CN101615293B - Device and method for VCT system parameter calibration - Google Patents

Abstract

The invention discloses a device and a method for VCT system parameter calibration. The device comprises a bracket, a fixed tube, a first sphere and a second sphere, wherein the fixed tube is vertically arranged on the bracket; and the first sphere; and the second sphere are fixed in the fixed tube and mutually separated. The method comprises the following steps: fixing the VCT system parameter calibration on an object stage, revolving the object stage for one circle, and acquiring a plurality of frames of images of the device formed on a flat panel detector during revolving; acquiring projection coordinates of sphere centers of the two spheres in each frame image; and evaluating and calibrating the parameters by using the projection coordinates. The device and the method can realize the calibration of VCT system mechanical parameters, and ensure the precision of the parameter calibration.

Description

VCT system parameter calibration device and method

Technical field

The present invention relates to VCT (Volumetric Computerized Tomography, the forthright and sincere three-dimensional computer fault imaging of high-resolution) system parameter calibration technology, be specifically related to a kind of VCT system parameter calibration device and method.

Background technology

There is crucial effect in the VCT system in fields such as Non-Destructive Testing, industrial flaw detection, reverse-engineerings.VCT is that with the key distinction of general CT its resolution is higher, and can reconstruct the inner structure etc. of three-dimensional body according to different scanning motion tracks in conjunction with corresponding reconstruction algorithm.Usually, the VCT system comprises X source, objective table and detector.

In 3D image reconstruction process to the actual scanning object, need obtain projected image some geometric parameters when taking, for example X source is to the distance of scanning object, to the distance of detector, the position of rotation centerline etc., have only these geometric parameters of acquisition, ability is 3 d objects reconstruction accurately, and the precision of these parameters, the directly precision of influence reconstruction.

In recent years, for Cone-Beam CT, reconstruction algorithm has obtained developing rapidly.Reconstruction algorithm mainly comprises analytical algorithm and iterative algorithm, and wherein, analytical algorithm divides approximate data and exact algorithm again.Therefore with respect to iterative algorithm, analytical algorithm has reconstruction speed faster, and takies less memory source, becomes the main flow algorithm in the actual CT system.And in analytical algorithm, approximate data is because simple on the mathematics, and it is easy to implement, and when cone angle is smaller, can obtains and rebuild effect preferably, so have a wide range of applications in practice.In various approximate datas based on filtered back projection, the algorithm of FDK type is the main flow in using always.

The FDK algorithm is a kind of approximate reconstruction algorithm based on circular orbit scanning, is proposed by Feldkamp, Davis and Kress.Why the FDK algorithm is a kind of approximate algorithm, is because the resolution when measuring, and reconstructed results and real-world object all can have departing from more or less.But, for the cone angle of appropriateness, this depart from very little.FDK is because its simplification becomes the algorithm that is most widely used in the actual cone-beam reconstruction.

The FDK algorithm is actually the three-dimensional extended of D fan filter back-projection algorithm.It comprises pre-weighting, three steps of one-dimensional filtering and back projection of data for projection.In the FDK process of reconstruction, need known following two parameters: the one, flat panel detector drops on the position on the turning axle; The 2nd, the position of light source.In the VCT system,, can not guarantee in the installation process definitely or accurately to make above-mentioned two positions reach desirable requirement because light source, turning axle and flat panel detector all are fixed on the same lathe.For this reason, mainly obtain above-mentioned two parameters in the prior art,, thereby obtain accurate reconstructed image in application so that the user proofreaies and correct the VCT system according to this parameter by parameter calibration.

Chinese patent 200510045796.3 discloses the calibration template of a kind of cone-beam X-ray CT system, and this template has two kinds of method for makings: a kind of is it to be laid respectively on foursquare four summits inlaying four point-like Metal Ball on the poly (methyl methacrylate) plate; Another kind is at four point-like apertures of brill on the sheet metal it to be laid respectively on foursquare four summits.

For above-mentioned first method, be strict square owing to can not guarantee square, manufacture craft requires high, and difficulty of processing is big, and error is big, is not suitable for high-precision fixed target requirement; For second method, except having above-mentioned shortcoming equally, be difficult to guarantee that the point-like aperture is proper circle, can't satisfy high-precision fixed target requirement equally.

Summary of the invention

The embodiment of the invention provides a kind of VCT system parameter calibration device and method, with the demarcation of realization to VCT system mechanics parameter, and the precision of assurance parameter calibration.

For this reason, the embodiment of the invention provides following technical scheme:

The forthright and sincere three-dimensional computer computed tomography (SPECT) system of a kind of high-resolution parameter calibration device comprises:

Support, vertically be installed on the stationary pipes of described support and be fixed in first spheroid and second spheroid that is separated from each other in the described stationary pipes, described support uses identical material with stationary pipes, described support and spheroid use different materials, so that X ray produces different decay when passing.

