CN104698849A - Gesture registering and imaging matching algorithm for co-orbit stripe seamless splicing of Agile Earth Observing Satellite - Google Patents

Abstract

The invention provides a gesture registering and imaging matching algorithm for co-orbit stripe seamless splicing of Agile Earth Observing Satellite, relates to the technical field of aerospace, an aims at solving the problem that the gesture and imaging under the stripe seamless splicing cannot be performed by the existing co-orbit stripe splicing technology. The algorithm comprises the steps of determining the longitude and latitude of a sub-satellite point of the satellite and the longitude and latitude of each stripe imaging starting point; calculating the position vectors in a WGS84 coordinate system, corresponding to the longitude and latitude of the sub-satellite point of the satellite and the longitude and latitude of each stripe imaging starting point; calculating a rolling angle and a pitching angle of the satellite, pointing to the first stripe imaging starting point according to the obtained position vectors; obtaining the ground width according to the earth radius and the geocentric angle corresponding to the ground area after the satellite rolls; iteratively calculating the rolling angle and the pitching angle of the satellite at other stripe imaging starting points according to the ground width formula and the stripe overlapping rate; solving a yaw angle of each stripe imaging starting point; achieving co-orbit stripe seamless splicing of the Agile Earth Observing Satellite according to the obtained rolling angle, the pitching angle and the yaw angle of each stripe. The method is high in accuracy, simple and reliable.

Description

Quick satellite is with the seamless spliced attitude registration of rail bar band and imaging matching algorithm

Technical field

The present invention relates to field of aerospace technology, be specifically related to a kind of quick satellite with the seamless spliced attitude registration of rail bar band and imaging matching algorithm.

Background technology

Satellite remote sensing obtains the important means of earth system data and cognitive earth system as the mankind, has been widely used in land resources remote sensing, disaster monitoring, environment mapping and the numerous areas such as atmospheric exploration, national defense and military.The relative size relation of foundation destination object and satellite borne sensor field of View, moonscope target can be divided into point target and regional aim two class.The point target region that normally an area is less, it can be covered by of a satellite borne sensor field of View; And the extensive area that regional aim is normally comprised by polygon or closed curve, need satellite borne sensor repeatedly to take through too much rail and just can complete observation mission, the image of regional aim is the splicing product of multiple bands.

Be utilize rapid attitude maneuver ability with rail multi-ribbon joining image-forming mode of operation, make satellite realize same track and repeatedly push away in the same way and sweep joining image-forming, to increase fabric width.This pattern requirements satellite completes after a band pushes away and sweep, continue to carry out the oppositely motor-driven of pitch orientation immediately in flight course at satellite, by the roll of certain angle, satellite is pointed to the distance of translation about fabric width simultaneously, once push away the starter strip of sweeping after making and front once to push away the starter strip of sweeping adjacent.Strips mosaic imaging pattern can obtain some the fabric widths image identical with band length, thus realizes the large fabric width imaging to regional aim, has broad application prospects in Future Satellite remotely sensed image cause.

Current China is still in the starting stage with rail strips mosaic, wherein how to ensure band seamless spliced be large technological difficulties.And existing satellite strips mosaic imaging technique is due to attitude maneuver complexity, often easily crack or overlap too much, waste the problems such as fabric width to stitching image.The patent No. is in " a kind of novel quick satellite motor-driven middle formation method " of CN 103983254 A, propose the computing formula of satellite to attitude during point target imaging, ensure that satellite optical axis points to point target all the time, but the method is inapplicable to regional aim.The patent No. is in " a kind of attitude adjusting method for quick satellite dynamic imaging " of CN 103217987 A, propose satellite to turn sequence according to the attitude of rolling-pitching-driftage and carry out attitude maneuver, deeply to derive the computing formula of roll angle, the angle of pitch and crab angle Changing Pattern, but do not related to the splicing of satellite band; The patent No. is in " the seamless spliced imaging electric systems of twin-lens nine planar array detectors " of CN 102905061 A, from optical design angle, provide the seamless spliced imaging electric system of a kind of twin-lens nine planar array detectors, but do not relate to the mode of operation of satellite, cannot solve band seamless spliced under attitude and imaging problem.

