CN102840861A - Navigational star screening method for star sensors - Google Patents

Navigational star screening method for star sensors Download PDF

Info

Publication number
CN102840861A
CN102840861A CN2012103445099A CN201210344509A CN102840861A CN 102840861 A CN102840861 A CN 102840861A CN 2012103445099 A CN2012103445099 A CN 2012103445099A CN 201210344509 A CN201210344509 A CN 201210344509A CN 102840861 A CN102840861 A CN 102840861A
Authority
CN
China
Prior art keywords
star
district
sub
nautical
connected domain
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN2012103445099A
Other languages
Chinese (zh)
Other versions
CN102840861B (en
Inventor
吴峰
沈为民
朱锡芳
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Suzhou University
Changzhou Institute of Technology
Original Assignee
Suzhou University
Changzhou Institute of Technology
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Suzhou University, Changzhou Institute of Technology filed Critical Suzhou University
Priority to CN201210344509.9A priority Critical patent/CN102840861B/en
Priority to CN201510107562.0A priority patent/CN106441280B/en
Publication of CN102840861A publication Critical patent/CN102840861A/en
Application granted granted Critical
Publication of CN102840861B publication Critical patent/CN102840861B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Abstract

The invention relates to a navigational star screening method for star sensors, which includes the following steps: (1) according to the limiting magnitude of a star sensor, the stars of the original star catalogue of the whole celestial sphere are filtered and a star number threshold Nth is determined; (2) the number of the remaining stars of the star sensor in the field of view of the current sky area is set as N; if N is less than or equal to Nth, then all the remaining stars are chosen as navigational stars, and step 3 is executed; and if N is greater than Nth, then the navigational stars in the field of view of the current sky area are segmented and screened by a multi-scale image plane; (3) after the navigational stars in the field of view of the current sky area are screened, the star sensor turns to the next direction to repeat navigational star screening in step 2 until the whole celestial sphere is traversed. The multi-scale image plane segmentation and screening method adopted in the invention can be adapted to the variation in the star numbers of different sky areas to delete the redundant stars of sky areas with high star distribution density and keep all the stars of low-density sky areas, and moreover, the distribution of the screened navigational stars is even.

