CN103267567A - Measuring device and method for vibration of flexible cantilever on basis of machine vision - Google Patents

Abstract

The invention discloses a measuring device and method for the vibration of a flexible cantilever on the basis of machine vision. The measuring device comprises a fixed bracket, the flexible cantilever, a camera bracket, a CCD (Charge Coupled Device) camera, a lens, a plurality of LED (Light-Emitting Diode) light-emitting tubes and a PC (Personal Computer), wherein one end of the flexible cantilever is fixed on the fixed bracket; all the LED light-emitting tubes are arranged on the upper surface of the flexible cantilever in sequence; the top of the fixed bracket is provided with the camera bracket; the CCD camera is fixed on the camera bracket and is provided with the lens; and an output port of the CCD camera is connected with the PC. The CCD camera measures each frame image of the vibration of the LED light-emitting tubes, and conveys the images to the PC; and the PC processes a detected image sequence, extracts the position of the mass center of the light spots of the LED light-emitting tubes and acquires the vibration displacement of all the LED light-emitting tubes and parameters for reflecting the low-frequency vibration of the flexible cantilever structure. The measuring device and method disclosed by the invention has the advantages that the non-contact effect is achieved, the measuring range is wide and the vibration characteristic of a measured object is not changed and the like, and can be widely applied.

Description

Measurement mechanism and method based on the flexible cantilever beam vibration of machine vision

Technical field

The present invention relates to the measuring method of large-scale flexible semi-girder vibration, particularly a kind of measurement mechanism and method of the flexible cantilever beam vibration based on machine vision.

Background technology

Solar energy sailboard is a kind of collection solar device, is generally used for the energy supply of satellite, spaceship.The modal damping of this class large-scale flexible structure is little, in case be subjected to the effect of certain exciting force, it significantly vibrates last very long.This not only can influence the work of space structure, also will make structure produce too early fatigure failure, influences the serviceable life of structure, or causes the damage of instrument in the structure.Therefore need to measure and suppress the vibration of this type of spacecraft flexible structure, particularly at the structural vibration of the flexible cantilever beam of analog solar windsurfing.

In the prior art, the vibration survey of the flexible cantilever beam of analog solar windsurfing mainly contains piezoelectric patches, acceierometer sensor, angular rate gyroscope sensor, photoelectrical position sensor (Position Sensitive Detector) and the fiber-optic grating sensor methods such as (Fiber Grating Sensor) of adopting: when piezoelectric patches is measured because multiple links such as signal amplifications, filtering can cause signal delay and phase place hysteresis, measuring speed slowly, be limited in scope; Acceierometer sensor and angular rate gyroscope sensor are to noise-sensitive, exist sluggish and temperature such as floats at its precision of problems affect, and can only measure on the object certain any displacement, want to obtain full detail, must on semi-girder, arrange several accelerometers or angular rate gyroscope respectively; The measurement range of photoelectrical position sensor is less, complex structure, operation easier is big, computation process is loaded down with trivial details and cost is expensive; The topmost problem of fiber-optic grating sensor is the demodulation of transducing signal, because fiber grating is relatively more fragile, is very easy to destroy in abominable working environment, thereby could uses packaging technology and safeguard measure complex structure after need encapsulating it.

Summary of the invention

Measurement mechanism and the method based on the flexible cantilever beam vibration of machine vision that the purpose of this invention is to provide a kind of simple in structure, stable performance, efficiently and accurately, realize to the flexible cantilever beam vibration displacement noncontact, accurately measure in real time.

The technical solution that realizes the object of the invention is, a kind of measurement mechanism of the flexible cantilever beam vibration based on machine vision, comprise fixed support, flexible cantilever beam, this measurement mechanism also comprises camera support, CCD camera, camera lens, a plurality of LED luminotron and PC; Described flexible cantilever beam one end is fixed on the fixed support, and the other end is free end, and each LED luminotron is along the upper surface that is successively set on flexible cantilever beam from the fixed support end to free-ended direction; The fixed support top is provided with camera support, and the CCD camera is fixed on the camera support, and the CCD camera configuration has camera lens, and the output port of CCD camera is connected with PC.

Measuring method based on the flexible cantilever beam vibration of machine vision is characterized in that step is as follows:

The 1st step: according to the parameter of CCD camera, each LED luminotron is tied to the horizontal transformation factor K of world coordinate system from image coordinate _x, vertical transformation factor K _yDemarcate;

The 2nd step: the CCD camera is gathered the vibrational image sequence of flexible cantilever beam, and exports the vibrational image sequence to PC by data line;

The 3rd step: according to the position of each LED luminotron in first two field picture, entire image is divided into the subregion corresponding with each LED luminotron, and determines the region of interest ROI of each subregion;

The 4th step: adopt the OTSU thresholding method to extract the hot spot of the interior corresponding LED luminotron of each region of interest ROI of flexible cantilever beam, and determine the centroid position of each hot spot;

