CN111538051A - Precision processing method for sweep large-width optical satellite - Google Patents

Abstract

The invention discloses a method for accurately processing a large-width optical satellite of a swinging sweep, which analyzes the mechanism of a scanning mechanism on the one hand according to the observation data and internal and external calibration parameters of the large-width optical satellite of the swinging sweep, further determining the positive and negative corresponding relation of the measured value of the induction synchronizer based on the scanning mechanism, establishing the conversion relation between the scanning mechanism and the satellite body to obtain a conversion matrix, on the other hand, modeling the orbit parameters and the attitude parameters by deducing the imaging process, calculating attitude and orbit parameters at the imaging moment, constructing an accurate processing geometric model according to the transformation matrix and the attitude and orbit parameters at the imaging moment, and then, constructing an RFM model, performing equivalent geometric model conversion on the accurate processing geometric model, and calculating RPCs parameters to obtain an equivalent geometric imaging model of the sweep distribution imaging process of the sweep large-width optical satellite, so as to realize accurate processing of each frame of earth observation image.

Description

A precise processing method for swing-sweep wide-width optical satellites

technical field

The invention relates to the technical field of aerospace, in particular to a method for precise processing of swing-sweep wide-width optical satellites.

Background technique

As an important means of obtaining spatial information, high-resolution optical remote sensing satellites play an important role in natural resource monitoring, military reconnaissance, and surveying and mapping. In the past, optical remote sensing satellites mainly used conventional linear array CCD push-broom methods, and in order to increase the width of Earth observation images, the number of physical device CCDs or camera loads was mainly increased, and optical splicing or non-collinear CCD splicing and dual-camera splicing were used. Equal field of view stitching method to obtain a larger overall image.

With the diversified development of optical payload imaging modes, the swing-scan area array optical payload has become an important observation method for obtaining large and wide images. It can scan one-dimensional and multi-step through the internal mechanism without increasing the number of detectors to increase earth observation. The size of the field of view can be reduced, so as to reduce the volume, mass, power consumption and development cost of the payload, and effectively improve the time resolution of satellite Earth observation. The common area array swing-sweep wide-width optical satellite camera payload includes three channels of visible light, mid-wave infrared and long-wave infrared. It can obtain three-channel images with a width of 120Km by swinging the mirror scanning mechanism in 8 steps in the vertical direction. Ability to acquire remote sensing data.

Compared with the previous conventional linear array push-broom imaging and area array imaging modes, the swing-sweep large-width optical satellite adopts the satellite attitude motorized condensation-sweep and swing-sweep mechanism to combine the multi-degree-of-freedom method for image data acquisition. This mode leads to the attitude of the satellite imaging process. The degree of nonlinearity of the model is high, the imaging mechanism is complex, and the geometric model has a high degree of freedom. Therefore, the construction and accurate processing of the high-precision geometric imaging model of the swing-sweep wide optical satellite is relatively complicated. There is no relevant research to provide a suitable method. This makes it difficult to implement subsequent image splicing, fusion and classification monitoring applications of swing-sweep wide optical satellites.

SUMMARY OF THE INVENTION

Aiming at some or all of the problems in the prior art, the present invention provides a precise processing method for a swing-sweep wide-width optical satellite. Precise processing of each frame of Earth observation images, including:

Determine the equivalent conversion matrix corresponding to the step-by-step imaging process of the swing-sweep wide optical satellite, including the conversion matrix between the camera load and the scanning mechanism, the conversion matrix between the scanning mechanism and the satellite body, and the relationship between the satellite body and the object space. conversion matrix between;

Obtain the orbit and attitude parameters at the imaging moment;

constructing an accurate processing geometric model of the swing-sweep wide-width optical satellite based on the equivalent transformation matrix and the orbit and attitude parameters at the imaging moment; and

Equivalent fitting is performed on the precise processing geometric model.

Further, the parameters of the equivalent geometric imaging model include satellite orbit, satellite imaging attitude, imaging time, internal calibration parameters of the camera, and installation parameters between different loads.

Further, the precise processing geometric model is obtained by modeling with a Lagrangian polynomial.

Further, the conversion matrix between the camera load and the scanning mechanism is a fixed value, which is determined by the camera load and the installation parameters of the satellite body.

Further, the conversion matrix between the scanning mechanism and the satellite body is determined by measuring parameters of the inductive synchronizer: it is defined that along the satellite flight direction, the clockwise imaging and recording angle of the inductive synchronizer is a negative value, and the counterclockwise imaging and recording angle is a positive value, When the imaging angle of the sub-satellite point is zero, the conversion matrix between the scanning mechanism and the satellite body at the imaging time t can be obtained.