Preferably, described first spheroid and second spheroid are the high-precision bearing steel ball.

Preferably, the diameter of described first spheroid and second spheroid is slightly less than the diameter of described stationary pipes, and is fixed in the described stationary pipes by outside winding bungee.

Preferably, described first spheroid and the adjustable positions of second spheroid in described stationary pipes.

Preferably, the material of described support and stationary pipes is an organic glass.

Preferably, offer a plurality of holes on the described support, the size and the stationary pipes in each hole are complementary, and described stationary pipes can vertically be inserted into wherein in any one hole, to regulate the spheroid moving radius.

Preferably, described support is circular.

Preferably, described device also comprises:

Be vertically installed in the fixture of described carriage center position, be used for described support is fixed in objective table.

Preferably, described fixture is for having externally threaded cylinder.

Alternatively, described fixture and described support are the one-shot forming structure.

Alternatively, described fixture is fixed on the described support by the mode of fusing.

A kind of method that realizes the forthright and sincere three-dimensional computer computed tomography (SPECT) system of high-resolution parameter calibration comprises:

The forthright and sincere three-dimensional computer computed tomography (SPECT) system of described high-resolution parameter calibration device is fixed on the objective table;

Described objective table is rotated a circle, and in rotary course, gather the image that the described device of multiframe forms on flat panel detector;

Obtain the projection coordinate of two spheroid centre ofs sphere in every two field picture;

Utilize described projection coordinate to estimate calibrating parameters.

Preferably, the described device of described collection multiframe comprises at the image that forms on the flat panel detector: equal angles is gathered the image that described device forms on flat panel detector.

Preferably, described origin is the center of planar detector, and X-axis is along the planar detector planar horizontal left, and the Z axle is along the planar detector plane straight up, Y-axis is the direction perpendicular to planar detector, and defines the direction of each coordinate axis according to right-hand rule;

The described process of obtaining the projection coordinate of two spheroid centre ofs sphere in every two field picture comprises:

By the gray scale clustering method background in the described image is separated with spheroid, obtain the spheroid image under the clean background;

To the gray-scale value summation of every capable pixel of described spheroid image, obtain gray scale vector h;

In described spheroid image, do level with the expectation value of gray scale vector h and cut apart, described spheroid image is divided into two images that only comprise a spheroid respectively;

To every image that only comprises a spheroid do level and expectation and vertically and expectation, obtain the trajectory coordinates of two spheroid gnomonic projections.

Preferably, the described described projection coordinate estimation calibrating parameters that utilizes comprises:

The trajectory coordinates of two spheroid gnomonic projections that utilization obtains is set up the least square objective function;

Estimate calibrating parameters by described objective function.

Preferably, the trajectory coordinates of two spheroid gnomonic projections obtaining of described utilization is set up the least square objective function and is comprised:

The trajectory coordinates of two spheroid gnomonic projections that utilization obtains and the parameter initial value of setting obtain the centre of sphere preimage coordinate of each spheroid;

Calculate parameter estimating error e1, the e2 of each spheroid and the distance error e3 of two spheroids respectively according to the centre of sphere preimage coordinate of described each spheroid;

Generate objective function mine=λ ₁e ₁+ λ ₂e ₂+ λ ₃e ₃, wherein, λ ₁, λ ₂, λ ₃Be the weights of optimizing, λ ₁+ λ ₂+ λ ₃=1.

Preferably, the trajectory coordinates of two spheroid gnomonic projections obtaining of the described utilization centre of sphere preimage coordinate that obtains each spheroid comprises:

Trajectory coordinates to two spheroid gnomonic projections is calculated respectively, obtains the projection of turning axle and the intersection of disappearance face and described flat panel detector;

With the intersection point of described two straight lines as the intersection point of main beam and described flat panel detector (Px, Py, Pz);

Obtain the rotation center coordinate of each spheroid according to the intersection point of the intersection point of main beam and rotating shaft and main beam and described flat panel detector;

According to the rotation center coordinate and the trajectory plane equation thereof of each spheroid, calculate the centre of sphere preimage coordinate of each spheroid;

Described main beam was represented the straight line of light source perpendicular to turning axle; Described disappearance face represented that described main beam was parallel to the plane of two spheroid Plane of rotations.

Preferably, adjust λ according to the size of spheroid track and the distance of spheroid ₁, λ ₂, λ ₃

Preferably, describedly estimate that by described objective function calibrating parameters comprises: utilize Gauss-Newton method that described calibrating parameters is optimized.

