CN103759645A - Aerostat capsule volume measurement method - Google Patents

Abstract

The invention discloses an aerostat capsule volume measurement method. Accurate space coordinates of a visual system calibration point are acquired through a laser distance measurement meter, then the visual system is accordingly calibrated, and a suspended aerostat capsule with a visual mark point bonded is shot through the calibrated visual system. Because a hull is huge, segmented shooting is carried out by pulling and moving an aerostat. Rebuilding of space coordinates of the mark point and three-dimensional reconstruction of partial capsule are carried out according to visual system parameters and the position of the shot mark point of the capsule on an image, a three-dimensional reconstruction model of the complete capsule is acquired by splicing capsule segments, and the volume of the capsule is worked out accordingly. The aerostat capsule volume measurement method provides an effective means for measuring the volume of the flexible capsule of the aerostat, thereby effectively guaranteeing development, production and safe operation of the aerostat; the method is further suitable for measuring space volumes of other large-sized objects and is significant in the large-size measurement field.

Description

A kind of aerostatics utricule volume measuring method

Technical field

The present invention relates to large-scale metrology field, space, particularly a kind of aerostatics utricule volume measuring method.

Background technology

Aerostatics hull is special-shaped flexible utricule, in process of manufacture, be subject to the impact of material behavior and processing technology, between aerostatics hull after moulding and design, there is certain deviation, if aerostatics utricule bulk and design load differ larger, will produce material impact to the balance quality of aerostatics, aerodynamic characteristic and flight quality.Utricule bulk is directly related with utricule volume again, and the volume of aerostatics has determined its useful load.Therefore, in aerostatics general assembly, total debugging, before testing, be necessary its utricule bulk and volume to measure.

Aerostatics bulky, the irregular existence abnormity of profile, be difficult to test indoor, and in the experimental enviroment of outfield due to the impact of wind field, air-flow, hull can not keep stationary state, be difficult to find suitable witness mark, therefore, be difficult to by existing routine techniques and stable large volume measuring method, its flexible bags body space size and volume be measured.

Summary of the invention

Technical matters to be solved by this invention is, for prior art deficiency, to provide a kind of aerostatics utricule volume measuring method, the volume of Measurement accuracy aerostatics utricule.

For solving the problems of the technologies described above, the technical solution adopted in the present invention is: a kind of aerostatics utricule volume measuring method, and the method is:

1) arrange vision measurement system: described vision measurement system comprises a plurality of camera mounting brackets and a plurality of tripod, and a plurality of camera pan-tilts are installed on described camera mounting bracket, on described camera pan-tilt, video camera is installed; Laser range finder The Cloud Terrace is installed on described tripod, on described laser range finder The Cloud Terrace, laser range finder is installed; Described camera pan-tilt, video camera, laser range finder The Cloud Terrace, laser range finder all access main control computer; Described a plurality of camera mounting bracket and a plurality of tripod are all evenly distributed on region for placing aerostatics utricule around;

2) determine world coordinate system: select a laser range finder A, the world coordinates of definition A is p _{laser, 1}=(0,0,0); Alternative is selected a laser range finder B, measures the distance L between A, B, and the world coordinates of definition B is p _{laser, 2}=(L, 0,0); Select again a laser range finder C, measure the distance l between A, C ₁and the distance l between B, C ₂, the world coordinates of definition laser range finder C is p _{laser, 3}=(x _c3, y _c3, 0), wherein

$\begin{array}{c}{x}_{c3}=\frac{{\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="HDA0000462318560000031.png"\; state="\{\{state\}\}">L</figure-callout>}}^2+({\mathrm{<figure-callout\; id="1"\; label="l"\; filenames="HDA0000462318560000013.png,HDA0000462318560000022.png"\; state="\{\{state\}\}">l</figure-callout>}}_1^2-{\mathrm{<figure-callout\; id="2"\; label="l"\; filenames="HDA0000462318560000031.png"\; state="\{\{state\}\}">l</figure-callout>}}_2^2)}{2L}\\ y_{c3}=\frac{\sqrt{({\mathrm{<figure-callout\; id="4"\; label="L"\; filenames="HDA0000462318560000021.png"\; state="\{\{state\}\}">L</figure-callout>}}^4+{\mathrm{<figure-callout\; id="1"\; label="l"\; filenames="HDA0000462318560000013.png,HDA0000462318560000022.png"\; state="\{\{state\}\}">l</figure-callout>}}_1^4+{\mathrm{<figure-callout\; id="2"\; label="l"\; filenames="HDA0000462318560000031.png"\; state="\{\{state\}\}">l</figure-callout>}}_2^4)-{(L^2-{\mathrm{<figure-callout\; id="1"\; label="l"\; filenames="HDA0000462318560000013.png,HDA0000462318560000022.png"\; state="\{\{state\}\}">l</figure-callout>}}_1^2)}^2-{(L^2-{\mathrm{<figure-callout\; id="2"\; label="l"\; filenames="HDA0000462318560000031.png"\; state="\{\{state\}\}">l</figure-callout>}}_2^2)}^2-{(l_1^2-{\mathrm{<figure-callout\; id="2"\; label="l"\; filenames="HDA0000462318560000031.png"\; state="\{\{state\}\}">l</figure-callout>}}_2^2)}^2}}{2L}\end{array};$