Further, the pose parameter modeling is performed using a sliding window fitting polynomial.

Further, a rational function model RFM is used to equivalently fit the swing-sweep wide-width optical satellite to accurately process the geometric model.

Further, the calculation of the rational polynomial coefficients RPCs of the RFM model includes:

Establish a global virtual grid for each frame of image;

Based on the swing-sweep wide-width optical satellite, the positive and negative transformation functions of the geometric model are accurately processed, and the virtual grid coordinates of the object square are obtained as control points; and

The parameter calculation is carried out using the principle of least squares adjustment.

Further, in the calculation of the rational polynomial coefficients RPCs, a method of ridge estimation is used to solve the ill-conditioned problem of the equation.

The present invention provides a method for precise processing of a swing-sweep wide-width optical satellite, which realizes the construction and precise processing of a high-precision and strict geometric imaging model. The method can solve the problem of high-precision ground observation for a swing-sweep wide-width optical satellite. The satellite follow-up processing and application lay the foundation.

Description of drawings

In order to further clarify the above and other advantages and features of the various embodiments of the present invention, a more specific description of the various embodiments of the present invention will be presented with reference to the accompanying drawings. It is understood that these drawings depict only typical embodiments of the invention and are therefore not to be considered limiting of its scope. In the drawings, the same or corresponding parts will be denoted by the same or similar numerals for clarity.

FIG. 1 shows a schematic flowchart of a method for precise processing of a swing-sweep wide-width optical satellite according to an embodiment of the present invention;

FIG. 2 shows a schematic flowchart of a method for precise processing of a swing-sweep wide-width optical satellite according to an embodiment of the present invention; and

FIG. 3 shows a schematic diagram of a simulation of a swing-sweep wide-width optical satellite imaging process according to an embodiment of the present invention.

Detailed ways

In the following description, the present invention is described with reference to various examples. However, one skilled in the art will recognize that the various embodiments may be practiced without one or more of the specific details or with other alternative and/or additional methods, materials or components. In other instances, well-known structures, materials, or operations are not shown or described in detail so as not to obscure the concepts of the present invention. Similarly, for purposes of explanation, specific quantities, materials and configurations are set forth in order to provide a thorough understanding of the embodiments of the invention. However, the invention is not limited to these specific details. Furthermore, it is to be understood that the various embodiments shown in the drawings are illustrative representations and have not necessarily been drawn to correct scale.

In this specification, reference to "one embodiment" or "the embodiment" means that a particular feature, structure or characteristic described in connection with the embodiment is included in at least one embodiment of the present invention. The appearances of the phrase "in one embodiment" in various places in this specification are not necessarily all referring to the same embodiment.

It should be noted that the embodiments of the present invention describe the process steps in a specific order, but this is only to illustrate the specific embodiment, rather than limiting the sequence of the steps. On the contrary, in different embodiments of the present invention, the sequence of each step can be adjusted according to the adjustment of the process.

The swing-sweep wide optical satellite camera payload has three channel detection functions of visible light, mid-wave infrared and long-wave infrared through spectroscopic technology, and the field of view of a single frame image is 1.6°. In order to solve the problem that the single imaging width is too small, the satellite mainly uses 1:3 ground speed reduction and its own installed scanning mechanism to perform swing and sweep imaging, and achieve 120-kilometer wide imaging in the vertical direction by controlling the overlap between frames and splicing. The scanning mechanism is mainly composed of scanning mirror, rotating shaft system, torque motor and induction synchronizer. The scanning mirror swings back and forth ±3.5° around the X-axis of the satellite body coordinate system (9 frames of images are obtained by 8 steps of one-way swing scanning in one cycle, and rotates 0.8° every 190ms, and no imaging is performed in the reverse direction), and each frame is recorded by the induction synchronizer. Image swing angle. The position measurement accuracy of the induction synchronizer is not less than ±3", and the repeatability is better than 0.2".