Embodiment of the invention VCT system parameter calibration device, not high for requirement on machining accuracy, make simple, with low cost.Based on the method for this device realization VCT system parameter calibration, can determine the various deviations of VCT mechanical system, comprise the location parameter of objective table, the straggling parameter of rotating shaft, the rotation of flat panel detector, pitching, the parameter that tilts, the location parameter of light source etc.; And parameter calibration precision height, can be applied to high resolving power VCT imaging, satisfy the demand of high precision imaging.

Description of drawings

Fig. 1 is the stereographic map of embodiment of the invention VCT system parameter calibration device;

Fig. 2 is the front view of embodiment of the invention VCT system parameter calibration device;

Fig. 3 is the left view of embodiment of the invention VCT system parameter calibration device;

Fig. 4 is the vertical view of embodiment of the invention VCT system parameter calibration device;

Fig. 5 is the process flow diagram of embodiment of the invention VCT system parameter calibration method;

Fig. 6 is the synoptic diagram of VCT system parameter calibration process in the embodiment of the invention;

Fig. 7 is a coordinate system synoptic diagram in the embodiment of the invention;

Fig. 8 is the gnomonic projection track synoptic diagram of two spheroids in the embodiment of the invention;

Fig. 9 is the perspective view of turning axle in the embodiment of the invention.

Embodiment

In order to make those skilled in the art person understand the scheme of the embodiment of the invention better, the embodiment of the invention is described in further detail below in conjunction with drawings and embodiments.

As shown in Figure 1, be the stereographic map of embodiment of the invention VCT system parameter calibration device, Fig. 2 is its front view, and Fig. 3 is its left view, and Fig. 4 is its vertical view.

This system comprises: support 11, vertically be installed on the stationary pipes 12 of described support and be fixed in first spheroid 13 and second spheroid 14 that is separated from each other in the described stationary pipes 12.

Preferably, described first spheroid 13 and second spheroid 14 are the high-precision bearing steel ball, such as being the high-precision bearing steel ball that meets the G3 standard.G3 standard steel ball is the highest standard of the steel bearing balls of standard GB/T 308-2002 regulation, and steel ball circumcircle and inscribed circle diameter difference are less than 0.8um.In embodiments of the present invention, the diameter of steel ball can be slightly less than the diameter of described stationary pipes 12, and twine bungee by the outside and be fixed in the described stationary pipes 12, certainly be not limited in this fixed form in the embodiment of the invention, also can adopt other modes to fix, such as, with a rubber through passing stationary pipes 12, again steel ball is filled in from two ends, because steel ball size approaches stationary pipes, as long as therefore have some resistances just can fix, perhaps at steel ball coated outside transparent adhesive tape, organic materials such as balloon perhaps intercept the stationary pipes of suitable length, and the stationary pipes of steel ball and intercepting is alternately filled in the vertical stationary pipes 12.Fixing principle is promptly will guarantee fixation, uses the low-density fixture again, and fixture will be penetrated by X ray easily, does not influence the imaging of steel ball.

In actual applications, described first spheroid 13 and the position of second spheroid 13 in described stationary pipes 12 can be fixed, and preferably, are designed to regulate according to application need.

The material of described support 11 and stationary pipes 12 is preferably organic glass, certainly, also can be other macromolecule materials that can be penetrated by X ray.

For the ease of in three-dimensional reconstruction, regulate the orbital radius of steel ball, also offer a plurality of holes 10 on the described support, the size in each hole and stationary pipes 12 are complementary, and described stationary pipes 12 can vertically be inserted into wherein in any one hole.The length of stationary pipes 12 can be processed into 200mm, can suitably adjust according to the size of scanned object in actual applications, and this embodiment of the invention is not limited.

The surface configuration of described support 11 is preferably circle, and diameter is 200mm, and thickness is 20mm.Can certainly be other shapes, such as square, ellipse etc.

In carrying out the parameter calibration process, described device need be fixed on the objective table, therefore, for the ease of fixing, described device also can further comprise: the fixture 15 that is vertically installed in described support home position, be used for described support 11 is fixed in objective table, described fixture 15 lays respectively on the not coplanar of described support 11 with described stationary pipes 12.

In embodiments of the present invention, described fixture 15 can have the different structure shape, and also can have multiple with the connected mode of support 11.

Such as, as shown in Figure 2, a kind of preferred structure of fixture 15 is for having externally threaded cylinder, and with described support be the one-shot forming structure.

For another example, described fixture 15 also can be fixed on the described support 11 by the mode of fusing.

In order to be applicable to the connected mode of different objective tables, be the interface of straight cuttings such as cylindrical, square or hexagon such as some objective table, therefore, described fixture 15 also can be designed to the corresponding interface shape as required.

