CN107941201A - The zero intersection optical satellite image simultaneous adjustment method and system that light is constrained with appearance - Google Patents

Abstract

The invention discloses a zero-intersection optical satellite image joint adjustment method and system constrained by the same attitude of rays. The method includes: S1 establishing an optical system according to the three-point collinear principle of the instantaneous projection center, the ground point and its corresponding image point. Strict imaging geometric model of satellite imagery; S2 introduces the translation and drift items of satellite attitude on the basis of satellite attitude measurement values, and constructs an attitude error compensation model of optical satellite imagery; The imaging ray corresponding to the point has the same attitude angle described in the two adjacent images, and the ray co-attitude constraint model of the optical satellite image is constructed; S4 uses the least square method to solve the attitude error compensation parameters. The invention utilizes the connection points between adjacent images to connect all zero-intersection images into a whole through the constraints of the same attitude of light rays inherent in the images, so that all zero-intersection optical satellite images can be realized at the same time under the condition of a small number of ground control points. precise positioning.

Description

Joint adjustment method and system for zero-crossing optical satellite imagery constrained by the same attitude of rays

technical field

The invention belongs to the technical field of photogrammetry and remote sensing, and in particular relates to a method and system for joint adjustment of optical satellite images with zero-crossing optical satellite images constrained by the same attitude of rays.

Background technique

High-resolution satellite remote sensing earth observation technology has become one of the important means for human beings to obtain geospatial information. A series of geospatial information products (such as digital elevation models and digital orthophotos) produced by high-resolution satellite images have been widely used. It is used in topographic surveying and mapping, land use investigation and renewal, geological exploration, agricultural and forestry resources investigation, earthquake relief and other application fields. In order to realize the productization of high-resolution satellite images more quickly, better serve my country's social economy and national defense construction, and create more social and economic benefits, it is necessary to solve the problem of precise target positioning of high-resolution satellite images. Accuracy directly determines the accuracy of geospatial information products.

In order to improve the accuracy of image positioning, high-resolution optical satellites are usually equipped with GPS receivers, star sensors, and gyroscopes to determine the position and attitude of satellites during satellite image collection. However, affected by the performance of satellite orbit determination and attitude measurement equipment, satellite position and attitude measurement values inevitably contain measurement errors. In the absence of ground control points, it is still difficult to obtain optimal positioning accuracy from high-resolution optical satellite images. At present, in order to eliminate the influence of these errors on image positioning and obtain optimal positioning accuracy, ground control points are still indispensable. As we all know, collecting ground control points in the field requires a lot of manpower, material and financial resources, especially in large areas covered by high-resolution optical satellite images. For the three-line array three-dimensional mapping satellites such as Tianhui-1 and Ziyuan-3, front-view, bottom-view and rear-view images covering the same area can be obtained from different angles. By making full use of the geometric constraints between the three-view images and performing block adjustment processing with the traditional beam method, with the assistance of a small number of ground control points, the precise positioning of the three-line array three-dimensional satellite images can be realized. However, for single-line array optical mapping satellites such as Gaofen-1, Gaofen-2, and Gaojing-1, the push-broom mode can only be used to collect single-view strip images along the track and divide the strip images into several Standard imagery with some overlap is distributed to users. After the strip image is divided into several standard images, the rays with the same name between adjacent standard images are essentially the same ray, and their intersection angle is 0° (that is, zero intersection). Affected by the zero intersection problem, it is difficult for adjacent standard images in the same strip to form an ideal stereo pair, which cannot satisfy the basic geometric constraint of "intersecting ray pairs with the same name under good intersection conditions" in the traditional bundle method block adjustment. That is to say, for zero-crossing optical satellite images in the same strip, it is difficult to use the traditional beam method block adjustment method to achieve precise positioning of satellite images assisted by a small number of ground control points.

In order to achieve precise positioning of zero-crossing optical satellite images in the same band, a certain number of ground control points need to be uniformly arranged on each scene image, which is often difficult to meet in the actual processing process. There are two main reasons: First, the ground coverage of single-scene high-resolution optical satellite images can usually reach 20×20km ² to 50×50km ² , and it requires a lot of cost investment to evenly arrange ground control points on each scene image; Second, affected by factors such as cloud cover, forest coverage, and lack of texture, it is difficult to obtain evenly distributed control points on each image. Therefore, how to achieve precise positioning of zero-crossing optical satellite images with the assistance of a small number of ground control points is of great importance for the application value of single-line array satellite images such as Gaofen-1, Gaofen-2, and Gaojing-1 in my country. significance.

