High-orbit satellite line-of-sight correction method based on earth curvature matching

Abstract

The application discloses a high-orbit satellite line-of-sight correction method based on earth curvature matching, which comprises the following steps: (1) constructing a target line-of-sight vector calculation model of a high-orbit satellite camera; (2) solving a coordinate set of a theoretical earth edge point and extracting a coordinate set of an actual earth edge point; (3) fitting a theoretical earth edge curve and an actual earth edge curve and constructing a coordinate point pair with consistent curvature characteristics; (4) constructing a line-of-sight compensation matrix and solving three-axis rotation compensation angles in the line-of-sight compensation matrix by using the coordinate point pair obtained in step (3); and (5) adding the line-of-sight compensation matrix in the target line-of-sight vector calculation model, realizing the unification of the theoretical earth edge and the actual earth edge and completing line-of-sight correction. According to the application, only a single frame of image containing the earth edge is needed to complete the line-of-sight correction with high precision.

Description

Technical Field

[0001] This invention relates to the field of aerospace remote sensing and satellite measurement technology, and in particular to a method for correcting the line of sight of high-orbit satellites based on Earth curvature matching. Background Technology

[0002] In space missions such as Earth observation and target positioning and tracking, the accuracy of target indication directly determines the effectiveness of mission execution. High-precision line-of-sight vectors are the core foundation for achieving high-precision positioning and tracking. During the operation of high-orbit satellites, various factors such as launch impact, alternating space temperatures, attitude disturbances, and aging of payload components can cause the installation reference of the payload optical system to shift, resulting in a deviation between the actual line of sight and the theoretical line of sight calibrated in the laboratory. This deviation directly reduces the positioning accuracy of satellite Earth observation, the accuracy of target identification, and the reliability of mapping data. Therefore, on-orbit line-of-sight correction is a key technical aspect to ensure the high performance of satellites.

[0003] In the past, various technical approaches have been developed for line-of-sight correction methods for high-orbit satellites, including calibration methods based on ground control points (GCPs), methods using stars for on-orbit calibration, and self-calibration methods.

[0004] One of the earliest applied methods was the calibration method based on ground control points (GCPs). This method pre-deploys ground control points with known coordinates and uses the deviation between the pixel positions of the control points in satellite-captured images and their theoretical positions to infer the line-of-sight correction parameters. Its core advantages are high calibration accuracy, simple and easy-to-understand principle, and applicability to areas accessible by ground control points. However, it is limited by environmental conditions such as region, weather, and lighting, and cannot achieve full global coverage calibration. Furthermore, the deployment and maintenance costs of ground control points are high, making it difficult to meet the needs of real-time and dynamic on-orbit correction by satellites.

[0005] Subsequently, a method using stars for in-orbit calibration was developed. This method uses the precise celestial coordinates of stars to collect star images through star sensors and combines them with satellite attitude information to infer line-of-sight deviation. Its advantage is that it does not rely on ground facilities, has a wide coverage area, and can achieve all-day calibration. However, this method is limited by the universality of star catalog spectral bands and is easily affected by factors such as satellite attitude stability and timeliness, and cannot meet the requirements for real-time line-of-sight correction for single-frame observation images.

[0006] In recent years, self-calibration methods have gradually attracted attention. This method completes calibration through the internal information of image data collected by the satellite's own payload, without the need for an external reference source. It has the advantages of strong real-time performance, high degree of autonomy, and flexible application scenarios. However, the disadvantage is that the calibration accuracy is easily affected by factors such as the richness of image texture and satellite attitude disturbances. In the case of a single texture area or a scenario with drastic attitude changes, the reliability of the calibration results is difficult to guarantee.

[0007] To address the above issues, considering that high-orbit satellites are prone to using reference features of the Earth's edge when observing space targets, ground texture features are significantly distorted due to the observation angle, and stellar observations are easily affected by atmospheric scattering and other factors.

