104 is used to optimize and process the feature trajectory to further improve the accuracy of the main point separation.

[0174] The residual parameter solving module 105 is used to comprehensively solve for the residual parameters after eliminating the initial alignment deviation between the star simulator and the turntable and separating the principal points, thereby achieving the separation of the intrinsic and extrinsic parameters of the star sensor. This invention also provides a memory storing multiple instructions for implementing the method as described in Embodiment 1.

[0175] like Figure 11 As shown, the present invention also provides an electronic device, including a processor 301 and a memory 302 connected to the processor 301. The memory 302 stores a plurality of instructions, which can be loaded and executed by the processor to enable the processor to perform the method as described in Embodiment 1.

[0176] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for separating intrinsic and extrinsic parameters of a star sensor based on feature trajectory imaging, characterized in that, include: S1. Analyze the imaging model of the star sensor and its calibration system, discover and propose the trajectory characteristics of the star sensor in the calibration system, and the imaging rules of the characteristic trajectory. S2, based on the imaging pattern of the characteristic trajectory in the calibration system, eliminate the initial alignment deviation between the star simulator and the turntable; wherein, the initial alignment deviation is part of the extrinsic parameters of the star sensor calibration system; S3, Based on the imaging pattern of the characteristic trajectory in the calibration system, separate the principal points; S4, optimize and process the feature trajectory to further improve the accuracy of the main point separation; S5. Based on eliminating the initial alignment deviation between the star simulator and the turntable and separating the principal points, the remaining parameters are solved comprehensively to achieve the separation of the internal and external parameters of the star sensor. S1 includes: S11. An imaging model of the star sensor and its calibration system was established for a star sensor calibration system using a high-precision three-axis turntable and star simulator as calibration equipment. S12, By analyzing the above model, we discovered and proposed the trajectory characteristics of star sensors in the calibration system, as well as the imaging rules of the characteristic trajectories. Based on the star sensor calibration system using a high-precision three-axis turntable and a star simulator as calibration equipment, the star sensor is mounted on the inner frame of the turntable. The outer and middle frames of the turntable are kept at arbitrary angular positions, and the angle between the star vector of the star simulator and the rotation axis of the turntable remains constant. When the inner frame is rotated, the star sensor rotates with the inner frame around the rotation axis of the turntable, and the star vector images onto the imaging plane. According to the principle of relative motion, this process can be equivalent to the star sensor remaining stationary while the star vector rotates around the rotation axis, creating a characteristic trajectory on the imaging plane. The imaging rules of the characteristic trajectory are analyzed and determined, based on the assumption that the star sensor mounting method remains unchanged. The imaging rules of the characteristic trajectory include: Rule 1: When the inner frame is rotated, the star vector from the star simulator rotates around the axis of rotation, and its imaging trajectory is the intersection of the imaging surface and the surface of the cone. This intersection is an ellipse. Rule 2: When the inner frame is rotated, the target imaging position E point remains unchanged and is uniquely determined. It is on the major axis of the ellipse. The ellipses formed by star vectors with different angles to the rotation axis on the imaging surface have different eccentricities. As the angle between the star vector and the rotation axis decreases, the elliptical trajectory of the star vector on the imaging surface gradually shrinks toward the target imaging position E point. Rule 3: The major axes of the elliptical trajectories formed by star vectors with different angles to the rotation axis on the imaging surface are collinear, and the line containing the major axis must pass through the principal point O of the star sensor. S2 includes: The method of eliminating the initial alignment deviation between the star simulator and the turntable based on the imaging pattern of the characteristic trajectory in the calibration system includes: adjusting the angle between the initial star vector of the star simulator and the rotation axis according to the imaging position of the initial star vector on the imaging plane, so that the angle is 0°, the initial star vector is parallel to the rotation axis, and the corresponding imaging point position is the target imaging position E. At this time, the initial alignment deviation α = 0°, β = 90°, thereby eliminating the initial alignment deviation between the star simulator and the turntable, including: Determining the coordinates of the target imaging position E includes: using multiple sets of elliptical trajectory imaging points (x i ,y i The parameters of the ellipse equation are solved using the least squares method, and the coordinates (x, y) of the target imaging position point E are determined based on these parameters. e y e ); Adjust the star simulator so that the imaging point of the initial star vector of the star simulator coincides with the target imaging position E. S3 includes: S31. The star sensor is installed on the inner frame of the turntable using two different installation methods. Under the two methods, the deviation angles between the star sensor's line of sight and the rotation axis are θ1 and θ2, respectively. The outer frame of the turntable is rotated to a suitable angle and kept constant. Rotating the inner frame, under the two installation methods, the star vector of the star simulator forms elliptical trajectories l1 and l2 in different directions on the imaging plane. Points E1 and E2 represent the intersection points of the rotation axis and the imaging plane under the two installation methods, respectively. The major axes P1 and P2 of the two ellipses are... ， The lines containing P3 and P4 both pass through the principal point O, and the intersection of the major axes of the two ellipses is the principal point O. S32, calculate the coordinates (x0, y0) of the principal point O, thereby separating the principal point O from the intrinsic parameters of the star sensor as a known parameter; including: calculating the slope k of the major axis line, determining the expression of the major axis line based on the slope k; and calculating the coordinates (x0, y0) of the principal point O according to the equations of the two major axis lines. S5 includes: Based on eliminating the initial alignment deviation between the star simulator and the turntable and separating the principal point (x0, y0), the optimization objective function is constructed by combining the star sensor and its calibration system model, as shown in Equation (16), where p is the parameter to be optimized, as shown in Equation (17). Where (f, k1, k2, p1, p2) are intrinsic parameters. For external parameters, (x ij y ij ) i＝1，2，3，4 For theoretical imaging coordinates, The actual observed imaging coordinates are given, and p is the parameter to be optimized. The Levenberg-Marquardt method is used to optimize the objective function. The parameter update iteration method is as shown in Equation (18), where J is the Jacobian matrix and G is the gradient. By using the above joint optimization method, the remaining intrinsic parameters (f,k1,k2,p1,p2) are solved in a comprehensive manner, thus achieving complete separation of intrinsic and extrinsic parameters.