In reality processing, embodiment of the invention VCT system parameter calibration device is not high for requirement on machining accuracy, makes simple, with low cost.The CT value of organic glass is little, separates with the spheroid projection in zone easily in the CT image.Described support only needs can two spheroids of secure, it is not swung get final product.Two spheroids can be selected the bearing steel of the G3 standard of selling on the market for use, diameter is 7.943mm, according to standard GB/T 308-2002, steel ball circumcircle and inscribed circle diameter difference are less than 0.8um, according to conventional VCT systematic parameter, at a distance of 1000mm, object distance 500mm, pixel size is that 127um calculates, and steel ball roundness error may cause the calculating of the bead centre of sphere to depart from＜0.01 pixel.

Based on the VCT system parameter calibration device that the invention described above embodiment provides, the embodiment of the invention also provides a kind of method of VCT system parameter calibration.

As shown in Figure 5, be the process flow diagram of the method for embodiment of the invention VCT system parameter calibration, may further comprise the steps:

Step 501 is fixed on described VCT system parameter calibration device on the objective table.

Step 502 rotates a circle described objective table, and in rotary course, gathers the image that the described device of multiframe forms on flat panel detector.

Particularly, in the calibration process, only need described device is placed on the objective table, according to getting final product conventional one week of CT scan, in scanning process, gather a two field picture every certain angle, gather 360 two field pictures such as gathering the image that the described device of a frame forms on flat panel detector every 1 degree, amounting to.

Step 503 is obtained the projection coordinate of two spheroid centre ofs sphere in every two field picture.

Particularly, can obtain the projection coordinate of two spheroid centre ofs sphere in every two field picture by following process:

(1) by the gray scale clustering method background in the described image is separated with spheroid, obtain the spheroid image under the clean background;

(2), obtain row gray scale vector h to the gray-scale value summation of every capable pixel of described spheroid image;

(3) in described spheroid image, do level with the expectation value of vectorial h and cut apart, described spheroid image is divided into two images that only comprise a spheroid respectively;

(4) to every image that only comprises a spheroid do level and expectation and vertically and expectation, obtain the trajectory coordinates of two spheroid gnomonic projections.

Step 504 utilizes described projection coordinate to estimate calibrating parameters.

Particularly, can utilize the trajectory coordinates of two spheroid gnomonic projections that obtain to set up the least square objective function; Estimate calibrating parameters by described objective function.

Further describe VCT system parameter calibration process in the embodiment of the invention below with reference to Fig. 6.

As shown in Figure 6, embodiment of the invention VCT system parameter calibration device 60 is fixed on the objective table 61, by the irradiation of X source 62, two spheroids in the described device 60 form projected image on flat panel detector.

In order to simplify follow-up computation process, can do following hypothesis to the VCT system:

(1) do not consider the out-of-flatness of detector, suppose that promptly detector is an ideal plane;

(2) do not consider the vibrations of light source and detector, suppose that promptly light source is relative with detector static;

(3) do not consider that object centers on beating of rotating shaft rotation, promptly suppose on the objective table object circular motion around the shaft;

The error that relevant above-mentioned hypothesis may exist will be carried out labor in the back.

As shown in Figure 7, can be defined as follows coordinate system:

The initial point of this coordinate system is the center of planar detector, and X-axis is along the planar detector planar horizontal left, and the Z axle is along the planar detector plane straight up, and Y-axis is the direction perpendicular to planar detector, and defines the direction of each coordinate axis according to right-hand rule.

If the coordinate of light source S is that (Sz), crossing light source is that (Pz), rotating shaft is Sp to the distance of light source for Px, Py perpendicular to the coordinate of the intersection point P of the light of rotating shaft and detector for Sx, Sy ₂, the distance of ordering with P is Sp ₁Definition rotating shaft position parametric t=Sp ₁/ Sp ₂, then in above-mentioned coordinate system, as long as P point coordinate, light source S coordinate and rotating shaft position parametric t have determined that whole mechanical system parameter has just been determined.

In embodiment of the invention VCT system parameter calibration device, described support and spheroid use different materials, so that X ray produces different decay when passing.

Be that organic glass support, stationary pipes are that plexi-glass tubular, described spheroid are that steel ball describes for example with described support below.

The decay of thin organic glass support is little because X ray penetrates, and the decay that penetrates steel ball is big, therefore can obtain the projection coordinate of two spheroid centre ofs sphere in every two field picture by technology such as image gray-scale transformation, image segmentation.

The image that described device forms on flat panel detector mainly comprises three class gray-scale values: background, organic glass matrix and steel ball, therefore can separate background by the method for gray scale cluster with spheroid.Like this, can obtain spheroid image under the clean background.

