CN102749061A - Steel rail abrasion measuring method based on dynamic template - Google Patents

Abstract

The invention provides a steel rail abrasion measuring method based on a dynamic template. The method comprises the steps of installing a charge coupled device (CCD) camera and a fan-shaped laser light source above the inside of two rails respectively; demarcating each CCD camera and an optical plane parameters; obtaining space coordinates of each pixel point of rail outlines in an image according to a rail outline space computation model; extracting feature point coordinates at the lower end of a rail waist and a rail head; generating a rail outline standard template based on feature points; and contrasting a rail measurement outline to the standard template to obtain abrasion value. The method can build a measurement coordinate system by determining two feature points of a measurement rail according to geometrical relationship of a standard rail outline, can fast and accurately obtain the standard template, does not need conduct aligning analysis of the measurement outline and a standard design outline, solves the problem that method based on a static template is difficult to match, greatly improves accuracy of abrasion measurement, reduces operation quantity of image analysis and processing and has good stability.

Description

Measurement of rail wear based method based on dynamic template

Technical field

The present invention relates to the track traffic technology, relate in particular to a kind of measurement of rail wear based method based on dynamic template.

Background technology

The general main rail head position that occurs in of the abrasion of rail is the decision rail principal element in serviceable life.Abrasion have been quickened the wearing and tearing of locomotive wheel on the one hand, have increased gauge, and the contact area of increase and locomotive wheel tread, and running resistance is increased, and have produced severe noise on the other hand.When wearing and tearing surpassed certain limit, rail head cross-section and wheel tread will lose matching, will have a strong impact on the high-speed railway stationarity of driving a vehicle, and traffic safety is caused great harm.Particularly for high-speed railway, when train speed is very high, even very little wearing and tearing also possibly cause derail.

China is at the maximum non-contact detection method that is based on machine vision of abrasion context of detection research at present.This method is utilized coordinate transform that image coordinate is mapped to object coordinates to carry out image rectification, mate with standard form more at last mostly, as adopting the ICP matching algorithm, thereby obtains the wearing valve of rail's end portion.Not only calculated amount is big, and template matches or Feature Points Matching difficulty, and shortcoming such as data processing speed is slow, and error is big make measurement result effectively for the maintenance of rail reliable foundation is not provided.Detect as for the scene abrasion, then mainly rely on the mechanical type contact measurement.This method measuring speed is slow, and choosing the measurement result influence of workplace is very big.Fixed and movable side head weares and teares after using a period of time, can make measurement result have certain error.

Summary of the invention

The purpose of this invention is to provide a kind of measurement of rail wear based method, to overcome the problem of existing method template coupling or Feature Points Matching difficulty based on dynamic template.

The technical scheme that the present invention is adopted for its technical matters of solution is,

A kind of measurement of rail wear based method based on dynamic template may further comprise the steps:

1) at two one steel rail inside fronts ccd video camera and fan-shaped LASER Light Source are installed respectively;

2) demarcate each ccd video camera and light-plane parameters;

3) obtain the volume coordinate of each pixel of rail profile in the image according to rail profile SPATIAL CALCULATION model;

4) extract the web of the rail and rail head lower end unique point coordinate;

5) generate rail profile standard form based on unique point;

6) contrast rail measuring wheel exterior feature and standard form obtain wearing valve.

Every side rail is provided with a ccd video camera and a fan-shaped LASER Light Source in the step 1), all is positioned at the rail inside front, and the optical plane that sends of fan-shaped LASER Light Source and rail is vertical vertical.

The detailed process of demarcating each ccd video camera parameter step 2) is following:

Video camera is fixed on the plane, takes the gridiron pattern scaling board image that is positioned at diverse location more than three, mate, calculate the mapping matrix between them through point and picture point on the scaling board:

