CN106023148A - Star image point position extraction method under sequence focusing observation mode - Google Patents

Abstract

The invention discloses a star image point position extraction method under a sequence focusing observation mode, which combines star movement laws with digital image processing technologies for solving the problems of non-obvious dispersed star spots of star image points under a focusing observation mode and not high positioning precision of sub-pixel stars. First of all, by use of a Gauss filtering method, star observation images are preprocessed, and high-frequency noise in the images are removed; through brightness distribution rules in the images, an adaptive threshold is calculated, for extracting star image point areas in the observation images; and based on this, through such digital image processing technologies as a star energy integration method, an energy curve fitting method, a star locus estimation method and the like, high-precision extraction of coordinates of the star image points is realized. According to the invention, a change trench of star image point energy is counted and analyzed, a star image point movement locus can be estimated to the maximum degree according to movement laws of stars, and a high-precision star mass center extraction capability is guaranteed under the focusing mode.

Description

A kind of sequence focuses on star image point position extracting method under observation mode

Technical field

The invention belongs to technical field of image processing, particularly relate to a kind of based on sequence image under focusing on observation mode Star image point position extracting method.

Background technology

Star image point position extraction accuracy directly determines the limit of star sensor attitude measure precision, therefore high-accuracy and constant Star center coordination technology occupies critical positions in star sensor technology development course.Along with sending out of planar array detector Exhibition, grid image location technology has become the basic fundamental of star place Detection And Tracking, and sub-pixed mapping interpolation location is the most general Time fixed star method for positioning mass center, by fixed star picture point image is carried out disperse on the detector so that picture point is diffused into detection Multiple pixels on device, utilize the intensity profile of different pixel, star image are carried out interpolation segmentation, can make fixed star center coordination precision Break through the yardstick of planar array detector pixel.Thus at present for improving the certainty of measurement of star sensor, segment frequently with interpolation Algorithm makes the center coordination precision of punctate opacity of the cornea reach sub-pixed mapping or higher.

But interpolation close classification requires that fixed star picture point must disperse guarantee fixed star matter on multiple pixels of detector The precision of heart location, under focusing on observation mode, the fixed star picture point spread the most on the detector is less, and main energetic often collects In in a pixel on planar array detector, this will have a strong impact on the positioning precision of interpolation close classification, if considering further that The effect of photoelectricity sampling noise, causes interpolation close classification to lose efficacy most probably；Interpolation algorithm of subdivision is observed mainly for single frames simultaneously Fixed star picture point in image positions, and under sequence observation mode, does not make full use of the star motions of adjacent interframe Information.

Summary of the invention

(1) goal of the invention: in view of this, present invention contemplates that a kind of sequence of offer focuses on star image point position under observation mode Extracting method, at least can solve existing focusing observation and cause degradation technical problem under fixed star barycenter extraction accuracy.

(2) technical scheme: the invention provides a kind of sequence and focus on star image point position extracting method under observation mode, should For the star observation image of geostationary orbit shooting, described method includes:

Step one: by the denoising of described star observation image is obtained denoising image；

Step 2: by the detection of the punctate opacity of the cornea of described denoising image being obtained the region of fixed star picture point；

Step 3: according to the region of described fixed star picture point respectively according to observation time to described denoising image row, column Carry out energy integral and obtain fixed star picture point energy integral curve；

Step 4: obtain peak of curve point by described fixed star picture point energy integral curve；

Step 5: obtain fixed star picture point movement locus according to described peak of curve point；

Step 6: obtain star image point position according to described fixed star picture point movement locus.

Wherein, " by the denoising of described star observation image is obtained denoising image " bag described in step one Include:

Construct the Gaussian filter with low-pass filter effect；

Utilize the Gaussian smoothing template constructed that described star observation image is carried out gaussian filtering, going after being smoothed Make an uproar image；

" by the detection of the punctate opacity of the cornea of described denoising image being obtained the region of fixed star picture point " bag described in step 2 Include:

Add up the Luminance Distribution in described denoising image；

Luminance mean value and the standard deviation of entire image is calculated from the Luminance Distribution of described denoising image；

The luminance mean value corresponding by described denoising image and standard deviation calculate the adaptive threshold of described denoising image；

Described denoising image is split as segmentation foundation by described adaptive threshold, retains brightness value higher than described The region of threshold value.