The 5th step: according to the centroid position of each hot spot, obtain horizontal shift Δ x and the vertical displacement delta y of corresponding LED luminotron in the image;

The 6th step: according to horizontal shift Δ x, the vertical displacement delta y of each LED luminotron in the image and the horizontal transformation factor K of the 1st step demarcation _x, vertical transformation factor K _y, obtain the real standard displacement x of each LED luminotron, actual perpendicular displacement y;

The 7th step: according to the acquisition order of the 2nd step vibrational image sequence, draw out the position curve of each LED luminotron, Fast Fourier Transform (FFT) is carried out in the displacement of each LED luminotron, obtain single order model frequency and the second-order modal frequency of flexible cantilever beam vibration.

The present invention compared with the prior art, have following remarkable advantage: (1) utilizes Machine Vision Detection can be easy to, obtain intuitively the low-frequency vibration information of flexible cantilever beam, the measurement noise is little, and adopts the simple, convenient enforcement of light source detection of following the tracks of the LED luminotron; (2) utilize vision-based detection that the advantage of multimetering is arranged, can measure multi-point displacement and deformation, real-time and accuracy height; (3) by picture portion, by each LED luminotron of row labels, choose suitable interesting image regions ROI rapidly, reduced image deal with data amount, greatly promoted the speed of image characteristics extraction; (4) under the situation of the vibration characteristics that does not change measured object, follow the tracks of light source by machine vision, detect the multiple spot dynamic displacement of flexible structure, have the advantages such as vibration characteristics that noncontact, measurement range are wide, do not change measured object.

Description of drawings

Fig. 1 is the structural representation of measurement mechanism that the present invention is based on the flexible cantilever beam vibration of machine vision.

Fig. 2 is the vertical view of flexible overarm arm upper surface.

Fig. 3 is CCD camera imaging principle schematic.

Fig. 4 is the level of vibration displacement curve figure of the 3rd LED luminotron in the embodiment of the invention.

Fig. 5 is the rumble spectrum figure of the 3rd LED luminotron in the embodiment of the invention.

Embodiment

The present invention is further illustrated below in conjunction with the drawings and specific embodiments.The present invention obtains the vibration displacement of flexible cantilever beam and the devices and methods therefor that carries out the low frequency modal vibration analysis for the method that adopts machine vision by the trace labelling light source.

As shown in Figure 1, the measurement mechanism based on the flexible cantilever beam vibration of machine vision comprises fixed support 1, flexible cantilever beam 5, camera support 2, CCD camera 3, camera lens 4, a plurality of LED luminotron 6 and PC 7; Described flexible cantilever beam 5 one ends are fixed on the fixed support 1, and the other end is free end, and each LED luminotron 6 is along the upper surface that is successively set on flexible cantilever beam 5 from the fixed support end to free-ended direction; Fixed support 1 top is provided with camera support 2, and CCD camera 3 is fixed on the camera support 2, and CCD camera 3 disposes camera lens 4, and the output port of CCD camera 3 is connected with PC 7.Described each LED luminotron equidistantly is fixed on the center line of flexible cantilever beam 5 upper surfaces, and all LED luminotrons 6 are all in the visual field of CCD camera 3.The quantity of LED luminotron 6 is determined according to the size of flexible cantilever beam 5 and the mode number of needs consideration and control.

Fig. 2 is the upper surface vertical view of flexible cantilever beam 5, comprises that flexible cantilever beam upper surface, flexible cantilever beam upper surface center line, CCD camera optical axis, optical axis are at the projection line of property semi-girder upper surface.The position of CCD camera 3 relative flexibility semi-girders 5 stiff ends and attitude are determined, the CCD plane of CCD camera 3 and the upper surface of flexible cantilever beam 5 13 angled α, α determines according to the visual angle of camera, the height that camera is placed, the length of flexible cantilever beam, the α scope is 0 °～90 °, and all LED luminotrons 6 are in the field range of CCD camera 3 all the time when making flexible cantilever beam 5 vibrations; As shown in Figure 2,15 one-tenth 90 °-α of flexible cantilever beam upper surface 13 and CCD camera optical axis angle when static, and CCD camera optical axis and last plane center line intersect, CCD camera optical axis at the projection line on plane on the flexible cantilever beam on the center line of last plane.