$z_{\mathrm{<figure-callout\; id="1"\; label="c\; u\; v"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">c</figure-callout>}}\left(\begin{array}{c}u\\ v\end{array}\right)\left(\begin{array}{c}\\ 1\end{array}\right)=\left(\begin{array}{cccc}{f}_x& 0& u_0& 0\\ 0& {\mathrm{<figure-callout\; id="0"\; label="f\; y\; v"\; filenames="HDA00001934594500041.png"\; state="\{\{state\}\}">f</figure-callout>}}_y& v_{}{}_0& 0\\ 0& 0& 1& 0\end{array}\right)\left(\begin{array}{cc}R& T\\ 0^T& 1\end{array}\right)\left(\begin{array}{c}{x}_w\\ y_{\mathrm{<figure-callout\; id="1"\; label="w\; z\; w"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">w</figure-callout>}}& \\ z_w{}_{}\\ 1\end{array}\right)=\left(\begin{array}{cccc}{m}_{11}& m_{12}& m_{13}& m_{14}\\ m_{21}& m_{22}& m_{23}& m_{24}\\ m_{31}& m_{32}& m_{33}& m_{34}\end{array}\right)\left(\begin{array}{c}{x}_w\\ y_{\mathrm{<figure-callout\; id="1"\; label="w\; z\; w"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">w</figure-callout>}}& \\ z_w{}_{}\\ 1\end{array}\right),$

And solve camera parameters through this matrix, z wherein _cBe the component of certain intersection point in camera coordinate system on the gridiron pattern, f _x, f _y, u ₀, v ₀Inner parameter for video camera; R is a rotation matrix with orthogonality, and T is a translation matrix, and they are external parameters of video camera.x _w, y _w, z _w, 1 is the coordinate of i the point in space; (u, v, 1) is the image coordinate of i point; m _IjThe capable j column element of i for projection matrix M; Decompose cancellation z _cCan obtain about m _IjLinear equation:

$\left\{\begin{array}{c}{x}_w{m}_{11}+y_w{m}_{12}+z_w{m}_{13}+m_{14}-{\mathrm{ux}}_w-{\mathrm{uy}}_w{m}_{32}-{\mathrm{uz}}_w{m}_{33}={\mathrm{um}}_{34}\\ x_w{m}_{21}+y_w{m}_{22}+z_w{m}_{23}+m_{24}-v{x}_w-v{y}_w{m}_{32}-{\mathrm{vz}}_w{m}_{32}={\mathrm{vm}}_{34}\end{array}\right.$

On the calibration piece n known point arranged, obtains 2n linear equation matrix about the Metzler matrix element:

$\left[\begin{array}{cccccccccc}{x}_{w1}& y_{w1}& {\mathrm{<figure-callout\; id="1"\; label="z\; w"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">z</figure-callout>}}_w& 1& 0& 0& 0& -u_1{x}_{w1}& -u_1{y}_{w1}& -u_1{\mathrm{<figure-callout\; id="1"\; label="z\; w"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">z</figure-callout>}}_w\\ 0& 0& 0& 0& x_{w1}& y_{w1}& z_{w1}& -v_1{x}_{w1}& -v_1{y}_{w1}& -v_1{\mathrm{<figure-callout\; id="1"\; label="z\; w"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">z</figure-callout>}}_w\\ & & & & ...& ...& ...& & & \\ & & & & ...& ...& ...& & & \\ x_{\mathrm{wn}}& y_{\mathrm{<figure-callout\; id="1"\; label="wn\; z\; wn"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">wn</figure-callout>}}& z_{\mathrm{wn}}{}_{}& 1& 0& 0& 0& -u_n{x}_{\mathrm{wn}}& -u_n{y}_{\mathrm{wn}}& -u_{\mathrm{<figure-callout\; id="0"\; label="n\; z\; wn"\; filenames="HDA00001934594500041.png"\; state="\{\{state\}\}">n</figure-callout>}}{z}_{\mathrm{wn}}{}_{}\\ 0& 0& 0& 0& x_{\mathrm{wn}}& y_{\mathrm{wn}}& z_{\mathrm{wn}}& -v_n{x}_{\mathrm{wn}}& -v_n{y}_{\mathrm{wn}}& -v_n{z}_{\mathrm{wn}}\end{array}\right]\&times;\left[\begin{array}{c}{m}_{11}\\ m_{12}\\ m_{13}\\ m_{14}\\ m_{21}\\ m_{22}\\ m_{23}\\ m_{24}\\ m_{31}\\ m_{32}\\ m_{33}\end{array}\right]=\left[\begin{array}{c}{u}_1{m}_{34}\\ v_1{m}_{34}\\ ...\\ ...\\ ...\\ \\ ...\\ ...\\ ...\\ u_n{m}_{34}\\ v_n{m}_{34}\end{array}\right]$

