CN106290256A - Quantitative background schlieren method based on video measuring - Google Patents

Abstract

The invention discloses a kind of quantitative background schlieren (BOS) method based on video measuring (VM), VM technology is combined with BOS light path, use uniform dot as the background dot of BOS, on the one hand circular markers image processing techniques ripe for VM is utilized, it is ensured that the spot placement accuracy of dot imaging reaches 0.02 pixel；On the other hand, the deviation displacement produced on background board through flow field from background dot is accurately calculated to the light beam at camera shooting center by collinearity equation, and provide accurate deflection angle computing formula, it is intended to accurately obtain deflection angle and the optical path difference of each background dot position, thus eliminate the error that existing BOS uses directional light deflection angle computing formula to produce, overcome image cross-correlation technique and existing BOS is surveyed quantitative limitation and the adverse effect to certainty of measurement thereof.

Description

Quantitative background schlieren method based on video measuring

Technical field

The present invention relates to FLOW VISUALIZATION and Pneumatic optical technical field of measurement and test, based on video measuring determine particularly to a kind of Amount background schlieren method.

Background technology

There is last 100 years as FLOW VISUALIZATION means in stration technique, but in actual applications, schlieren is the most only determined Property observe and analyze, seldom as the means of testing of quantification.This mainly has two reasons:

1) factor affecting schlieren gradation of image is a lot, such as the intensity of light source, glass transmission, test section thickness etc., causes The quantitative relationship being difficult to set up between image and measured physical quantity；

2) the first derivative field of schlieren image reflection refractive index, extracts difficulty data.

To this end, existing background schlieren (BOS) technology is on the basis of tradition schlieren principle, incorporate Particle Image Velocimetry The Particle-beam Tracing of (Particle Image Velocimetry, PIV) and particle picture treatment technology carry out flow field survey, and it is excellent Point is as follows:

1) as it is shown in figure 1, different from tradition stration technique, BOS quantifies deflection of light amount not by light intensity, but logical Cross PIV technology and ask for background speckle side-play amount on camera CCD, quantitatively obtain the given light deflection of light position through flow field Move, ambient light can be avoided or measure the flow field self-luminous impact on measurement result.

2) BOS is without using a large amount of optical instruments in traditional schlieren, such as the equipment such as diaphragm and lens, not only cost Low, and eliminate the restriction to measuring field scope of lens or reflecting mirror.

But, existing BOS technology still there is problems in that

1) the deflection angle computing formula of existing directional light is not suitable for the BOS light path of Fig. 1.Reason is: as in figure 2 it is shown, For the background speckle A of given position, from an A to the light of camera shooting center OOn CCD, imaging point is a, it is clear that Not parallel with z-axis, when the little deflection angle computing formula using directional light obtains Error is obvious, and especially the most remote when background speckle A distance z-axis, error is the biggest.

2) the light deflection displacement of certain set point position can not be accurately obtained: owing to using PIV technology to ask for background speckle In the deflection of light displacement of imaging plane, i.e. obtain the displacement of whole window by asking for the cross correlation of iterative query window Variable quantity, therefore, what it obtained is the average displacement amount of whole iterative window.

3) precision of existing BOS is affected big by the size of iterative query window: will if image cross-correlation window selection is too small Can not get correct side-play amount, and window selection is excessive, can reduce again the precision of side-play amount, especially when flowing change is violent, During such as light beam through supersonic speed shock wave, the background speckle deviation Displacement Saltation amplitude at shock wave is big, PIV image cross-correlation window technique Asking for side-play amount then can be because of image change acutely failure, poor robustness.

4) existing BOS use random image is as background, if background image occurs the structure with space periodicity, PIV algorithm will be unable to correct interpretation background displacement vector.

Summary of the invention

In order to overcome the disadvantages mentioned above of prior art, the present invention proposes a kind of quantitative background schlieren based on video measuring Method, combines (Videogrammetric measurement, VM) technology with BOS light path, uses uniform dot to make For the background dot of BOS, on the one hand utilize circular markers image processing techniques ripe for VM, it is ensured that the point location of dot imaging Precision reaches 0.02 pixel；On the other hand, accurately calculated from background dot to camera shooting center by collinearity equation The deviation displacement that the light beam at (i.e. entrance pupil center) produces on background board through flow field, and provide accurate deflection angle calculating public affairs Formula, it is intended to accurately obtain deflection angle and the optical path difference of each background dot position, thus eliminate existing BOS and use directional light deflection angle The error that computing formula produces, overcomes image cross-correlation technique and existing BOS is surveyed quantitative limitation and to certainty of measurement not Profit impact.