Based on the measuring method of the flexible cantilever beam vibration of machine vision, step is as follows:

The 1st step: according to the parameter of CCD camera 3, each LED luminotron 6 is tied to the horizontal transformation factor K of world coordinate system from image coordinate _x, vertical transformation factor K _yDemarcate, concrete steps are as follows:

(1.1) with the camera model approximate expression actual imaging of no perspective distortion, as shown in Figure 3, set up coordinate system: with the mid point O of flexible cantilever beam 5 upper surface fixed edges _wBe initial point, set up world coordinate system O _w-X _wY _wZ _wPhotocentre O with CCD camera 3 _CBe initial point, setting up camera coordinates is O _c-X _cY _cZ _cBe initial point O with first pixel of the image upper left corner, set up image coordinate system O-UV; Central point O with image _dBe initial point, setting up the image planes physical coordinates is O _d-X _dY _d, O wherein _dCoordinate be (c _x, c _y); The optical axis of CCD camera 3 and X _cO _cY _c, UOV, X _dO _dY _dThe plane is all vertical, the angled α of upper surface of the CCD imaging plane of CCD camera 3 and flexible cantilever beam 5, the optical axis of CCD camera 3 and X _WO _WY _WThe angled 90 °-α in plane, wherein X _WAxle, X _CAxle, X _dAxle is parallel, and CCD camera 3 positions and attitude are fixed, camera coordinates system and world coordinate system only exist around X-axis rotate, translation relation;

(1.2) world coordinate system O _w-X _wY _wZ _wBe O to camera coordinates _c-X _cY _cZ _cTransformation relation be:

$\left[\begin{array}{c}{x}_c\\ y_c\\ z_c\end{array}\right]=R\left[\begin{array}{c}{x}_w\\ y_w\\ z_w\end{array}\right]+T---\left(1\right)$

(x wherein _c, y _c, z _c) for camera coordinates be O _c-X _cY _cZ _cCoordinate, (x _w, y _w, z _w) be world coordinate system O _w-X _wY _wZ _wCoordinate,

$R=\left[\begin{array}{ccc}1& 0& 0\\ 0& \cos \mathrm{\θ}& \sin \mathrm{\θ}\\ 0& -\sin \mathrm{\θ}& \cos \mathrm{\θ}\end{array}\right],T=\left[\begin{array}{c}0\\ 0\\ b\end{array}\right]$

In the formula, R is 3 * 3 rotational transform matrix, and θ is the anglec of rotation around X-axis; T is 3 * 1 translation vector, and b is shift value;

(1.3) camera coordinates is O _c-X _cY _cZ _cBe O to the image planes physical coordinates _d-X _dY _dCoordinate transform be:

$x_d=f\frac{x_c}{z_c},y_d=f\frac{y_c}{z_c}---\left(2\right)$

In the formula, (x _d, y _d) for the image planes physical coordinates be O _d-X _dY _dCoordinate, f is that effective focal length is the distance of the plane of delineation and optical center of lens;

(1.4) the image planes physical coordinates is O _d-X _dY _dTo being transformed to of image coordinate system O-UV:

$u=\frac{x_d}{d_x}+c_x,v=\frac{y_d}{d_y}+c_y---\left(3\right)$

In the formula, (u v) is the coordinate of image coordinate system O-UV, d _xBe in the image pixel along the physical size of x axle, d _yBe in the image pixel along the physical size of y axle;

(1.5) composite type (1)～(3) are by world coordinate system O _w-X _wY _wZ _wHomogeneous coordinates to image coordinate system O-UV conversion are expressed as: $s\left[\begin{array}{c}u\\ v\\ 1\end{array}\right]=\left[\begin{array}{ccc}\frac{f}{d_x}& 0& {\mathrm{sc}}_x\\ 0& \frac{f}{d_{\mathrm{<figure-callout\; id="0"\; label="y\; sc\; y"\; filenames="HDA00003303139200021.png"\; state="\{\{state\}\}">y</figure-callout>}}}& {\mathrm{sc}}_y{}_{}\\ 0& 0& s\end{array}\right]\left[\begin{array}{cc}R& T\end{array}\right]\left[\begin{array}{c}{x}_w\\ y_w\\ z_w\\ 1\end{array}\right]---\left(4\right)$

In the formula, S is scale-up factor, and formula (4) is transformed to:

$\left[\begin{array}{c}u\\ v\\ 1\end{array}\right]=\left[\begin{array}{cccc}\frac{f}{d_{\mathrm{<figure-callout\; id="0"\; label="s\; z\; c"\; filenames="HDA00003303139200021.png"\; state="\{\{state\}\}">s</figure-callout>}}{z}_c{}_{}}& 0& 0& c_x\\ 0& \frac{f}{d_y{z}_c}& 0& {\mathrm{<figure-callout\; id="0"\; label="c\; y"\; filenames="HDA00003303139200021.png"\; state="\{\{state\}\}">c</figure-callout>}}_y\\ 0& 0& 0& 1\end{array}\right]\left[\begin{array}{cccc}1& 0& 0& 0\\ 0& \cos \mathrm{\&theta;}& \sin \mathrm{\&theta;}& 0\\ 0& -\sin \mathrm{\&theta;}& \cos \mathrm{\&theta;}& \mathrm{<figure-callout\; id="0"\; label="b"\; filenames="HDA00003303139200021.png"\; state="\{\{state\}\}">b</figure-callout>}\\ 0& 0& 0& 1\end{array}\right]\left[\begin{array}{c}{x}_w\\ y_w\\ z_w\\ 1\end{array}\right]=\left[\begin{array}{cccc}\frac{f}{d_x{\mathrm{<figure-callout\; id="0"\; label="z\; c"\; filenames="HDA00003303139200021.png"\; state="\{\{state\}\}">z</figure-callout>}}_c}& 0& 0& c_x\\ 0& \frac{f\cos \mathrm{\&theta;}}{d_y{z}_c}& \frac{f\sin \mathrm{\&theta;}}{d_y{z}_{\mathrm{<figure-callout\; id="0"\; label="c\; c\; y"\; filenames="HDA00003303139200021.png"\; state="\{\{state\}\}">c</figure-callout>}}}& c_y{}_{}\\ 0& 0& 0& 1\end{array}\right]\left[\begin{array}{c}{x}_w\\ y_w\\ z_w\\ 1\end{array}\right]$