In order to lay a foundation for the subsequent processing and application of the swing-sweep wide-width optical satellite, the present invention provides a precise processing method for the swing-sweep wide-width optical satellite, as shown in FIG. 2 and FIG. Observation data, internal and external calibration parameters, on the one hand, analyze the scanning mechanism mechanism, and then determine the positive and negative correspondence between the measured values of the induction synchronizer based on the scanning mechanism mechanism, establish the conversion relationship between the scanning mechanism and the satellite body, and obtain the conversion matrix, On the other hand, by deducing its imaging process, it can realize the modeling of orbit parameters and attitude parameters, and calculate the attitude and orbit parameters at the imaging time. convert the precise processing geometric model into an equivalent geometric model and calculate the RPCs parameters to obtain the equivalent geometric imaging model of the sweep distribution imaging process of the sweeping wide optical satellite, so as to realize the image quality of each frame of Earth observation images. Precise processing. The technical solutions of the present invention are further described below.

FIG. 1 shows a schematic flowchart of a method for precise processing of a wide-width optical satellite swept and swept according to an embodiment of the present invention. As shown in Fig. 1, an accurate processing method for a swing-sweep wide-width optical satellite is constructed by constructing an equivalent geometric imaging model of the swing-sweep distribution imaging process of the swing-sweep wide-width optical satellite, so as to achieve the accuracy of each frame of Earth observation images. processing, including:

扫描机构与卫星本体之间的转换矩阵

以及卫星本体与物方空间的转换矩阵，其中：Step 101, determine the transformation matrix. In order to achieve accurate processing of each frame of Earth observation images, a strict equivalent geometric imaging model needs to be constructed. In an embodiment of the present invention, the model parameters involved in the equivalent geometric imaging model include satellite orbit, satellite imaging attitude, Imaging time, camera internal calibration parameters, installation parameters between different loads, etc. The step-by-step imaging process of the satellite through the scanning mirror swing can be equivalent to establishing the transformation matrix between the camera load and the scanning mechanism

Conversion matrix between scanning mechanism and satellite body

And the transformation matrix of the satellite ontology and the object space, where:

采用相机载荷与卫星本体安装参数作为固定值；以及Transformation matrix between camera payload and scanning mechanism

The camera payload and satellite body mounting parameters are used as fixed values; and

通过感应同步器测量参数进行确定：定义沿着卫星飞行方向，感应同步器顺时针成像记录角度为负值，逆时针成像记录角度为正值，星下点成像角度为零。则在某成像时刻t，记感应同步器记录角度为θ，扫描机构与卫星本体之间转换矩阵

表示为：Conversion matrix between scanning mechanism and satellite body

Determined by the measurement parameters of the inductive synchronizer: define along the satellite flight direction, the inductive synchronizer's clockwise imaging recording angle is negative, the counterclockwise imaging recording angle is positive, and the sub-satellite imaging angle is zero. Then at a certain imaging time t, the recording angle of the induction synchronizer is θ, and the conversion matrix between the scanning mechanism and the satellite body is

Expressed as:

Next, in step 102, the orbit and attitude parameters are acquired. According to the imaging process and mechanism of the swing-sweep wide-width optical satellite, in order to realize the 120-kilometer vertical orbit swing-sweep imaging, the satellite needs to reduce the ground speed by 1:3, because the attitude of the satellite adopts the attitude maneuver control during the ground speed imaging process. , the attitude changes rapidly at different times and becomes strongly nonlinear. Therefore, in order to ensure that high-precision external orientation parameters can be obtained for each frame of image, it is necessary to fine-tune the orbit and attitude of the satellite at different imaging times for subsequent accurate processing. In an embodiment of the present invention, considering that there is no movement of the orbit during the satellite imaging process, the Lagrangian polynomial is used to accurately process the modeling of the geometric model, and then the sliding window fitting polynomial is used to calculate the difference of the satellite at different imaging moments. The trajectory and attitude are refined and modeled, and the attitude parameter modeling process includes:

Let the satellite attitude observation value group include n time series output values (q ₁ , q ₂ , q ₃ ,...,q _n-1 , q _n ), t _k is the imaging time, and denote the attitude quaternion of n epochs For (q ₀ i,q ₁ i,q ₂ i,q ₃ i),i=1,2,...,n, the corresponding m-1 optimal orthogonal polynomial P _qri (t) is fitted as follows:

P _qri (t)=a ₀ +a ₁ t+a ₂ t ² +...+a _m-1 t ^m-1 , (m≤n,r=1,2,3),

Among them, t represents the time, a _j , j=0,1,...,m-1 represents the polynomial coefficient, and the above formula is set as the linear combination of each orthogonal polynomial δ _j (t):

P _qri (t)=c ₀ δ _o (t)+c ₁ δ ₁ (t)+…+c _m-1 δ _m-1 (t),(r=1,2,3),

Among them, c _j ,j=0,1,...,m-1 represents the orthogonal polynomial coefficient, then according to the principle of least squares, the fitting value of the attitude quaternion at the imaging moment t _k can be obtained as follows:

表示t_k时刻四元数矢量部分拟合值，

表示t_k时刻四元数标量部分拟合值，

表示四元数矢量部分正交多项式拟合系数，

表示四元数矢量部分正交多项式；in,

represents the fitting value of the quaternion vector part at time t _k ,

represents the fitted value of the quaternion scalar part at time t _k ,

represents the partially orthogonal polynomial fitting coefficients of the quaternion vector,

represents a quaternion vector partially orthogonal polynomial;

Next, in step 103, an accurate processing geometric model is constructed. Based on the transformation matrix and the orbit and attitude parameters, the precise processing geometric model of the swing-sweep wide-width optical satellite is constructed:

表示相机载荷定标系数，R_broadsensor表示广义安装矩阵，所述相机载荷定标系数以及广义安装矩阵通过在轨定标方式计算得到；t表示成像时刻，[X Y Z]^T表示目标点的物方坐标，(Ψ_x(l,s),Ψ_y(l,s))表示CCD探元号(l,s)的指向角大小，[X_s(t) Y_s(t) Z_s(t)]^T表示摄影中心的物方坐标，所述坐标通过轨道参数内插得到；λ表示成像比例系数，

分别表示由扫描机构到相机载荷测量坐标系的旋转矩阵、卫星本体到扫描机构的旋转矩阵、由J2000坐标系到卫星本体坐标系旋转矩阵、以及WGS84坐标系到J2000坐标系旋转矩阵；以及in,

Represents the camera load calibration coefficient, R _broadsensor represents the generalized installation matrix, and the camera load calibration coefficient and the generalized installation matrix are calculated by on-orbit calibration; t represents the imaging time, [XYZ] ^T represents the object coordinate of the target point , (Ψ _x (l,s),Ψ _y (l,s)) indicates the pointing angle of the CCD detector number (l,s), [X _s (t) Y _s (t) Z _s (t)] ^T represents the object coordinate of the photographing center, which is obtained by interpolation of the orbit parameters; λ represents the imaging scale coefficient,

respectively represent the rotation matrix from the scanning mechanism to the camera load measurement coordinate system, the rotation matrix from the satellite body to the scanning mechanism, the rotation matrix from the J2000 coordinate system to the satellite body coordinate system, and the WGS84 coordinate system to the J2000 coordinate system rotation matrix; and

Finally, in step 104, an equivalent geometric imaging model is obtained. Although the precise processing geometric model can establish the relationship between the pixel coordinates of each frame of image and the geographic coordinates of the corresponding object square points, it is not versatile in subsequent sensor calibration, image fusion and other application processes, and the calculation efficiency is low and the coordinates are inverse. Calculation requires multiple iterations. Therefore, in order to achieve accurate processing of each frame of Earth observation images, it is also necessary to perform equivalent fitting. In an embodiment of the present invention, a rational function model RFM (Rational Function Model) is used to accurately process the geometric model. Do an equivalent fit, including:

Regularize the image point image coordinates (l, s), latitude and longitude coordinates (B, L) and the ellipsoid height H, so that the coordinate range is between [-1, 1], then the image point image coordinates (l, s ) corresponding to the normalized coordinates (l _n , s _n ) of the image side and the normalized coordinates (U, V, W) of the object coordinates (B, L, H) are respectively expressed as:

Among them, LineOff and SampleOff respectively represent the translation value of the image coordinate; LineScale and SampleScale respectively represent the zoom value of the image coordinate; LonOff, LatOff and HeiOff respectively represent the translation value of the object coordinate; and LonScale, LatScale and HeiScale respectively represent the object coordinate The scaling value of the coordinates;

Then, for each scene image, the relationship between the image coordinate and the object coordinate can be expressed as a polynomial ratio as follows:

in,

Num _L (U,V,W)

=a ₁ +a ₂ V+a ₃ U+a ₄ W+a ₅ VU+a ₆ VW+a ₇ UW+a ₈ V ² +a ₉ U ²

+a ₁₀ W ² +a ₁₁ VUW+a ₁₂ V ³ +a ₁₃ VU ² +a ₁₄ VW ² +a ₁₅ V ² U+a ₁₆ U ³

+a ₁₇ UW ² +a ₁₈ V ² W+a ₁₉ U ² W+a ₂₀ W ³

Den _L (U,V,W)