" according to observation time, described fixed star is seen respectively according to the region of described fixed star picture point described in step 3 Altimetric image row, column carries out energy integral and obtains fixed star picture point energy integral curve " including:

The described star observation image each row energy in region to described fixed star picture point successively according to observation time Integration obtains line direction energy integral curve；

The described star observation image each row energy in region to described fixed star picture point successively according to observation time Integration obtains column direction energy integral curve.

" obtaining peak of curve point by described fixed star picture point energy integral curve " described in step 4 including:

Fixed star picture point energy integral curve described in each row, column is carried out nonlinear least square fitting obtain in punctate opacity of the cornea region The parametric equation of each row, column integral curve represents；

Each row, column peak of curve point is obtained by the parametric equation of described each row, column integral curve.

" obtaining fixed star picture point movement locus according to described peak of curve point " described in step 5 including:

The fixed star moment through each row, column center is obtained according to peak of curve point described in each row, column；

Fixed star is combined with the corresponding moment through the coordinate of each row, column center, constructs respectively and see about fixed star Survey the row, column position sample points collection in moment；

Fixed star picture point is obtained by the row, column position sample points collection in star observation moment is carried out linear least square fitting Movement locus.

The calculating process bag of " the obtaining star image point position according to described fixed star picture point movement locus " described in step 6 Include:

$\left\{\begin{array}{c}x\left(t\right)=x_0+v_x\×(t-t_x)\\ y\left(t\right)=y_0+v_y\×(t-t_y)\end{array}\right.$

Wherein, x (t), y (t) represent fixed star picture point coordinate on image planes column and row direction respectively；t_x、t_yRepresent that fixed star is seen Survey the moment；x₀、y₀Represent that fixed star is at t respectively_xWith t_yThe image planes column and row coordinate that moment is corresponding；v_x、v_yFor fixed star picture point in image planes Movement velocity on column and row direction.

By above step, present invention achieves a kind of sequence and focus on star image point position extracting method under observation mode.

(3) advantage: star image point position extracting method provided by the present invention, to fixed star picture point energy in sequence observed image Amount variation tendency is added up, and can analyze star motions track according to energy curves；Utilize method of least square that fixed star is transported Dynamic track is fitted, and the fixed star center-of-mass coordinate obtained by the present invention carries relative to the fixed star barycenter under traditional focus observation mode Take result, there is higher barycenter extraction accuracy.Method proposed by the invention is applicable to be positioned on geostationary orbit continuously The star observation image of shooting.

Accompanying drawing explanation

Fig. 1 focuses on the flow process of the star image point position extracting method under observation mode for the sequence that the embodiment of the present invention provides Figure；

Fig. 2 is the principle schematic of energy integral process in the embodiment of the present invention；

Fig. 3 is the star observation image that simulation geostationary orbit focuses on shooting.

In order to enable clearly to realize the structure of embodiments of the invention, it is labelled with certain size, structure and device in the drawings, But it is only for signal needs, is not intended to limit the invention in this specific dimensions, structure, device and environment, according to specifically Needing, these devices and environment can be adjusted or revise by those of ordinary skill in the art, the adjustment that carried out or Person's amendment is still included in the scope of appended claims.

Detailed description of the invention

In the following description, the multiple different aspect of the present invention will be described, but, for the common skill in this area For art personnel, the present invention can be implemented just with the some or all structures of the present invention or flow process.In order to explain Definition for, elaborate specific number, configuration and order, however, it will be apparent that there is no the situation of these specific detail Under can also implement the present invention.In other cases, in order to not obscure the present invention, will no longer for some well-known features It is described in detail.

Embodiment 1

In order to solve to focus on the technical problems such as fixed star picture point positioning precision is low under observation mode, embodiments provide A kind of sequence focuses on star image point position extracting method under observation mode, is applied to the star observation figure of geostationary orbit shooting Picture, as it is shown in figure 1, said method comprising the steps of:

Step S101: by the denoising of described star observation image is obtained denoising image；

Star observation image described in the present embodiment refers to the star chart picture gathered by star sensor, and this image is also referred to as star Figure.Space background noise, the serious interference of electronic imaging noise on image in star chart, can make under focusing on observation mode simultaneously Punctate opacity of the cornea does not has obvious disperse star speckle on CCD imaging array, brings for follow-up high-precision star image point position extraction work Great difficulty.