(5)

Because LED luminotron 6 is only on the plane $\left[\begin{array}{c}{x}_{\mathrm{<figure-callout\; id="0"\; label="w\; y\; w"\; filenames="HDA00003303139200021.png"\; state="\{\{state\}\}">w</figure-callout>}}& \\ y_w{}_{}\\ 0\end{array}\right]$ Internal vibration, so following formula (5) is deformed into:

$\left[\begin{array}{c}{x}_w\\ y_w\end{array}\right]=\frac{z_c}{f}\left[\begin{array}{cc}{d}_x& 0\\ 0& \frac{d_y}{\cos \mathrm{\&theta;}}\end{array}\right]\left[\begin{array}{c}u\\ v\end{array}\right]-\frac{z_c}{f}\left[\begin{array}{c}{c}_x{d}_x\\ c_y\frac{d_y}{\cos \mathrm{\&theta;}}\end{array}\right]---\left(6\right)$

Get image coordinate system O-UV to world coordinate system O by formula (6) _w-X _wY _wZ _wThe horizontal transformation factor K _xWith vertical transformation factor K _y, be shown below respectively:

$K_x=\frac{z_c{d}_x}{f},K_y=\frac{z_c{d}_y}{f\cos \mathrm{\&theta;}}---\left(7\right)$

In the formula, d _x, d _y, f, c _x, c _yBe CCD camera inner parameter, θ, z _cBe external parameter, inner parameter is relevant with the camera inner structure, and external parameter obtains by demarcation.

The 2nd step: CCD camera 3 is gathered the vibrational image sequence of flexible cantilever beams 5, and exports the vibrational image sequence to PC 7 by data line; By the PC program to the vibrational image sequence identification of being correlated with handle the displacement of obtaining LED luminotron 6 places on the flexible cantilever beam 5.

The 3rd step: according to the position of each LED luminotron 6 in first two field picture, entire image is divided into the subregion corresponding with each LED luminotron 6, and determines the region of interest ROI of each subregion; According to the maximum perpendicular displacement of LED luminotron 6 each subregion is chosen the region of interest ROI of image, reduce the data volume of handling, the LED luminotron mainly is horizontal displacement x, and y is less for the perpendicular displacement amount.

The 4th step: adopt the OTSU thresholding method to extract the hot spot of the interior corresponding LED luminotron 6 of each region of interest ROI of flexible cantilever beam 5, and determine the centroid position of each hot spot; The OTSU thresholding method is a kind of method of determining threshold value automatically that makes the inter-class variance maximum, and its basic thought is that image pixel is divided into background and target two classes, calculates the inter-class variance maximal value by search, obtains optimal threshold, and the specific implementation step is as follows:

(4.1) number of pixel in interior each gray level of statistics region of interest ROI;

(4.2) number of pixels that calculates each gray level accounts for ratio and the total gray average of this region of interest ROI of this region of interest ROI;

(4.3) the traversal gray level is calculated and is determined maximum inter-class variance σ, and the gray-scale value of corresponding this maximum between-cluster variance is threshold value T:

σ ²＝p _A(ω _A-ω ₀) ²+p _B(ω _B-ω ₀) ² (8)

In the formula, p _A, ω _ABe respectively probability and gray average that background classes A occurs, p _B, ω _BBe respectively probability and gray average that target class B occurs, ω ₀Be the total gray average of this region of interest ROI;

(4.4) according to threshold value T with image binaryzation, obtain the hot spot of corresponding LED luminotron 6 in this region of interest ROI.