The technical solution adopted for the present invention to solve the technical problems is: a kind of quantitative background schlieren based on video measuring Method, comprises the steps:

Step one, build BOS system: background board is vertical with the Z axis of coordinate system O-XYZ, background board is arranged between ranks Every equal round dot；

Step 2, utilizing the uniform round dot coordinate of ranks on background board, use video measuring scaling method, calibration for cameras exists Location parameter (the X of coordinate system O-XYZ_s,Y_s,Z_s) and attitude parameter (φ, ω, κ), and camera intrinsic parameter；

Step 3, when having flowing, with camera shooting background plate sequential chart picture；

The deviation displacement of set point A, deflection angle, light path on the background board image of the t shooting that step 4, calculating give Difference:

If the picture point that is angled to that A is on the camera CCD of t is a',It is A' with the intersection point of background board,

1) utilizing collinearity equation to calculate A' (X, Z) coordinate figure under coordinate system O-XYZ, recycling collinearity equation is further Calculate A' X under wind tunnel axis system and Y-coordinate value；

2) the deviation displacement of calculating A point:

If A is the round dot of the i-th row jth row on background board, the deviation displacement of its X Yu Y direction is designated as Δ X respectively^(i,j) With Δ Y^(i,j), then:

ΔX^(i,j)The X-coordinate value of the X-coordinate value-A of=A'；

ΔY^(i,j)The Y-coordinate value of the Y-coordinate value-A of=A'；

3) deflection angle of calculating A point:

The deflection angle of A point X with Y direction is designated as respectivelyWithThen:

${\mathrm{\ϵ}}_x^{(i,j)}=\arctan \left(\frac{{\mathrm{\ΔX}}^{(i,j)}}{b+b{\tan }^2{\mathrm{\γ}}_x^{(i,j)}+{\mathrm{\ΔX}}^{(i,j)}{\mathrm{tan\γ}}_x^{(i,j)}}\right);$

${\mathrm{\ϵ}}_y^{(i,j)}=\arctan \left(\frac{{\mathrm{\ΔY}}^{(i,j)}}{b+b{\tan }^2{\mathrm{\γ}}_y^{(i,j)}+{\mathrm{\ΔY}}^{(i,j)}{\mathrm{tan\γ}}_y^{(i,j)}}\right);$

In formula, b is the distance between center, flow field and background board,

${\mathrm{\γ}}_{x^{(i,j)}}=\arctan \left(\frac{x^{(i,j)}}{f}\right);$

${\mathrm{\γ}}_{y^{(i,j)}}=\arctan \left(\frac{y^{(i,j)}}{f}\right);$

In formula, x^(i,j)And y^(i,j)For X under coordinate system O-XYZ of the imaging round dot a of A point and Y-coordinate value；

4) optical path difference of calculating A point:

Compared with prior art, the positive effect of the present invention is: different from existing BOS method, and video is surveyed by the present invention Amount (VM) technology combines with BOS light path, uses uniform dot as the background dot of BOS, obtains following innovative point:

1) deflection angle of non-parallel light and the precise calculation of optical path difference are given；

2) background round dot diameter based on BOS light path parameter and spacing parameter computing formula are established, to guarantee dot The spot placement accuracy of imaging reaches 0.02 pixel；

3) deviation displacement quantitative scaling formula is established, in order to by changing the background board distance to center, flow field, scaling Light deflection displacement at round dot on background board, adjusts the sensitivity of quantitative background schlieren based on video measuring.

Therefore, can accurately calculate the light beam from each background dot to camera shooting center (i.e. entrance pupil center) to wear Cross deviation displacement, deflection angle and optical path difference that flow field produces on background board, it is achieved the quantitative measurement of background schlieren.