Contents of the invention

Aiming at the current situation that a large number of ground control points are required for accurate positioning of multi-view zero-intersection optical satellite images in the same strip, the present invention proposes a joint adjustment method and system for zero-intersection optical satellite images constrained by the same attitude of rays. The same attitude constraint is used to realize the precise positioning of zero-crossing optical satellite images assisted by a small number of ground control points.

The present invention provides a joint adjustment method for zero-crossing optical satellite images constrained by the same attitude of rays, including:

S1 establishes a rigorous imaging geometric model of optical satellite images based on the three-point collinear principle of the instantaneous projection center, the ground point and its corresponding image point;

S2 introduces the translation item and the drift item of satellite attitude on the basis of satellite attitude measurement value, constructs the attitude error compensation model of optical satellite image; Described translation item and drift item constitute the attitude error compensation parameter of optical satellite image;

S3 According to the attitude angles described by the imaging rays corresponding to the image points on the two adjacent images in the same strip in the two adjacent images are the same, construct the ray same-attitude constraint model of the optical satellite image;

S4 uses the least square method to solve the attitude error compensation parameters, specifically:

4.1 According to the ground control points on the first scene image and the last scene image in the same strip, use the rigorous imaging geometric model and attitude error compensation model to establish the error equations of the first scene image and the last scene image respectively;

4.2 According to the connection points between two adjacent images in the same strip, the error equations between the two adjacent images are respectively established by using the constraint model of the same attitude of light between the two adjacent images;

4.3 Based on the error equation established in substep 4.1 and substep 4.2, the normal equation is established according to the principle of least squares adjustment;

4.4 Iteratively solve the method equation to obtain the attitude error compensation parameters.

Further, the attitude error compensation model is:

Among them, (φ, ω, κ) represents the satellite attitude; Indicates the satellite attitude measurement; A translation term representing the attitude of the satellite; Indicates the drift item of the satellite attitude; l and l ₀ respectively represent the row coordinates of the image point and the central scanning line image in the image plane coordinate system;

And the ray same attitude constraint model is:

in, and Denote the satellite attitude measurement values described by the imaging ray in scene i and scene i+1 respectively; and respectively represent the attitude error compensation parameters described by the imaging light in the i-th scene and the i+1-th scene image; l ⁱ and l ⁱ⁺¹ respectively represent the image point in the image plane coordinate system of the i-th scene and i+1 scene image row coordinates; and Respectively represent the line coordinates of the central scanning row images of the i-th scene and the i+1 scene images in their respective image plane coordinate systems; the i-th scene and the i+1 scene images are two adjacent images in the same strip.

Further, the error equation established in sub-step 4.1 is:

V ₁ =A ₁ X ₁ -L ₁

V _k ＝ A _k X _k -L _k

Among them, the vectors V ₁ and V _k respectively represent the correction numbers of the observed values of the image point coordinates on the first scene image and the last scene image; the matrices A ₁ and A _k represent the attitude error compensation of the first scene image and the last scene image respectively The design matrix formed by the partial derivatives of the parameters, the design matrix is obtained according to the rigorous imaging geometric model and the attitude error compensation model; the vectors X ₁ and X _k represent the correction numbers of the attitude error compensation parameters of the first scene image and the last scene image respectively; The vectors L ₁ and L _k represent the residuals of the image point coordinates on the first scene image and the last scene image respectively.

Further, the error equation established in sub-step 4.2 is:

Among them, the vector V _i,i+1 represents the correction number of the inconsistent value of the light attitude between the two adjacent images i and i+1, and the inconsistent value of the light attitude represents the satellite described by the same imaging light in the adjacent two images. The difference of attitude; Vector Xi and Xi ₊₁ respectively represent the correction number of the attitude error compensation parameter of adjacent two scene images _i and i+1; and respectively represent the design matrix formed by the partial derivatives of the attitude error compensation parameters of the images i and i+1 in the adjacent two scene images i and i+1, and the design matrix is obtained according to the ray same-attitude constraint model.