The present embodiment utilize matrix form represent star observation image, in matrix in the numerical value representative image of each element The brightness value of respective pixel position, i.e. gray value；Then by constructing the Gaussian filter with low-pass filter effect, right Described image array carries out gaussian filtering, the denoising image after being smoothed.

Step S102: by the detection of the punctate opacity of the cornea of described denoising image being obtained the region of fixed star picture point；

The frequency distribution of statistical picture gray value on given denoising image, and calculate entire image based on intensity profile The meansigma methods of gray scale and standard deviation, and then adaptively calculate present frame according to the gray-scale watermark of every width observed image The segmentation threshold of image；Recycling adaptive threshold according to splitting denoising image as segmentation, retains brightness value and is higher than The region of described threshold value is as fixed star region.

Step S103: according to the region of described fixed star picture point respectively according to observation time to described denoising image line, Row carry out energy integral and obtain fixed star picture point energy integral curve, as shown in Figure 2；

The generally image planes size of observed image is relatively big, not only can drop if carrying out energy integral for view picture star observation image The processing speed of low fixed star barycenter extraction algorithm, but also may be carried by the interference effect fixed star barycenter of system Banded improvement The precision taken；Therefore, detect, according to punctate opacity of the cornea, the fixed star picture point region that obtains, according to observation time on denoising image only for picture The gradation of image in some region carries out row, column integration, obtains each row, column in punctate opacity of the cornea region and observes moment punctate opacity of the cornea energy about difference Integral curve.

Step S104: obtain peak of curve point by described fixed star picture point energy integral curve；

Fixed star picture point energy integral curve owing to directly obtaining row, column integration is vulnerable to the impact of imaging noise, bent Line fluctuates and changes greatly, it is difficult to estimate the peak point of integral curve the most exactly；It is thus desirable to it is bent to energy integral Line carries out nonlinear least square fitting, is obtained the mathematic(al) representation of curve by the minimum error after Optimal Fitting, and then Auditory localization cues peak point.

Step S105: obtain fixed star picture point movement locus according to described peak of curve point；

The precise moments striding across corresponding row, column center is can determine that, thus by the peak point of each row, column integral curve Construct each row, column center point set coordinate about the observation moment；By respectively to row, column position about the observation moment Movement locus is fitted, and can obtain star image equation of locus on CCD imaging array.

Step S106: obtain star image point position according to described fixed star picture point movement locus.

By the given observation moment, calculate star image point position according to the star image equation of locus obtained, thus can obtain Whole sequence focuses on the fixed star picpointed coordinate under observation mode.

Star motions information in the present embodiment method binding sequence image, when can stride across each pixel according to fixed star picture point, as The variation tendency of energy in unit, estimates star motions track, promotes the precision that star image point position extracts, and eliminates focusing observation The problem such as under pattern precision is the highest.

Concrete, step S101 includes:

Step S1011: structure low pass Gaussian filter；

The standard deviation of given two-dimensional Gaussian function, structure template size is the Gaussian filter of 3 × 3, template center Position coordinates is (0,0), and the rest may be inferred, and each position coordinate in template substitutes into the power calculating respective element in two-dimensional Gaussian function Value:

$w_{i,j}=e^{-\frac{i^2+j^2}{2{\mathrm{\σ}}^2}}$

Wherein, w_i,jRepresenting the Gauss weights being specified element in Gaussian template by i, j, σ represents the standard deviation of Gaussian function. I, j represent the row, column index of element relative image center in image array.

After complete Gaussian template weights to be calculated, more whole Gaussian template is normalized:

$w_{i,j}=\frac{w_{i,j}}{\underset{i}{\Σ}\underset{j}{\Σ}{w}_{i,j}}$

Wherein, w_i,jRepresent that former weights are w_i,jThe new Gaussian template weights that obtain after normalized of element.

Step S1012: utilize the low-pass Gaussian filter of structure that described image array is carried out gaussian filtering, obtain denoising After image.

The process that original star observation image carries out gaussian filtering is:

$\begin{array}{c}{I}^{\′}(i,j)=I(i,j)*W\\ =\underset{k=-1}{\overset{1}{\Σ}}\underset{l=-1}{\overset{1}{\Σ}}I(i-k,j-l)\×w_{k,l}\end{array}$

Wherein, (i j) and I'(i, j) represents the element of the i-th row jth row, W table in the image array before and after filtering to I respectively Show Gaussian filter；* convolution algorithm is represented.K, l are integer.