The 5th step: according to the centroid position of each hot spot, obtain horizontal shift Δ x and the vertical displacement delta y of corresponding LED luminotron 6 in the image; Barycenter horizontal level M _x, upright position M _yComputing formula be:

$M_x=\frac{\underset{i=0}{\overset{\mathrm{height}-1}{\mathrm{\&Sigma;}}}\underset{j=0}{\overset{\mathrm{width}-1}{\mathrm{\&Sigma;}}}f(j,i).j}{\underset{i=0}{\overset{\mathrm{height}-1}{\mathrm{\&Sigma;}}}\underset{j=0}{\overset{\mathrm{width}-1}{\mathrm{\&Sigma;}}}f(j,i)}---\left(9\right)$

$M_y=\frac{\underset{i=0}{\overset{\mathrm{height}-1}{\mathrm{\&Sigma;}}}\underset{j=0}{\overset{\mathrm{width}-1}{\mathrm{\&Sigma;}}}f(j,i).i}{\underset{i=0}{\overset{\mathrm{height}-1}{\mathrm{\&Sigma;}}}\underset{j=0}{\overset{\mathrm{width}-1}{\mathrm{\&Sigma;}}}f(j,i)}---\left(10\right)$

In the formula, (j is to be positioned at that (height represents the pixels tall of region of interest ROI for j, i) gray values of pixel points, and width represents the pixel wide of region of interest ROI i) to f.

Determine horizontal shift Δ x, the vertical displacement delta y of flexible cantilever beam 5 upper end LED luminotrons 6 in the image:

Δx＝M _x-x ₀,Δy＝M _y-y ₀ (11)

Wherein, x ₀The horizontal coordinate of representing flexible cantilever beam 5 upper end LED luminotrons 6 when static, y ₀The vertical coordinate of representing flexible cantilever beam 5 upper end LED luminotrons 6 when static.Horizontal shift Δ x, vertical displacement delta y represent on level after 6 imagings of LED luminotron, the vertical direction the shared pixel size of displacement in image respectively.

The 6th step: according to horizontal shift Δ x, the vertical displacement delta y of each LED luminotron 6 in the image and the horizontal transformation factor K of the 1st step demarcation _x, vertical transformation factor K _y, obtain the real standard displacement x of each LED luminotron 6, actual perpendicular displacement y:

x＝K _xΔx,y＝K _yΔy (12)

The 7th step: according to the acquisition order of the 2nd step vibrational image sequence, draw out the position curve of each LED luminotron 6, Fast Fourier Transform (FFT) is carried out in the displacement of each LED luminotron 6, obtain the single order mode frequencies omega of flexible cantilever beam 5 vibrations ₁With the second-order modal frequencies omega ₂

Embodiment 1

Flexible cantilever girder construction with the analog solar windsurfing is example, and measurement mechanism and the method for flexible cantilever beam vibration that the present invention is based on machine vision is specific as follows.

In conjunction with Fig. 1, the present invention is based on the measurement mechanism of the flexible cantilever beam vibration of machine vision, comprise fixed support 1, flexible cantilever beam 5, camera support 2, CCD camera 3, camera lens 4,6 LED luminotrons 6 and PC 7; Described flexible cantilever beam 5 one ends are fixed on the fixed support 1, and the other end is free end, and the dimensioning modest ability of flexible cantilever beam 5 * wide * height is 1200mm * 80mm * 2mm; LED luminotron 6 is along be respectively a LED luminotron 6-1, the 2nd LED luminotron 6-2, the 3rd LED luminotron 6-3, the 4th LED luminotron 6-4, the 5th LED luminotron 6-5 and the 6th LED luminotron 6-6 from the fixed support end to free-ended direction, be successively set on the upper surface of flexible cantilever beam 5, equidistantly paste 6 red LED luminotrons along the every 200mm of the center line of flexible cantilever beam upper surface.Fixed support 1 top is provided with camera support 2, and CCD camera 3 is fixed on the camera support 2, and the model of CCD camera 3 is Cognex In-Sight5400, sensor type is 1/3 inch CCD, resolution is 640 * 480, dark 256 grey levels in position, per second 60 two field pictures; CCD camera 3 disposes camera lens 4, and the C mouth mirror head 4 of 8mm focal length is installed on the CCD camera; Under this configuration, the visual angle of camera is 38.5 °.The output port of CCD camera 3 is connected with PC 7 by data line.Utilize CCD camera 3 to detect the low-frequency vibration image of flexible cantilever beam 5 as sensor.The CCD camera that adopts in the present embodiment, d _x=d _ySo pixel is identical with the physical size of y axle along x in the image.

Fig. 2 is the upper surface vertical view of flexible cantilever beam 5, comprises that flexible cantilever beam upper surface 13, flexible cantilever beam upper surface center line 14, CCD camera optical axis 15, optical axis are at the projection line 16 of property semi-girder upper surface.The position of CCD camera 3 relative flexibility semi-girders 5 stiff ends and attitude are determined, all LED luminotrons 6 are in the field range of CCD camera 3 all the time when making flexible cantilever beam 5 vibrations, the CCD plane of CCD camera 3 and the upper surface of flexible cantilever beam 5 13 angled α, α=60 °, as shown in Figure 2, flexible cantilever beam upper surface 13 and CCD camera optical axis ° angle, 15 one-tenth 90 °-α=30 when static, and CCD camera optical axis and last plane center line intersect, CCD camera optical axis at the projection line on plane on the flexible cantilever beam on the center line of last plane.