Summary of the invention

The present invention be solve existing same rail bar band splicing cannot realize band seamless spliced under attitude and imaging problem, provide a kind of quick satellite with the seamless spliced attitude registration of rail bar band and imaging matching algorithm.

Quick satellite is with the seamless spliced attitude registration of rail bar band and imaging matching algorithm, and this algorithm is realized by following steps:

Step one, determine sub-satellite point longitude and latitude and each band imaging starting point longitude and latitude; And calculate described sub-satellite point longitude and latitude and the corresponding position vector under WGS84 coordinate system of each band imaging starting point longitude and latitude;

Step 2, calculate according to the position vector that obtains in step 2 the roll angle that satellite points to satellite band one imaging starting point and pitching angle theta ₁,

Roll angle

At triangle Δ AG ₁in S:

$G_{1}S=\sqrt{{(x_{G1}-x_S)}^2+{(y_{G1}-y_S)}^2+{(z_{G1}-z_S)}^2}$

${\mathrm{AG}}_1=\sqrt{{(x_A-x_{G1})}^2+{(y_A-y_{G1})}^2+{(z_A-z_{G1})}^2}$

Pitching angle theta ₁for in formula, R _efor earth radius, H is orbit altitude, and the position of satellite is the coordinate of A, A is (xA, yA, zA), G ₁for the imaging starting point of satellite band one, S is the imaging starting point G of star band one ₁the perpendicular intersection of point along sub-satellite track direction to sub-satellite point, the position vector of S point is (x _s, y _s, z _s), O is geocentric angle, and ∠ AOS is the angle of OA and OS in triangle Δ OAS;

Base area radius of a ball R _eand the geocentric angle β that after satellite roll, ground region is corresponding ₂, obtain ground fabric width W ₁, i.e. the width of satellite band one imaging;

W ₁＝β ₂*R _e；

Step 3, according to the ground fabric width formula obtained in each band Duplication and step 2, the roll angle of iterative computation satellite band n imaging starting point and the angle of pitch; Described n be more than or equal to 2 positive integer;

The geocentric angle that (n-1)th satellite band ground fabric width is corresponding is β _(2n-2),

In formula, for the roll angle of band n-1, the geocentric angle β of the non-overlapped part correspondence of band n _(2n-1)for: β _(2n-1)=β _(2n-2)* (1-m%);

The roll angle of satellite band n for:

The pitching angle theta of satellite band n _nfor:

In formula, v is satellite flight speed, t _(n-1)that satellite is from G _(n-1)point points to G again through push-scanning image _npoint elapsed time; M is band Duplication, and α is viewing field of camera angle;

Step 4, ask for the crab angle of each band imaging starting point, and according to the roll angle of each band obtained, the angle of pitch and crab angle, realize quick satellite seamless spliced with rail bar band;

The crab angle ψ of described band one ₁for:

In formula, ω _efor rotational-angular velocity of the earth, λ _g1imaging starting point G ₁latitude, λ _k1for substar K during satellite sensing band one imaging starting point ₁latitude, μ is Gravitational coefficient of the Earth, t ₀for when to the imaging at the beginning of satellite band when, t _Ωfor spending the southbound node moment,

The crab angle ψ of satellite band n _nfor:

In formula, λ _gnimaging starting point G _nlatitude, λ _knfor substar K during satellite sensing band n imaging starting point _nlatitude, μ is Gravitational coefficient of the Earth, t _(n-1)that satellite is from G _(n-1)point points to G again through push-scanning image _npoint elapsed time.

Beneficial effect of the present invention: method of the present invention the present invention is directed to the demand of rapid attitude maneuver satellite, solves the problem of satellite with the seamless spliced imaging of rail bar band.The imaging of satellite strips mosaic is due to attitude maneuver complexity, and often easily crack or overlap too much, waste fabric width to stitching image.The present invention is based on seamless angle to set out and calculate attitude of satellite angle, not only ensure that the complete of stitching image but also be unlikely to cause the waste of fabric width, greatly improved observation scope and the observed efficiency of rapid attitude maneuver satellite joining image-forming pattern.And the present invention, fully in conjunction with earth curved surface, can not only try to achieve more accurate solution, and computing velocity is fast, can meet the constraint of Large-scale Optimization Problems for Algorithms T-cbmplexity.