Then, to the gray-scale value of every capable pixel of this image do and, obtain vectorial h, the value h (i) of i the element of h is the capable gray-scale value sum of i, does horizontal cut-off rule with the expectation value of h in image, and former figure is divided into two images that have only a ball.

To every figure do level and expectation and vertically and expectation, thereby obtain the trajectory coordinates of two steel ball gnomonic projections, as shown in Figure 8.

The track that gnomonic projection has been arranged just can be according to this track estimating system parameter.Particularly, can utilize the trajectory coordinates of two spheroid gnomonic projections that obtain to set up the least square objective function; Estimate calibrating parameters by described objective function.

The detailed process that the trajectory coordinates of two spheroid gnomonic projections that utilization obtains is set up the least square objective function can may further comprise the steps:

The trajectory coordinates of two spheroid gnomonic projections that utilization obtains obtains the centre of sphere preimage coordinate of each spheroid;

Calculate the parameter estimating error e of each spheroid respectively according to the centre of sphere preimage coordinate of described each spheroid ₁, e ₂, and the distance error e of two spheroids ₃

Generate objective function mine=λ ₁e ₁+ λ ₂e ₂+ λ ₃e ₃, wherein, λ ₁, λ ₂, λ ₃Be the weights of optimizing, λ ₁+ λ ₂+ λ ₃=1.

To be elaborated to said process below.

Before parameter estimation, at first provide as giving a definition and theorem:

Theorem 1: in desirable VCT system, the steel ball rotation center is the position of θ and is projection A ', the B ' three point on a straight line of position on detector of θ+π in angle in angle at the projection O ' on the detector and steel ball.

Definition 1: light source S does the straight line perpendicular to turning axle excessively, and establishing intersection point is F, claims that straight line SF is a main beam.

Definition 2: cross the plane that main beam makes to be parallel to the bead Plane of rotation, be referred to as disappearance face.

Theorem 2: any two the projection straight line ls of parallel lines on detector on the steel ball Plane of rotation ₁And l ₂Intersection point, one fixes on the intersection of disappearance face and detector.

Can prove above-mentioned theorem by mathematical computations, not repeat them here.

According to above-mentioned theorem 1 as can be known,, can calculate the projection of the rotation center of each steel ball, connect the projection of the rotation center of two steel balls, can obtain the projection of turning axle on detector by each gnomonic projection.

In actual computation, because the existence of various errors in order to reduce error as much as possible, can adopt following method to calculate the projection of turning axle:

(1) determines the straight-line equation of 2 lines of steel ball gnomonic projection at the steel ball gnomonic projection at angle θ place and angle θ+π place, and determine

The place the steel ball gnomonic projection with

The straight-line equation of 2 lines of steel ball gnomonic projection at place, two equations of simultaneous calculate the intersecting point coordinate of two straight lines;

(2) make θ=θ+Δ, Δ is a constant, according to first group of coordinate of above-mentioned (1) step calculating;

(3) calculate the mean value of the many groups coordinate obtain previously, with it as some projection on detector on the axis.

(4) in like manner,, calculate the another one subpoint of axial line on detector, thereby determine the projection equation of shaft axis according to the projection of another steel ball.

Can calculate the projection of turning axle according to perspective view by preceding step, as shown in Figure 9.

Then,, utilize the projection of the turning axle that calculates, can determine the intersection of disappearance face and detector according to above-mentioned theorem 2.

Bead is at θ, θ+π,

The centre of sphere of four positions constitutes a square, and θ,

The line L of place's centre of sphere ₁With

θ+π centre of sphere line L ₂Parallel, according to theorem 2, L ₁With L ₂Projection P on detector ₁, P ₂Intersection point be positioned on the intersection L of disappearance face and detector, can obtain different parallel lines by changing θ, thereby obtain a plurality of intersection points, and then definite intersection L.

Consider the actual machine error, can determine intersection L according to following algorithm:

(1) calculate steel ball at θ,

The straight line L that place's projection is linked to be ₁With

The straight line L that θ+π place projection is linked to be ₂, and calculate L ₁With L ₂Intersection point M ₁

(2) make θ _I+1=θ _i+ Δ repeats above-mentioned steps (1), obtains M _I+1

(3) calculate another steel ball after the same method at θ ₁, θ ₁The straight line that the projection of+π place is linked to be with The intersection point N of the straight line that place's projection is linked to be _i

(4) with the M that calculates _i, N _iCarry out linear fit, obtain straight-line equation L, L is the intersection of disappearance face and detector.

By said process, can utilize the perspective view of the steel ball centre of sphere, calculate the projection of turning axle and the intersection of disappearance face and detector, the intersection point of these two lines be main beam and detector intersection point (Px, Py, Pz).