Based on the measuring method of the flexible cantilever beam vibration of machine vision, step is as follows:

The 1st step: according to the parameter of CCD camera 3,6 LED luminotrons 6 are tied to the horizontal transformation factor K of world coordinate system from image coordinate _x, vertical transformation factor K _yDemarcate, because in the actual imaging, the K of diverse location place on the flexible cantilever beam 5 _x, K _yValue changes, so 66 of LED luminotrons are located all will demarcate, obtains and need demarcate 6 LED luminotron places, obtains (K _X1, K _Y1), (K _X2, K _Y2), (K _X3, K _Y3), (K _X4, K _Y4), (K _X5, K _Y5), (K _X6, K _Y6), because d _x=d _y, then formula (7) changes into:

$K_x=\frac{z_c{d}_x}{f},K_y=\frac{z_c{d}_x}{f\cos \mathrm{\&theta;}}$

In the image on the flexible cantilever beam plane midline a bit at image coordinate system O-UV and world coordinate system O _w-X _wY _wZ _wIn initial position be respectively (u ₁, v ₁) and (x _W1, y _W1, 0), the position in a certain moment is (u during vibration ₂, v ₂) and (x _W2, y _W2, 0), this pass between the displacement under image coordinate system and the world coordinate system is:

$\sqrt{{(x_{w2}-x_{w1})}^2+{(y_{w2}-y_{w1})}^2}=\frac{z_c{d}_x}{f\left|\cos \mathrm{\&theta;}\right|}\sqrt{{\cos }^2\mathrm{\&theta;}{(u_2-u_1)}^2+{(v_2-v_1)}^2}$

The 2nd step: CCD camera 3 is gathered the vibrational image sequence of flexible cantilever beams 5, and exports the vibrational image sequence to PC 12 by data line; By the PC program to the vibrational image sequence identification of being correlated with handle the displacement of obtaining LED luminotron 6 places on the flexible cantilever beam 5.

The 3rd step: according to the position of each LED luminotron 6 in first two field picture, entire image is divided into 6 sub regions corresponding with each LED luminotron 6, and determines the region of interest ROI of each subregion; According to the maximum perpendicular displacement of LED luminotron 6 each subregion is chosen the region of interest ROI of image, reduce the data volume of handling, the LED luminotron mainly is horizontal displacement x, and y is less for the perpendicular displacement amount.The ROI that the whole horizontal directions of each LED luminotron sub-region processes, vertical 50 pixel datas are formed, namely the ROI size is 640 * 50.

The 4th step: adopt the OTSU thresholding method to extract the hot spot of the interior corresponding LED luminotron 6 of each region of interest ROI of flexible cantilever beam 5, and determine the centroid position of each hot spot.

The 5th step: according to the centroid position of each hot spot, obtain horizontal shift Δ x and the vertical displacement delta y of corresponding LED luminotron 6 in the image, acquisition order according to the 2nd step vibrational image sequence, the horizontal shift curve of the 3rd LED luminotron 6-3 as shown in Figure 4, horizontal ordinate is that the time, (unit: second), ordinate was displacement (unit: pixel).

The 6th step: according to horizontal shift Δ x, the vertical displacement delta y of each LED luminotron 6 in the image and the horizontal transformation factor K of the 1st step demarcation _x, vertical transformation factor K _y, obtain the real standard displacement x of each LED luminotron 6, actual perpendicular displacement y.

The 7th step: according to the acquisition order of the 2nd step vibrational image sequence, draw out the position curve of each LED luminotron 6, Fast Fourier Transform (FFT) is carried out in the displacement of each LED luminotron 6, obtain the single order mode frequencies omega of flexible cantilever beam 5 vibrations ₁With the second-order modal frequencies omega ₂The horizontal shift of the 3rd LED luminotron 6-3 is carried out curve that Fast Fourier Transform (FFT) obtains as shown in Figure 5, and horizontal ordinate is frequency (unit: hertz), and ordinate is amplitude.

In sum, the constructed machine vision vibration measurement device of method that the present invention proposes can be followed the tracks of the method for light source by machine vision under the situation of the vibration characteristics that does not change measured object, detect the multiple spot dynamic displacement of flexible structure, and obtain the low-frequency vibration modal information.This method has the plurality of advantages such as vibration characteristics that noncontact, measurement range are wide, do not change measured object.