3), to all the other arbitrary laser range finder X, measure respectively above-mentioned laser range finder A, B, C to the distance l between laser range finder X _a, l _band l _c, obtain the world coordinates (x, y, z) of laser range finder X, wherein x, y, z meets system of equations:

$\left\{\begin{array}{c}{x}^2+{\mathrm{<figure-callout\; id="2"\; label="y"\; filenames="HDA0000462318560000031.png"\; state="\{\{state\}\}">y</figure-callout>}}^2+z^2=l_A^2\\ {(x-L)}^2+{\mathrm{<figure-callout\; id="2"\; label="y"\; filenames="HDA0000462318560000031.png"\; state="\{\{state\}\}">y</figure-callout>}}^2+z^2=l_B^2\\ {(x-x_{c3})}^2+{(y-y_{c3})}^2+z^2={\mathrm{<figure-callout\; id="2"\; label="l\; C"\; filenames="HDA0000462318560000031.png"\; state="\{\{state\}\}">l</figure-callout>}}_C^{}\end{array}\right.;$

The like, obtain the world coordinates of all the other all laser range finders;

4) video camera is taken to be arranged in and is placed the calibration point plate array that a plurality of calibration point plates in aerostatics utricule region form; Determine that i platform laser range finder is to the distance l between j calibration point plate _ij; By solving the least square solution of following system of equations, determine the world coordinates (x of j calibration point plate _j, y _j, z _j):

${(x_j-x_{\mathrm{laser},i})}^2+{(y_j-y_{\mathrm{laser},i})}^2+{(z_j-z_{\mathrm{laser},i})}^2={\mathrm{<figure-callout\; id="2"\; label="l\; ij"\; filenames="HDA0000462318560000031.png"\; state="\{\{state\}\}">l</figure-callout>}}_{\mathrm{ij}}^{};$

Wherein, i=1 ..., N _laser; J=1 ..., N _cali; N _laserfor laser range finder number of units; N _caliquantity for calibration point plate; (x _{laser, i}, y _{laser, i}, z _{laser, i}be the world coordinates p of i platform laser range finder _{laser, i};

5) repetition above-mentioned steps 4) 4～9 times, every video camera all obtains the image of 4～9 width calibration point plate arrays, thereby forms n _calin _camthe calibration maps image set of width calibration point plate array image construction, wherein n _cali=4～9, N _camfor the video camera number of units in vision measurement system; According to standard camera scaling method, utilize the world coordinates of described calibration maps image set and each calibration point plate, determine the parameter c of r platform video camera in measure field _r=(α _r, β _r, γ _r, t _x,r, t _y, _r, t _z,r, f _r, κ _r, s _x,r, s _y,r, c _x,r, c _y,r), (α wherein _r, β _r, γ _r) represent that the space of r platform video camera in world coordinate system is towards angle, (t _x,r, t _y,r, t _z,r) coordinate of expression r platform video camera in world coordinate system, f _rthe main distance that represents r platform video camera, κ _rthe radial distortion factor that represents r platform video camera, (s _x,r, s _y,r) represent the scaling factor of r platform video camera, (c _x,r, c _y,r) represent the principal point coordinate of r platform video camera;

6) aerostatics utricule is placed in aerostatics utricule region, on the aerostatics utricule surface of unaerated, a plurality of visual indicia points are set, aerostatics utricule is inflated, deflection arch by aerostatics utricule starts, make aerostatics utricule by the shooting area of described vision measurement system, described in computer control, vision measurement system carries out segmentation shooting to described aerostatics utricule one side, and in two adjacent segmentations, has at least 3 identical visual indicia points to appear in photographic images corresponding to these two segmentations simultaneously; Complete after the shooting of aerostatics utricule one side, according to identical method, complete the shooting of aerostatics utricule opposite side;

7) coordinate of setting some segmentations is aerostatics utricule coordinate system, and the visual indicia point transformation in all the other segmentations, in described aerostatics utricule coordinate system, is obtained to the aerostatics utricule coordinate of all visual indicia points;

8) utilize the aerostatics utricule coordinate of all visual indicia points, by the three-dimensional model on standard triangle method for reconstructing reconstruct aerostatics utricule surface;