The equation of main beam is:

$\frac{x-\mathrm{Px}}{\mathrm{Sx}-\mathrm{Px}}=\frac{y-\mathrm{Py}}{\mathrm{Sy}-\mathrm{Py}}=\frac{z-\mathrm{Pz}}{\mathrm{Sz}-\mathrm{Pz}}$

If the intersection point of main beam and rotating shaft be (Fx, Fy, Fz):

$\left\{\begin{array}{c}\mathrm{Fx}=t(\mathrm{Sx}-\mathrm{Fx})+\mathrm{Px};\\ \mathrm{Fy}=t(\mathrm{Sy}-\mathrm{Fy})+\mathrm{Py};\\ \mathrm{Fy}=t(\mathrm{Sz}-\mathrm{Fz})+\mathrm{Pz};\end{array}\right.$

Rotating shaft through intersection point (Fz), and perpendicular to main beam, the rotation center of top steel ball is positioned in the rotating shaft for Fx, Fy, and light source with above steel ball rotation center projection (Px ₁, Py ₁, Pz ₁) line on, so the coordinate of top steel ball rotation center satisfies following formula:

$\left\{\begin{array}{c}(x-\mathrm{Fx})(\mathrm{Sx}-\mathrm{Px})+(y-\mathrm{Fy})(\mathrm{Sy}-\mathrm{Py})+(z-\mathrm{Fz})(\mathrm{Sz}-\mathrm{Pz})=0\\ \frac{(x-P{x}_1)}{\mathrm{Sx}-{\mathrm{<figure-callout\; id="1"\; label="Px"\; filenames="H2009101623297E00042.png"\; state="\{\{state\}\}">Px</figure-callout>}}_1}=\frac{\left({y-\mathrm{Py}}_1\right)}{\mathrm{Sy}-\mathrm{<figure-callout\; id="1"\; label="P\; y"\; filenames="H2009101623297E00042.png"\; state="\{\{state\}\}">P</figure-callout>}{y}_{}{}_1}=\frac{(z-P{z}_1)}{\mathrm{Sz}-{\mathrm{<figure-callout\; id="1"\; label="Pz"\; filenames="H2009101623297E00042.png"\; state="\{\{state\}\}">Pz</figure-callout>}}_1}\end{array}\right.$

Find the solution above-mentioned system of equations can obtain the top steel ball rotation center coordinate (x ₁, y ₁, z ₁).

In like manner, can obtain the rotation center coordinate (x of below steel ball ₂, y ₂, z ₂).

Because the steel ball trajectory plane is perpendicular to turning axle and process track center, so can obtain the trajectory plane equation of top steel ball:

(x-x ₁)(x ₁-x ₂)+(y-y ₁)(y ₁-y ₂)+(z-z ₁)(z ₁-z ₂)＝0

In like manner, can obtain the below steel ball the trajectory plane equation be:

(x-x ₂)(x ₁-x ₂)+(y-y ₂)(y ₁-y ₂)+(z-z ₂)(z ₁-z ₂)＝0

The preimage of top steel ball is positioned at top steel ball projection On the intersection point of the line of light source and trajectory plane, wherein, i represents the i frame, promptly satisfies following system of equations:

$\left\{\begin{array}{c}(x-x_1)(x_1-x_2)+(y-y_1)(y_1-y_2)+(z-z_1)(z_1-z_2)=0\\ \frac{(x-P{x}_1^i)}{(\mathrm{Sx}-P{x}_1^i)}=\frac{\left(y-P{y}_1^i\right)}{(\mathrm{Sy}-P{y}_1^i)}=\frac{(z-P{z}_1^i)}{(\mathrm{Sz}-P{z}_1^i)}\end{array}\right.$

Find the solution above-mentioned equation and can obtain the preimage coordinate of i frame steel ball projection

In like manner, can calculate the preimage coordinate of below steel ball at the i frame

Can obtain the preimage coordinate of each steel ball projection by said process, under the correct parameter estimation, the preimage coordinate of each steel ball projection should be on a circle, therefore, can obtain i frame steel ball and to the distance of rotation center is: $r_1^i={[{(x_1^i-x_1)}^2+{(y_1^i-y_1)}^2+{(z_1^i-z_1)}^2]}^{\frac{1}{2}}.$

In order to eliminate of the influence of orbital radius size, the distance error of i frame top steel ball rotation center can be defined as relative error to error:

$e_1=\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{\left(\frac{r_1^i-{\mathrm{<figure-callout\; id="1"\; label="Mr"\; filenames="H2009101623297E00042.png"\; state="\{\{state\}\}">Mr</figure-callout>}}_1}{{\mathrm{Mr}}_1}\right)}^2$

In the following formula N is the frame number of images acquired.In like manner, obtain the parameter estimating error of below steel ball:

$e_2=\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{\left(\frac{r_2^i-{\mathrm{Mr}}_2}{{\mathrm{Mr}}_2}\right)}^2$

In the following formula N is the frame number of images acquired.