Claims (6)

1. measurement mechanism based on the flexible cantilever beam vibration of machine vision, comprise fixed support (1), flexible cantilever beam (5), it is characterized in that this measurement mechanism also comprises camera support (2), CCD camera (3), camera lens (4), a plurality of LED luminotron (6) and PC (7); Described flexible cantilever beam (5) one ends are fixed on the fixed support (1), and the other end is free end, and each LED luminotron (6) is along the upper surface that is successively set on flexible cantilever beam (5) from the fixed support end to free-ended direction; Fixed support (1) top is provided with camera support (2), and CCD camera (3) is fixed on the camera support (2), and CCD camera (3) disposes camera lens (4), and the output port of CCD camera (3) is connected with PC (7).

2. the measurement mechanism of the flexible cantilever beam vibration based on machine vision according to claim 1, it is characterized in that: the quantity of described LED luminotron (6) is 6, be respectively a LED luminotron (6-1), the 2nd LED luminotron (6-2), the 3rd LED luminotron (6-3), the 4th LED luminotron (6-4), the 5th LED luminotron (6-5) and the 6th LED luminotron (6-6), and each LED luminotron is fixed on equidistantly on the center line of flexible cantilever beam (5) upper surface.

3. the measurement mechanism of the flexible cantilever beam vibration based on machine vision according to claim 1, it is characterized in that: described CCD camera (3) setting angle meets the following conditions: each LED luminotron (6) is all in the visual field of CCD camera (3).

4. measuring method based on the flexible cantilever beam vibration of machine vision is characterized in that step is as follows:

The 1st step: according to the parameter of CCD camera (3), each LED luminotron (6) is tied to the horizontal transformation factor K of world coordinate system from image coordinate _x, vertical transformation factor K _yDemarcate;

The 2nd step: CCD camera (3) is gathered the vibrational image sequence of flexible cantilever beam (5), and exports the vibrational image sequence to PC (12) by data line;

The 3rd step: according to the position of each LED luminotron (6) in first two field picture, entire image is divided into the subregion corresponding with each LED luminotron (6), and determines the region of interest ROI of each subregion;

The 4th step: adopt the OTSU thresholding method to extract the hot spot of the interior corresponding LED luminotron (6) of each region of interest ROI of flexible cantilever beam (5), and determine the centroid position of each hot spot;

The 5th step: according to the centroid position of each hot spot, obtain horizontal shift Δ x and the vertical displacement delta y of corresponding LED luminotron (6) in the image;

The 6th step: according to horizontal shift Δ x, the vertical displacement delta y of each LED luminotron (6) in the image and the horizontal transformation factor K of the 1st step demarcation _x, vertical transformation factor K _y, obtain the real standard displacement x of each LED luminotron (6), actual perpendicular displacement y;

The 7th step: according to the acquisition order of the 2nd step vibrational image sequence, draw out the position curve of each LED luminotron (6), Fast Fourier Transform (FFT) is carried out in displacement to each LED luminotron (6), obtains single order model frequency and the second-order modal frequency of flexible cantilever beam (5) vibration.

5. the measuring method of the flexible cantilever beam vibration based on machine vision according to claim 4 is characterized in that, the described horizontal transformation factor K that each LED luminotron (6) is tied to world coordinate system from image coordinate of the 1st step _x, vertical transformation factor K _yDemarcate, concrete steps are as follows:

(1.1) set up coordinate system: with the mid point O of flexible cantilever beam (5) upper surface fixed edge _wBe initial point, set up world coordinate system O _w-X _wY _wZ _wPhotocentre O with CCD camera (3) _CBe initial point, setting up camera coordinates is O _c-X _cY _cZ _cBe initial point O with first pixel of the image upper left corner, set up image coordinate system O-UV; Central point O with image _dBe initial point, setting up the image planes physical coordinates is O _d-X _dY _d, O wherein _dCoordinate be (c _x, c _y); Optical axis and the X of CCD camera (3) _cO _cY _c, UOV, X _dO _dY _dThe plane is all vertical, the angled α of upper surface of the CCD imaging plane of CCD camera (3) and flexible cantilever beam (5), 0 °＜α＜90 °, optical axis and the X of CCD camera (3) _WO _WY _WThe angled 90 °-α in plane, wherein X _WAxle, X _CAxle, X _dAxle is parallel;

(1.2) world coordinate system O _w-X _wY _wZ _wBe O to camera coordinates _c-X _cY _cZ _cTransformation relation be:

$\left[\begin{array}{c}{x}_c\\ y_c\\ z_c\end{array}\right]=R\left[\begin{array}{c}{x}_w\\ y_w\\ z_w\end{array}\right]+T---\left(1\right)$

(x wherein _c, y _c, z _c) for camera coordinates be O _c-X _cY _cZ _cCoordinate, (x _w, y _w, z _w) be world coordinate system O _w-X _wY _wZ _wCoordinate,

$R=\left[\begin{array}{ccc}1& 0& 0\\ 0& \cos \mathrm{\&theta;}& \sin \mathrm{\&theta;}\\ 0& -\sin \mathrm{\&theta;}& \cos \mathrm{\&theta;}\end{array}\right],T=\left[\begin{array}{c}0\\ 0\\ b\end{array}\right]$