=b ₁ +b ₂ V+b ₃ U+b ₄ W+b ₅ VU+b ₆ VW+b ₇ UW+b ₈ V ² +b ₉ U ²

+b ₁₀ W ² +b ₁₁ VUW+b ₁₂ V ³ +b ₁₃ VU ² +b ₁₄ VW ² +b ₁₅ V ² U+b ₁₆ U ³

+b ₁₇ UW ² +b ₁₈ V ² W+b ₁₉ U ² W+b ₂₀ W ³

Num _S (U,V,W)

=c ₁ +c ₂ V+c ₃ U+c ₄ W+c ₅ VU+c ₆ VW+c ₇ UW+c ₈ V ² +c ₉ U ²

+c ₁₀ W ² +c ₁₁ VUW+c ₁₂ V ³ +c ₁₃ VU ² +c ₁₄ VW ² +c ₁₅ V ² U+c ₁₆ U ³

+c ₁₇ UW ² +c ₁₈ V ² W+c ₁₉ U ² W+c ₂₀ W ³

Den _S (U,V,W)

=d ₁ +d ₂ V+d ₃ U+d ₄ W+d ₅ VU+d ₆ VW+d ₇ UW+d ₈ V ² +d ₉ U ²

+d ₁₀ W ² +d ₁₁ VUW+d ₁₂ V ³ +d ₁₃ VU ² +d ₁₄ VW ² +d ₁₅ V ² U

+d ₁₆ U ³ +d ₁₇ UW ² +d ₁₈ V ² W+d ₁₉ U ² W+d ₂₀ W ³

Among them, a _i , b _i , c _i , d _i (i=1,2,...,20) are rational polynomial coefficients RPCs (Rational PolynomialCoefficients);

In an embodiment of the present invention, the calculation of the rational polynomial coefficients RPCs includes:

Establish a global virtual grid for each frame of image;

Based on the swing-sweep wide-width optical satellite, the positive and negative transformation functions of the geometric model are accurately processed, and the virtual grid coordinates of the object square are obtained as control points; and

The parameter calculation is carried out using the principle of least squares adjustment.

In yet another embodiment of the present invention, in the calculation of the RPCs, the ridge estimation method is used to solve the ill-conditioned equation problem, so as to overcome the matrix singularity caused by the non-uniform distribution of control points or the over-parameterization of the model for solving the RPCs.

While various embodiments of the present invention have been described above, it should be understood that they have been presented by way of example only, and not limitation. It will be apparent to those skilled in the relevant art that various combinations, modifications and changes can be made therein without departing from the spirit and scope of the present invention. Therefore, the breadth and scope of the invention disclosed herein should not be limited by the above-disclosed exemplary embodiments, but should be defined only in accordance with the appended claims and their equivalents.

Claims (9)

1. a pendulum sweeping large-width optical satellite precise processing method, is characterized in that, comprises the steps: Determine the equivalent conversion matrix corresponding to the step-by-step imaging process of the swing-sweep wide optical satellite, the equivalent conversion matrix includes the conversion matrix between the camera payload and the scanning mechanism, the conversion matrix between the scanning mechanism and the satellite body, and the satellite The transformation matrix between the ontology and the object space; Obtain the orbit and attitude parameters at the imaging moment; constructing an accurate processing geometric model of the swing-sweep wide-width optical satellite based on the equivalent transformation matrix and the orbit and attitude parameters at the imaging moment; and Equivalent fitting is performed on the precise processing geometric model. 2 . The method of claim 1 , wherein the conversion matrix between the camera payload and the scanning mechanism is a fixed value, and the fixed value is determined by the camera payload and installation parameters of the satellite body. 3 . 3. The method according to claim 1, wherein the conversion matrix between the scanning mechanism and the satellite body is determined by measuring parameters of the inductive synchronizer, and the determination comprises the following steps: It is defined that along the flight direction of the satellite, the clockwise imaging recording angle of the induction synchronizer is negative, the counterclockwise imaging recording angle is positive, and the sub-satellite imaging angle is zero, then the relationship between the scanning mechanism and the satellite body at the imaging time t can be obtained. conversion matrix

as follows:

Among them, θ is the recording angle of the inductive synchronizer at the imaging time t. 4. method as claimed in claim 1 is characterized in that, the orbit and attitude parameter of imaging moment are obtained by adopting sliding window fitting polynomial modeling, and the attitude quaternion fitting value of t _k imaging moment is as follows:

in,

represents the fitting value of the quaternion vector part at time t _k ,

represents the fitted value of the quaternion scalar part at time t _k ,

represents the partially orthogonal polynomial fitting coefficients of the quaternion vector,