[0008] Therefore, developing a technology for efficient line-of-sight correction under conditions with the Earth's edge has significant application advantages and practical necessity in the current field of aerospace remote sensing and satellite measurement technology. Summary of the Invention

[0009] This invention provides a high-orbit satellite line-of-sight correction method based on Earth curvature matching, which can achieve high-precision line-of-sight correction with only a single frame containing the Earth's edge.

[0010] A method for correcting the line-of-sight of high-orbit satellites based on Earth curvature matching includes the following steps:

[0011] (1) Construct a target line-of-sight vector solution model for high-orbit satellite cameras;

[0012] (2) Solve for the coordinate set of theoretical Earth edge points and extract the coordinate set of actual Earth edge points;

[0013] (3) Fit the theoretical Earth edge curve and the actual Earth edge curve to construct coordinate point pairs with consistent curvature characteristics;

[0014] (4) Construct the line-of-sight compensation matrix and use the coordinate point pairs obtained in step (3) to solve the three-axis rotation compensation angle in the line-of-sight compensation matrix;

[0015] (5) Add a line of sight compensation matrix to the target line of sight vector solution model to unify the theoretical edge of the earth and the actual edge of the earth and complete the line of sight correction.

[0016] In step (1), the target line-of-sight vector solution model is used to realize the mutual conversion between image plane coordinates and the target line-of-sight vector in the geocentric geofixed coordinate system.

[0017] In step (2), the binary search method is used to solve for the coordinate set of theoretical Earth edge points, and the LoG operator is used to extract the coordinate set of actual Earth edge points.

[0018] The bisection method is used to solve for the coordinate set of points at the theoretical Earth's edge, specifically including:

[0019] (2-1) In On the image plane, enumerate columns from 1 to N, for the... The column, the two endpoints and the midpoint of the search area are respectively , and ;

[0020] (2-2) Obtain the line-of-sight vectors of the three points in the geocentric coordinate system through the target line-of-sight vector solution model;

[0021] (2-3) Combine the satellite's position in the geocentric coordinate system, the satellite's line-of-sight vector in the geocentric coordinate system, and the unknowns in the Earth WGS-84 model. The quadratic equation in one variable is solved by substituting the line-of-sight vectors corresponding to the two endpoints and one midpoint of the search area into the equation. ;

[0022] (2-4) If the line-of-sight vectors corresponding to a certain endpoint and the midpoint both point to the interior of the Earth or the cold space, take the midpoint as the endpoint, retake the midpoint, and repeat steps (2-3) and (2-4) to obtain the image coordinates corresponding to the edge of the Earth on that column;

[0023] (2-5) Repeat steps (2-1) to (2-4) to obtain the set of coordinates of the theoretical Earth edge points.

[0024] The LoG operator is used to extract the set of coordinates of actual points on the edge of the Earth, specifically including:

[0025] Substituting the two-dimensional Gaussian function into the two-dimensional Laplace operator yields the LoG operator;

[0026] The LoG operator is convolved with the Earth's edge image captured by a high-orbit satellite camera; the zero-crossing points of the output image are found to obtain the edge positions of the input image; the set of points that close to the outermost image boundary is extracted, which is the actual set of Earth's edge points on the image.

[0027] In step (3), the theoretical Earth's edge curve and the actual Earth's edge curve are fitted, specifically including:

[0028] Parametric equations were fitted using the coordinates of theoretical and actual Earth edge points, and the coefficients of the fitted equations were solved using the least squares method.

[0029] In step (3), constructing coordinate point pairs with consistent curvature characteristics specifically includes:

[0030] Based on the fitted theoretical and actual Earth edge curves, the curvature at each point on the theoretical Earth edge is calculated. This curvature is then transferred back to the actual Earth edge curve. Under the condition that the curvature error is less than a preset error value, the corresponding coordinates of the actual Earth edge in the image are calculated, forming pairs of coordinate points with consistent curvature characteristics. Preferably, the preset error value is... .