Description

A kind of method that is used for the screening nautical star of star sensor
Technical field
The invention belongs to the celestial navigation technical field, relate to a kind of method that is used for the screening nautical star of star sensor.
Background technology
Star sensor passes through importance in star map recognition; The characteristic of comparative observation star group and navigation star group, identification observation star is confirmed their coordinates in body coordinate system and inertial coordinates system; Thereby the attitude of satellite of measuring is the highest attitude of satellite surveying instrument of a kind of precision in the modern space industry.Importance in star map recognition is the core technology of star sensor; Setting up navigation star storehouse is the important prerequisite of identification star chart; The choose reasonable nautical star is for reducing navigation star stack features similarity; Improve importance in star map recognition speed and importance in star map recognition success ratio, strengthen the anti-pseudo-star interference performance of star sensor, it is significant to improve attitude measurement accuracy.
When nautical star was evenly distributed on the whole day ball, the star stack features redundancy of nautical star was little, and importance in star map recognition stability is high, and usually with preferred (screening) algorithm of nautical star distributing homogeneity evaluation, preferred (screening) algorithm of present nautical star roughly can be divided into two big types.
First kind algorithm evenly is distributed as starting point with nautical star at the whole day ball.The orthogonal grid method of propositions such as the soughing of the wind in forest trees in 1998 projects to the unit celestial sphere on the plane, and quadrature is cut apart this projection plane, and the whole day ball is divided into a lot of mutual Uncrossed homalographic sky districts, and choosing a fixed star in distinguishing in each day is nautical star.Because day district's length breadth ratio is along with latitude changes, nautical star density is also inhomogeneous.Samaan in 2004; Sphere piecemeal method (The Spherical Patches method), fixed elevation spiral collimation method (The Fixed-Slope Spiral method) and charged particle method (the The Charged Particles method) scheduling algorithm of propositions such as Malak A are divided equally celestial sphere; Each day district length breadth ratio is little with the relation of present position, and the nautical star distribution that obtains is also more even.Be published in the nautical star optimization algorithm based on Boltzmann entropy of ELECTRONICS LETTERS on the 40th the 2nd phase of volume in 2004; From two selected nautical stars; Choose other nautical stars one by one; Make all Boltzmann entropys that selected nautical star minimum, this algorithm can effectively be deleted redundant star, obtains uniform whole day ball nautical star and distributes.The number of nautical star in the visual field of the less consideration star sensor of this type of algorithm and each Tian Qu visual field, though can realize that nautical star evenly distributes, when the visual field was very big, the nautical star that at every turn can observe still had redundancy.
Second type of algorithm is implemented in the even distribution on the whole day ball from the even distribution of nautical star on local celestial sphere.Li Lihong in 2000 etc. propose the magnitude method of weighting; Give different weights for every fixed star according to magnitude, the fixed star of low magnitude has high weight, and the fixed star of high magnitude has low weights; Choose nautical star according to weights; Algorithm is superior to the orthogonal grid method, but the less consideration star place of this algorithm, the nautical star distributing homogeneity has much room for improvement.The Hye-Young Kim of Texas A&M university in 2002 etc. have proposed the self-organization nautical star and have chosen algorithm; In the visual field of satisfying the arbitrary axis sensing, reach under the prerequisite of certain navigation star number; Position relation according to fixed star; Select nautical star one by one, nautical star is distributed on local and the whole day ball all more even.The recurrence of propositions such as Zheng Sheng in 2004 is chosen algorithm according to the fixed star number that can observe in the visual field; Method based on SVMs; Generate dynamic magnitude threshold value, the observation star of not distinguishing on the same day in the visual field according to this threshold value screening obtains nautical star, and this method can obtain more uniform nautical star and distribute; But, return and to choose the nautical star that algorithm obtains and distribute still even inadequately for the star sensor that the fixed limit magnitude is arranged.
Summary of the invention
The technical matters that the present invention will solve provides a kind of method that is suitable for evenly filtering out nautical star that is used for star sensor.
Basic thought of the present invention is; Because the sky district in the visual field of star sensor only occupies the very little part of whole day ball; So the sky district in this visual field can regard plane domain as; If the image planes of the nautical star imaging arbitrarily in the visual field evenly distribute, nautical star is also approximate on the whole day ball so evenly distributes.Like this, can screen nautical star as density, convert the distribution problem of nautical star on the whole day ball distribution problem of its star picture on image planes into according to the star on the image planes.
Under said basic thought, the invention provides a kind of method that is used for the screening nautical star of star sensor, comprising:
Step 1, according to the limiting magnitude of star sensor, the original star catalogue of whole day ball is made the star filtration treatment, promptly delete the fixed star that double star, variable and magnitude are higher than limiting magnitude; And confirm star number threshold value N according to star Pattern Recognition Algorithm Th
Step 2, said star sensor are made as N in the quantity of the residue star in visual field, district's day before yesterday, if N≤N Th, then said residue star is all elected nautical star as, execution in step three;
If N>N Th, then cut apart the said nautical star in visual field, district's day before yesterday of screening through multiple dimensioned image planes, its step is following:
Step (1) is imaged onto image planes with said residue star, and being divided into line number to these image planes is that p, columns are the orthogonal grid of q; Each grid in the said orthogonal grid is a sub-district;
Step (2) travels through each sub-district successively, and the quantity of inspection residue star wherein wherein, if the quantity of sub-district residue star has many, then keeps a wherein the brightest star, deletes all the other stars; Judge simultaneously to remain the quantity of star this moment, if N≤N Th, then establishing the current residual star is nautical star, traversal finishes, execution in step three; If N>N Th, then continue traversal; If behind all sub-districts of traversal, N is still greater than N Th, then be used as pixel to the sub-district, if in the sub-district star is arranged; Then the gray-scale value of this pixel is non-0, if no star in the sub-district, then the gray-scale value of this pixel is 0; The neighbor cell with residue star after the traversal is divided into connected domain, calculates the center-of-mass coordinate and the sub-district number of each connected domain;
Step (3) is chosen the wherein maximum connected domain of sub-district number, and being located at a nearest star of center-of-mass coordinate that leaves this connected domain in this connected domain is redundant star; If have many from the nearest star of center-of-mass coordinate in this connected domain, a then wherein the darkest star is a redundant star; If it is a plurality of that number maximum connected domain in sub-district has, then selecting the darkest in these a connected domains star is redundant star; Delete said redundant star; Judge to remain the quantity of star this moment, if N≤N Th, then establishing the current residual star is nautical star, execution in step three; If N>N Th, then repeat this step (3);
Step (4) is if after no longer including connected domain; N is still greater than N Th, the value of then said p and q all subtracts 1, and repeating step (1) is to (4); Up to N≤N Th
After step 3, said nautical star screening when visual field, district's day before yesterday finished, said star sensor forwarded next orientation repeating step two to the screening nautical star, until traversal whole day ball.
Further; If remaining the quantity of star, a sub-district has many described in the said step (2); Then keep a wherein the brightest star, the method for deleting all the other stars comprises: in said orthogonal grid, define three two-dimensional array Marray, Idarray and MAGarray in advance; If m is capable in the said orthogonal grid, the sub-districts of n row have and have a star at least, Marray [m] [n]=1 then, otherwise be zero; If in the said sub-district many stars are arranged, then also write down the asterisk and the magnitude of a brightest star respectively, and delete all the other stars with IDarray [m] [n] and MAGarray [m] [n].
Further, in order to calculate the center-of-mass coordinate of said connected domain fast, the center-of-mass coordinate computing method in the said step (3) are:
x c = Σ x i k , y c = Σ y i k ,