Specify m ₃₄=1, obtain 2n linear equation about other elements of Metzler matrix, the number of unknown element is 11, is designated as 11 dimensional vector m, writes a Chinese character in simplified form: Km=U, K are left side 2n * 11 matrixes; M is 11 unknown dimensional vectors; U is the 2n dimensional vector on the right; K, U are known vector; Obtain > with least square method as 2n; 11 o'clock above-mentioned linear equations separate for:

m＝K ^TK ^-1K ^TU

M vector and m ₃₄=1 constituted the institute find the solution Metzler matrix; Therefore, by 6 above known points in space and their picture point coordinate, can obtain Metzler matrix.

After obtaining Metzler matrix, can divide the whole inside and outside parameter that calculate video camera by relation.

The detailed process of demarcating light-plane parameters step 2) is following:

Increase auxiliary camera C during calibration ₂, with video camera C ₁Form the binocular scaling system, the common shooting is positioned at the light belt image on the scaling board, and be based upon video camera C with the world coordinate system initial point this moment ₁The photocentre place, through respectively to C ₁, C ₂Calibration can be confirmed its inner parameter A ₁, A ₂And external parameter B ₁, B ₂, then their projection matrix is respectively M ₁=A ₁B ₁(because world coordinate system is based upon C ₁So the photocentre place is B ₁=I, I are unit matrix)=A ₁, M ₂=A ₂B ₂=A ₂[R ₁₂, T ₁₂] ([R ₁₂, T ₁₂] be video camera C ₁, C ₂Between the position relational matrix), CCD picked-up light belt forms after the RGB image, it is extracted Flame Image Process such as component, gray processing, the computing of difference shadow, binaryzation, refinement, to extract the pixel coordinate of each point on the striation axis, projection has following form:

$Z_1\left(\begin{array}{c}{u}_1\\ v_1\\ 1\end{array}\right)=M_1\left(\begin{array}{c}{x}_w\\ y_{\mathrm{<figure-callout\; id="1"\; label="w\; z\; w"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">w</figure-callout>}}& \\ z_w{}_{}\\ 1\end{array}\right)=\left(\begin{array}{cccc}{m}_{11}^1& m_{12}^1& m_{13}^1& m_{14}^1\\ m_{21}^1& m_{22}^1& m_{23}^1& m_{24}^1\\ m_{31}^1& m_{32}^1& m_{33}^1& m_{34}^1\end{array}\right)\left(\begin{array}{c}{x}_w\\ y_{\mathrm{<figure-callout\; id="1"\; label="w\; z\; w"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">w</figure-callout>}}& \\ z_w{}_{}\\ 1\end{array}\right)$

$Z_2\left(\begin{array}{c}{u}_2\\ v_2\\ 1\end{array}\right)=M_2\left(\begin{array}{c}{x}_w\\ y_{\mathrm{<figure-callout\; id="1"\; label="w\; z\; w"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">w</figure-callout>}}& \\ z_w{}_{}\\ 1\end{array}\right)=\left(\begin{array}{cccc}{m}_{11}^2& m_{12}^2& m_{13}^2& m_{14}^2\\ m_{21}^2& m_{22}^2& m_{23}^2& m_{24}^2\\ m_{31}^2& m_{32}^2& m_{33}^2& m_{34}^2\end{array}\right)\left(\begin{array}{c}{x}_w\\ y_{\mathrm{<figure-callout\; id="1"\; label="w\; z\; w"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">w</figure-callout>}}& \\ z_w{}_{}\\ 1\end{array}\right)$

Wherein, Z ₁, Z ₂Be the component of some P in two camera coordinates on the scaling board light belt, but cancellation during calculating; (u ₁, v ₁), (u ₂, v ₂) be respectively a P at C ₁, C ₂In pixel coordinate.Above two formulas of simultaneous obtain:

$\left\{\begin{array}{c}(u_1{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{31}-{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{11})x_w+(u_1{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{32}-{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{12})y_w+(u_1{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{33}-{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{33})z_w={{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{14}-u_1{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{34}\\ (v_1{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{31}-{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{11})x_w+(v_1{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{32}-{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{22})y_w+(v_1{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{33}-{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{23})z_w={{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{24}-v_1{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{34}\\ (u_2{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{31}-{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{11})x_w+(u_2{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{32}-{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{12})y_w+(u_2{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{33}-{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{13})z_w={{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{14}-u_2{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{34}\\ (v_2{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{31}-{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{21})x_w+(v_2{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{22}-{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{22})y_w+(v_2{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{33}-{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{23})z_w={{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{24}-v_2{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{34}\end{array}\right.$

Thus, can obtain the world coordinates of a P.In like manner, constantly move the locus of scaling board, can obtain (the m > of m on the light belt; 3) world coordinates (x of individual point _Wi, y _Wi, z _Wi) (i=1,2,3 ... m).