In step S102, the punctate opacity of the cornea detection process to described denoising image is:

Step S1021: the frequency that each gray value of statistical picture occurs on given denoising image, obtains star observation The intensity profile of image；

Step S1022: meansigma methods based on intensity profile calculating entire image gray scale and standard deviation:

$\stackrel{\‾}{I}=\frac{1}{m\×n}\underset{i=1}{\overset{m}{\Σ}}\underset{j=1}{\overset{n}{\Σ}}I(i,j)$

$I_{std}=\sqrt{\frac{1}{m\×n}\underset{i=1}{\overset{m}{\Σ}}\underset{j=1}{\overset{n}{\Σ}}{\left(I\right(i,j)-\stackrel{\‾}{I})}^2}$

Wherein,Represent gradation of image average, I_stdRepresenting gradation of image standard deviation, (i j) represents in image array i-th to I The numerical value of row jth column matrix element；The row, column number of m Yu n representative image matrix respectively.

Step S1023: according to gradation of image average and standard deviation calculating star observation image adaptive threshold value:

$T=\stackrel{\‾}{I}+\mathrm{\α}\×I_{std}$

Wherein T represents that adaptive threshold, α represent predetermined threshold value amplitude and adjust parameter, and α value is 3 in the present embodiment, From there through Gauss distribution 3 σ criterion, the false alarm rate of detection is controlled.

Step S1024: utilize adaptive threshold as segmentation according to splitting denoising image, retains brightness value and is higher than The region of described threshold value is as fixed star region:

$B(i,j)=\left\{\begin{array}{cc}{I}^{\&prime;}(i,j),& I^{\&prime;}(i,j)\&GreaterEqual;T\\ 0,& I^{\&prime;}(i,j)<T\end{array}\right.$

Wherein, I'(i, j) (i, j) represents the element of the i-th row jth row in the denoising image that segmentation is forward and backward respectively, and T is with B Adaptive threshold.

Described step S103 specifically includes:

Step S1031: be integrated for image each row pixel grey scale in fixed star region according to the observation moment, add up each row The curve that pixel grey scale integration changed with the moment:

$\begin{array}{cc}I(t;i)={\&Sigma;}_{j\&Element;S_{col}}I(i,j,t),& i\&Element;S_{row},t\&Element;P_{time}\end{array}$

Wherein, S_row、S_colRepresent fixed star track respectively and streak the set of image row, column；T represents observation moment, P_timeIt is The set being made up of all observation moment；(i, j t) represent the grey scale pixel value of t observed image the i-th row jth row to I；I(t； I) represent and the i-th row is carried out the integral curve with the change of observation moment that energy integral obtains.

Step S1032: be integrated for image each row pixel grey scale in fixed star region according to the observation moment, add up each row The curve that pixel grey scale integration changed with the moment:

$\begin{array}{cc}I(t;j)={\&Sigma;}_{i\&Element;S_{row}}I(i,j,t),& j\&Element;S_{col},t\&Element;P_{time}\end{array}$

Wherein, i, j represent row, column index, S in image respectively_row、S_colRepresent fixed star track respectively and streak image row, column Set；T represents observation moment, P_timeThe set being made up of all observation moment；I(t；J) represent jth is arranged and carry out energy The integral curve with the change of observation moment that integration obtains.

Described step S104 specifically includes:

Step S1041: utilize nonlinear least square method to be fitted energy integral curve, by minimizing error letter NumberObtain the integral curve equation after matching:

$\mathrm{\&theta;}=\underset{\mathrm{\&theta;}}{\arg \min }{\&Sigma;}_{t\&Element;P_{time}}{\&lsqb;I\left(t\right)-f(t;\mathrm{\&theta;})\&rsqb;}^2$

$f(t;\mathrm{\&theta;})=\mathrm{\&lambda;}\&times;e^{-\frac{{(t-\mathrm{\&beta;})}^2}{{\mathrm{\&eta;}}^2}},\mathrm{\&theta;}=\{\mathrm{\&lambda;},\mathrm{\&beta;},\mathrm{\&eta;}\}$

Wherein, θ represents curvilinear equation parameter to be estimated, and θ represents the curvilinear equation parameter after estimation, specifically comprises λ, β, η, β represent the position of integral curve peak point, the moment of energy peak i.e. occur；λ represents peak value；η represents the expansion of energy integral Divergence；I (t) represents the energy product score value that t integral curve is corresponding, and during concrete calculating, I (t) can be replaced I (t；I) or I(t；J) the i-th row or the energy integral curve of jth row are represented.