[0031] In step (4), the line-of-sight compensation matrix is ​​constructed as follows:

[0032] ;

[0033] in, For the line-of-sight compensation matrix, This is the three-axis rotation compensation angle.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] 1. It does not rely on external reference sources such as ground control points or stars. It can achieve correction by simply using the natural feature of the Earth's edge that is commonly found in high-orbit satellite observation images. This greatly improves the method's scene adaptability and autonomy, and breaks through the geographical and environmental limitations of traditional methods.

[0036] 2. Only a single frame containing the edge of the Earth is needed to complete the line-of-sight correction, which significantly simplifies the calibration process, improves the real-time performance of on-orbit correction, and can quickly respond to line-of-sight deviations caused by dynamic changes such as satellite attitude disturbances on orbit.

[0037] 3. By precisely matching the curvature of the Earth to achieve line-of-sight correction, there is no need to add additional high-precision sensors or other payloads, which reduces the hardware cost and complexity of the satellite system. At the same time, it can ensure high correction accuracy, providing reliable technical support for high-precision Earth observation, target positioning and tracking missions of high-orbit satellites. Attached Figure Description

[0038] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0039] Figure 1 This is a flowchart of a high-orbit satellite line-of-sight correction method based on Earth curvature matching, according to an embodiment of the present invention.

[0040] Figure 2 This is a schematic diagram illustrating the principle of the high-orbit satellite line-of-sight correction method based on Earth curvature matching, as described in an embodiment of the present invention. Detailed Implementation

[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. 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.

[0042] It should be noted that, unless otherwise specified, the features in the following embodiments and implementation methods can be combined with each other.

[0043] like Figure 1 As shown, a high-orbit satellite line-of-sight correction method based on Earth curvature matching includes: calculating the theoretical row and column numbers of the Earth's edge in the image at the current imaging moment by using an established on-orbit payload target line-of-sight vector solution model and combining satellite attitude and orbit information; extracting the actual row and column numbers of the Earth's edge in the image according to the LoG operator; fitting the theoretical and actual Earth's edge; calculating the curvature at each point on the two Earth's edges and constructing feature point pairs with consistent curvature characteristics; and then calculating the exterior orientation line-of-sight compensation matrix based on the above control point pairs to achieve line-of-sight correction.

[0044] Specifically, it mainly includes the following steps:

[0045] S1, Establish a line-of-sight vector solution model for high-orbit satellite payload targets.

[0046] Based on the geometric imaging principle of high-orbit satellite payloads, the target line-of-sight vector calculation model of high-orbit satellite cameras is constructed as follows:

[0047] ;

[0048] in, Image point in pixel coordinate system The corresponding line-of-sight vector in the geocentric coordinate system; This is the rotation matrix for transforming from the geocentric inertial coordinate system to the geocentric fixed coordinate system. For the current Julian time; This is the rotation matrix for transforming the satellite's body coordinate system to the geocentric inertial coordinate system. The parameters are the satellite's three-axis attitude angles, namely the pitch angle. Roll angle Yaw angle ; This is the rotation matrix for transforming the camera coordinate system to the satellite body coordinate system; For the angle of the mirror The corresponding reflection matrix; These are the rotation angles of the mirrors in the east-west and north-south directions, respectively. The principal point of the camera; Principal point error; , This refers to the centroid extraction error of image points; , The size of the pixel in the x and y directions; The principal distance of the camera; Principal distance error; These are the coordinates of the image point; This represents the normalized unit vector.

[0049] The conversion between image plane coordinates and line-of-sight vectors in the Earth-centered Earth-fixed system can be achieved based on the target line-of-sight vector solution model.

[0050] S2, the bisection method is used to solve for the theoretical Earth's edge coordinates.