2. The method for separating intrinsic and extrinsic parameters of a star sensor based on feature trajectory imaging according to claim 1, characterized in that, S11 includes: The star sensor is a high-precision attitude measurement instrument targeting a star. Its imaging system can be simplified to a pinhole imaging model, where P(u,v) represents the ideal imaging point coordinates of star S on the image sensor, O(x0,y0) is the principal point, f is the focal length, and star S can be represented by the star vector V in the star sensor coordinate system. s To measure, the specific expression is shown in (1); Considering distortion, the star vector V s It can be expressed in the form of equation (2), where δ x ,δ y These represent distortions in the X and Y directions, respectively. Distortion δ in both X and Y directions x ,δ y The expression is shown in equation (3), where k1 and k2 are the second-order radial distortion coefficients, and p1 and p2 are the second-order tangential distortions. Based on equations (1) to (3), since the coordinates of the star's imaging point on the image sensor are known, its star vector V in the star sensor coordinate system is... s Determined by the internal parameters (f, x0, y0, k1, k2, p1, p2); In the star sensor calibration system that uses a high-precision three-axis turntable and a star simulator as calibration equipment, due to installation deviations, the initial star vector V0 of the star simulator is not parallel to the turntable coordinate system Z. r The initial alignment deviation between the axis, the star simulator and the turntable, and the fact that the turntable coordinate system and the star sensor coordinate system are not completely consistent; The initial alignment deviation of the star simulator is represented by two external parameters α and β. The initial vector V0 of the star simulator can be expressed as Equation (4) in the turntable coordinate system, where α and β represent pitch and yaw angles, respectively. V0=[cosαcosβ,sinαsinβ,sinβ] T (4) Installation deviation matrix R between the turntable coordinate system and the star sensor coordinate system m The expression is given by three external parameters, as shown in equation (5), where V r R represents the star vector in the turntable coordinate system. m This represents the rotation matrix from the turntable coordinate system to the star sensor coordinate system. These are the Euler angles of the three axes of the coordinate system around the turntable; Therefore, in the star sensor calibration system using a high-precision three-axis turntable and star simulator as calibration equipment, the imaging model of the star sensor and its calibration system includes intrinsic parameters (f, x0, y0, k1, k2, p1, p2) and extrinsic parameters.

3. The method for separating intrinsic and extrinsic parameters of a star sensor based on feature trajectory imaging according to claim 2, characterized in that, S4 includes: S41, with a fixed installation method, rotate the inner frame, and star vectors with different angles to the rotation axis form elliptical trajectories with different eccentricities on the imaging surface, the major axes of the elliptical trajectories being collinear; under the same installation method, change the angle between the star vector and the rotation axis to generate multiple elliptical trajectories, and use the multiple elliptical trajectories to fit the straight line of the major axis together; S42, using multiple different installation methods to generate corresponding elliptical trajectories, and taking the average of the intersection points of multiple major axis lines of the corresponding elliptical trajectory as the final principal point coordinates, wherein: the corresponding elliptical trajectories collected under each installation method are diagonally distributed, and the deviation angle θ between the line of sight and the rotation axis of the star sensor under each installation method is between 4° and 20°.

4. A star sensor intrinsic and extrinsic parameter separation system based on feature trajectory imaging, used to implement the separation method according to any one of claims 1-3, characterized in that, include: The model building module (101) is used to analyze the imaging model of the star sensor and its calibration system, discover and propose the trajectory characteristics of the star sensor in the calibration system, and the imaging rules of the characteristic trajectory. The initial alignment deviation elimination module (102) is used to eliminate the initial alignment deviation between the star simulator and the turntable based on the imaging law of the characteristic trajectory in the calibration system; wherein, the initial alignment deviation is part of the extrinsic parameters of the star sensor calibration system; The principal point separation module (103) is used to separate principal points based on the imaging rules of the feature trajectories in the calibration system; The feature trajectory optimization design module (104) is used to optimize and process the feature trajectory in order to further improve the accuracy of the main point separation. The residual parameter solving module (105) is used to solve the residual parameters comprehensively based on eliminating the initial alignment deviation between the star simulator and the turntable and separating the principal points, so as to realize the separation of the internal and external parameters of the star sensor.

5. An electronic device, characterized in that, It includes a processor and a memory, the memory storing multiple instructions, and the processor being used to read the instructions and execute the method as described in any one of claims 1-3.

6. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a plurality of instructions, which can be read by a processor and executed as described in any one of claims 1-3.

Images