In the formula, R is 3 * 3 rotational transform matrix, and θ is the anglec of rotation around X-axis; T is 3 * 1 translation vector, and b is shift value;

(1.3) camera coordinates is O _c-X _cY _cZ _cBe O to the image planes physical coordinates _d-X _dY _dCoordinate transform be:

$x_d=f\frac{x_c}{z_c},y_d=f\frac{y_c}{z_c}---\left(2\right)$

In the formula, (x _d, y _d) for the image planes physical coordinates be O _d-X _dY _dCoordinate, f is that effective focal length is the distance of the plane of delineation and optical center of lens;

(1.4) the image planes physical coordinates is O _d-X _dY _dTo being transformed to of image coordinate system O-UV:

$u=\frac{x_d}{d_x}+c_x,v=\frac{y_d}{d_y}+c_y---\left(3\right)$

In the formula, (u v) is the coordinate of image coordinate system O-UV, d _xBe in the image pixel along the physical size of x axle, d _yBe in the image pixel along the physical size of y axle;

(1.5) composite type (1)～(3) are by world coordinate system O _w-X _wY _wZ _wHomogeneous coordinates to image coordinate system O-UV conversion are expressed as: $s\left[\begin{array}{c}u\\ v\\ 1\end{array}\right]=\left[\begin{array}{ccc}\frac{f}{d_x}& 0& {\mathrm{sc}}_x\\ 0& \frac{f}{d_y}& {\mathrm{sc}}_y\\ 0& 0& s\end{array}\right]\left[\begin{array}{cc}R& T\end{array}\right]\left[\begin{array}{c}{x}_w\\ y_w\\ z_w\\ 1\end{array}\right]---\left(4\right)$

In the formula, S is scale-up factor, and formula (4) is transformed to:

$\left[\begin{array}{c}u\\ v\\ 1\end{array}\right]=\left[\begin{array}{cccc}\frac{f}{d_s{z}_c}& 0& 0& c_x\\ 0& \frac{f}{d_y{z}_c}& 0& c_y\\ 0& 0& 0& 1\end{array}\right]\left[\begin{array}{cccc}1& 0& 0& 0\\ 0& \cos \mathrm{\&theta;}& \sin \mathrm{\&theta;}& 0\\ 0& -\sin \mathrm{\&theta;}& \cos \mathrm{\&theta;}& b\\ 0& 0& 0& 1\end{array}\right]\left[\begin{array}{c}{x}_w\\ y_w\\ z_w\\ 1\end{array}\right]=\left[\begin{array}{cccc}\frac{f}{d_x{z}_c}& 0& 0& c_x\\ 0& \frac{f\cos \mathrm{\&theta;}}{d_y{z}_c}& \frac{f\sin \mathrm{\&theta;}}{d_y{z}_c}& c_y\\ 0& 0& 0& 1\end{array}\right]\left[\begin{array}{c}{x}_w\\ y_w\\ z_w\\ 1\end{array}\right]---\left(5\right)$

Because LED luminotron (6) is only on the plane $\left[\begin{array}{c}{x}_w\\ y_w\\ 0\end{array}\right]$ Internal vibration, so following formula (5) is deformed into:

$\left[\begin{array}{c}{x}_w\\ y_w\end{array}\right]=\frac{z_c}{f}\left[\begin{array}{cc}{d}_x& 0\\ 0& \frac{d_y}{\cos \mathrm{\&theta;}}\end{array}\right]\left[\begin{array}{c}u\\ v\end{array}\right]-\frac{z_c}{f}\left[\begin{array}{c}{c}_x{d}_x\\ c_y\frac{d_y}{\cos \mathrm{\&theta;}}\end{array}\right]---\left(6\right)$

Get image coordinate system O-UV to world coordinate system O by formula (6) _w-X _wY _wZ _wThe horizontal transformation factor K _xWith vertical transformation factor K _y, be shown below respectively:

$K_x=\frac{z_c{d}_x}{f},K_y=\frac{z_c{d}_y}{f\cos \mathrm{\&theta;}}---\left(7\right)$

6. the measuring method of the flexible cantilever beam vibration based on machine vision according to claim 4 is characterized in that, described OTSU thresholding method of the 4th step, and step is as follows:

(4.1) number of pixel in interior each gray level of statistics region of interest ROI;

(4.2) number of pixels that calculates each gray level accounts for ratio and the total gray average of this region of interest ROI of this region of interest ROI;

(4.3) the traversal gray level is calculated and is determined maximum inter-class variance, and the gray-scale value of corresponding this maximum between-cluster variance is threshold value;

(4.4) according to threshold value with image binaryzation, obtain the hot spot of corresponding LED luminotron (6) in this region of interest ROI.

Images