9) determine 2 points of three-dimensional surface curvature maximum on described three-dimensional model, the axis that the line of take between these 2 is aerostatics, the length l using the length of described axis as aerostatics utricule, carries out discrete sampling with step-length △ d to described axis, obtains N _sindividual axis sampled point; The value of △ d is no more than 10cm; In each axis sample point, according to the three-dimensional model on aerostatics utricule surface, determine utricule surface and the intersecting lens of crossing the axis vertical plane of current axis sampled point, thereby determine the border in utricule cross section, then on the border in utricule cross section, found the long-chord in cross section of current axis sampled point, travel through after each axis sampled point, using the direction of the string of length maximum in the long-chord in all cross sections as Width, and the width w using the length of the string of this length maximum as aerostatics utricule; Again travel through each axis sampled point, in each axis sample point, calculate utricule cross section and in the direction with described Width quadrature, cross the chord length of this axis sampled point, and get the chord length of length minimum in all chord lengths, using the direction of chord length of length minimum as short transverse, and the height h using the length of the chord length of length minimum as aerostatics utricule; Again travel through all axis sampled points, determine the utricule cross section of g axis sample point, then the rectangular node that utilizes the length of side to be no more than 10cm carries out discretize to the utricule cross section of g axis sample point, utilizes rectangular node number in the utricule cross section drop on g axis sample point as the approximate value A of area of section _g, the volume V of calculating aerostatics utricule: $V=\underset{g=1}{\overset{N_s}{\mathrm{\&Sigma;}}}{A}_g\&CenterDot;\mathrm{\&Delta;d}.$

In two adjacent segmentations, there are 5～10 identical visual indicia points to appear in photographic images corresponding to these two segmentations simultaneously.

In described step 8), △ d gets 1cm, and the length of side of described rectangular node is 1cm.

Compared with prior art, the beneficial effect that the present invention has is: the present invention provides effective means for the flexible utricule volume of Measurement accuracy aerostatics, thereby for development, production and the safe operation of aerostatics provides sound assurance, method of the present invention is also applicable to the spatial volume of other large-sized object and measures, and in large-scale metrology field, has great importance.

Accompanying drawing explanation

Fig. 1 is the approximate range schematic diagram of one embodiment of the invention ship first point and utricule afterbody point;

Fig. 2 is two some schematic diagram of curvature maximum on one embodiment of the invention three-dimensional surface;

Fig. 3 is the axis schematic diagram of one embodiment of the invention aerostatics;

Fig. 4 is the long-chord schematic diagram in cross section that one embodiment of the invention is crossed aerostatics axis;

Fig. 5 is the short transverse schematic diagram of one embodiment of the invention aerostatics utricule;

Fig. 6 is one embodiment of the invention vision measurement system structural representation;

Fig. 7 is one embodiment of the invention support and video camera scheme of installation;

Fig. 8 is one embodiment of the invention tripod and laser range finder scheme of installation;

Fig. 9 is one embodiment of the invention calibration point plate schematic diagram.

Embodiment

The principle of the invention is as follows:

(1) vision measurement of utricule bulk

Because aerostatics utricule size is large, so some video camera in vision measurement system need to be installed on higher support.

While utilizing vision system to measure aerostatics, what can record is each bulk volume coordinate that is pasted in advance aerostatics appearance, and will obtain the parameters such as aerostatics volume, also needs according to the three-dimensional surface of bulk coordinate reconstruct aerostatics.Three-dimensional reconstruction adopts the triangular net on k-arest neighbors method construct aerostatics surface, in conjunction with the common shape deviation of introducing in the design outline of aerostatics and aerostatics manufacture process and the analysis of distribution, obtain each leg-of-mutton priori curved surface form, construct or select suitable interpolation curved surface, as bilinear interpolation or bicubic B-spline surface interpolation etc., rebuild aerostatics surface.

Because aerostatics utricule size is larger, generally also need the rational vision system of employing scale to carry out piecemeal measurement to aerostatics, then the three-dimensional point set of each piecemeal of aerostatics is spliced.Suppose to consider two piecemeal A and B, the mark point set on these two piecemeals is respectively

$P_A=\{p_i^A=(x_i^A,y_i^A,z_i^A)|1\&le;i\&le;n_A\}---\left(1\right)$

$P_B=\{p_j^B=(x_j^B,y_j^B,z_j^B)|1\&le;j\&le;n_B\}---\left(2\right)$

The coordinate system of stereo visual system when the concentrated coordinate of these two gauge points is all measured corresponding to piecemeal.Suppose that certain gauge point of hull appears in piecemeal A and B simultaneously, while taking due to piecemeal there is movement in hull, make this gauge point in different piecemeal shooting process, there is change in the position corresponding to visual angle system coordinate system, be that the same gauge point of hull is different with the coordinate in B at piecemeal A, be expressed as:

P=(x ^a, y ^a, z ^a) and q=(x ^b, y ^b, z ^b) (3)

Obviously, the hypothesis that is rigid body based on hull, the coordinate p that this gauge point obtains when different piecemeals are taken and q can be contacted by a rigid body rotational transform and a translation of rigid body conversion, and by these conversion, by gauge point, the coordinate transform in a piecemeal is the coordinate in another piecemeal.Between p and q, there is following relation

p=Rq+t (4)

