CN103293333B - A kind of flowing two-dimension speed field measurement method of staggered scanning CCD and device - Google Patents

Abstract

The invention discloses flowing two-dimension speed field measurement method and the device of a kind of staggered scanning CCD, the present invention implants trace particle in flowing object to be measured (gas or liquid), trace particle can be followed fluid due to viscosity and together be moved, this measurement mechanism adopts the gas laser (as He-Ne laser instrument) of continuous luminous as light source and common staggered scanning ccd sensor as image detector, transient state captures the flow image that a frame contains trace particle, utilize filtering afterwards, binaryzation processes flow image, be about to an interleaved two field picture according to odevity again and be separated into two width images, cross-correlation technique is adopted to carry out computing, flowing two-dimension speed field can be obtained.When the synchronous device of the present invention without the need to powerful laser instrument, high-resolution CCD and complexity, the velocity survey of the PIV whole audience can be realized; Equipment is simple, cheap, is applicable to different occasion Fluid field and measures.

Description

A kind of flowing two-dimension speed field measurement method of staggered scanning CCD and device

Technical field

The invention belongs to flow field survey technical field, particularly relate to the measuring technique of the two-dimentional instantaneous flow velocity field of staggered scanning ccd sensor.

Background technology

For the measurement of two dimensional surface velocity field, mainly contain Particle-beam Tracing tachometric method (Particle Tracer Velocimetry is called for short PTV) and particle image ve locity method (Particle Image Velocimetry is called for short PIV).PTV/PIV is the very important instrument of one of the Study of Fluid mechanics that the eighties in 20th century grows up, it is by scattering followability and the good trace particle of reflective in flow field, the tangent plane region of institute's flow measurement field is irradiated with sheet laser, record by imaging system is adopted to obtain continuously twice or the particle picture of multiexposure, multiple exposure, recycling image cross-correlation method obtains the average displacement of particle picture in each zonule, can obtain the velocity distribution of the whole audience.PIV technology overcomes the limitation of Single-point velocity determination technology in the past, achieve stream field transient state, the whole audience, noncontact, glitch-free measurement, just be used widely in very short time after appearance, become most important surveying instrument in fluid mechanics and thermal conduction study research, be widely used in the every field relevant to flow measurement, greatly promote the scientific development of the association areas such as fluid mechanics, thermal conduction study, Combustion.

But current commercial PIV instrument and equipment is complicated, expensive.Business PIV instrument will reach 20-30 ten thousand U.S. dollar in international market price, significantly limit the application and development of PIV measuring instrument.Therefore, research and development equipment is simple, reasonable price, and the measurement range flow field survey instrument that can meet again in most cases is also one of direction of PIV technical development simultaneously.

Summary of the invention

Goal of the invention: for above-mentioned existing Problems existing and deficiency, the invention provides flowing two-dimension speed field measurement method and the device of a kind of staggered scanning CCD.Compare existing commercial PIV instrument, the inventive method and dress only need utilize the gas laser of low-power, small-sized continuous luminous as light source, and become pulsed light without the need to modulation, utilize the image of staggered scanning ccd video camera record fluid, therefore utilize the velocity field of a two field picture measurable flow body, have that equipment is simple, low cost and other advantages.

Technical scheme: for achieving the above object, the present invention by the following technical solutions: a kind of flowing two-dimension speed field measurement method of staggered scanning CCD, comprises the following steps:

A, treating to add trace particle in fluid measured, laser emitting light forms fan-shaped light sheets under the effect of cylindrical mirror; The each self-forming scatter spot of trace particle under the irradiation of sheet light source, and the flow image g (x utilizing staggered scanning ccd sensor to obtain a frame to contain trace particle, y), comprise two field picture, utilize image pick-up card collection and be stored in computing machine;

Particle flow image g (x, y) that b, image pick-up card collect is the image of degenerating, and carries out filtering Recovery processing by formula (1):

F(x,y)＝g(x,y)*h _w(x,y) (1)

Wherein, F (x, y) is the image after filtering after Recovery processing, h _w(x, y) is the point spread function recovered based on Wiener filtering image;

C, binary conversion treatment is carried out to the gray scale of the particle picture of Recovery processing after filtering, is divided into two groups:

$f^{*}(x,y)=\left\{\begin{array}{cc}0,& F(x,y)\∈{[0,L}_1)\∪(L_2,256]\\ 1,& F(x,y)\∈[L_1,L_2]\end{array}\right.---\left(2\right)$

Wherein, L ₁, L ₂for the thresholding of setting;

D, by the f of binary conversion treatment ^*(x, y) one two field picture become parity field image according to the law separation that parity rows combines, respectively using odd even image as a two field picture, therefore can obtain two two field pictures with