The present invention proposes BOS based on VM, and first give the deflection angle of non-parallel light and optical path difference accurately calculates public affairs Formula, eliminates the error that existing BOS uses directional light deflection angle computing formula to produce, and can accurately calculate from each background little Round dot sets out deviation displacement that the light beam at camera shooting center (i.e. entrance pupil center) produces through flow field on background board, deviation Angle and optical path difference, it is achieved the accurate measurement of background schlieren.

Secondly, based on the deviation displacement quantitative scaling formula set up, by changing the distance between center, flow field and background board, Adjust the sensitivity of quantitative background schlieren based on video measuring, to guarantee to record deviation displacement field and the deviation of preferable signal to noise ratio Angle field data.

Finally, present invention background based on BOS light path parameter round dot diameter and spacing parameter computational methods, in conjunction with maturation Video measuring technology, uses uniform dot as the background dot of BOS, accurately calculates each background dot with collinearity equation Deviation displacement, in conjunction with the present invention deviation displacement quantitative scale formula, promote BOS robustness, overcome image cross-correlation Existing BOS is surveyed quantitative limitation and the adverse effect to certainty of measurement thereof by technology, even if when flowing change is violent, as light beam is worn When crossing supersonic speed shock wave, remain to effectively obtain high-precision measurement result.

Accompanying drawing explanation

Examples of the present invention will be described by way of reference to the accompanying drawings, wherein:

Fig. 1 is BOS schematic diagram；

Fig. 2 is deflection angle schematic diagram calculation；

Fig. 3 is background board round dot parameter determination schematic diagram.

Detailed description of the invention

A kind of quantitative background schlieren method based on video measuring, comprises the steps:

1, principle builds BOS system as shown in Figure 1, it is desirable to background board is vertical with the Z axis shown in Fig. 2, and background board is arranged Ranks interval is had to be the dot of S, a diameter of Φ of dot_D.On background board, the method for designing of given round dot A is as follows:

1) as it is shown on figure 3, put the diameter of A

Φ_D=u Φ_d/f (1)

Φ in formula_dFor the diameter of imaging round dot a on camera of the round dot on background board, it is desirable to Φ_dMore than or equal to 6 pixels, To guarantee that the spot placement accuracy of dot imaging is up to 0.02 pixel.U is the background board distance to camera shooting center O, and f is burnt Away from.

2) the ranks interval of round dot

S=us/f (2)

In formula, s is the spacing distance on camera between imaging round dot of the round dot on background board, typically takes s equal to 1.5 to 2 times Φ_d。

3) the round dot sharp on background board is required, such as: white with black round dot or white background bullet.

2, Fig. 2 is the three-dimensional eulerian coordinate system O-XYZ of standard geometrical relationship on YOZ perspective plane, utilizes on background board The circular dot coordinate that ranks are uniform, uses ripe video measuring scaling method (as camera calibration based on plane template is calculated Method), calibration for cameras is at the location parameter (X of coordinate system O-XYZ as shown in Figure 2_s,Y_s,Z_s) and attitude parameter (φ, ω, κ), and Camera intrinsic parameter.

When 3, having flowing, with camera shooting background plate sequential chart picture.On the background board image of wherein given t shooting The deviation displacement of set point A, deflection angle, optical path difference (OPD) computational methods are as follows:

As in figure 2 it is shown, owing to flowing causes lightThere is deviation, if A is angled to picture on the camera CCD of t Point is a',It is that A'(i.e. A is because producing light deflection phenomenon deviation position on background board with the intersection point of background board)

1) collinearity equation is utilized to calculate A' (X, Z) coordinate figure under coordinate system shown in Fig. 2.Collinearity equation describes VM phase Round dot and the mathematical model of picture point three thereof on machine, examination background board, expression formula is

$\left\{\begin{array}{c}x+f\frac{a_1(X-X_s)+b_1(Y-Y_s)+c_1(Z-Z_s)}{a_3(X-X_S)+b_3(Y-Y_s)+c_3(Z-Z_s)}=x_0\\ y+f\frac{a_2(X-X_s)+b_2(Y-Y_s)+c_2(Z-Z_s)}{a_3(X-X_S)+b_3(Y-Y_s)+c_3(Z-Z_s)}=y_0\end{array}\right.---\left(3\right)$

(x in formula₀,y₀) it is respectively camera image plane center, (x, y) is the image plane coordinate of tested point, and (X, Y, Z) is for treating Measuring point three-dimensional coordinate under coordinate system shown in Fig. 2, (a₁,a₂,a₃,b₁,b₂,b₃,c₁,c₂,c₃) be camera attitude angle (φ, ω, 9 direction cosines in spin matrix R κ) formed.