Wherein

$R=\left[\begin{array}{ccc}1& 0& 0\\ 0& \cos \mathrm{\&alpha;}& -\sin \mathrm{\&alpha;}\\ 0& \sin \mathrm{\&alpha;}& \cos \mathrm{\&alpha;}\end{array}\right]\&CenterDot;\left[\begin{array}{ccc}\cos \mathrm{\&beta;}& 0& \sin \mathrm{\&beta;}\\ 0& 1& 0\\ -\sin \mathrm{\&beta;}& 0& \cos \mathrm{\&beta;}\end{array}\right]\&CenterDot;\left[\begin{array}{ccc}\cos \mathrm{\&gamma;}& -\sin \mathrm{\&gamma;}& 0\\ \sin \mathrm{\&gamma;}& \cos \mathrm{\&gamma;}& 0\\ 0& 0& 1\end{array}\right]---\left(5\right)$

For rigid body rotational transform matrix,

t=(t _x,t _y,t _z) ^T (6)

For translation of rigid body vector.Visible, for rigid body translation, there are 3 translation freedoms and 3 rotary freedoms, there are 6 degree of freedom.The splicing of three-dimensional point set is exactly the different coordinate figures that obtain in different piecemeals are measured according to known same gauge point, calculates 6 degree of freedom parameter values describing rigid body translation completely.

In order to realize three-dimensional point set splicing, need in different piecemeals, all include abundant same tag point.If the gauge point that two piecemeals obtain is concentrated, only include 1 identical gauge point, only can determine 3 degree of freedom of translation of rigid body, cannot determine 3 rotary freedoms; If only comprise 2 identical gauge points, these 2 gauge points can determine 3 translation freedoms and 2 rotary freedoms jointly, cannot determine the vertical direction rotary freedom of these 2 lines.Therefore, in two piecemeals, at least need to comprise 3 or 3 above same tag points, could determine required rigid body translation by Simultaneous Equations.

If consider measurements and calculations error, equation (4) is put concentrated registration mark point to two piecemeals and is not strictly set up, and needs more registration mark point to make their matching error minimum.

Suppose that A and B piecemeal have m identical gauge point V, the coordinate of these gauge points in two piecemeals is

$V_A=\{P_{i_k}=(x_{i_k}^A,y_{i_k}^A,z_{i_k}^A)|1\&le;i_k\&le;n_A,1\&le;k\&le;m\}---\left(7\right)$

$V_B=\{q_{j_k}=(x_{j_k}^B,y_{j_k}^B,z_{j_k}^B)|1\&le;j_k\&le;n_B,1\&le;k\&le;m\}---\left(8\right)$

Our target is to determine R and t, makes matching error objective function

$\underset{k=1}{\overset{m}{\mathrm{\&Sigma;}}}{\left|\right|{\mathrm{Rq}}_{j_k}+t-p_{i_k}\left|\right|}^2---\left(9\right)$

Minimum.

By setting and the piecemeal shooting process of suitable design gauge point, just can divide interblock to obtain abundant same tag point in difference, then utilize optimization method to solve objective function (9), the gauge point coordinate of each different piecemeal gained is all transformed in a common coordinate reference system, and then according to the three-dimensional model on the coordinate reconstruct hull surface of all hull gauge points in common coordinate reference system, complete the committed step of measurement.

(2) scene of calibration point volume coordinate is quick and precisely measured

By being no less than the pin-point accuracy laser range finder of three, the three-dimensional coordinate position of calibration point and video camera is measured.Utilize servomotor to coordinate reducing gear, the position angle of stadimeter mounting platform and the angle of pitch are accurately controlled.For can automatically control laser range finder one by one tested point more to quantity, that space distribution scope is wider measure, should be every stadimeter and be equipped with an auxiliary camera, utilize the scene image of shot by camera to control the pose of stadimeter.

(3) calibration of body vision measuring system

For vision system, every video camera need to be demarcated 11～12 parameters, and the video camera of this vision system is estimated more than ten platform, also need to determine the relative pose parameter between different cameras, cause the number of parameters of the required demarcation of whole vision measurement system many, cause the difficulty of solving-optimizing problem to strengthen.Therefore, problem of calibrating should be decomposed into the less optimization subproblem of some parameters and solve, and make full use of volume coordinate measuring technique, further reduce the difficulty that solves of optimizing subproblem, improve the reliability of result.

Specific implementation process of the present invention is as follows:

(1) vision measurement device of aerostatics utricule volume (as shown in Figure 6 to 8) comprises support 1, N _cam(N _cam=14～20) platform can by external connection control the video camera 5 taken and equal number can computer-controlled camera pan-tilt 6, main control computer 4, N _laser(N _laser=4～6) platform laser range finder 7, tripod 2, laser range finder The Cloud Terrace 8, during measurement, also need N _cali(N _cali=36～100) individual calibration point plate 3(disc gauge point plate, the desirable 30～50cm of diameter, the desirable 5～10cm of black disk diameter at center, as shown in Figure 9) and abundant adhesive type visual indicia point (quantity is looked tested aerostatics utricule size and determined); In the present embodiment, camera pan-tilt and laser range finder The Cloud Terrace all adopt the three-dimensional scientific and technological 810 type turntables in Qingdao;