E, two two field pictures obtained in parity field with on basis, by formula (3) to two two field pictures with carry out cross-correlation calculation,

$C_{\mathrm{ff}*}=\frac{L}{\sqrt{\mathrm{nm}}}---\left(3\right)$

Wherein, C _ff* be related coefficient, L is with being respectively diagnostic window overlay elements in two two field picture gray matrixs is all the number of 1, n and m is respectively with in matrix, element is the number of 1; According to the displacement (d finding population in related coefficient maximal value way determination diagnostic window in certain area _x, d _y);

F, at the displacement (d obtaining population in diagnostic window _x, d _y) basis on, calculate two-dimension speed field by formula (4) and (5),

$v_x=\frac{d_x}{\mathrm{\τ}}---\left(4\right)$

$v_y=\frac{d_y}{\mathrm{\τ}}---\left(5\right)$

Wherein, τ is the time interval between two field picture.

Further, the described point spread function h recovered based on Wiener filtering image _wthe transfer function H of (x, y) _w(u, v) is calculated by formula (6),

$H_w(u,v)=\frac{H^{*}(u,v)}{{\left|H(u,v)\right|}^2+S_n(u,v)/S_f(u,v)}---\left(6\right)$

Wherein, H (u, v) is because the relative motion between camera and object causes the transport function of image blurring point spread function h (x, y); H* (u, v) is H (u, v) complex conjugate; S _n(u, v) and S _f(u, v) is respectively the power spectrum of noise and picture signal; U and v is respectively x, spatial frequency on y direction;

Then to H _w(u, v) carries out inversefouriertransform, obtains because the relative motion between camera and object causes image blurring point spread function h _w(x, y).

Further, for image due to the relative motion between camera and object cause fuzzy, the transfer function H (u, v) of imaging system is obtained by formula (7),

$H(u,v)=\frac{\sin \mathrm{\π}(\mathrm{\αu}+\mathrm{\βv})T}{\mathrm{\π}(\mathrm{\αu}+\mathrm{\βv})T}{e}^{-\mathrm{j\π}(\mathrm{\αu}+\mathrm{\βv})T}---\left(7\right)$

In formula, T is the time shutter, and α, β are that constant represents movement velocity component in the x and y direction, and the mean value of α and β utilizes the flow of fluids within pipes and pipeline section to amass estimation and obtains.

Further, the power spectrum S of described noise _f(u, v) is calculated by formula (8):

S _f(u，v)＝S _g(u，v)/|H(u，v)| ²(8)

In formula (8), S _gthe power spectrum that (u, v) is degraded image; S _f(u, v) is the power spectrum of original undegraded moving image; For noise normally white noise, its power is by the autocorrelation function of calculating one amplitude and noise acoustic image n (x, y), then rated output spectrum obtains S _n(u, v).

Further, described threshold value L ₁and L ₂obtain by the following method: add up acquisition, L by the gray-scale value side figure of the actual particle picture recorded ₀represent the gray-scale value that the gray shade scale frequency is maximum, P (L ₀) be the maximum frequency occurred, in computing, 0.1*P (L is less than for gray-scale value ₀) point unified be labeled as uninterested point, be 0; Otherwise be labeled as 1.

Further, described laser instrument adopts the gas laser of continuous luminous.

Beneficial effect: compared with prior art, the present invention has the following advantages: when synchronous device without the need to powerful laser instrument, high-resolution CCD and complexity, can realize the velocity survey of the PIV whole audience; Equipment is simple, cheap, is applicable to different occasion Fluid field and measures.

Accompanying drawing explanation

Fig. 1 is the structural representation of the flowing two-dimension speed field measurement device of staggered scanning CCD of the present invention;

Fig. 2 is the schematic flow sheet of measuring method of the present invention.

Wherein, laser instrument 1, lens combination 2, staggered scanning CCD3, computing machine 4, cylindrical mirror 5.

Embodiment

Below in conjunction with the drawings and specific embodiments, illustrate the present invention further.Should understand these embodiments to be only not used in for illustration of the present invention and to limit the scope of the invention, after having read the present invention, the amendment of those skilled in the art to the various equivalent form of value of the present invention has all fallen within the application's claims limited range.

Principle of work of the present invention: to treating to plant trace particle in fluid measured, trace particle is subject to laser light sheet and irradiates, and staggered scanning CCD3, under the control of software for calculation, obtains flow field figure picture by image pick-up card collection, and is stored in computing machine 4; In computing machine 4, first data processing software carries out filtering, binary conversion treatment to the image obtained, and a two field picture is separated into two field picture, carries out relevant treatment, can obtain and treat the field distribution of fluid measured two-dimension speed isolated two field picture.