The equation on spatial light plane can be represented as follows:

Ax _w+By _w+Cz _w+1=0

Wherein, A, B, C are 3 components of this planar process vector n.

Because the world coordinates of unique point also satisfies optic plane equations on the light belt, overdetermined equation group of then available m unique point structure, its matrix form is:

$\left[\begin{array}{ccc}{x}_{w1}& y_{w1}& z_{w1}\\ x_{w2}& y_{w2}& {\mathrm{<figure-callout\; id="2"\; label="z\; w"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">z</figure-callout>}}_w\\ .& .& .\\ .& .& .\\ .& .& .\\ x_{\mathrm{wm}}& y_{\mathrm{wm}}& z_{\mathrm{wm}}\end{array}\right]\left[\begin{array}{c}A\\ B\\ C\end{array}\right]=\left[\begin{array}{c}-1\\ -1\\ .\\ .\\ .\\ -1\end{array}\right]$

Or be abbreviated as:

GS=L

Wherein, G is the matrix of coefficients on the equality left side, S=[A, B, C] ^T, L=[1,1 ..., 1] ^T, utilize least square method can obtain S=(G then ^TG) ^-1G ^TL and coefficient A, B, C.

It is following to obtain the detailed process of the volume coordinate of each pixel of rail profile in the image according to rail profile SPATIAL CALCULATION model in the step 3):

World coordinate system is based upon camera coordinates fastens, make both overlap fully, then do not have rotation and translation relation between the two, at this moment

$\left(\begin{array}{cc}R& T\\ 0^{\mathrm{<figure-callout\; id="1"\; label="T"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">T</figure-callout>}}& 1\end{array}\right)=I$

For unit matrix and z is arranged _c=z _w, then

$z_{\mathrm{<figure-callout\; id="1"\; label="c\; u\; v"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">c</figure-callout>}}\left(\begin{array}{c}u\\ v\end{array}\right)\left(\begin{array}{c}\\ 1\end{array}\right)=\left(\begin{array}{cccc}{f}_x& 0& u_0& 0\\ 0& f_y& v_0& 0\\ 0& 0& 1& 0\end{array}\right)\left(\begin{array}{c}{x}_w\\ y_{\mathrm{<figure-callout\; id="1"\; label="w\; z\; w"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">w</figure-callout>}}& \\ z_w{}_{}\\ 1\end{array}\right)$

The p point also in the plane that laser throwed, therefore satisfies optic plane equations simultaneously:

Ax _w+By _w+Cz _w+1=0

In the formula, A, B, C are the optical plane coefficient.Simultaneous can obtain:

$\left(\begin{array}{ccc}{f}_x& 0& u_0-u\\ 0& f_y& v_0-v\\ a& b& c\end{array}\right)\left(\begin{array}{c}{x}_w\\ y_w\\ z_w\end{array}\right)=\left(\begin{array}{c}0\\ 0\\ -1\end{array}\right)$

Can get the world coordinates (x of any point on the rail profile outline line _w, y _w, z _w).

The detailed process of extracting the web of the rail and rail head lower end unique point coordinate in the step 4) is following,

(1) web of the rail feature point extraction:

Appoint in the web of the rail circular arc behind three-dimensional reconstruction and get 1 X _i(x _Wi, y _Wi, z _Wi), make the space sphere with radius R, note space spherical equation is:

X ^TQX=0

Wherein x is the coordinate of putting on the sphere, X=[x _w, y _w, z _w, 1] ^T, Q is 4 * 4 symmetric matrix: $Q=\left[\begin{array}{cccc}1& 0& 0& -x_{\mathrm{<figure-callout\; id="0"\; label="Wi"\; filenames="HDA00001934594500041.png"\; state="\{\{state\}\}">Wi</figure-callout>}}\\ 0& 1& 0& -{\mathrm{<figure-callout\; id="0"\; label="y\; Wi"\; filenames="HDA00001934594500041.png"\; state="\{\{state\}\}">y</figure-callout>}}_{\mathrm{Wi}}\\ 0& 0& 1& -z_{\mathrm{Wi}}\\ -x_{\mathrm{Wi}}& -y_{\mathrm{Wi}}& -{\mathrm{<figure-callout\; id="2"\; label="z\; Wi\; l"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">z</figure-callout>}}_{\mathrm{Wi}}& l^{}{}^2\end{array}\right],$ l ²=x _Wi ²+ y _Wi ²+ z _Wi ²-R ²

Because the point on the rail outline line makes x in optical plane _t=x _w, y _t=y _w, then optic plane equations can be write as:

X＝Mt

Wherein $M=\left[\begin{array}{ccc}-b/a& -c/a& -1/a\\ 1& 0& 0\\ 0& 1& 0\\ 0& 0& 1\end{array}\right],$ $t=\left[\begin{array}{c}{x}_{\mathrm{<figure-callout\; id="1"\; label="t\; y\; t"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">t</figure-callout>}}& \\ y_t{}_{}\\ 1\end{array}\right],$ Can get X ^TQX=t ^TM ^TQMt=0,

Order

$C=M^T\mathrm{QM}=\left[\begin{array}{ccc}{(b/a)}^2+1& \mathrm{bc}/a^2& b/a^2+(b/c)x_{\mathrm{wi}}-y_{\mathrm{wi}}\\ \mathrm{bc}/a^2& {(c/a)}^2+1& c/a^2+(c/a)x_{\mathrm{wi}}-z_{\mathrm{wi}}\\ b/a^2+(b/a)x_{\mathrm{wi}}-y_{\mathrm{wi}}& c/a^2+(c/a)x_{\mathrm{wi}}-{\mathrm{<figure-callout\; id="1"\; label="z\; wi"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">z</figure-callout>}}_{\mathrm{wi}}& 1/a^2+(2/a)x_{\mathrm{wi}}+{\mathrm{<figure-callout\; id="2"\; label="l"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">l</figure-callout>}}^2\end{array}\right]$

T is promptly arranged ^TCt=0, C are symmetric matrix, so t is the point on the quafric curve.To get a pair of point be the space sphere of R as radius if in the same circular arc of the web of the rail, appoint, and here R equals the radius of web of the rail arc section, then can obtain two intersection points:

$\left\{\begin{array}{c}{t}^T{C}_{i}t=0\\ t^T{C}_{j}t=0\end{array}\right.$

Can know that by detecting principle the near point of two intersection point middle distance video camera photocentres (just apart from the world coordinate system initial point) is the centre point of the arc section of asking, it can calculate acquisition at three-dimensional coordinate figure.In actual engineering,, can obtain out m center of circle X because every section arc profile has m to individual point in the image ₁, X ₂..., X _mCoordinate, the order:

d _k=(x _wo-x _wk) ²+(y _wo-y _wk) ²+(z _wo-z _wk)2

X wherein _Wo, y _Wo, z _WoBe center of circle optimum point X _oCoordinate components, x _Wk, y _Wk, z _WkBe X _k(k=1,2 ..., coordinate components m) can be obtained X through following formula _oOptimum solution:

$\left\{\begin{array}{c}\underset{k=1}{\overset{m}{\mathrm{\&Sigma;}}}{d}_k=L\\ \frac{\&PartialD;L}{{\&PartialD;x}_{\mathrm{wo}}}=\frac{\&PartialD;L}{\&PartialD;y_{\mathrm{wo}}}=\frac{\&PartialD;L}{\&PartialD;z_{\mathrm{wo}}}=0\end{array}\right.$

(2) rail head lower end feature point extraction:

Because structured light and position for video camera are above the rail side on the detection principle, the real image rail head and the web of the rail exist and significantly cut apart, and therefore can search for the location fast.On the other hand, because can there be the deviation of a pixel in image mostly through micronization processes, so the rail head lower extreme point that obtains of actual search location and 8 adjoint points thereof all might be actual lower extreme points, { X _B1, X _B2..., X _B9Coordinate can calculate.Ask itself and web of the rail arc section centre point X _oDistance { d _Ob1, d _Ob2..., d _Ob9, the error of note theoretical and actual range is ε i ₌| d _Obi-d _Ob|, i=1,2 ..., 9, d wherein _ObBe theoretical.Get ε _Min=min{ ε ₁, ε ₂..., ε ₉, keep in mind ε _MinThe time pairing X _BiBe X _b, be the rail head lower extreme point of being asked.