(2) in measure field (aerostatics utricule peripheral region, this region need guarantee that video camera and laser range finder can measure aerostatics utricule) set up support, install The Cloud Terrace and camera, laid the cable between The Cloud Terrace and camera and communication unit and between communication unit and computing machine, The Cloud Terrace and camera have been adjusted into suitable attitude;

(3) vision measurement is arranged laser range finder in the place of measuring, and its putting position should be as far as possible peripheral in the effective range of carrying out vision system, and equal clear is blocked between any two;

(4) determine as follows world coordinate system: select a laser range finder A, the world coordinates of definition A is p _{laser, 1}=(0,0,0); Alternative is selected a laser range finder B, utilizes laser ranging to measure the distance L between A, B, and the world coordinates of definition B is p _{laser, 2}=(L, 0,0); Select again a laser range finder C, utilize laser ranging to measure the distance l between A, C ₁and the distance l between B, C ₂, the world coordinates of definition C is p _{laser, 3}=(x _c3, y _c3, 0), wherein

$\begin{array}{c}{x}_{c3}=\frac{{\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="HDA0000462318560000031.png"\; state="\{\{state\}\}">L</figure-callout>}}^2+({\mathrm{<figure-callout\; id="1"\; label="l"\; filenames="HDA0000462318560000013.png,HDA0000462318560000022.png"\; state="\{\{state\}\}">l</figure-callout>}}_1^2-{\mathrm{<figure-callout\; id="2"\; label="l"\; filenames="HDA0000462318560000031.png"\; state="\{\{state\}\}">l</figure-callout>}}_2^2)}{2L}\\ y_{c3}=\frac{\sqrt{({\mathrm{<figure-callout\; id="4"\; label="L"\; filenames="HDA0000462318560000021.png"\; state="\{\{state\}\}">L</figure-callout>}}^4+{\mathrm{<figure-callout\; id="1"\; label="l"\; filenames="HDA0000462318560000013.png,HDA0000462318560000022.png"\; state="\{\{state\}\}">l</figure-callout>}}_1^4+{\mathrm{<figure-callout\; id="2"\; label="l"\; filenames="HDA0000462318560000031.png"\; state="\{\{state\}\}">l</figure-callout>}}_2^4)-{(L^2-{\mathrm{<figure-callout\; id="1"\; label="l"\; filenames="HDA0000462318560000013.png,HDA0000462318560000022.png"\; state="\{\{state\}\}">l</figure-callout>}}_1^2)}^2-{(L^2-{\mathrm{<figure-callout\; id="2"\; label="l"\; filenames="HDA0000462318560000031.png"\; state="\{\{state\}\}">l</figure-callout>}}_2^2)}^2-{({\mathrm{<figure-callout\; id="1"\; label="l"\; filenames="HDA0000462318560000013.png,HDA0000462318560000022.png"\; state="\{\{state\}\}">l</figure-callout>}}_1^2-{\mathrm{<figure-callout\; id="2"\; label="l"\; filenames="HDA0000462318560000031.png"\; state="\{\{state\}\}">l</figure-callout>}}_2^2)}^2}}{2L}\end{array}$

(5), to all the other arbitrary laser range finder X, utilize laser range finder A, B, C in (2) to measure respectively them separately to the distance l between X _a, l _band l _c, then calculate the world coordinates (x, y, z) of X, wherein x, y, z meets system of equations

$\left\{\begin{array}{c}{x}^2+{\mathrm{<figure-callout\; id="2"\; label="y"\; filenames="HDA0000462318560000031.png"\; state="\{\{state\}\}">y</figure-callout>}}^2+{\mathrm{<figure-callout\; id="2"\; label="z"\; filenames="HDA0000462318560000031.png"\; state="\{\{state\}\}">z</figure-callout>}}^2=l_A^2\\ {(x-L)}^2+{\mathrm{<figure-callout\; id="2"\; label="y"\; filenames="HDA0000462318560000031.png"\; state="\{\{state\}\}">y</figure-callout>}}^2+{\mathrm{<figure-callout\; id="2"\; label="z"\; filenames="HDA0000462318560000031.png"\; state="\{\{state\}\}">z</figure-callout>}}^2=l_B^2\\ {(x-x_{c3})}^2+{(y-y_{c3})}^2+{\mathrm{<figure-callout\; id="2"\; label="z"\; filenames="HDA0000462318560000031.png"\; state="\{\{state\}\}">z</figure-callout>}}^2=l_C^2\end{array}\right.---\left(1\right)$

Can obtain thus the world coordinates p of all laser range finders _{laser, i}(i=1 ..., N _laser);

(6) in measuring place, the diverse location place of (being in aerostatics utricule put area) lays calibration point plate, utilize vision measurement system to take calibration point plate array, and determine that by laser ranging every laser range finder i is to the distance l between each calibration point plate j _ij(i=1 ..., N _laser, j=1 ..., N _cali), by asking for the least square solution of following system of equations, determine the world coordinates (x of calibration point plate j _j, y _j, z _j):