[0051] exist The method for solving the theoretical Earth's edge using the bisection method on the image plane is as follows:

[0052] ① Column by column from 1 to Perform enumeration, for the th The column, the endpoints and midpoints of the search area are respectively and ;

[0053] ② The line-of-sight vectors corresponding to the three points in ① in the geocentric coordinate system are obtained through the target line-of-sight vector solution model. ;

[0054] ③ Let the satellite's position in the ECEF coordinate system be... The satellite's line-of-sight vector in the ECEF coordinate system is The semi-major axis of the Earth WGS-84 model is The minor semi-axis is By combining the satellite position, satellite line-of-sight vector, and Earth WGS-84 model, we can obtain...

[0055] ;

[0056] This is an unknown. The quadratic equation of This represents the distance coefficient. If the equation has two positive roots, it means that the line of sight intersects the Earth at two points; if the equation has one positive root, it means that the line of sight is tangent to the Earth, and the point of tangency is the edge of the Earth; if the equation has no positive roots, it means that the line of sight is moving away from the Earth.

[0057] Substitute the line-of-sight vectors corresponding to the endpoints and midpoints of the search area into the equation to calculate... .

[0058] ④ If the solution of a certain endpoint and the midpoint has the same properties, it means that the geographical locations corresponding to the endpoint and the midpoint are both cold space or the interior of the Earth. Update the endpoint to the midpoint, and repeat steps ③ and ④ to obtain the image coordinates corresponding to the edge of the Earth on that column.

[0059] ⑤ Repeat steps ①②③④ to obtain the theoretical set of Earth's edge points on the image. As attached Figure 2 As shown by the green line.

[0060] S3, the LoG operator extracts the actual Earth's edge coordinates.

[0061] The two-dimensional Laplace operator is: ;

[0062] The two-dimensional Gaussian function is: ,

[0063] Substituting the two-dimensional Gaussian function into the two-dimensional Laplacian operator yields the LoG operator. :

[0064] ;

[0065] Convolution of the LoG operator with the image , To output the image, The input image is used as the reference point. Finding the zero-crossing points in the output image yields the edge positions of the input image. Extracting the outermost set of points that close to the image boundary represents the actual Earth's edge point set on the image. As attached Figure 2 As shown by the red line.

[0066] S4, fits the Earth's edge, and constructs feature point pairs with consistent curvature characteristics.

[0067] Parametric equations are fitted to both theoretical and actual Earth's edge points, with the following parameter forms: .

[0068] ;

[0069] Solving the coefficients using the least squares method and The objective function is:

[0070] ;

[0071] Calculate the curvature at each point on the theoretical edge of the Earth, then substitute the curvature at each point back to the actual curve of the Earth's edge, minimizing the curvature error. Under the given conditions, the actual Earth's edge coordinates in the image are calculated, forming pairs of coordinate points with consistent curvature characteristics:

[0072] ;

[0073] These are the coordinates of a point on the theoretical edge of the Earth. These correspond to the coordinates of points with the same curvature on the actual edge of the Earth.

[0074] S5 calculates the exterior orientation sight compensation matrix to achieve sight correction.

[0075] Compensation matrix Represented as:

[0076] ;

[0077] The least squares method is used to calculate the triaxial rotation compensation angle using the coordinate points obtained in step S4. The formula is:

[0078] ;

[0079] The line-of-sight vector calculated by the target line-of-sight vector solution model with added line-of-sight compensation matrix is ​​shown in the following formula. The line-of-sight vector is calculated from the uncompensated target line-of-sight vector solution model.

[0080] ;

[0081] The compensated target line-of-sight vector solution model achieves the unification of the theoretical Earth's edge and the actual Earth's edge, thus completing the line-of-sight correction.