The zero-crossing optical satellite image joint adjustment system of the same attitude constraint of rays provided by the present invention includes:

The rigorous imaging geometric model building module is used to establish a rigorous imaging geometric model of optical satellite images according to the three-point collinear principle of the instantaneous projection center, the ground point and its corresponding image point;

The attitude error compensation model building module is used to introduce the translation item and the drift item of the satellite attitude on the basis of the satellite attitude measurement value, and constructs the attitude error compensation model of the optical satellite image; the translation item and the drift item constitute the attitude of the optical satellite image Error compensation parameters;

The ray same-attitude constraint model building module is used to construct the ray same-attitude of the optical satellite image according to the same attitude angle described by the imaging rays corresponding to the image points on the two adjacent images in the same strip. constraint model;

Attitude error compensation parameter solving module is used to solve the attitude error compensation parameters by least square method;

The attitude error compensation parameter solving module further includes:

The first error equation building module is used to establish the first scene image and the last scene image respectively by using the rigorous imaging geometry model and the attitude error compensation model according to the ground control points on the first scene image and the last scene image in the same strip The error equation of the scene image;

The second error equation building module is used to establish the error between each adjacent two scene images according to the connection points between the adjacent two scene images in the same strip, using the light beam same attitude constraint model between the adjacent two scene images equation;

A normal equation building module is used to set up a normal equation based on the error equation established by the first error equation building module and the second error equation building module according to the principle of least squares adjustment;

The iterative solution module is used to iteratively solve the method equation to obtain the attitude error compensation parameters.

The present invention has following advantage and beneficial effect:

Starting from the imaging mechanism of the optical satellite sensor, the present invention establishes a light-attitude constraint model for optical satellite images, and on this basis, combines the rigorous imaging geometric model of optical satellite images and the attitude error compensation model to propose a zero-angle constraint model for light-attitude constraints A joint adjustment method and system for rendezvous optical satellite images. The invention can simultaneously realize precise positioning of multi-view zero-intersection optical satellite images in the same strip with a small number of ground control points, thereby providing technical support for the wide application of single-line array optical satellite images in my country.

Description of drawings

Fig. 1 is the method flowchart of the embodiment of the present invention;

Fig. 2 is a schematic diagram of segmentation of a single linear array strip image according to an embodiment of the present invention;

Fig. 3 is a distribution map of control points in the Wuhan-Xianning test area adopted in the embodiment of the present invention.

Detailed ways

In order to more clearly illustrate the embodiments of the present invention and/or the technical solutions in the prior art, the specific implementation manners of the present invention will be described below with reference to the accompanying drawings. Obviously, the drawings in the following description are only embodiments of the present invention, and those skilled in the art can also obtain other drawings based on these drawings and obtain other implementation.

The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

The invention utilizes the connection points between the images, and logically connects all zero-crossing optical satellite images in the same strip into a complete strip image through the constraint of the same attitude of the light rays, so that in the case of a small number of ground control points, At the same time, the joint accurate solution of the attitude error compensation parameters of all satellite images in the strip is realized. Therefore, it is expected to greatly reduce the economic, human and material costs required for field measurement of ground control points by making full use of the same-attitude constraints of light rays when performing precise positioning of zero-crossing optical satellite images in the same strip.

The process flow of the method provided by the embodiment of the present invention is shown in Figure 1, including the steps: (1) establishing a rigorous imaging geometric model of the optical satellite image; (2) establishing an attitude error compensation model of the optical satellite image; (3) establishing an optical satellite image The ray-attitude constraint model of the image; (4) Solve the attitude error compensation parameters of the optical satellite image.

The specific implementation process of each step will be described below.

In step (1), a rigorous imaging geometric model of optical satellite images is established.

Let (X, Y, Z) and (X _S , Y _S , Z _S ) be the object space coordinates of the ground point P and the instantaneous projection center S respectively; (0, y) is the image point p corresponding to the ground point P at Image square coordinates in the instantaneous image coordinate system; (x ₀ , y ₀ ) is the principal point coordinates of the image; (Δx, Δy) is the correction value of camera lens distortion; f is the principal distance of the camera; λ is the scale factor; The attitude angle corresponding to the scan line image is The rotation matrix formed by it is R.