Accompanying drawing explanation

Fig. 1 is quick satellite of the present invention with the process flow diagram of the seamless spliced attitude registration of rail bar band and imaging matching algorithm method;

Fig. 2 is that quick satellite of the present invention is with ellipsoid (plane figure) radius and geocentric latitude relation schematic diagram in the seamless spliced attitude registration of rail bar band and imaging matching algorithm;

Fig. 3 is that quick satellite of the present invention is with the seamless spliced attitude registration of rail bar band and imaging matching algorithm Satellite roll angle angle of pitch schematic diagram;

Fig. 4 is that quick satellite of the present invention is with the seamless spliced attitude registration of rail bar band and imaging matching algorithm Satellite roll ground fabric width schematic diagram;

Fig. 5 is that quick satellite of the present invention is with rail bar band seamless spliced attitude registration and the overlapping schematic diagram of imaging matching algorithm Satellite imaging strips mosaic;

Fig. 6 is that quick satellite of the present invention is with the seamless spliced attitude registration of rail bar band and imaging matching algorithm Satellite crab angle schematic diagram;

Fig. 7 is that quick satellite of the present invention is with the imaging point length velocity relation figure in the seamless spliced attitude registration of rail bar band and imaging matching algorithm.

Embodiment

Embodiment one, composition graphs 1 to Fig. 7 illustrate present embodiment, and quick satellite is with the seamless spliced attitude registration of rail bar band and imaging matching algorithm, and the method is realized by following steps:

Determine sub-satellite point longitude and latitude and each band imaging starting point longitude and latitude; And calculate described sub-satellite point longitude and latitude and the corresponding position vector under WGS84 coordinate system of each band imaging starting point longitude and latitude;

One, sub-satellite point longitude and latitude and each band imaging starting point longitude and latitude is determined; Composition graphs 2, if the longitude and latitude of 1 X in ground is (γ, λ), the semimajor axis of ellipsoid a under WGS84 coordinate system _earth=6378.137km, earth eccentricity e _earthunder the model of=0.08181919, the position vector coordinate (x, y, z) of this point is

$\left\{\begin{array}{c}x=a_{\mathrm{earth}}\×\cos \left(\mathrm{\λ}\right)\×\cos \left(\mathrm{\γ}\right)\\ y=a_{\mathrm{earth}}\×\cos \left(\mathrm{\λ}\right)\×\sin \left(\mathrm{\γ}\right)\\ z=a_{\mathrm{earth}}\×\left(\sqrt{1-{e_{\mathrm{earth}}}^2}\right)\×\sin \left(\mathrm{\λ}\right)\end{array}\right.$

And then obtain satellite band one imaging starting point G ₁point and substar K ₁position vector (the x of point _g1, y _g1, z _g1), (x _k1, y _k1, z _k1).

Composition graphs 3, if earth radius is R _e, orbit altitude is H, and the coordinate of satellite position A is (x _a, y _a, z _a), corresponding substar position is K ₁, satellite flight speed is v, calculates and points to G ₁the satellite roll angle of point and pitching angle theta ₁.

By draw the position vector (x of S point _s, y _s, z _s)

Solving a triangle Δ OAS obtains corresponding central angle is:

$\∠\mathrm{AOS}=\arccos \left(\frac{x_A{x}_S+y_A{y}_S+z_A{z}_S}{\sqrt{{x_A}^2+{y_A}^2+{z_A}^2}*\sqrt{{x_S}^2+{y_S}^2+{z_S}^2}}\right)$