X wherein c, y cBe the gray scale center-of-mass coordinate of each connected domain, x i, y iThe sequence number of representing the row and column at place, i sub-district in this connected domain, k representes the sub-district sum in the connected domain.
Further, in order better image planes to be cut apart; The ratio of the initial value of said p and q and the row of said image planes, row size are than identical.
Compared with prior art; The present invention has following advantage: the multiple dimensioned image planes of employing among (1) the present invention are cut apart the redundant star that method for screening can adapt to not the star distribution high density sky district of the star number in district variation deletion on the same day; All stars that keep low-density sky district, and the nautical star of screening is evenly distributed; (2) the present invention is used as pixel to the sub-district; Have in the sub-district star then the gray-scale value of this pixel be non-0, in the sub-district no star then the gray-scale value of this pixel be 0, the neighbor cell with residue star after the traversal is divided into connected domain; Can calculate the center-of-mass coordinate of this connected domain fast through the ranks of this connected domain; And can effectively delete the redundant star of high density area through this center-of-mass coordinate, the distribution of residue star is trended towards evenly, the even nautical star of promptly last acquisition; (3) nautical star through star sensor completion whole day ball screens, and the nautical star of the whole day ball of screening distributes even too; (4) ratio of the initial value of said p and q makes cutting apart of sub-district more reasonable with row, the row size of said image planes than identical, is convenient to successively decreasing of p and q and cuts apart to accomplish multiple dimensioned image planes.
Description of drawings
For content of the present invention is more clearly understood, below basis specific embodiment and combine accompanying drawing, the present invention is done further detailed explanation, wherein
Fig. 1 is that the present invention is based upon the body coordinate system synoptic diagram on the optical system;
Fig. 2 is the rotation relationship synoptic diagram of inertial coordinates system and body coordinate system;
Fig. 3 is the process flow diagram of the method for screening nautical star of the present invention;
Fig. 4 is the original star chart that provides for example;
Fig. 5 is cut apart after the image planes distribution plan of fixed star in the current visual field according to 5 * 5;
Fig. 6 is the distribution plan that remains star behind each sub-district faint star of deletion in the current visual field;
Fig. 7 is the largest connected territory of deletion distance barycenter remains star the most nearby behind the star in the current visual field a distribution plan;
Fig. 8 is than the distribution plan that remains star in the current visual field behind the faint star in 2 largest connected territories of deletion;
Fig. 9 is the distribution plan of nautical star behind the fixed star in the current visual field of screening;
Figure 10 is cut apart after the image planes distribution plan of fixed star in the current visual field according to 8 * 8;
Figure 11 is the distribution plan that remains star behind each sub-district faint star of deletion in the current visual field;
Figure 12 is deletion apart from the connected domain barycenter distribution plan of residue star in the current visual field behind the star the most nearby;
Figure 13 is cut apart after the image planes distribution plan of residue star in the current visual field according to 7 * 7;
Figure 14 is cut apart after the image planes distribution plan of residue star in the current visual field according to 6 * 6;
Figure 15 is cut apart after the image planes distribution plan of residue star in the current visual field according to 5 * 5;
Figure 16 is cut apart the distribution plan that image planes are handled nautical star in the current visual field, back according to 5 * 5;
Figure 17 is when limiting magnitude is 5.2 grades, and star filters the distribution plan of back, the preceding nautical star of screening;
Figure 18 be when limiting magnitude be 5.2 grades and visual field when being 21.91 ° * 16.47 °, utilization the present invention screens the distribution plan of back nautical star on the whole day ball;
Figure 19 be when limiting magnitude be 5.2 grades and visual field when being 21.91 ° * 16.47 °, screen the probability distribution graph of nautical star star number in the preceding visual field;
Figure 20 be when limiting magnitude be 5.2 grades and visual field when being 21.91 ° * 16.47 °, the probability distribution graph of nautical star star number in the visual field, screening back;
Figure 21 be before limiting magnitude is be 21.91 ° * 16.47 ° time screening of 5.2 grades and visual field with visual field, screening back in the cumulative probability distribution plan of nautical star star number.
Embodiment
Below in conjunction with accompanying drawing and embodiment the present invention is elaborated:
Embodiment 1
Star image height density region has two characteristics, and the one, exist star picture number many in the equal areas, the 2nd, the distance between the star picture is near.The present invention designs the method for the screening nautical star that is used for star sensor according to these two characteristics.
At first,, image planes are divided into the rectangular area of a plurality of homalographics, form an orthogonal grid along the focal plane row of image planes detector, column direction according to first characteristics.Be to simplify narration, each rectangular area is called a sub-district (be in the said orthogonal grid each grid be a sub-district), and the five equilibrium of row, column direction is called yardstick at interval.Select certain yardstick that image planes (also claiming as the plane) are cut apart, in the practical operation, adopting and setting the line number that image planes are cut apart is that p, columns are that q realizes again; If star image height density region contains many stars, the probability that is detected according to brighter star is big, star is as the high characteristics of signal to noise ratio (S/N ratio), so a wherein the brightest star of reservation is handled all sub-districts by this method.
Then, according to second characteristic, the residue star of those close together must be in each other in the neighborhood, also is that the sub-district at their places is communicated with, and forms a connected domain, and star density is high more, and the connected domain scope is big more.If deletion near the star of the barycenter in largest connected territory, then can make connected domain split into a plurality of little connected domains, this regional star density descends.And then from result, select next largest connected territory, handle again according to similar approach, up to no longer including connected domain.
Utilize this yardstick can't determine can also delete which star on earth again, then increase waits branch at interval, reduces isodisperse, and promptly the value of p and q all subtracts 1.
For convenient explanation principle of the present invention, suppose that at present the star sensor body coordinate system is based upon on the optical system, as shown in Figure 1.Make optical system equivalence be the ideal image system, H and H ' be respectively its thing, as square principal point, and f is the focal length of optical system, and the initial point of star sensor body coordinate system is being located as square principal point H ', X bAxle, Y bAxle is parallel to the row and column of image planes detector focal plane respectively, Z in picture side's interarea bAxle is along optical axis, and its forward is as shown in Figure 1, and three constitute right-handed coordinate system.The direction cosine vector of fixed star S in this coordinate system is V b, at X b, Y bField angle XFLD on the direction, YFLD, the coordinate on image planes are (x b, y b).
Be located in the inertial coordinates system, optical axis points to (α c, δ c), can obtain body coordinate system by certain way rotatory inertia coordinate system.As shown in Figure 2, inertial coordinates system earlier around the Z axle by+X axis+Y axle rotation alpha c, obtain X ' Y ' Z ' coordinate system, new coordinate system again around Y ' axle by+Z ' axial+X ' axle half-twist-δ c, obtain X " Y " Z " coordinate system, this coordinate system is around Z " axle rotation φ, obtain body coordinate system X bY bZ b
According to the conversion of above-mentioned star sensor setting and inertial coordinate and body coordinate, the method that is used for the screening nautical star of star sensor of the present invention comprises:
Step 1, according to the limiting magnitude of star sensor, the original star catalogue of whole day ball is made the star filtration treatment, promptly delete the fixed star that double star, variable and magnitude are higher than limiting magnitude; And confirm star number threshold value N according to star Pattern Recognition Algorithm Th
Step 2, said star sensor are made as N in the quantity of the residue star in visual field, district's day before yesterday, if N≤N Th, then said residue star is all elected nautical star as, execution in step three;
If N>N Th, then cut apart the said nautical star in visual field, district's day before yesterday of screening through multiple dimensioned image planes, its step is following:
Step (1) is imaged onto image planes with said residue star, and being divided into line number to these image planes is that p, columns are the orthogonal grid of q; Each grid in the said orthogonal grid is a sub-district;
Step (2) travels through each sub-district successively, and the quantity of inspection residue star wherein wherein, if the quantity of sub-district residue star has many, then keeps a wherein the brightest star, deletes all the other stars; Judge simultaneously to remain the quantity of star this moment, if N≤N Th, then establishing the current residual star is nautical star, traversal finishes, execution in step three; If N>N Th, then continue traversal; If behind all sub-districts of traversal, N is still greater than N Th, then be used as pixel to the sub-district, if in the sub-district star is arranged; Then the gray-scale value of this pixel is non-0, if no star in the sub-district, then the gray-scale value of this pixel is 0; The neighbor cell with residue star after the traversal is divided into connected domain, calculates the center-of-mass coordinate and the sub-district number of each connected domain;
Step (3) is chosen the wherein maximum connected domain of sub-district number, and being located at a nearest star of center-of-mass coordinate that leaves this connected domain in this connected domain is redundant star; If have many from the nearest star of center-of-mass coordinate in this connected domain, a then wherein the darkest star is a redundant star; If it is a plurality of that number maximum connected domain in sub-district has, then selecting the darkest in these a connected domains star is redundant star; Delete said redundant star; Judge to remain the quantity of star this moment, if N≤N Th, then establishing the current residual star is nautical star, execution in step three; If N>N Th, then repeat this step (3);
Step (4) is if after no longer including connected domain; N is still greater than N Th, the value of then said p and q all subtracts 1, and repeating step (1) is to (4); Up to N≤N Th
After step 3, said nautical star screening when visual field, district's day before yesterday finished, said star sensor forwarded next orientation repeating step two screening nautical star to, until traversal whole day ball.
Center-of-mass coordinate computing method in the said step (3) are:
x c = Σ x i k , y c = Σ y i k ,
X wherein c, y cBe the gray scale center-of-mass coordinate of each connected domain, x i, y iThe sequence number of representing the row and column at place, i sub-district in this connected domain, k representes the sub-district sum in the connected domain.
The ratio of the initial value of said p and q and the row of said image planes, row size are than identical.
Embodiment 2
On the basis of embodiment 1, realize the said method that is used for the screening nautical star of star sensor, its practical implementation process is following:
The first step is made the star filtration treatment to original star catalogue.According to limiting magnitude deletion faint star, delete variable, double star simultaneously, for making things convenient for subsequent treatment, the residue sing data with ascending series arrangement, and is confirmed star number threshold value N according to star Pattern Recognition Algorithm according to declination Th, promptly select corresponding star Pattern Recognition Algorithm to confirm star number threshold value N according to actual needs Th, also can oneself set as required.
In second step, the star sensor optical axis points to coordinate (α on the whole day ball i, δ i) the position, α iOr δ i1 ° of each change, traversal whole day ball.Extract each and point to the residue star in the visual field, calculate residue star sum N.If N≤N Th, then said residue star is all elected nautical star as; The star sensor optical axis forwards next orientation to, judges again, greater than threshold value, begins following steps up to star number.
The method of extracting the residue star in the visual field is that (α δ) satisfies at first to pick out coordinate
|δ-δ c|≤w m
Star, w wherein mThe corresponding field angle of expression star sensor image planes detector diagonal line.
Following formula limits the upper and lower bound of residue star declination in the current visual field.Because declination δ span is-90 °-90 °, works as δ i-w mDuring less than-90 °, should be provided with down and be limited to-90 °, similarly, work as δ i-w mDuring greater than 90 °, the upper limit should be set to 90 °, obtains
Figure BDA00002148219400081
The residue sing data confirms that with dichotomy method the declination value is just greater than δ by the declination ordering BotThe position of star is read subsequent data then, extracts the residue star, up to the declination value greater than δ Top
Then, calculating the orientation of residue star in body coordinate system that has extracted, is that (α, star δ) has for right ascension and declination
V bx V by V bz = cos φ sin φ 0 - sin φ cos φ 0 0 0 1 cos ( 90 - δ i ) 0 - sin ( 90 - δ i ) 0 1 0 sin ( 90 - δ i ) 0 cos ( 90 - δ i ) ×
cos α i sin α i 0 - sin α i cos α i 0 0 0 1 cos α cos δ sin α cos δ sin δ
It is at X b, Y bField angle XFLD on the direction, YFLD do
XFLD = - tg - 1 ( V bx V bz ) , YFLD = - tg - 1 ( V by V bz )
If establish optical system at X b, Y bMaximum field of view angle on the direction is W AAnd W B, have only satisfied
|XFLD|≤W A/2、|YFLD|≤W B/2
Fixed star could be observed.Obtain the residue star in the current visual field through the following formula screening, also obtain their field angle XFLD, YFLD simultaneously, and their total N.
The 3rd step was imaged onto image planes with all stars in the visual field, pressed
x b=ftan(XFLD),y b=ftan(YFLD)
The position of calculating and writing down each star picture.
The 4th step; Row, column direction along the focal plane; Cut apart the grid that image planes are p * q; Set up and two-dimensional array Marray, Idarray and MAGarray that the capable q of three p that initialization is corresponding with grid is listed as, the ratio of p and q should try one's best with focal plane row, column direction size than consistent, identical with the yardstick of assurance both direction.During beginning, p and q should get bigger value, so that examine or check the star distribution density in detail.Array Marray and Idarray are initialized as 0, and MAGarray is initialized as-99.99.
In the 5th step, travel through each sub-district to calculate the sub-district at the star place of being extracted.For coordinate is (x b, y b) the star picture, it is in, and m on the image planes is capable, there is the star picture sub-district of n row, so
n = floor ( p 2 f tan ( w A 2 ) x b + p 2 - 1 )
m = floor ( q 2 f tan ( w B 2 ) y b + q 2 - 1 )
Wherein a nearest integer littler than x, Marray [m] [n]=1 are got in floor (x) expression.If there is the star picture of many stars this sub-district, only keep ludcida, the fixed star asterisk that IDarray [m] [n] record remains, MAGarray [m] [n] writes down its magnitude, upgrades residue star sum N in the current visual field, i.e. the star of many stars sub-district screening.
If the 6th step is N≤N Th, returned for second step, otherwise adopt pixel clustering algorithm (pixel clustering algorithm referring to author Yang Fan, " Digital Image Processing and analysis ", ISBN 978-7-5124-0188-4) connection field calculating center-of-mass coordinate, and do further to screen.Promptly be used as pixel to the sub-district, the star number in the sub-district is used as gray-scale value, adopts eight connections that the image planes grid is divided into a plurality of connected domains, calculates the center-of-mass coordinate and the sub-district number of each connected domain.Choose that maximum connected domain of sub-district number, deletion is from the nearest star of center-of-mass coordinate.If many stars are all nearest from center-of-mass coordinate, delete wherein the darkest star.If it is a plurality of that the connected domain that the sub-district number is maximum has, then delete the darkest star in these connected domains.Simultaneously, upgrade N, Marray, and the value of IDarray, MAGarray.If N>N Th, repeat this step, the next maximum connected domain of the scope of seeking again.As N≤N ThThe time, the nautical star screening in the current visual field finishes.If N>N Th, and all no longer connections (promptly not having connected domain) of the sub-district at any two stars place, then carry out next step.
In the 7th step, p and q reduce 1, increase and divide equally yardstick, and the sub-district area of image planes has a small amount of increase, and distance several star pictures far away possibly can be communicated with again, again since the execution of the 4th step, up to N≤N Th
The 8th step finished when the whole day ball travels through, and the nautical star screening finishes.
Above-mentioned steps is simplified to process flow diagram, and is as shown in Figure 3.
Embodiment 3
On the basis of embodiment 1 and embodiment 2, cut apart the embodiment of yardstick (being the different values of q and p) and further specify content of the present invention compared with prior art through two kinds.
Following embodiment with the SAO star catalogue as original star catalogue; SAO star catalogue (The Smithsonian Astrophysical Observatory Star Catalog/ Shi Misong astrophysics platform star catalogue) is a uranometry star catalogue, is published by Shi Misong astrophysics platform in 1966, comprises 258,997 fixed stars altogether.This star catalogue by before some star catalogues compile and form, but it is above and accurately measured fixed star voluntarily only to include 9.0 grades.Star names in the SAO star catalogue is followed digital number by alphabetical SAO beginning and is represented that fixed star is with the declination subregion, and per 10 degree are a district, and being divided into is 18 districts, and the fixed star in each district sorts according to the right ascension position.