According to the three-point collinear principle of the instantaneous projection center, the ground point P and its corresponding image point p, a rigorous imaging geometric model for precise positioning of optical satellite images is established, and its mathematical expression is:

In step (2), the attitude error compensation model of the optical satellite image is established.

High-resolution optical satellites are usually equipped with GPS receivers, star sensors and gyroscopes to measure (X _S , Y _S , Z _S ) and Analyzing formula (1), we can see that the errors that affect the positioning accuracy of high-resolution optical satellite images can be divided into two categories: one is static errors, which mainly include the errors of principal point, principal distance and lens distortion; the other is time The variable error mainly includes the measurement error of satellite position and attitude.

For ease of understanding, various errors will be introduced separately below.

1) The error of image principal point, principal distance and lens distortion.

In recent years, my country's high-resolution optical satellite sensor on-orbit geometric calibration technology has made a qualitative leap. Through periodic on-orbit geometric calibration, the precise values of principal point, principal distance and lens distortion parameters can be obtained. Therefore, in the process of precise positioning of high-resolution optical satellite images, the principal point, principal distance and lens distortion can be regarded as known values.

2) The measurement error of the satellite position.

With the continuous development of my country's optical satellite orbit determination technology, the position measurement accuracy of my country's optical remote sensing satellites has reached the sub-meter level or even higher. Therefore, in the process of precise positioning of high-resolution optical satellite images, the measurement error of satellite position can be ignored, or compensated by satellite attitude parameters.

3) Measurement error of satellite attitude.

At present, the attitude measurement accuracy of my country's optical remote sensing satellites can only reach a few arcseconds or even tens of arcseconds. Compared with the measurement error of the satellite position, the measurement error of the satellite attitude has a more obvious impact on the image positioning accuracy. For satellite attitude measurement errors, ground control points must be used to eliminate them in order to obtain the optimal positioning accuracy of optical satellite images.

The present invention is in satellite attitude measurement value Introducing a translation term based on and drift term Establish an attitude error compensation model for optical satellite images, namely:

In formula (2), l and l ₀ are the row coordinates of the image point p and the central scanning line image in the image plane coordinate system, respectively.

In step (3), the ray co-attitude constraint model of the optical satellite image is established.

The imaging sensors carried on high-resolution optical remote sensing satellites are usually line array sensors. Through the push-broom mode along the track, the single line array sensor can collect a complete strip image, as shown in Figure 2. In the figure, o-ls represents the image plane coordinate system of the strip image, point o is defined as the first pixel point of the first scanning line, the l axis is along the direction of the satellite orbit, and the s axis is perpendicular to the direction of the satellite orbit.

Point P is any ground point in object space, and point p is the corresponding image point of ground point P on the strip image.

Usually, satellite image suppliers will divide the complete strip image into several standard images with a certain degree of overlap, and provide the standard images to image users. The i-th scene and i+1 scene images in Figure 2 are two adjacent standard images after segmentation, and o _i -l _i s _i and o _i+1 -l _i+1 s _i+1 represent the i-th scene respectively The image plane coordinate system of scene and i+1 scene images is defined the same as o-ls; the image point p is located in the overlapping area of the two scene images.

In fact, after the strip image is divided into standard images, the image point p on the i-th scene and the i+1 scene image is still the conformation of the ground point P, and the imaging ray pP corresponding to the image point p is in the object space The pose in the coordinate system does not change. That is to say, the attitude angle described by the imaging ray pP in the i-th scene image Instead of the attitude angle described in the i+1th scene image Are the same.

According to the ray same-pose constraint, we can get:

Substituting equation (2) into equation (3), the ray co-attitude constraint model of zero-crossing optical satellite images is established:

In formula (4), the meanings of the parameters are the same as those in formula (2), and the superscripts i and i+1 represent the images of the i-th scene and the i+1 scene respectively.

more specific:

The image point p represents the connection point between two adjacent zero-crossing images in the same strip;

and Denote the satellite attitude measurement values described by the imaging ray pP in scene i and scene i+1 respectively;

and Respectively represent the translation items of the satellite attitude of the i-th scene and the i+1 scene image, and Respectively represent the drift items of the satellite attitude of the i-th scene and the i+1 scene images;

That is, the attitude error compensation parameters described by the imaging ray pP in the i-th scene image, That is, the attitude error compensation parameters described by the imaging ray pP in the i+1th scene image;

l ⁱ and l ⁱ⁺¹ are the row coordinates of image point p in the image plane coordinate system of scene i and scene i+1 respectively;

and are the row coordinates of the central scanning row images of the i-th scene and the i+1 scene images in their respective image plane coordinate systems.