Roll angle

At triangle Δ AG ₁have in S

$G_{1}S=\sqrt{{(x_{G1}-x_S)}^2+{(y_{G1}-y_S)}^2+{(z_{G1}-z_S)}^2}$

${\mathrm{AG}}_1=\sqrt{{(x_A-x_{G1})}^2+{(y_A-y_{G1})}^2+{(z_A-z_{G1})}^2}$

The angle of pitch in formula, S is the imaging starting point G of star band one ₁the perpendicular intersection of point along sub-satellite track direction to sub-satellite point, O is geocentric angle, and ∠ AOS is the angle of OA and OS in triangle Δ OAS;

Composition graphs 4, setting α is the field angle of Satellite Camera, and B, C are respectively the right boundary of ground fabric width after satellite roll, β ₁be corresponding geocentric angle, β ₂be corresponding geocentric angle.Roll angle after the roll of calculating satellite corresponding ground fabric width, i.e. the width W of satellite band 1 ₁.

In Δ OAB and Δ OAC, obtain according to sine,

Satellite roll the geocentric angle β that ground fabric width behind angle is corresponding ₂for:

Satellite roll angle after ground fabric width W ₁for:

W ₁＝β ₂*R _e

Composition graphs 5, BC is respectively the right boundary of satellite band 1, and DE is respectively the right boundary of band 2, and FI is respectively the right boundary of band 3, β ₃for corresponding geocentric angle, β ₄for corresponding geocentric angle, β ₅for corresponding geocentric angle.Setting band Duplication is m%, and for saving fabric width and realizing seamless spliced, general m value is 5≤m≤10.Satellite is from G ₁point points to G again through push-scanning image ₂point elapsed time is t ₁, G ₂put through push-scanning image to sensing G ₃point elapsed time is t ₂.Calculate satellite respectively and point to region, sight spot band 2 band 3 starting point G ₂, G ₃roll angle and pitching angle theta ₂, θ ₃.

corresponding geocentric angle ${\mathrm{\β}}_3=\frac{W_1*(1-m\%)}{R_e}={\mathrm{\β}}_2*(1-m\%)$

In triangle Δ OAD, solution

$\∠\mathrm{OAD}=\arcsin \left(\frac{R_e*\sin ({\mathrm{\β}}_1+{\mathrm{\β}}_3)}{\sqrt{{R_e}^2+{(R_e+H)}^2-2R_2*(R_e+H)*\cos ({\mathrm{\β}}_1+{\mathrm{\β}}_3)}}\right)$

The roll angle of satellite band 2 is:

The angle of pitch of satellite band 2 is

In like manner,

β ₅＝β ₄*(1-m％)

The roll angle of satellite band 3 is

The angle of pitch of satellite band 3 is

Composition graphs 6 and Fig. 7, arctic point is N, and earth radius is R _e, ω _efor rotational-angular velocity of the earth, satellite is at t ₀moment to band 1 imaging, G ₁band 1 starting point, G ₁the declination of point is λ _g1, imaging start time satellite is positioned at an A, and substar is K ₁, P is Track of Sub-Satellite Point and equatorial node, i.e. satellite southbound node, and spending the southbound node moment is t _Ω, connect NK and hand over equator in a Q, calculate the crab angle ψ that satellite points to band n imaging starting point _n.

By solving the equation of motion of satellite, ascending node argument can be drawn:

$u=\sqrt{\frac{\mathrm{\μ}}{a^3}}*(t_0-t_{\mathrm{\Ω}})$

In formula: μ=398600.44km ³/ s ², be Gravitational coefficient of the Earth, a is satellite orbit semi-major axis.In Δ AOS, obtained by sine

In Δ PKQ trirectangular spherical triangle, utilize trirectangular spherical triangle formula, the deflection of the convected velocity that must be caused by satellite motion

∠PKQ＝arccos(tanλ _K1*cotu)

The convected velocity caused by earth rotation is

V _e＝ω _e*R _e*cosλ _G1

The convected velocity caused by satellite motion is

$V_s=\sqrt{\frac{\mathrm{\μ}}{a^3}}*R_e*\cos \∠\mathrm{AOS}$

According to the length velocity relation of imaging point, composition graphs 7, can draw satellite roll add pitch attitude motor-driven after crab angle be