Embodiment one
Choosing said SAO star catalogue is original star catalogue; Limiting magnitude is that 5.5 grades, field angle are 20 ° * 20 °, and after original star catalogue star filtration treatment, the star picture of residue star on image planes that certain optical axis points in the visual field distributes as shown in Figure 4; Have 18 residue stars this moment, these stars distribute even inadequately.Get N Th=7, p and q initial value all are taken as 5, and be as shown in Figure 5 after image planes are cut apart by 5 * 5, has to have many stars in a plurality of sub-districts, behind each sub-district deletion faint star, the reservation ludcida, obtains Fig. 6, remains 10 stars.Among Fig. 6, the 1st to 5 row, the 1st, 2 row are wherein formed a connected domain by five sub-districts, and it is maximum to comprise star picture number, and this zone star density is high.Near the star of this connected domain barycenter, this regional star density descends in deletion.The coordinate of these five stars is (1,1), (2,1), (3,2), (4,1), (5,1), and their barycenter is (3,1.25), and it is nearest to (3,2), and that star of deletion (3,2) obtains Fig. 7 so.At this moment, have 3 connected domains and all contain 2 sub-districts, their connected domain scope is maximum.Wherein the star of the 2nd row the 1st row is the darkest, then deletes it.Remaining 2 connected domains contain 2 sub-districts, and scope is maximum, and the star of the 1st row the 3rd row is the darkest stars in these 2 connected domains, deletes it again.The residue star number is not more than threshold value N like this Th, The selection result obtains like Fig. 8 and shown in Figure 9.
Embodiment two
To screening N like the residue star in the corresponding visual field of Fig. 4 ThStill be taken as 7, p and q initial value all are taken as 8, and be shown in figure 10 after image planes are cut apart by 8 * 8, has to have many stars in a plurality of sub-districts, behind each sub-district deletion faint star, the reservation ludcida, obtains Figure 11.
In Figure 11, the 1st to 4 row, the 1st, 2 row are wherein formed a connected domain by four sub-districts, and it is maximum to comprise star picture number, and this zone star density is high.Near the star of the barycenter of this connected domain, this regional star density descends in deletion.The coordinate of these four stars is (1,1), (2,2), (3,1), (4,1), and their barycenter is (2.5,1.25), and it is nearest to (3,1), so that star of deletion (3,1).Then from result, select largest connected territory again, handle again, all no longer be communicated with up to the sub-district at any two residue star places according to similar approach.Result behind processing Figure 11 is shown in figure 12, and the residue star distributes more evenly in this yardstick.
Utilize this yardstick can't determine can also delete which star on earth again, then increase waits branch at interval, reduces isodisperse.Image planes are divided into 7 * 7 totally 49 homalographic zones, obtain Figure 13, do similar processing according to the method for a last yardstick.Increase yardstick again like Figure 14, image planes are divided into 6 * 6 handle, obtain Figure 15 and 16, visible residue star distributes very even, and these remaining stars promptly can be used as nautical star.
Embodiment three
For the method for screening nautical star of the present invention and orthogonal grid method and Boltzmann entropy algorithm are made comparisons; Table 1 provides when the visual field and gets 11.5 ° * 11.5 °, 14 ° * 14 ° respectively; When limiting magnitude got 6 and 7.5 etc. respectively, screening nautical star result's data relatively.The data of orthogonal grid method and Boltzmann entropy algorithm derive from the paper " A General Method of the automatically selection of guide star " and the publication in 2004 that are published in Proceedings of ICSP ' 98 meetings respectively and roll up the paper " Boltzmann entropy-based guide star selection algorithm for star tracker " of the 2nd phase at " ELECTRONICS LETTERS " the 40th.Adopt the present invention, star number threshold value N ThBe taken as 6, row, column direction isodisperse p and q initial value all are taken as 8, and the navigation star storehouse Boltzmann entropy of setting up thus is minimum, and whole day ball homogeneity is best.From local celestial sphere homogeneity, the nautical star star number maximal value that orthogonal grid method and Boltzmann entropy algorithm obtain is bigger, and minimum value is less, and homogeneity is also slightly poor.
The comparison of table 1 the present invention and orthogonal grid method and Boltzmann entropy algorithm
Figure BDA00002148219400131
For choosing algorithm with recurrence, magnitude weighting algorithm and self-organized algorithm are 8 ° * 8 ° when getting the visual field relatively, and limiting magnitude gets 6.5 between 7.9 during totally 7 values, and the result who uses the present invention to screen behind the nautical star is as shown in table 2.Algorithm is chosen in recurrence; Magnitude weighting algorithm and self-organized algorithm data derive from Zheng Sheng in 2004 etc. respectively and are published in " aerospace journal " the 25th the 1st phase of volume " a kind of new nautical star is chosen algorithm research "; Li Lihong in 2000 etc. are published in " optical technology " the 26th the 4th phase of volume " the autonomous triangle star map recognizer of a kind of improved whole day ", and Hye-Young Kim in 2002 etc. are published in the paper " Self-organizing Guide Star Selection Algorithm for Star Trackers:Thinning Method " in the IEEE on aerospace conference proceedings meeting.
Table 2 shows that the present invention can effectively reduce the star number in high density sky district, and very little to the influence of low-density sky district star number, 95% to navigate star number between 5 to 12 in the visual field, Heaven district.When limiting magnitude was lower, the nautical star number was slightly more, mainly be this moment some Tian Qu visual field in star number less than N ThEven there is redundant star in a day district on every side, can not delete.
When limiting magnitude was 6.5 and 7.3, recurrence was chosen algorithm and has been deleted too much star in a part day district, and day district's navigation star number above 10% is less than 5; When limiting magnitude during greater than 7.5 grades; The navigation star storehouse that this algorithm is set up still has in a lot of Tian Qu visual field star number too many, and the star number maximal value is bigger.Utilization magnitude weighting algorithm and self-organized algorithm screening nautical star, the large percentage in low-density sky district.Along with limiting magnitude increases, the navigation star storehouse that the utilization self-organized algorithm is set up, redundancy is increasing.The standard deviation in the navigation star storehouse that the present invention obtains is minimum, and it is the most even to distribute, and is superior to magnitude weighting algorithm, self-organized algorithm and regression algorithm.
Nautical star filtering algorithm contrast during 8 ° * 8 ° visual fields of table 2
Figure BDA00002148219400141
Figure BDA00002148219400151
Embodiment four
Getting the SAO star catalogue is original star catalogue, and limiting magnitude is 5.2 etc., and star sensor detector length breadth ratio is 4:3, and the visual field is 21.91 ° * 16.47 °, and the initial value of row, column direction isodisperse p and q gets 12 and 9, chooses star number threshold value N ThBe 6.When utilization the present invention sets up navigation star storehouse, delete 529 stars altogether, kept 1078 stars.Figure 17 and Figure 18 are respectively before the nautical star screening and the distribution of screening back on the whole day ball, and the Boltzmann entropy drops to 1.3643 * 10 by original 0.0119 -4, nautical star is more evenly distributed.
Through whole day ball traversal, the nautical star star number that statistics is screened after preceding and the screening distributes result such as Figure 19 and shown in Figure 20.Nautical star star number maximal value is by original 47 in the visual field; Be reduced to 18, and minimum is 2 to remain unchanged, the star number standard deviation that calculates is reduced to 1.87 by original 6.15; Average star number is reduced to 9.37 by 13.75, and the homogeneity of nautical star on local celestial sphere improves.Figure 21 is that the cumulative probability of nautical star star number in the visual field distributes; When star number less than 4 the time; Two curves before and after the nautical star screening overlap; The probability that occurs 4 above nautical stars in the visual field is 99.94% all, shows that the present invention can effectively reduce the star number amount in star distribution high density sky district, reduces nautical star characteristic redundancy.
Obviously, the foregoing description only be for clearly the present invention is described and is done for example, and be not to be qualification to embodiment of the present invention.For the those of ordinary skill in affiliated field, on the basis of above-mentioned explanation, can also make other multi-form variation or change.Here need not also can't give exhaustive to all embodiments.And these belong to conspicuous variation or the change that spirit of the present invention extended out and still are among protection scope of the present invention.