${(x_j-x_{\mathrm{laser},i})}^2+{(y_j-y_{\mathrm{laser},i})}^2+{(z_j-z_{\mathrm{laser},i})}^2={\mathrm{<figure-callout\; id="2"\; label="l\; ij"\; filenames="HDA0000462318560000031.png"\; state="\{\{state\}\}">l</figure-callout>}}_{\mathrm{ij}}^{},i=1,...,N_{\mathrm{laser}}$

(x wherein _{laser, i}, y _{laser, i}, z _{laser, i}) be the world coordinates p of i platform laser range finder _{laser, i};

(7) step in (6) is repeated to N _cf(N _cf=4～9) inferior, obtain by N _cfone group of uncalibrated image that width calibration point plate array image forms, then utilizes existing standard camera scaling method, according to this group uncalibrated image and wherein the world coordinates of each calibration point plate obtain the parameter c of each camera _i=(α _i, β _i, γ _i, t _x,i, t _y,i, t _z,i, f _i, κ _i, s _x,i, s _y,i, c _x,i, c _y,i), (α wherein _i, β _i, γ _i) represent that the space of i platform camera in world coordinate system is towards angle, (t _x,i, t _y,i, t _z,i) coordinate of expression i platform camera in world coordinate system, f _ithe main distance that represents i platform camera, κ _ithe radial distortion factor that represents i platform camera, (s _x,i, s _y,i) represent the scaling factor (calculate and obtain by the calibration process of standard) of i platform camera, (c _x,i, c _y,i) represent the principal point coordinate of i platform camera.Carry out timing signal, can to the position of camera, measure according to the method in (6), for arbitrary video camera Y, utilizing each laser range finder to measure them separately to the distance between Y, then by least square method, solving the system of equations in formula (1); Using the world coordinates of the video camera i trying to achieve as (t _x,i, t _y,i, t _z,i) initial value, to reduce to demarcate difficulty, improve the accuracy of parameters obtained;

(8) before aerostatics utricule inflation, on utricule surface, with suitable density, paste visual indicia point, main principle is the significantly bending that do not have of the utricule surface between adjacent visual indicia point.Then inflate.After having inflated, by deflection arch, started, under the pulling of towing vehicle, hull, gradually by effective shooting area of vision measurement system, carries out segmentation shooting by computer control vision measurement system, complete the shooting of a side utricule, and then turn around to take opposite side utricule.When carrying out segmentation and take, require adjacent two to take in segmentations, all at least to have to be no less than 3 identical gauge points and to appear among captured image simultaneously, preferably can there be 5～10 same tag points that simultaneously appear in adjacent two segmentations;

(9) to each segmentation, utilize the three-dimensional reconstruction method of standard, according to the vision system parameter of having demarcated, reconstruct the three-dimensional world coordinate of each gauge point in this segmentation.Suppose that in the k of gained and k+1 segmentation, each gauge point set is respectively P _kand P _k+1:

P _k={(x _k,i,y _k,i,z _k,i) ^T|i=1,…,n _k},P _k+1={(x _k+1,i,y _k+1,i,z _k+1,i) ^T|i=1,…,n _k+1}

N wherein _kand n _k+1represent respectively the gauge point quantity in k and k+1 segmentation, x, y, z represents respectively the three-dimensional coordinate component of respective point, A ^tthe transposition of representing matrix A.And suppose that common gauge point in k and k+1 the segmentation coordinate set in two segmentations is respectively:

$C_k=\left\{{(x_{k,u_j},y_{k,u_j},z_{k,u_j})}^T\right|1\&le;u_j\&le;n_k\},C_{k+1}\{{(x_{k+1,v_j},y_{k+1,v_j},z_{k+1,v_j})}^T|1\&le;v_j\&le;n_{k+1}\},j=1,...,n_c^{\left(k\right)}$

In formula,

the number that represents same tag point in k and k+1 segmentation.By least square method, solve following system of equations to obtain transform matrix M _k:

$M_k=\left[\begin{array}{ccc}{a}_{11}^{\left(k\right)}& a_{12}^{\left(k\right)}& a_{13}^{\left(k\right)}\\ a_{21}^{\left(k\right)}& a_{22}^{\left(k\right)}& a_{23}^{\left(k\right)}\\ a_{31}^{\left(k\right)}& a_{32}^{\left(k\right)}& a_{33}^{\left(k\right)}\end{array}\right]$

$\left\{\begin{array}{c}{a}_{11}^{\left(k\right)}{x}_{k+1,v_j}+a_{12}^{\left(k\right)}{y}_{k+1,v_j}+a_{13}^{\left(k\right)}{z}_{k+1,v_j}=x_{k,u_j}\\ a_{21}^{\left(k\right)}{x}_{k+1,v_j}+a_{22}^{\left(k\right)}{y}_{k+1,v_j}+a_{23}^{\left(k\right)}{z}_{k+1,v_j}=y_{k,u_j},j=1,...,n_c^{\left(k\right)}\\ a_{31}^{\left(k\right)}{x}_{k+1,v_j}+a_{32}^{\left(k\right)}{y}_{k+1,v_j}+a_{33}^{\left(k\right)}{z}_{k+1,v_j}=z_{k,u_j}\end{array}\right.$