As shown in Figure 1, the flowing two-dimension speed field measurement device based on staggered scanning CCD3 sensor of the present invention mainly comprises light source, lens combination 2, staggered scanning CCD3, image pick-up card, computing machine 4.Light source can adopt continuous wave laser 1 (as He-Ne laser instrument 1) or pulsed laser 1, and laser instrument 1 exit end is connected with cylindrical mirror 5 by optical bracket.Lens combination 2 comprise two ordinary optical camera lenses (magnification 0.501 of lens combination 2, focal length 225mm, f number 5.9, design parameter can experimentally demand select), be connected with staggered scanning CCD3 by C bayonet socket.Staggered scanning CCD3 is connected with Image Data Acquisition Card by signal cable, and data collecting card is embedded in computing machine 4.Under image data processing system in computing machine 4 controls, the total tunes such as image acquisition, storage and process can be completed.

As shown in Figure 2, below the course of work of the present invention and data calculating are illustrated:

Treat to add trace particle in fluid measured, laser instrument 1 emergent light forms fan-shaped light sheets under the effect of cylindrical mirror 5, under the irradiation of sheet light source, the scatter spot that each self-forming of trace particle one is little, staggered scanning CCD3 sensor is utilized to obtain a frame moving image g (x, y), comprise two field picture, utilize image pick-up card collection and be stored in computing machine 4.

Frame particle picture g (x, y) collected from image card is a secondary image of degenerating, and need carry out filtering Recovery processing, can process according to following formula it:

f ^*(x，y)＝g(x，y)*h _w(x，y) (1)

Wherein, f ^*(x, y) is the image after Recovery processing, h _w(x, y) is the point spread function recovered based on Wiener filtering image.

To the particle picture f recovered after filtering ^*the gray scale of (x, y) carries out binary conversion treatment, is divided into two groups:

$f^{*}(x,y)=\left\{\begin{array}{cc}0,& f^{*}(x,y)\∈[0,L_1]\∪[L_2,256]\\ 1,& f^{*}(x,y)\∈[L_1,L_2]\end{array}\right.---\left(2\right)$

Wherein L ₁, L ₂it is the thresholding by certain standard setting.Through the f of binary conversion treatment ^*(x, y), the matrix be made up of 1 and 0.

By the f of binary conversion treatment ^*(x, y) one two field picture become parity field image according to the law separation that parity rows combines, respectively using odd even image as a two field picture, therefore can obtain two two field pictures with

At two two field pictures that parity field obtains with on basis, to two two field pictures with carry out cross-correlation calculation, computing method are:

$C_{\mathrm{ff}*}=\frac{L}{\sqrt{\mathrm{nm}}}---\left(3\right)$

Wherein, C _ff* be related coefficient, L is with being respectively diagnostic window overlay elements in two two field picture gray matrixs is all the number of 1, n and m is respectively with in matrix, element is the number of 1.From correlation theory, two identical image correlation coefficients are maximum, so the way that can be worth greatly according to the amount finding related coefficient in certain area determines the displacement (d of population in diagnostic window _x, d _y).

Obtain the displacement (d of population in diagnostic window _x, d _y), in conjunction with the time interval τ between two field picture, two-dimension speed field through type (4) and formula (5) can calculate and obtain:

$v_x=\frac{d_x}{\mathrm{\τ}}---\left(4\right)$

$v_y=\frac{d_y}{\mathrm{\τ}}---\left(5\right)$

The described point spread function h recovered based on Wiener filtering image _wthe transfer function H of (x, y) _w(u, v) can be expressed as:

$H_w(u,v)=\frac{H^{*}(u,v)}{{\left|H(u,v)\right|}^2+S_n(u,v)/S_f(u,v)}---\left(6\right)$

In formula Chinese style, H (u, v) is that H* (u, v) is H (u, v) complex conjugate, S because the relative motion between camera and object causes the transport function of image blurring point spread function h (x, y) _n(u, v) and S _f(u, v) is respectively the power spectrum of noise and picture signal, u and v is respectively x, spatial frequency on y direction.Obtain H _wafter (u, v), by inverse-Fourier transform, h can be obtained _w(x, y).

For fuzzy due to the relative motion between camera and object of image, the transfer function H (u, v) of imaging system can be expressed as:

$H(u,v)=\frac{\sin \mathrm{\π}(\mathrm{\αu}+\mathrm{\βv})T}{\mathrm{\π}(\mathrm{\αu}+\mathrm{\βv})T}{e}^{-\mathrm{j\π}(\mathrm{\αu}+\mathrm{\βv})T}---\left(7\right)$

In formula, T is the time shutter, and α, β are that constant represents movement velocity component in the x and y direction, can calculate in an estimated value substitution formula.

S _f(u, v) can by the power spectrum S of degraded image and original image _g(u, v) and S _fmeet between (u, v):

s _f(u，v)＝s _g(u，v)/|H(u，v)| ²(8)

Degenerate Graphs is known, can estimate that its average correlation function obtains power spectrum S _g(u, v).