Step (4), solving the attitude error compensation parameters of the optical satellite image.

For the sake of brevity, here only three adjacent scenes (the first scene, the second scene and the third scene) zero-intersection optical satellite images in the same strip are taken as examples to illustrate the solution method of the attitude error compensation parameters of the present invention. For more zero-crossing optical satellite images in the same strip, the same can be deduced. Among them, the ground control points are only distributed on the first scene (that is, the first scene) image and the last scene (that is, the third scene) image in the strip, and the middle scene (that is, the second scene) image is connected with other The images are linked.

The solution process of attitude error compensation parameters is realized as follows:

1) For each ground control point on the images of the first scene and the third scene, according to the rigorous imaging geometric model constructed in step (1) and the attitude error compensation model constructed in step (2), respectively establish the first scene and the second scene The error equation of the scene image is as follows:

V ₁ =A ₁ X ₁ -L ₁ (5)

V ₃ =A ₃ X ₃ -L ₃ (6)

In formula (5)～(6):

Vectors V ₁ and V ₃ respectively represent the correction numbers of the observed values of the image point coordinates on the images of the first scene and the third scene;

Matrices A ₁ and A ₃ respectively represent the design matrix formed by the partial derivatives of the attitude error compensation parameters of the first scene and the third scene images, and the design matrix is obtained according to the rigorous imaging geometric model and the attitude error compensation model;

Vectors X ₁ and X ₃ respectively represent the correction numbers of the attitude error compensation parameters of the first scene and the third scene images;

Vectors L ₁ and L ₃ represent the residuals of image point coordinates on the images of the first scene and the third scene, respectively.

2) For each connection point between the images of the first scene and the second scene, and between the images of the second scene and the third scene, the error equations are respectively established according to the constraint model of the same attitude of light constructed in step (3), as follows:

In formula (7)～(8):

The vectors V _1,2 and V _2,3 are the correction numbers of the inconsistent value of the light attitude between the first scene and the second scene image, and between the second scene and the third scene image, and the light attitude inconsistent value represents the same imaging light The difference between the satellite attitudes described in two adjacent images;

matrix and Respectively represent the design matrix formed by the partial derivatives of the attitude error compensation parameters of the first scene and the second scene image, and the design matrix is obtained according to the ray same-attitude constraint model;

and Respectively, the design matrices formed by the partial derivatives of the attitude error compensation parameters of the second scene and the third scene images;

The vector X ₂ represents the correction number of the attitude error compensation parameter of the second scene image.

3) Based on the error equations (5) to (8), the normal equation is formed according to the principle of least squares adjustment:

4) Iteratively solve the unknowns X ₁ , X ₂ and X ₃ in the method equation, so as to obtain the attitude error compensation parameters of the images of the first scene, the second scene and the third scene.

Example

In this example, the downward-looking images (image 1, image 2, image 3, and image 4) of Sijing Resource No. 3 covering the area from Wuhan to Xianning, Hubei Province were selected for experimentation. Adjacent images have a small amount of overlap and belong to the same band. Zero-crossing optical satellite imagery of . The basic information of the test area is shown in Table 1, and the distribution of ground control points is shown in Figure 3.

Table 1 Basic Information of Wuhan-Xianning Experimental Area

In order to verify the effectiveness and practicability of the present invention, this embodiment first uses four ground control points to solve the attitude error compensation parameters of each scene image respectively, and uses the remaining ground control points as checkpoints to count the image positioning accuracy, which is listed in Table 2.

Table 2 Positioning accuracy of ZY-3 satellite images

Analyzing the test results in Table 2, it can be seen that the positioning accuracy of each Jing Zi Zi No. 3 satellite image is better than 0.9 pixels. However, in order to obtain such a high image positioning accuracy, it is necessary to evenly distribute 4 ground control points on each scene image. Therefore, without considering the inherent geometric constraints of zero-crossing optical satellite images, when the number of images increases, the number of ground control points required for image positioning will also increase, which will undoubtedly significantly increase the field measurement of ground control points. workload, increasing human and financial costs. Moreover, due to factors such as cloud cover, forest coverage, and lack of texture, it is difficult to obtain evenly distributed control points on each image, which brings difficulties to the precise positioning of optical satellite images.