[0082] The embodiments described above provide a detailed explanation of the technical solutions and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, additions, and equivalent substitutions made within the scope of the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for correcting the line-of-sight of high-orbit satellites based on Earth curvature matching, characterized in that, Includes the following steps: (1) Construct a target line-of-sight vector solution model for high-orbit satellite cameras; (2) Solve for the coordinate set of theoretical Earth edge points and extract the coordinate set of actual Earth edge points; (3) Fit the theoretical Earth edge curve and the actual Earth edge curve to construct coordinate point pairs with consistent curvature characteristics; (4) Construct the line-of-sight compensation matrix and use the coordinate point pairs obtained in step (3) to solve the three-axis rotation compensation angle in the line-of-sight compensation matrix; (5) Add a line of sight compensation matrix to the target line of sight vector solution model to unify the theoretical edge of the earth and the actual edge of the earth and complete the line of sight correction.

2. The high-orbit satellite line-of-sight correction method based on Earth curvature matching according to claim 1, characterized in that, In step (1), the target line-of-sight vector solution model is used to realize the mutual conversion between image plane coordinates and the target line-of-sight vector in the geocentric geofixed coordinate system.

3. The high-orbit satellite line-of-sight correction method based on Earth curvature matching according to claim 1, characterized in that, In step (2), the binary search method is used to solve for the coordinate set of theoretical Earth edge points, and the LoG operator is used to extract the coordinate set of actual Earth edge points.

4. The high-orbit satellite line-of-sight correction method based on Earth curvature matching according to claim 3, characterized in that, The bisection method is used to solve for the coordinate set of points at the theoretical Earth's edge, specifically including: (2-1) In On the image plane, enumerate columns from 1 to N, for the... The column, the two endpoints and the midpoint of the search area are respectively , and ; (2-2) Obtain the line-of-sight vectors of the three points in the geocentric coordinate system through the target line-of-sight vector solution model; (2-3) Combine the satellite's position in the geocentric coordinate system, the satellite's line-of-sight vector in the geocentric coordinate system, and the unknowns in the Earth WGS-84 model. The quadratic equation in one variable is solved by substituting the line-of-sight vectors corresponding to the two endpoints and one midpoint of the search area into the equation. ; (2-4) If the line-of-sight vectors corresponding to a certain endpoint and the midpoint both point to the interior of the Earth or the cold space, take the midpoint as the endpoint, retake the midpoint, and repeat steps (2-3) and (2-4) to obtain the image coordinates corresponding to the edge of the Earth on that column; (2-5) Repeat steps (2-1) to (2-4) to obtain the set of coordinates of the theoretical Earth edge points.

5. The high-orbit satellite line-of-sight correction method based on Earth curvature matching according to claim 3, characterized in that, The LoG operator is used to extract the set of coordinates of actual points on the edge of the Earth, specifically including: Substituting the two-dimensional Gaussian function into the two-dimensional Laplace operator yields the LoG operator; The LoG operator is convolved with the Earth's edge image captured by a high-orbit satellite camera; the zero-crossing points of the output image are found to obtain the edge positions of the input image; the set of points that close to the outermost image boundary is extracted, which is the actual set of Earth's edge points on the image.

6. The high-orbit satellite line-of-sight correction method based on Earth curvature matching according to claim 1, characterized in that, In step (3), the theoretical Earth's edge curve and the actual Earth's edge curve are fitted, specifically including: Parametric equations were fitted using the coordinates of theoretical and actual Earth edge points, and the coefficients of the fitted equations were solved using the least squares method.

7. The high-orbit satellite line-of-sight correction method based on Earth curvature matching according to claim 1, characterized in that, In step (3), constructing coordinate point pairs with consistent curvature characteristics specifically includes: Based on the fitted theoretical Earth edge curve and the actual Earth edge curve, the curvature at each point on the theoretical Earth edge is calculated. The curvature at each point is then brought back to the actual Earth edge curve. Under the condition that the curvature error is less than the preset error value, the corresponding coordinates of the actual Earth edge in the image are calculated, forming a pair of coordinate points with consistent curvature characteristics.

8. The high-orbit satellite line-of-sight correction method based on Earth curvature matching according to claim 1, characterized in that, In step (4), the line-of-sight compensation matrix is ​​constructed as follows: ； in, For the line-of-sight compensation matrix, This is the three-axis rotation compensation angle.

Images