Step S1042: according to the peak of curve point position of the energy integral curve acquisition correspondence row, column after matching.

Described step S105 specifically includes:

Step S1051: fixed star picture point strides across in corresponding row, column to utilize the peak point position of each row, column integral curve to determine The precise moments during heart, each row, column center and the coordinate points observing the moment in thus constructing fixed star picture point moving region Collection

$S_{row}^t=\left\{({\mathrm{\&beta;}}_i,i-0.5)\right|i\&Element;S_{row}\}$

$S_{col}^t=\left\{({\mathrm{\&beta;}}_j,j-0.5)\right|j\&Element;S_{col}\}$

Wherein, β_iRepresent the peak point position obtained according to the i-th row energy integral curve matching；β_jRepresent according to jth row energy The peak point position that amount integral curve matching obtains.

Step S1052: by the row, column position got about the coordinate point set observing the momentCome respectively to perseverance Star movement locus is fitted, acquisition star image Movement Locus Equation on CCD imaging array:

$\left\{\begin{array}{c}{\mathrm{\&theta;}}_x=\underset{{\mathrm{\&theta;}}_x}{\mathrm{argmin}}{\&Sigma;}_{t\&Element;P_{time}}{\&lsqb;j\left(t\right)-x(t;{\mathrm{\&theta;}}_x)\&rsqb;}^2\\ \begin{array}{cc}x(t;{\mathrm{\&theta;}}_x)=x_0+v_x\&times;(t-t_x),& {\mathrm{\&theta;}}_x=\{x_0,t_x,v_x\}\end{array}\end{array}\right.$

$\left\{\begin{array}{c}{\mathrm{\&theta;}}_y={\mathrm{argmin}}_{{\mathrm{\&theta;}}_y}{\&Sigma;}_{t\&Element;P_{time}}{\&lsqb;i\left(t\right)-y(t;{\mathrm{\&theta;}}_y)\&rsqb;}^2\\ \begin{array}{cc}y(t;{\mathrm{\&theta;}}_y)=y_0+v_y\&times;(t-t_y),& {\mathrm{\&theta;}}_y=\{y_0,t_y,v_y\}\end{array}\end{array}\right.$

Wherein, θ_xWith θ_yIt is the fixed star column and row position the treating matching equation of locus parameter about the moment respectively, θ_x、θ_yRespectively It it is the equation of locus parameter about the moment of the fixed star column and row position after matching；x(t；θ_x) represent fixed star column direction equation of locus, y (t；θ_y) it is fixed star line direction equation of locus；I (t), j (t) are the row, column coordinates of t fixed star picture point.t_x、t_yRepresent that fixed star is seen Survey the moment；x₀、y₀Represent that fixed star is at t respectively_xWith t_yThe image planes column and row coordinate that moment is corresponding；v_x、v_yFor fixed star picture point in image planes Movement velocity on column and row direction.

The characteristics of motion of fixed star picture point is showed in the way of energy integral variation tendency by the present embodiment method, and with Time avoid directly calculating on star chart punctate opacity of the cornea center-of-mass coordinate in single-frame images, make full use of star motions letter abundant in sequence frame Breath makes up the too small problem causing fixed star barycenter extraction accuracy to reduce of the star speckle spread, the Movement Locus Equation after matching come Calculate fixed star picpointed coordinate, promote the precision of star image point position extracting method.

Embodiment 2

The present invention is described in detail according to an actual scene for the present embodiment.

The present embodiment method comprises the following steps:

(1) gray matrix of star observation image is obtained

The high bit image got by star sensor is input in computer, by the form of imagery exploitation gray matrix Represent；

Wherein, I_m×nMatrix representation forms for star observation image；I_mnGradation of image for the n-th row of m row in image Value.

(2) star observation image is carried out denoising

This step includes:

A, constructs low pass Gaussian filter；

B, utilizes Gaussian filter that image array is carried out convolution algorithm, reaches the effect of image denoising.

(3) fixed star picture point region is obtained in the detection of denoising image enterprising planet point

The frequency that on statistics denoising image, each gray value occurs obtains the intensity profile of star observation image, further according to ash Degree distribution is adaptively calculated gradation of image threshold value, utilizes gray threshold that denoising image carries out binary segmentation, gained binary map As the pixel value position more than zero represents fixed star picture point region.