For noise normally white noise, its power spectrum by the autocorrelation function of the secondary noise image n (x, y) of calculating one, can obtain S in rated output spectrum _n(u, v).

Described threshold value L ₁and L ₂acquisition methods: can by the grey level histogram of the actual particle image recorded, statistics obtains.L ₀represent the gray-scale value that the gray shade scale frequency is maximum, P (L ₀) the maximum frequency that occurs, in computing, 0.1*P (L is less than for gray-scale value ₀) point unified be labeled as uninterested point, be 0; Otherwise be labeled as 1.

Claims (6)

1. a flowing two-dimension speed field measurement method of staggered scanning CCD, comprises the following steps:

A, treating to add trace particle in fluid measured, laser emitting light forms fan-shaped light sheets under the effect of cylindrical mirror; The each self-forming scatter spot of trace particle under the irradiation of sheet light source, and the flow image g (x utilizing staggered scanning ccd sensor to obtain a frame to contain trace particle, y), comprise two field picture, utilize image pick-up card collection and be stored in computing machine;

Particle flow image g (x, y) that b, image pick-up card collect is the image of degenerating, and carries out filtering Recovery processing by formula (1):

F(x,y)＝g(x,y)*h _w(x,y) (1)

Wherein, F (x, y) is the image after filtering after Recovery processing, h _w(x, y) is the point spread function recovered based on Wiener filtering image;

C, binary conversion treatment is carried out to the gray scale of the particle picture of Recovery processing after filtering, is divided into two groups:

Wherein, L ₁, L ₂for the threshold value of setting;

D, by the f of binary conversion treatment ^*(x, y) one two field picture become parity field image according to the law separation that parity rows combines, respectively using odd even image as a two field picture, therefore can obtain two two field picture f ₁ ^*(x, y) and f ₂ ^*(x, y);

E, the two two field picture f obtained in parity field ₁ ^*(x, y) and f ₂ ^*on (x, y) basis, by formula (3) to two two field picture f ₁ ^*(x, y) and f ₂ ^*(x, y) carries out cross-correlation calculation,

Wherein, C _ff* be related coefficient, L is f ₁ ^*(x, y) and f ₂ ^*(x, y) is respectively diagnostic window overlay elements in two two field picture gray matrixs is all the number of 1, n and m is respectively f ₁ ^*(x, y) and f ₂ ^*in (x, y) matrix, element is the number of 1; According to the displacement (d finding population in related coefficient maximal value way determination diagnostic window in certain area _x, d _y);

F, at the displacement (d obtaining population in diagnostic window _x, d _y) basis on, calculate two-dimension speed field by formula (4) and (5),

Wherein, τ is the time interval between two field picture.

2. method according to claim 1, is characterized in that: the described point spread function h recovered based on Wiener filtering image _wthe transfer function H of (x, y) _w(u, v) is calculated by formula (6),

Wherein, H (u, v) is because the relative motion between camera and object causes the transport function of image blurring point spread function h (x, y); H* (u, v) is H (u, v) complex conjugate; S _n(u, v) and S _f(u, v) is respectively the power spectrum of noise and picture signal; U and v is respectively x, spatial frequency on y direction;

Then to H _w(u, v) carries out inversefouriertransform, obtains because the relative motion between camera and object causes image blurring point spread function h _w(x, y).

3. method according to claim 2, is characterized in that: for image due to the relative motion between camera and object cause fuzzy, the transfer function H (u, v) of imaging system is obtained by formula (7),

In formula, T is the time shutter, and α, β are that constant represents movement velocity component in the x and y direction, and the mean value of α and β utilizes the flow of fluids within pipes and pipeline section to amass estimation and obtains.

4. method according to claim 3, is characterized in that: the power spectrum S of described noise _f(u, v) is calculated by formula (8):

In formula (8), S _gthe power spectrum that (u, v) is degraded image; S _f(u, v) is the power spectrum of original undegraded moving image; For noise normally white noise, its power is by the autocorrelation function of calculating one amplitude and noise acoustic image n (x, y), then rated output spectrum obtains S _n(u, v).

5. method according to claim 1, is characterized in that: described threshold value L ₁and L ₂obtain by the following method: add up acquisition, L by the gray-scale value side figure of the actual particle picture recorded ₀represent the gray-scale value that the gray shade scale frequency is maximum, P (L ₀) be the maximum frequency occurred, in computing, 0.1*P (L is less than for gray-scale value ₀) point unified be labeled as uninterested point, be 0; Otherwise be labeled as 1.

6. method according to claim 1, is characterized in that: described laser instrument adopts the gas laser of continuous luminous.