In addition, under the desirable parameter estimation, the distance of two steel balls should be always Bd, therefore defines the third error, the i.e. distance error of two steel balls:

$e_3=\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{\{{[{(x_1^i-x_2^i)}^2+{(y_1^i-x_2^i)}^2+{(z_1^i-x_2^i)}^2]}^{\frac{1}{2}}-\mathrm{Bd}\}}^2$

Comprehensive above-mentioned three kinds of errors, the objective function that generates optimization is:

mine＝λ ₁e ₁+λ ₂e ₂+λ ₃e ₃，λ ₁+λ ₂+λ ₃＝1

Wherein, λ ₁, λ ₂, λ ₃For the weights of optimizing, can suitably adjust according to the size of little sphere path curve and the distance of bead in the practical application.

Can obtain the mechanical parameter of VCT system by finding the solution above-mentioned objective function, provide the method for solving of this planning problem below.

Analyze when defining coordinate system in front and obtain, whole mechanical system is by the P point coordinate, and light source S point coordinate and rotating shaft position coefficient t are unique to be determined, in the process of above-mentioned generation objective function, provided the P point coordinate, therefore parameter to be optimized have only rotating shaft position parametric t and light source coordinate (sx, sy, sz), promptly the decision vector of this optimization problem is X=(t, sx, sy, sz).This optimization problem belongs to least squares problem, can adopt Gauss-Newton (Gauss-newton) algorithm computation, and its algorithm is as follows:

1. two balls are constant apart from Bd, remain unchanged in the optimizing process, with the mechanical system reading that has error as the algorithm initial value.

2. calculate the Jacobi matrix J of the k time iteration,

Be the local derviation of i error of error of calculation vector about j parameter of decision vector.

3. calculated gains matrix:

$\mathrm{\&Delta;X}=-{\left(J_k^T{J}_k\right)}^{-1}{J}_k^T{e}_k;$

X _K+1=X _k+ Δ X, undated parameter is up to arriving end condition.End condition generally can be set to iteration round k＜N or | Δ X|＜ε.

Certainly, in embodiments of the present invention, finding the solution of described objective function is not limited in the above-mentioned Gauss-Newton algorithm of employing, can also adopts other algorithms to calculate, such as, the L-M algorithm, genetic algorithm, simulated annealing etc., detailed process does not repeat them here.

The method of embodiment of the invention VCT system parameter calibration can be determined the various deviations of VCT mechanical system to comprise the location parameter of objective table, the straggling parameter of rotating shaft, the rotation of flat panel detector, pitching, left and right sides parameter, the location parameter of light source etc.; And parameter calibration precision height, can be applied to the high-resolution ct imaging.Test findings shows that the parameter calibration precision can satisfy the demand that imaging precision is 30um.

Utilize the method for the embodiment of the invention, can be at different scanning objects, set different fixed position (being the relative position of light source, objective table and detector), parameter when calibrating diverse location, offer the user, after making the user to object scanning,, rebuild the 3-D view of scanned object exactly according to these calibrating parameters.Experimental result shows that the parameter calibration precision can satisfy the demand that imaging precision is 30um.

One of ordinary skill in the art will appreciate that all or part of step that realizes in the foregoing description method is to instruct relevant hardware to finish by program, described program can be stored in the computer read/write memory medium, described storage medium, as: ROM/RAM, magnetic disc, CD etc.

More than the embodiment of the invention is described in detail, used embodiment herein the present invention set forth, the explanation of above embodiment just is used for help understanding method and apparatus of the present invention; Simultaneously, for one of ordinary skill in the art, according to thought of the present invention, the part that all can change in specific embodiments and applications, in sum, this description should not be construed as limitation of the present invention.

Claims (19)

1. the forthright and sincere three-dimensional computer computed tomography (SPECT) system of a high-resolution parameter calibration device is characterized in that, comprising:

Support, vertically be installed on the stationary pipes of described support and be fixed in first spheroid and second spheroid that is separated from each other in the described stationary pipes, described support uses identical material with stationary pipes, described support and spheroid use different materials, so that X ray produces different decay when passing.

2. device according to claim 1 is characterized in that, described first spheroid and second spheroid are the high-precision bearing steel ball.

3. device according to claim 1 is characterized in that, the diameter of described first spheroid and second spheroid is slightly less than the diameter of described stationary pipes, and is fixed in the described stationary pipes by outside winding bungee.

4. device according to claim 3 is characterized in that, described first spheroid and the adjustable positions of second spheroid in described stationary pipes.