Then to P _k+1in each point (x, y, z) ^tall pass through

$\left[\begin{array}{c}{x}^{\&prime;}\\ y^{\&prime;}\\ z^{\&prime;}\end{array}\right]=M_k\left[\begin{array}{c}x\\ y\\ z\end{array}\right]P$

Carry out coordinate transform, be transformed into coordinate in the coordinate system of k segmentation (x ', y ', z ') ^t.Select the 1st segmentation as unified aerostatics ship coordinate system, to the gauge point in k segmentation (x, y, z) ^t, can pass through

$\left[\begin{array}{c}{x}^{\&prime;}\\ y^{\&prime;}\\ z^{\&prime;}\end{array}\right]=M_1{\mathrm{<figure-callout\; id="2"\; label="M"\; filenames="HDA0000462318560000031.png"\; state="\{\{state\}\}">M</figure-callout>}}_2...M_k\left[\begin{array}{c}x\\ y\\ z\end{array}\right]$

Be converted to aerostatics coordinate system coordinate (x ', y ', z ') ^t;

(10) utilize the aerostatics coordinate (x ', y ', z ') of all gauge points ^t, by the three-dimensional model on standard triangle method for reconstructing reconstruct aerostatics utricule surface;

(11) utilize man-machine interactively to determine that aerostatics utricule puts the approximate range with utricule afterbody point foremost, as the H region in Fig. 1 and T region; Then within the scope of this, find two points of curvature maximum on three-dimensional surface, as the A in Fig. 2, B 2 points; The line of usining between these 2 is as the axis of aerostatics, as shown in the dotted line in Fig. 3.Using the length l of AL as aerostatics utricule; With enough little step-length △ d, axis is carried out to discrete sampling, obtain Ns axis sampled point; In each sample point, according to utricule Three-dimension Reconstruction Model, determine utricule surface and the intersecting lens of crossing the axis vertical plane of current axis sampled point, these intersecting lenses have just provided the border in utricule cross section, then on the border in utricule cross section, found the long-chord of current axis sampled point, as shown in Figure 4.Travel through after each axis sampled point, get the direction of the elder in the long-chord in all cross sections as Width, its length is as the width w of aerostatics utricule; Again travel through each axis sampled point, at every, locate to calculate utricule cross section and in the direction with Width quadrature, cross the chord length of axis sampled point, and get the elder in all this kind of strings, using its direction as short transverse, its length is as the height h of aerostatics utricule, as shown in Figure 5; At this, travel through all axis sampled points, obtain the utricule cross section of i sample point, then utilize enough little square net pair cross-section to carry out discretize, utilization drops on square net number in cross section as the approximate value A of area of section _i, finally with

volume as aerostatics utricule.

Claims (4)

1. an aerostatics utricule volume measuring method, is characterized in that, the method is:

1) arrange vision measurement system: described vision measurement system comprises a plurality of camera mounting brackets and a plurality of tripod, and a plurality of camera pan-tilts are installed on described camera mounting bracket, on described camera pan-tilt, video camera is installed; Laser range finder The Cloud Terrace is installed on described tripod, on described laser range finder The Cloud Terrace, laser range finder is installed; Described camera pan-tilt, video camera, laser range finder The Cloud Terrace, laser range finder all access main control computer; Described a plurality of camera mounting bracket and a plurality of tripod are all evenly distributed on for placing around the region of aerostatics utricule;

2) determine world coordinate system: select a laser range finder A, the world coordinates of definition A is p _{laser, 1}=(0,0,0); Alternative is selected a laser range finder B, measures the distance L between A, B, and the world coordinates of definition B is p _{laser, 2}=(L, 0,0); Select again a laser range finder C, measure the distance l between A, C ₁and the distance l between B, C ₂, the world coordinates of definition laser range finder C is p _{laser, 3}=(x _c3, y _c3, 0), wherein

$\begin{array}{c}{x}_{c3}=\frac{L^2+(l_1^2-l_2^2)}{2L}\\ y_{c3}=\frac{\sqrt{(L^4+l_1^4+l_2^4)-{(L^2-l_1^2)}^2-{(L^2-l_2^2)}^2-{(l_1^2-l_2^2)}^2}}{2L}\end{array};$

3), to all the other arbitrary laser range finder X, measure respectively above-mentioned laser range finder A, B, C to the distance l between laser range finder X _a, l _band l _c, obtain the world coordinates (x, y, z) of laser range finder X, wherein x, y, z meets system of equations:

$\left\{\begin{array}{c}{x}^2+y^2+z^2=l_A^2\\ {(x-L)}^2+y^2+z^2=l_B^2\\ {(x-x_{c3})}^2+{(y-y_{c3})}^2+z^2=l_C^2\end{array}\right.;$