Claims (3)

1. method that is used for the screening nautical star of star sensor comprises:
Step 1, according to the limiting magnitude of star sensor, the original star catalogue of whole day ball is made the star filtration treatment, promptly delete the fixed star that double star, variable and magnitude are higher than limiting magnitude; And confirm star number threshold value N according to star Pattern Recognition Algorithm Th
Step 2, said star sensor are made as N in the quantity of the residue star in visual field, district's day before yesterday, if N≤N Th, then said residue star is all elected nautical star as, execution in step three;
If N>N Th, then cut apart the said nautical star in visual field, district's day before yesterday of screening through multiple dimensioned image planes, its step is following:
Step (1) is imaged onto image planes with said residue star, and being divided into line number to these image planes is that p, columns are the orthogonal grid of q; Each grid in the said orthogonal grid is a sub-district;
Step (2) travels through each sub-district successively, and the quantity of inspection residue star wherein wherein, if the quantity of sub-district residue star has many, then keeps a wherein the brightest star, deletes all the other stars; Judge simultaneously to remain the quantity of star this moment, if N≤N Th, then establishing the current residual star is nautical star, traversal finishes, execution in step three; If N>N Th, then continue traversal; If behind all sub-districts of traversal, N is still greater than N Th, then be used as pixel to the sub-district, if in the sub-district star is arranged; Then the gray-scale value of this pixel is non-0, if no star in the sub-district, then the gray-scale value of this pixel is 0; The neighbor cell with residue star after the traversal is divided into connected domain, calculates the center-of-mass coordinate and the sub-district number of each connected domain;
Step (3) is chosen the wherein maximum connected domain of sub-district number, and being located at a nearest star of center-of-mass coordinate that leaves this connected domain in this connected domain is redundant star; If have many from the nearest star of center-of-mass coordinate in this connected domain, a then wherein the darkest star is a redundant star; If it is a plurality of that number maximum connected domain in sub-district has, then selecting the darkest in these a connected domains star is redundant star; Delete said redundant star; Judge to remain the quantity of star this moment, if N≤N Th, then establishing the current residual star is nautical star, execution in step three; If N>N Th, then repeat this step (3);
Step (4) is if after no longer including connected domain; N is still greater than N Th, the value of then said p and q all subtracts 1, and repeating step (1) is to (4); Up to N≤N Th
After step 3, said nautical star screening when visual field, district's day before yesterday finished, said star sensor forwarded next orientation repeating step two to the screening nautical star, until traversal whole day ball.
2. the method for screening nautical star according to claim 1 is characterized in that: the center-of-mass coordinate computing method in the said step (3) are:
x c = Σ x i k , y c = Σ y i k
X wherein c, y cBe the gray scale center-of-mass coordinate of each connected domain, x i, y iThe sequence number of representing the row and column at place, i sub-district in this connected domain, k representes the sub-district sum in the connected domain.
3. the method for screening nautical star according to claim 1 and 2 is characterized in that: the ratio of the initial value of said p and q and the row of said image planes, row size are than identical.
CN201210344509.9A 2012-09-17 2012-09-17 Navigational star screening method for star sensors Expired - Fee Related CN102840861B (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201210344509.9A CN102840861B (en) 2012-09-17 2012-09-17 Navigational star screening method for star sensors
CN201510107562.0A CN106441280B (en) 2012-09-17 2012-09-17 The method of screening nautical star for star sensor