Pose parameter (X due to camera_s,Y_s,Z_s) and the Z seat of (φ, ω, κ), the image plane coordinate of some a' and some A' Scale value (being background board vertical installation position on Z axis) is it is known that can be obtained A' X under wind tunnel axis system by formula (3) With Y-coordinate value.

2) the deviation displacement of A point is calculated.If A is the circular dot of the i-th row jth row on background board, by its X and Y direction Deviation displacement is designated as Δ X respectively^(i,j)With Δ Y^(i,j), then calculating formula is as follows

ΔX^(i,j)The X-coordinate value (4) of the X-coordinate value-A of=A'

ΔY^(i,j)The Y-coordinate value (5) of the Y-coordinate value-A of=A'

3) deflection angle of A point is calculated.X is as follows with the deflection angle computing formula of Y direction:

${\mathrm{\ϵ}}_x^{(i,j)}=\mathrm{arc}tan\left(\frac{{\mathrm{\ΔX}}^{(i,j)}}{b+b{\tan }^2{\mathrm{\γ}}_x^{(i,j)}+{\mathrm{\ΔX}}^{(i,j)}{\mathrm{tan\γ}}_x^{(i,j)}}\right)---\left(7\right)$

${\mathrm{\ϵ}}_y^{(i,j)}=\mathrm{arc}tan\left(\frac{{\mathrm{\ΔY}}^{(i,j)}}{b+b{\tan }^2{\mathrm{\γ}}_y^{(i,j)}+{\mathrm{\ΔY}}^{(i,j)}{\mathrm{tan\γ}}_y^{(i,j)}}\right)---\left(8\right)$

In formula, b is the distance (shown in Fig. 2) between center, flow field and background board, and

${\mathrm{\γ}}_{x^{(i,j)}}=\mathrm{arc}tan\left(\frac{x^{(i,j)}}{f}\right)---\left(9\right)$

${\mathrm{\γ}}_{y^{(i,j)}}=\mathrm{arc}tan\left(\frac{y^{(i,j)}}{f}\right)---\left(10\right)$

X in formula^(i,j)And y^(i,j)For X under coordinate system O-XYZ shown in Fig. 2 of the imaging round dot a of A point and Y-coordinate value.

4) optical path difference of A point is calculated.Calculating formula is as follows:

4, by adjusting distance b between center, flow field and background board, round dot deviation displacement on scaling background board, in order to adjust The sensitivity of quantitative background schlieren based on video measuring.Set point A on the background board image of wherein given t shooting The calculating formula with b scaling of deviation displacement is as follows:

${\mathrm{\ΔX}}^{(i,j)}=b\frac{(1+{\tan }^2{\mathrm{\γ}}_{x^{(i,j)}})}{(1-{\mathrm{tan\ϵ}}_x^{(i,j)}{\mathrm{tan\γ}}_{x^{(i,j)}})}{\mathrm{tan\ϵ}}_x^{(i,j)}---\left(12\right)$

${\mathrm{\ΔY}}^{(i,j)}=b\frac{(1+{\tan }^2{\mathrm{\γ}}_{y^{(i,j)}})}{(1-{\mathrm{tan\ϵ}}_y^{(i,j)}{\mathrm{tan\γ}}_{y^{(i,j)}})}{\mathrm{tan\ϵ}}_y^{(i,j)}---\left(13\right)$

Claims (7)

1. a quantitative background schlieren method based on video measuring, it is characterised in that: comprise the steps:

Step one, build BOS system: background board is vertical with the Z axis of coordinate system O-XYZ, background board is arranged ranks interval phase Deng round dot；

Step 2, utilizing the uniform round dot coordinate of ranks on background board, use video measuring scaling method, calibration for cameras is at coordinate It is the location parameter (X of O-XYZ_s,Y_s,Z_s) and attitude parameter (φ, ω, κ), and camera intrinsic parameter；

Step 3, when having flowing, with camera shooting background plate sequential chart picture；