Table 3 has listed the positioning accuracy of the No. 3 satellite image obtained by the method of the present invention, that is, two ground planes are respectively arranged on the first scene image (image 1) and the last scene image (image 4) in the same strip Control points, adjacent images are connected by connecting points, and the joint adjustment processing of the zero-crossing optical satellite images constrained by the same attitude of rays is performed, and then the remaining ground control points on each scene image are used as checkpoints to count each scene separately Image positioning accuracy.

Table 3 Positioning accuracy of ZY-3 satellite images

Analyzing the test results in Table 3, it can be seen that the four satellite images in the same strip can be logically connected into one strip image by using the inherent ray-attitude constraint between the zero-crossing optical satellite images. When two ground control points are respectively distributed on the first scene and the last scene image, the precise positioning of the four scene images can be realized at the same time, and the obtained image positioning accuracy is better than 0.8 pixels. It can be seen that the method of the present invention can greatly reduce the number of ground control points required for precise positioning of zero-crossing optical satellite images under the premise of ensuring the accuracy of image positioning, thereby effectively reducing the economic cost of field measurement of ground control points.

In summary, it can be seen that the zero-crossing optical satellite image joint adjustment method proposed by the present invention is practicable. In view of the current situation that the precise positioning of traditional zero-intersection optical satellite images needs to evenly distribute ground control points on each satellite image, the method of the present invention only needs to evenly arrange a small number of ground control points on the first scene and the last scene image, and can simultaneously Accurately solve the attitude error compensation parameters of each satellite image, and then realize the precise positioning of the zero-crossing optical satellite image.

During specific implementation, the method provided by the present invention can realize the automatic operation process based on software technology, and can also realize the corresponding system in a modular manner.

The above-mentioned embodiments are used to explain the present invention, rather than to limit the present invention. Within the spirit of the present invention and the protection scope of the claims, any modification and change made to the present invention will fall into the protection scope of the present invention.

Claims (5)

1. The light same-attitude constrained zero-intersection optical satellite image joint adjustment method is characterized by comprising the following steps of:

s1, establishing a strict imaging geometric model of the optical satellite image according to the three-point collinear principle of the instantaneous projection center, the ground point and the corresponding image point;

s2, introducing a translation term and a drift term of the satellite attitude on the basis of the satellite attitude measurement value, and constructing an attitude error compensation model of the optical satellite image; the translation term and the drift term form an attitude error compensation parameter of the optical satellite image;

s3, constructing a light homography constraint model of the optical satellite image according to the same attitude angles described in the two adjacent images by the imaging light corresponding to the image points on the two adjacent images in the same strip;

s4, solving attitude error compensation parameters by using a least square method, specifically:

4.1 respectively establishing error equations of the 1 st scene image and the last 1 st scene image by using a rigorous imaging geometric model and an attitude error compensation model according to ground control points on the 1 st scene image and the last 1 st scene image in the same strip;

4.2 according to each connection point between two adjacent scene images in the same strip, respectively establishing an error equation between each two adjacent scene images by using a light homography constraint model between the two adjacent scene images;

4.3 establishing a normal equation according to the least square adjustment principle based on the error equations established in the substep 4.1 and the substep 4.2;

4.4, iteratively solving a method equation to obtain an attitude error compensation parameter.

2. The method of claim 1, wherein the method comprises:

the attitude error compensation model is as follows:

wherein (phi, omega, kappa) represents the satellite attitude;representing satellite attitude measurements;a translation term representing the satellite attitude;a drift term representing a satellite attitude; l and l₀Respectively representing the line coordinates of the image point and the central scanning line image under an image plane coordinate system;

and the light ray same-posture constraint model is as follows:

wherein,andrespectively representing the satellite attitude measurement values described by the imaging light in the ith scene and the i +1 scene images;andrespectively representing attitude error compensation parameters described by imaging light in the ith scene and the (i + 1) th scene images; lⁱAnd lⁱ⁺¹Respectively representing the line coordinates of the image points under the image plane coordinate systems of the ith scene and the i +1 th scene images;andrespectively representing the line coordinates of the central scanning line images of the ith scene and the i +1 th scene in respective image plane coordinate systems; the ith scene and the i +1 scene are adjacent two scenes in the same strip.