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201210344509.9A CN102840861B (en) 2012-09-17 2012-09-17 Navigational star screening method for star sensors

Related Child Applications (1)

Application Number Title Priority Date Filing Date
CN201510107562.0A Division CN106441280B (en) 2012-09-17 2012-09-17 The method of screening nautical star for star sensor

Publications (2)

Publication Number Publication Date
CN102840861A true CN102840861A (en) 2012-12-26
CN102840861B CN102840861B (en) 2015-04-29

Family

ID=47368440

Family Applications (2)

Application Number Title Priority Date Filing Date
CN201510107562.0A Expired - Fee Related CN106441280B (en) 2012-09-17 2012-09-17 The method of screening nautical star for star sensor
CN201210344509.9A Expired - Fee Related CN102840861B (en) 2012-09-17 2012-09-17 Navigational star screening method for star sensors

Family Applications Before (1)

Application Number Title Priority Date Filing Date
CN201510107562.0A Expired - Fee Related CN106441280B (en) 2012-09-17 2012-09-17 The method of screening nautical star for star sensor

Country Status (1)

Country Link
CN (2) CN106441280B (en)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103438905A (en) * 2013-08-30 2013-12-11 中国人民解放军第二炮兵工程大学 Method for estimating completeness of star catalogue of star sensor
CN105424034A (en) * 2015-10-30 2016-03-23 北京控制工程研究所 Shipborne all-time starlight and inertia combined navigation system
CN106123891A (en) * 2016-06-22 2016-11-16 中国人民解放军63680部队 Faint star autonomous classification method based on girth feature
CN106595645A (en) * 2016-11-02 2017-04-26 上海航天控制技术研究所 Method for making guide star database based on output accuracy of star sensors
CN104154929B (en) * 2014-07-08 2017-05-24 南京航空航天大学 Optimal selection method of navigational stars of star map simulator based on star density
CN107816986A (en) * 2017-09-05 2018-03-20 湖北航天技术研究院总体设计所 A kind of star sensor whole day ball fast star identification method
CN108106612A (en) * 2017-12-13 2018-06-01 常州工学院 Star sensor nautical star selection method
CN109813301A (en) * 2019-01-29 2019-05-28 中国人民解放军国防科技大学 Method for quickly determining optimal navigation star direction

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107389057B (en) * 2017-06-26 2019-08-09 北京控制工程研究所 A kind of daytime environment navigation star recognition methods
CN108507569B (en) * 2017-11-10 2020-04-17 中国人民解放军国防科技大学 Missile-borne fixed star library rapid generation method for starlight/inertia composite guidance
CN112034340B (en) * 2019-06-03 2023-05-09 中国人民解放军63756部队 Method for screening fault characteristics of measurement and control antenna motor
CN111578934B (en) * 2020-04-30 2022-07-29 中国人民解放军国防科技大学 Refraction star optimization method and system based on inertia/astronomical combined navigation application
CN113284160B (en) * 2021-04-23 2024-03-12 北京天智航医疗科技股份有限公司 Method, device and equipment for identifying surgical navigation mark beads
CN113670299B (en) * 2021-08-16 2022-10-25 北京航空航天大学 Serial and parallel combined view field gating imaging method for star sensor
CN115951589B (en) * 2023-03-15 2023-06-06 中科院南京天文仪器有限公司 Star uniform selection method based on maximized Kozachenko-Leonenko entropy