The like, obtain the world coordinates of all the other all laser range finders;

4) video camera is taken to be arranged in and is placed the calibration point plate array that a plurality of calibration point plates in aerostatics utricule region form; Determine that i platform laser range finder is to the distance l between j calibration point plate _ij; By solving the least square solution of following system of equations, determine the world coordinates (x of j calibration point plate _j, y _j, z _j):

${(x_j-x_{\mathrm{laser},i})}^2+{(y_j-y_{\mathrm{laser},i})}^2+{(z_j-z_{\mathrm{laser},i})}^2=l_{\mathrm{ij}}^2;$

Wherein, i=1 ..., N _laser; J=1 ..., N _cali; N _laserfor laser range finder number of units; N _caliquantity for calibration point plate; (x _{laser, i}, y _{laser, i}, z _{laser, i}be the world coordinates p of i platform laser range finder _{laser, i};

5) repetition above-mentioned steps 4) 4～9 times, every video camera all obtains the image of 4～9 width calibration point plate arrays, thereby forms n _calin _camthe calibration maps image set of width calibration point plate array image construction, wherein n _cali=4～9, N _camfor the video camera number of units in vision measurement system; According to standard camera scaling method, utilize the world coordinates of described calibration maps image set and each calibration point plate, determine the parameter c of r platform video camera in measure field _r=(α _r, β _r, γ _r, t _x,r, t _y,r, t _z,r, f _r, κ _r, s _x,r, s _y,r, c _x,r, c _y,r), (α wherein _r, β _r, γ _r) represent that the space of r platform video camera in world coordinate system is towards angle, (t _x,r, t _y,r, t _z,r) coordinate of expression r platform video camera in world coordinate system, f _rthe main distance that represents r platform video camera, κ _rthe radial distortion factor that represents r platform video camera, (s _x,r, s _y,r) represent the scaling factor of r platform video camera, (c _x,r, c _y,r) represent the principal point coordinate of r platform video camera;

6) tested aerostatics utricule is placed in aerostatics utricule region, on the aerostatics utricule surface of unaerated, a plurality of visual indicia points are set, aerostatics utricule is inflated, deflection arch by aerostatics utricule starts, make aerostatics utricule by the shooting area of described vision measurement system, described in computer control, vision measurement system carries out segmentation shooting to described aerostatics utricule one side, and in two adjacent segmentations, has at least 3 identical visual indicia points to appear in photographic images corresponding to these two segmentations simultaneously; Complete after the shooting of aerostatics utricule one side, according to identical method, complete the shooting of aerostatics utricule opposite side;

7) coordinate of setting some segmentations is aerostatics utricule coordinate system, and the visual indicia point transformation in all the other segmentations, in described aerostatics utricule coordinate system, is obtained to the aerostatics utricule coordinate of all visual indicia points;

8) utilize the aerostatics utricule coordinate of all visual indicia points, by the three-dimensional model on standard triangle method for reconstructing reconstruct aerostatics utricule surface;

9) determine 2 points of three-dimensional surface curvature maximum on described three-dimensional model, the axis that the line of take between these 2 is aerostatics, the length l using the length of described axis as aerostatics utricule, carries out discrete sampling with step-length △ d to described axis, obtains N _sindividual axis sampled point; The value of △ d is no more than 10cm; In each axis sample point, according to the three-dimensional model on aerostatics utricule surface, determine utricule surface and the intersecting lens of crossing the axis vertical plane of current axis sampled point, thereby determine the border in utricule cross section, then on the border in utricule cross section, found the long-chord in cross section of current axis sampled point, travel through after each axis sampled point, using the direction of the string of length maximum in the long-chord in all cross sections as Width, and the width w using the length of the string of this length maximum as aerostatics utricule; Again travel through each axis sampled point, in each axis sample point, calculate utricule cross section and in the direction with described Width quadrature, cross the chord length of this axis sampled point, and get the chord length of length minimum in all chord lengths, using the direction of chord length of length minimum as short transverse, and the height h using the length of the chord length of length minimum as aerostatics utricule; Again travel through all axis sampled points, determine the utricule cross section of g axis sample point, then the rectangular node that utilizes the length of side to be no more than 10cm carries out discretize to the utricule cross section of g axis sample point, utilizes rectangular node number in the utricule cross section drop on g axis sample point as the approximate value A of area of section _g, the volume V of calculating aerostatics utricule: $V=\underset{g=1}{\overset{N_s}{\mathrm{\&Sigma;}}}{A}_g\&CenterDot;\mathrm{\&Delta;d}.$

2. aerostatics utricule volume measuring method according to claim 1, is characterized in that, in described step 5), has 5～10 identical visual indicia points to appear in photographic images corresponding to these two segmentations in two adjacent segmentations simultaneously.

3. aerostatics utricule volume measuring method according to claim 1 and 2, is characterized in that, in described step 8), △ d gets 1cm.

4. aerostatics utricule volume measuring method according to claim 1 and 2, is characterized in that, in described step 8), the length of side of described rectangular node is 1cm.

Images