(4) successively the row, column being positioned at fixed star region on all observed images is carried out energy integral, obtain fixed star picture point The energy integral curve of each row, column.

This step includes:

A, reads image after denoising, is integrated obtaining perseverance to each row of each two field picture from first row in fixed star region The curve time of integration of constellation territory Nei Gelie；The time of integration, the abscissa of curve was the moment, and vertical coordinate is that this is listed in the corresponding moment Pixel value cumulative integral and.

B, reads image after denoising, is integrated obtaining perseverance to each row of each two field picture from the first row in fixed star region The curve time of integration of each row in constellation territory；The time of integration, the abscissa of curve was the moment, and vertical coordinate is that this row is in the corresponding moment Pixel value cumulative integral and.

(5) the energy integral curve to row, column each in fixed star region is fitted obtaining peak of curve point position

Utilize Gaussian function to carry out the optical point spread function of approximate representation star sensor, use Gaussian curve that integration is changed Trend is modeled.In modeling process, by the model parameter of Nonlinear least squares optimization Algorithm for Solving matched curve, and Peak of curve point position is calculated from model parameter.

(6) according to the peak point of energy integral curve in row, column each in fixed star region, fixed star picture point movement locus is entered Row is estimated

Assuming that the movement locus that fixed star picture point is on CCD imaging array is linear uniform motion, and in (5) in punctate opacity of the cornea region Star image coordinates on the position of each row, column energy integral peak point is corresponding different observation moment star motions tracks.With peak value The star image coordinate that point is corresponding is sample point, utilize method of least square respectively to punctate opacity of the cornea level, upright position about the observation moment The parametric equation of the straight line parameter solves, and thus estimates punctate opacity of the cornea track.

(7) by the high-precision star image point position of punctate opacity of the cornea trajectory calculation

Utilize the parametric equation of the straight line estimated in (6) that the position coordinates of each imaging moment fixed star picture point can be calculated

$\left\{\begin{array}{c}x\left(t\right)=x_0+v_x\&times;(t-t_x)\\ y\left(t\right)=y_0+v_y\&times;(t-t_y)\end{array}\right.$

Wherein, x (t), y (t) represent fixed star picture point coordinate on image planes column and row direction respectively.

Experimental technique: in the case of star sensor is 1～5ms the time of integration, when whole CCD image planes are streaked in star motions Time, respectively the observed image data obtained by the stars such as observation 1,2,3,4,5,6 are tested.Experimental result is as shown in the table.

By the present embodiment being carried out emulation experiment test assessment.The present embodiment method under observation mode that focuses on extracts fixed star The precision of image point position remains to be maintained at sub-pixel, and contrasts with fixed star barycenter true value coordinate, and error is relatively low, shows this enforcement Example utilizes sequence frame observed image data to have higher extraction accuracy under focusing on observation mode.

The present invention is directed to focus under observation mode that fixed star picture point is without obvious disperse star speckle, sub-pix fixed star positioning precision is the highest Problem, star motions rule is combined with digital image processing techniques, utilizes gaussian filtering method that star observation image is entered Row pretreatment, removes the high-frequency noise in image；And calculated adaptive threshold by the Luminance Distribution rule in image, in order to extract Fixed star picture point region in observed image；On this basis by the energy in fixed star picture point region is integrated, analyze star image The trend of energy variation, matching energy curves, estimate that fixed star picture point movement locus on CCD imaging array realizes height The fixed star picpointed coordinate of precision extracts.The present invention is directed to be positioned on geostationary orbit and focus on the star observation image of shooting and carry out Star image positions, and can estimate fixed star picture point movement locus to greatest extent according to the characteristics of motion of fixed star, and protect under focusing mode Demonstrate,proving high-precision fixed star barycenter extractability, simulation geostationary orbit focuses on the star observation image of shooting as shown in Figure 3.

In several embodiments provided herein, it should be understood that disclosed equipment and method, can be passed through it Its mode realizes.Apparatus embodiments described above is only schematically, such as, the division of described unit, it is only A kind of logic function divides, and actual can have other dividing mode, such as when realizing: multiple unit or assembly can be in conjunction with, or It is desirably integrated into another system, or some features can be ignored, or do not perform.It addition, shown or discussed each composition portion Dividing coupling each other or direct-coupling or communication connection can be the INDIRECT COUPLING by some interfaces, equipment or unit Or communication connection, can be electrical, machinery or other form.