5. device according to claim 1 is characterized in that, the material of described support and stationary pipes is an organic glass.

6. according to each described device of claim 1 to 5, it is characterized in that offer a plurality of holes on the described support, the size and the stationary pipes in each hole are complementary, described stationary pipes can vertically be inserted into wherein in any one hole, to regulate the spheroid moving radius.

7. device according to claim 6, it is held to levy and is, and described support is circular.

8. device according to claim 7 is characterized in that, described device also comprises:

Be vertically installed in the fixture of described carriage center position, be used for described support is fixed in objective table.

9. device according to claim 8 is characterized in that, described fixture is for having externally threaded cylinder.

10. device according to claim 9 is characterized in that, described fixture and described support are the one-shot forming structure.

11. device according to claim 9 is characterized in that, described fixture is fixed on the described support by the mode of fusing.

12. a method of utilizing the described device of claim 1 to 11 to realize the forthright and sincere three-dimensional computer computed tomography (SPECT) system of high-resolution parameter calibration is characterized in that, comprising:

The forthright and sincere three-dimensional computer computed tomography (SPECT) system of described high-resolution parameter calibration device is fixed on the objective table;

Described objective table is rotated a circle, and in rotary course, gather the image that the described device of multiframe forms on flat panel detector;

Obtain the projection coordinate of two spheroid centre ofs sphere in every two field picture;

Utilize described projection coordinate to estimate calibrating parameters.

13. method according to claim 12 is characterized in that, the described device of described collection multiframe comprises at the image that forms on the flat panel detector: equal angles is gathered the image that described device forms on flat panel detector.

14. method according to claim 12, it is characterized in that, described origin is the center of described flat panel detector, X-axis is along described flat panel detector planar horizontal left, the Z axle is along described flat panel detector plane straight up, Y-axis is the direction perpendicular to described flat panel detector, and defines the direction of each coordinate axis according to right-hand rule;

The described process of obtaining the projection coordinate of two spheroid centre ofs sphere in every two field picture comprises:

By the gray scale clustering method background in the described image is separated with spheroid, obtain the spheroid image under the clean background;

To the gray-scale value summation of every capable pixel of described spheroid image, obtain gray scale vector h;

In described spheroid image, do level with the expectation value of gray scale vector h and cut apart, described spheroid image is divided into two images that only comprise a spheroid respectively;

To every image that only comprises a spheroid do level and expectation and vertically and expectation, obtain the trajectory coordinates of two spheroid gnomonic projections.

15. method according to claim 14 is characterized in that, the described described projection coordinate estimation calibrating parameters that utilizes comprises:

The trajectory coordinates of two spheroid gnomonic projections that utilization obtains is set up the least square objective function;

Estimate calibrating parameters by described objective function.

16. method according to claim 15 is characterized in that, the trajectory coordinates of two spheroid gnomonic projections that described utilization obtains is set up the least square objective function and is comprised:

The trajectory coordinates of two spheroid gnomonic projections that utilization obtains and the parameter initial value of setting obtain the centre of sphere preimage coordinate of each spheroid;

Calculate parameter estimating error e1, the e2 of each spheroid and the distance error e3 of two spheroids respectively according to the centre of sphere preimage coordinate of described each spheroid;

Generate objective function mine=λ ₁e ₁+ λ ₂e ₂+ λ ₃e ₃, wherein, λ ₁, λ ₂, λ ₃Be the weights of optimizing, λ ₁+ λ ₂+ λ ₃=1.

17. method according to claim 16 is characterized in that, the centre of sphere preimage coordinate that the trajectory coordinates of two spheroid gnomonic projections that described utilization obtains obtains each spheroid comprises:

Trajectory coordinates to two spheroid gnomonic projections is calculated respectively, obtains the projection of turning axle, and the intersection of the disappearance face of obtaining and described flat panel detector;

With the intersection point of described projection and described intersection as the intersection point of main beam and described flat panel detector (Px, Py, Pz);

Obtain the rotation center coordinate of each spheroid according to the intersection point of the intersection point of main beam and rotating shaft and main beam and described flat panel detector;

According to the rotation center coordinate and the trajectory plane equation thereof of each spheroid, calculate the centre of sphere preimage coordinate of each spheroid;

Described main beam was represented the straight line of light source perpendicular to turning axle; Described disappearance face represented that described main beam was parallel to the plane of two spheroid Plane of rotations.

18. method according to claim 16 is characterized in that, adjusts λ according to the size of spheroid track and the distance of spheroid ₁, λ ₂, λ ₃

19. method according to claim 15 is characterized in that, describedly estimates that by described objective function calibrating parameters comprises:

Utilize Gauss-Newton method that described calibrating parameters is optimized.

Images