${\mathrm{\ψ}}_1=\arctan \left(\frac{V_e*\cos (\∠\mathrm{PKQ})}{V_s+V_e*\sin (\∠\mathrm{PKQ})}\right)$

Substar during setting satellite sensing band 2 imaging starting point is K ₂, substar when pointing to band 3 imaging starting point is K ₃, then the crab angle ψ of band 2 ₂for

The crab angle ψ of band 3 ₃for

Quick satellite three-axis attitude angle described in present embodiment is all obtained, and completes with the seamless spliced algorithm of rail bar band.In present embodiment, algorithm is simple and practical, is applicable to the calculating of multiple band (being not limited to three), and has taken into full account earth curved surface, calculates more accurate, provides more reliable foundation for realizing the seamless spliced imaging of quick satellite.

Claims (1)

1. quick satellite is with the seamless spliced attitude registration of rail bar band and imaging matching algorithm, and it is characterized in that, this algorithm is realized by following steps:

Step one, determine sub-satellite point longitude and latitude and each band imaging starting point longitude and latitude; And calculate described sub-satellite point longitude and latitude and the corresponding position vector under WGS84 coordinate system of each band imaging starting point longitude and latitude;

Step 2, calculate according to the position vector that obtains in step 2 the roll angle that satellite points to satellite band one imaging starting point and pitching angle theta ₁,

Roll angle

At triangle Δ AG ₁in S:

$G_{1}S=\sqrt{{(x_{G1}-x_S)}^2+{(y_{G1}-y_s)}^2+{(z_{G1}-z_S)}^2}$

${\mathrm{AG}}_1=\sqrt{{(x_A-x_{G1})}^2+{(y_A-y_{G1})}^2+{(z_A-z_{G1})}^2}$

Pitching angle theta ₁for in formula, R _efor earth radius, H is orbit altitude, and the position of satellite is the coordinate of A, A is (xA, yA, zA), G ₁for the imaging starting point of satellite band one, S is the imaging starting point G of star band one ₁the perpendicular intersection of point along sub-satellite track direction to sub-satellite point, the position vector of S point is (x _s, y _s, z _s), O is geocentric angle, and ∠ AOS is the angle of OA and OS in triangle Δ OAS;

Base area radius of a ball R _eand the geocentric angle β that after satellite roll, ground region is corresponding ₂, obtain ground fabric width W ₁, i.e. the width of satellite band one imaging;

W ₁＝β ₂*R _e；

Step 3, according to the ground fabric width formula obtained in each band Duplication and step 2, the roll angle of iterative computation satellite band n imaging starting point and the angle of pitch; Described n be more than or equal to 2 positive integer;

The geocentric angle that (n-1)th satellite band ground fabric width is corresponding is β _(2n-2),

In formula, for the roll angle of band n-1, the geocentric angle β of the non-overlapped part correspondence of band n _(2n-1)for: β _(2n-1)=β _(2n-2)* (1-m%);

The roll angle of satellite band n for:

The pitching angle theta of satellite band n _nfor:

In formula, v is satellite flight speed, t _(n-1)that satellite is from G _(n-1)point points to G again through push-scanning image _npoint elapsed time; M is band Duplication, and α is viewing field of camera angle;

Step 4, ask for the crab angle of each band imaging starting point, and according to the roll angle of each band obtained, the angle of pitch and crab angle, realize quick satellite seamless spliced with rail bar band;

The crab angle ψ of described band one ₁for:

In formula, ω _efor rotational-angular velocity of the earth, λ _g1imaging starting point G ₁latitude, λ _k1for substar K during satellite sensing band one imaging starting point ₁latitude, μ is Gravitational coefficient of the Earth, t ₀for when to the imaging at the beginning of satellite band when, t _Ωfor spending the southbound node moment,

The crab angle ψ of satellite band n _nfor:

In formula, λ _gnimaging starting point G _nlatitude, λ _knfor substar K during satellite sensing band n imaging starting point _nlatitude, μ is Gravitational coefficient of the Earth, t _(n-1)that satellite is from G _(n-1)point points to G again through push-scanning image _npoint elapsed time.