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH01237411A (en) * 1988-03-17 1989-09-21 Toshiba Corp Fixed star sensor
CN101236084A (en) * 2008-01-31 2008-08-06 北京控制工程研究所 Method for eliminating close-packed stellar field from star catalogue
CN101246590A (en) * 2008-03-03 2008-08-20 北京航空航天大学 Star loaded camera spacing deformation geometric image correction method

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
RU2033949C1 (en) * 1993-02-09 1995-04-30 Севастиян Дмитриевич Гнатюк Self-contained on-board control system for space vehicle
CN100357705C (en) * 2004-12-28 2007-12-26 北京航空航天大学 Method for recognising star map based on LPT transforms
CN101441082B (en) * 2008-12-22 2010-08-11 中国科学院长春光学精密机械与物理研究所 Rapid triangle star map recognition method
CN102313549B (en) * 2011-07-28 2013-07-24 北京航空航天大学 Identification method for triangular star atlas based on characteristic of inertia ratio

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH01237411A (en) * 1988-03-17 1989-09-21 Toshiba Corp Fixed star sensor
CN101236084A (en) * 2008-01-31 2008-08-06 北京控制工程研究所 Method for eliminating close-packed stellar field from star catalogue
CN101246590A (en) * 2008-03-03 2008-08-20 北京航空航天大学 Star loaded camera spacing deformation geometric image correction method

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
LINTAO等: "《A general method of the automatical selection of guide star》", 《PROCEEDINGS OF 1998 FOURTH INTERNATIONAL CONFERENCE SIGNAL PROCESSING 》 *
李光蕊等: "《适用于星敏感器的导航星表构造算法研究 》", 《光学技术 》 *
田宏等: "《星敏感器导航星表建立》", 《空间控制技术与应用》 *
郑胜等: "《一种新的导航星选取算法研究》", 《宇航学报》 *

Cited By (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103438905A (en) * 2013-08-30 2013-12-11 中国人民解放军第二炮兵工程大学 Method for estimating completeness of star catalogue of star sensor
CN103438905B (en) * 2013-08-30 2016-01-20 中国人民解放军第二炮兵工程大学 A kind of star sensor star catalogue complete evaluation method
CN104154929B (en) * 2014-07-08 2017-05-24 南京航空航天大学 Optimal selection method of navigational stars of star map simulator based on star density
CN105424034A (en) * 2015-10-30 2016-03-23 北京控制工程研究所 Shipborne all-time starlight and inertia combined navigation system
CN105424034B (en) * 2015-10-30 2018-06-19 北京控制工程研究所 A kind of boat-carrying round-the-clock starlight INS Integrated Navigation System
CN106123891A (en) * 2016-06-22 2016-11-16 中国人民解放军63680部队 Faint star autonomous classification method based on girth feature
CN106123891B (en) * 2016-06-22 2018-11-16 中国人民解放军63680部队 Faint star autonomous classification method based on perimeter feature
CN106595645A (en) * 2016-11-02 2017-04-26 上海航天控制技术研究所 Method for making guide star database based on output accuracy of star sensors
CN107816986A (en) * 2017-09-05 2018-03-20 湖北航天技术研究院总体设计所 A kind of star sensor whole day ball fast star identification method
CN107816986B (en) * 2017-09-05 2020-05-19 湖北航天技术研究院总体设计所 Rapid star map identification method for all celestial sphere of star sensor
CN108106612A (en) * 2017-12-13 2018-06-01 常州工学院 Star sensor nautical star selection method
CN109813301A (en) * 2019-01-29 2019-05-28 中国人民解放军国防科技大学 Method for quickly determining optimal navigation star direction

Also Published As

Publication number Publication date
CN106441280B (en) 2019-05-14
CN102840861B (en) 2015-04-29
CN106441280A (en) 2017-02-22

Similar Documents

Publication Publication Date Title
CN102840861A (en) Navigational star screening method for star sensors
CN102865865B (en) Star sensor navigational star screening method by orthogonal list
Zhang Star identification
CN107133325B (en) Internet photo geographic space positioning method based on street view map
CN100541232C (en) The thick bearing calibration of aviation multiple spectrum scanner geometric under the no attitude information condition
CN106680798B (en) A kind of identification of airborne LIDAR air strips overlay region redundancy and removing method
CN104154929B (en) Optimal selection method of navigational stars of star map simulator based on star density
Tampubolon et al. UAV data processing for large scale topographical mapping
CN104751451B (en) Point off density cloud extracting method based on unmanned plane low latitude high resolution image
CN103020934A (en) Seamless automatic image splicing method resistant to subtitle interference
Zhang et al. UAV‐derived imagery for vegetation structure estimation in rangelands: validation and application
Stepinski et al. Digital topography models for Martian surfaces
CN103925922A (en) Method for identifying fixed stars in ICCD star atlas under high-dynamic condition
Carr et al. ISEA discrete global grids
CN104298667A (en) Navigation database building method applicable to high level background star pattern recognition
Lumban-Gaol et al. Investigations on the bundle adjustment results from sfm-based software for mapping purposes
CN103884336A (en) Establishment method of daytime star sensor-based infrared star detection navigation star catalogue
Ferreira et al. Mapping Local Climate Zones for São Paulo Metropolitan Region: a comparison between the local climate zone map and two other local maps
Zink et al. Photogrammetric point clouds for GIS-based high-resolution estimation of solar radiation for roof-top solar systems
CN105676191A (en) Marine microwave remote sensing technology-based scatterometer data converting method and device
Niederheiser et al. Mapping alpine vegetation location properties by dense matching
Zhou et al. A pyramid image matching algorithm for chang'e-2 CCD images of the moon
Gotovac et al. A model for automatic geomapping of aerial images mosaic acquired by UAV
Peng et al. Topographic mapping capability analysis of moderate resolution imaging camera (MoRIC) imagery of Tianwen-1 Mars mission
Kozhevnikov Clouds as Manifestation of Wave Disturbance above Mountain Ridges

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20150429

Termination date: 20170917

CF01 Termination of patent right due to non-payment of annual fee