Represents a quaternion vector partially orthogonal polynomial. 5. The method of claim 1 , wherein the precise processing geometric model is obtained by modeling using Lagrangian polynomials, and the precise processing geometric model is expressed as follows:

in:

Indicates the camera load calibration coefficient; R _broadsensor represents a generalized installation matrix; t represents the imaging time; [XYZ] ^T represents the object coordinate of the target point; (Ψ _x (l, s), Ψ _y (l, s)) represents the size of the pointing angle of the CCD detector number (l, s); [X _s (t) Y _s (t) Z _s (t)] ^T represents the object coordinate of the photography center; λ represents the imaging scale factor;

They represent the rotation matrix from the scanning mechanism to the camera load measurement coordinate system, the rotation matrix from the satellite body to the scanning mechanism, the rotation matrix from the J2000 coordinate system to the satellite body coordinate system, and the WGS84 coordinate system to the J2000 coordinate system rotation matrix. 6 . The method according to claim 5 , wherein the camera load calibration coefficient and the generalized installation matrix are calculated by an on-orbit calibration method, and the object coordinate of the imaging center is obtained by interpolation of orbit parameters. 7 . 7. The method according to claim 1, wherein a rational function model RFM is used to perform equivalent fitting on the precise processing geometric model, and the equivalent fitting comprises: Regularize the image coordinates (l, s), latitude and longitude coordinates (B, L) and ellipsoid height H to make the coordinates range between [-1, 1], and obtain the image coordinates (l _n , s _n ) and the normalized coordinates of the object coordinates (U, V, W); and Establish the RFM model of the image coordinates and the object coordinates:

in: Num _L (U, V, W) =a ₁ +a ₂ V+a ₃ U+a ₄ W+a ₅ VU+a ₆ VW+a ₇ UW+a ₈ V ² +a ₉ U ² +a ₁₀ W ² +a ₁₁ VUW+a ₁₂ V ³ +a ₁₃ VU ² +a ₁₄ VW ² +a ₁₅ V ² U+a ₁₆ U ³ +a ₁₇ UW ² +a ₁₈ V ² W+a ₁₉ U ² W+a ₂₀ W ³ Den _L (U, V, W) =b ₁ +b ₂ V+b ₃ U+b ₄ W+b ₅ VU+b ₆ VW+b ₇ UW+b ₈ V ² +b ₉ U ² +b ₁₀ W ² +b ₁₁ VUW+b ₁₂ V ³ +b ₁₃ VU ² +b ₁₄ VW ² +b ₁₅ V ² U+b ₁₆ U ³ +b ₁₇ UW ² +b ₁₈ V ² W+b ₁₉ U ² W+b ₂₀ W ³ Num _s (U, V, W) =c ₁ +c ₂ V+c ₃ U+c ₄ W+c ₅ VU+c ₆ VW+c ₇ UW+c ₈ V ² +c ₉ U ² +c ₁₀ W ² +c ₁₁ VUW+c ₁₂ V ³ +c ₁₃ VU ² +c ₁₄ VW ² +c ₁₅ V ² U+c ₁₆ U ³ +c ₁₇ UW ² +c ₁₈ V ² W+c ₁₉ U ² W+c ₂₀ W ³ _Dens (U, V, W) =d ₁ +d ₂ V+d ₃ U+d ₄ W+d ₅ VU+d ₆ VW+d ₇ UW+d ₈ V ² +d ₉ U ² +d ₁₀ W ² +d ₁₁ VUW+d ₁₂ V ³ +d ₁₃ VU ² +d ₁₄ VW ² +d ₁₅ V ² U+d ₁₆ U ³ +d ₁₇ UW ² +d ₁₈ V ² W+d ₁₉ U ² W+d ₂₀ W ³ Among them, a _i , b _i , c _i , d _i (i=1, 2, . . . , 20) are rational polynomial coefficients RPCs. 8. The method of claim 7, wherein the calculation of the rational polynomial coefficients RPCs comprises the steps of: Establish a global virtual grid for each frame of image; Based on the swing-sweep wide-width optical satellite, the positive and negative transformation functions of the geometric model are accurately processed, and the virtual grid coordinates of the object square are obtained as control points; and The parameter calculation is carried out using the principle of least squares adjustment. 9 . The method of claim 8 , wherein, in the calculation of the rational polynomial coefficients RPCs, a method of ridge estimation is used to solve the ill-conditioned equation. 10 .

Images