3. The method of claim 1, wherein the method comprises:

the error equation established in substep 4.1 is:

V₁＝A₁X₁-L₁

V_k＝A_kX_k-L_k

wherein, the vector V₁And V_kRespectively representing the correction numbers of the coordinate observed values of the image points on the 1 st scene image and the last 1 st scene image; matrix A₁And A_kRespectively representing a design matrix formed by partial derivatives of attitude error compensation parameters of the 1 st scene image and the last 1 st scene image, wherein the design matrix is obtained according to the rigorous imaging geometric model and the attitude error compensation model; vector X₁And X_kRespectively representing the correction numbers of the attitude error compensation parameters of the 1 st scene image and the last 1 st scene image; vector L₁And L_kAnd respectively representing the residual error of the coordinate of the image point on the 1 st scene image and the last 1 st scene image.

4. The method of claim 1, wherein the method comprises:

the error equation established in substep 4.2 is:

<mrow> <msub> <mi>V</mi> <mrow> <mi>i</mi> <mo>,</mo> <mi>i</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> <mo>=</mo> <msubsup> <mi>B</mi> <mi>i</mi> <mrow> <mi>i</mi> <mrow> <mo>(</mo> <mi>i</mi> <mo>+</mo> <mn>1</mn> <mo>)</mo> </mrow> </mrow> </msubsup> <msub> <mi>X</mi> <mi>i</mi> </msub> <mo>-</mo> <msubsup> <mi>B</mi> <mrow> <mi>i</mi> <mo>+</mo> <mn>1</mn> </mrow> <mrow> <mi>i</mi> <mrow> <mo>(</mo> <mi>i</mi> <mo>+</mo> <mn>1</mn> <mo>)</mo> </mrow> </mrow> </msubsup> <msub> <mi>X</mi> <mrow> <mi>i</mi> <mo>+</mo> <mn>1</mn> </mrow> </msub> </mrow>

wherein, the vector V_i,i+1The correction number represents the light attitude inconsistency value between two adjacent scene images i and i +1, and the light attitude inconsistency value represents the difference of satellite attitudes described by the same imaging light in the two adjacent scene images; vector X_iAnd X_i+1Respectively representing the correction numbers of the attitude error compensation parameters of two adjacent scenes of the images i and i + 1;andrespectively representing a design matrix formed by partial derivatives of attitude error compensation parameters of images i and i +1 in two adjacent scenes of images i and i +1, wherein the design matrix is obtained according to a light homography constraint model.

5. The optical satellite image of zero intersection of the light with the appearance constraint unites the adjustment system, its characteristic is, including:

the rigorous imaging geometric model building module is used for building a rigorous imaging geometric model of the optical satellite image according to the three-point collinear principle of the instantaneous projection center, the ground point and the corresponding image point;

the attitude error compensation model establishing module is used for introducing a translation term and a drift term of the satellite attitude on the basis of the satellite attitude measurement value and establishing an attitude error compensation model of the optical satellite image; the translation term and the drift term form an attitude error compensation parameter of the optical satellite image;

the light same-attitude constraint model establishing module is used for establishing a light same-attitude constraint model of the optical satellite image according to the same attitude angles described by the imaging light corresponding to the image points on the two adjacent images in the same strip;

the attitude error compensation parameter solving module is used for solving the attitude error compensation parameters by adopting a least square method;

the attitude error compensation parameter solving module further comprises:

the first error equation establishing module is used for respectively establishing error equations of the 1 st scene image and the last 1 st scene image by utilizing a rigorous imaging geometric model and an attitude error compensation model according to ground control points on the 1 st scene image and the last 1 st scene image in the same strip;

the second error equation establishing module is used for respectively establishing an error equation between each two adjacent scene images by utilizing a light homomorphic constraint model between the two adjacent scene images according to each connecting point between the two adjacent scene images in the same strip;

the normal equation establishing module is used for establishing a normal equation based on the error equations established by the first error equation establishing module and the second error equation establishing module and establishing a normal equation according to the least square adjustment principle;

and the iterative solution module is used for iteratively solving the equation of the method to obtain the attitude error compensation parameter.