The deviation displacement of set point A, deflection angle, optical path difference on the background board image of the t shooting that step 4, calculating give:

If the picture point that is angled to that A is on the camera CCD of t is a',It is A' with the intersection point of background board,

1) utilizing collinearity equation to calculate A' (X, Z) coordinate figure under coordinate system O-XYZ, recycling collinearity equation calculates further Go out A' X under wind tunnel axis system and Y-coordinate value；

2) the deviation displacement of calculating A point:

If A is the round dot of the i-th row jth row on background board, the deviation displacement of its X Yu Y direction is designated as Δ X respectively^(i,j)With Δ Y^(i,j), then:

ΔX^(i,j)The X-coordinate value of the X-coordinate value-A of=A'；

ΔY^(i,j)The Y-coordinate value of the Y-coordinate value-A of=A'；

3) deflection angle of calculating A point:

The deflection angle of A point X with Y direction is designated as respectivelyWithThen:

${\mathrm{\ϵ}}_x^{(i,j)}=\arctan \left(\frac{{\mathrm{\ΔX}}^{(i,j)}}{b+b{\tan }^2{\mathrm{\γ}}_x^{\left(i,j\right)}+{\mathrm{\ΔX}}^{(i,j)}{\mathrm{tan\γ}}_x^{(i,j)}}\right);$

${\mathrm{\ϵ}}_y^{\left(i,j\right)}=\arctan \left(\frac{{\mathrm{\ΔY}}^{\left(i,j\right)}}{b+b{\tan }^2{\mathrm{\γ}}_y^{\left(i,j\right)}+{\mathrm{\ΔY}}^{\left(i,j\right)}{\mathrm{tan\γ}}_y^{\left(i,j\right)}}\right);$

In formula, b is the distance between center, flow field and background board,

${\mathrm{\γ}}_{x^{\left(i,j\right)}}=\arctan \left(\frac{x^{(i,j)}}{f}\right);$

${\mathrm{\γ}}_{y^{\left(i,j\right)}}=\arctan \left(\frac{y^{(i,j)}}{f}\right);$

In formula, x^(i,j)And y^(i,j)For X under coordinate system O-XYZ of the imaging round dot a of A point and Y-coordinate value；

4) optical path difference of calculating A point:

Quantitative background schlieren method based on video measuring the most according to claim 1, it is characterised in that: described round dot A diameter of Φ_DDetermine as follows:

Φ_D=u Φ_d/f

In formula, Φ_dFor the diameter of imaging round dot a on camera of the round dot on background board, u is background board to camera shooting center O Distance, f is focal length.

Quantitative background schlieren method based on video measuring the most according to claim 2, it is characterised in that: described Φ_dIt is more than Equal to 6 pixels.

Quantitative background schlieren method based on video measuring the most according to claim 2, it is characterised in that: described round dot Ranks interval S determines as follows:

S=us/f

In formula, s is the spacing distance on camera between imaging round dot of the round dot on background board.

Quantitative background schlieren method based on video measuring the most according to claim 4, it is characterised in that: s arrives equal to 1.5 The Φ of 2 times_d。

Quantitative background schlieren method based on video measuring the most according to claim 1, it is characterised in that: as follows Adjust distance b between center, flow field and background board, round dot deviation displacement on background board zoomed in and out:

${\mathrm{\ΔX}}^{(i,j)}=b\frac{(1+{\tan }^2{\mathrm{\γ}}_{x^{\left(i,j\right)}})}{(1-{\mathrm{tan\ϵ}}_x^{(i,j)}{\mathrm{tan\γ}}_{x^{\left(i,j\right)}})}{\mathrm{tan\ϵ}}_x^{(i,j)};$

${\mathrm{\ΔY}}^{(i,j)}=b\frac{(1+{\tan }^2{\mathrm{\γ}}_{y^{\left(i,j\right)}})}{(1-{\mathrm{tan\ϵ}}_y^{(i,j)}{\mathrm{tan\γ}}_{y^{\left(i,j\right)}})}{\mathrm{tan\ϵ}}_y^{(i,j)}.$

Quantitative background schlieren method based on video measuring the most according to claim 1, it is characterised in that: described round dot is White with black round dot or white background bullet.