The above-mentioned unit illustrated as separating component can be or may not be physically separate, shows as unit The parts shown can be or may not be physical location, i.e. may be located at a place, it is also possible to be distributed to multiple network list In unit；Part or all of unit therein can be selected according to the actual needs to realize the purpose of the present embodiment scheme.

It addition, each functional unit in various embodiments of the present invention can be fully integrated in a processing module, it is possible to Being that each unit is individually as a unit, it is also possible to two or more unit are integrated in a unit；Above-mentioned Integrated unit both can realize to use the form of hardware, it would however also be possible to employ hardware adds the form of SFU software functional unit and realizes.

One of ordinary skill in the art will appreciate that: all or part of step realizing said method embodiment can be passed through The hardware that programmed instruction is relevant completes, and aforesaid program can be stored in a computer read/write memory medium, this program Upon execution, perform to include the step of said method embodiment；And aforesaid storage medium includes: movable storage device, read-only Memorizer (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disc or The various media that can store program code such as person's CD.

The above, the only detailed description of the invention of the present invention, but protection scope of the present invention is not limited thereto, and any Those familiar with the art, in the technical scope that the invention discloses, can readily occur in change or replace, should contain Cover within protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with described scope of the claims.

Claims (7)

1. a star image point position extracting method under sequence focuses on observation mode, is applied to the fixed star of geostationary orbit shooting Observed image, it is characterised in that described method includes:

By the denoising of described star observation image is obtained denoising image；

By the detection of the punctate opacity of the cornea of described denoising image being obtained the region of fixed star picture point；

Region according to described fixed star picture point carries out energy integral according to observation time to described denoising image row, column respectively Obtain fixed star picture point energy integral curve；

Peak of curve point is obtained by described fixed star picture point energy integral curve；

Fixed star picture point movement locus is obtained according to described peak of curve point；

Star image point position is obtained according to described fixed star picture point movement locus.

Method the most according to claim 1, it is characterised in that described by the denoising to described star observation image Obtain denoising image to include:

Construct the Gaussian smoothing template with low-pass filter effect；

Utilize the Gaussian smoothing template constructed that described star observation image is carried out gaussian filtering, the denoising figure after being smoothed Picture.

Method the most according to claim 1, it is characterised in that the process of described punctate opacity of the cornea detection is:

Add up the Luminance Distribution in described denoising image；

Luminance mean value and the standard deviation of entire image is calculated from the Luminance Distribution of described denoising image；

The luminance mean value corresponding by described denoising image and standard deviation calculate the adaptive threshold of described denoising image；

Described denoising image is split as segmentation foundation by described adaptive threshold, retains brightness value higher than described adaptive Answer the region of threshold value.

Method the most according to claim 1, it is characterised in that the described region according to fixed star picture point is to described denoising Image row, column carries out energy integral and obtains fixed star picture point energy integral curve and include:

According to observation time, the described denoising image each row energy integral in region to described fixed star picture point obtains successively Line direction energy integral curve；

According to observation time, the described denoising image each row energy integral in region to described fixed star picture point obtains successively Column direction energy integral curve.

Method the most according to claim 1, it is characterised in that described obtained by described fixed star picture point energy integral curve Peak of curve point includes:

Fixed star picture point energy integral curve described in each row, column is carried out nonlinear least square fitting and obtains in punctate opacity of the cornea region each The parametric equation of row, column integral curve represents；

Each row, column peak of curve point is obtained by the parametric equation of described each row, column integral curve.

Method the most according to claim 1, it is characterised in that obtained fixed star picture point movement locus by described peak of curve point Calculating process include:

The fixed star moment through each row, column center is obtained according to peak of curve point described in each row, column；

Fixed star is combined with the corresponding moment through the coordinate of each row, column center, when constructing respectively about star observation The row, column position sample points collection carved；

The motion of fixed star picture point is obtained by the row, column position sample points collection in star observation moment being carried out linear least square fitting Track.

Method the most according to any one of claim 1 to 6, it is characterised in that described according to the motion of described fixed star picture point Track obtains the calculating process of star image point position:

Wherein, x (t), y (t) represent fixed star picture point coordinate on image planes column and row direction respectively；t_x、t_yWhen representing star observation Carve；x₀、y₀Represent that fixed star is at t respectively_xWith t_yThe image planes column and row coordinate that moment is corresponding；v_x、v_yFor fixed star picture point in image planes column and row Movement velocity on direction.