Detailed process based on unique point generation rail profile standard form in the step 5) is following:

Establishing space coordinates needs not three points on same straight line, after web of the rail arc section unique point and rail head lower extreme point are definite, take up an official post at the rail profile and to get not and X _oAnd X _b1 X of conllinear _r, get vector

With X _oPoint is set up cartesian coordinate system for initial point, and the vector of x ' direction does

Then the vector of z ' direction does $\stackrel{\&RightArrow;}{o^{\&prime;}{z}^{\&prime;}}=\stackrel{\&RightArrow;}{o^{\&prime;}{x}^{\&prime;}}\&times;\stackrel{\&RightArrow;}{X_o{X}_r},$ The vector of y ' direction does

In case this coordinate system is set up, can know that by the space profiles geometric relationship of standard rail the space profiled outline of whole rail is able to set up, therefore, can in every frame measurement image, generate the standard form of a rail profile dynamically.

To obtain the detailed process of wearing valve following for the wide and standard form of contrast rail measuring wheel in the step 6):

The measurement image that photographs is handled the measurement light belt that obtains the rail profile, calculate to set up and be positioned at the video camera photocentre rail profile 3 d space coordinate under boundary's coordinate system of conducting oneself in society.Set up new world coordinates; Make its initial point be positioned at the center of rail foot, x " axle overlaps y with the rail base " spool and rail profile central lines; Z " perpendicular to x " y " plane;, therefore can the volume coordinate on the rail profile be mapped to coordinate system o " x " y " x " down, thereby the wearing valve of calculating because the coordinate system o ' x ' y ' z ' that sets up through unique point exists clear and definite rotation and translation relation with o " x " y " z ".

Because adopted above technical scheme, the present invention compared with prior art has following advantage and good effect:

1. do not need to measure the analysis of aliging of profile and reference design profile, broken traditionally, improved the accuracy of measurement of wear greatly based on the difficult problem of static template coupling;

2. reduce the operand of image analysis processing and had good stable property.

Certainly, any one specific embodiment of embodiment of the present invention content might not reach above whole technique effect simultaneously.

Description of drawings

Fig. 1 is the process flow diagram of the measurement of rail wear based method that proposes of the present invention;

Fig. 2 is the measurement of wear systematic schematic diagram of this method;

Fig. 3 is the vision measurement mathematical model of this method;

Fig. 4 is a rail head lower extreme point identification synoptic diagram;

Fig. 5 is a web of the rail center of arc synoptic diagram;

Fig. 6 is a standard rail contour images among the embodiment;

Fig. 7 a is the rail profile that measures and the three dimensions contour curve comparison diagram of nominal contour;

Fig. 7 b is the face profile curve map after Fig. 7 a coordinate conversion.

Embodiment

For technological means, creation characteristic that the present invention is realized, reach purpose and effect and be easy to understand and understand, below in conjunction with diagram and specific embodiment, further set forth the present invention.

As shown in Figure 1, the measurement of rail wear based method based on dynamic template that the present invention proposes may further comprise the steps:

1) at two one steel rail inside fronts ccd video camera and fan-shaped LASER Light Source are installed respectively, the optical plane that fan-shaped LASER Light Source sends is vertical perpendicular to rail, and concrete mounting means is as shown in Figure 2;

2) demarcate each ccd video camera and light-plane parameters;

3) obtain the volume coordinate of each pixel of rail profile in the image according to rail profile SPATIAL CALCULATION model;

4) extract the web of the rail and rail head lower end unique point coordinate;

5) generate rail profile standard form based on unique point;

6) contrast rail measuring wheel exterior feature and standard form obtain wearing valve.

Its principle is:

At first ccd video camera and light-plane parameters are demarcated, wherein the camera calibration process is:

Video camera is fixed on the plane, takes the gridiron pattern scaling board image that is positioned at diverse location more than three, mate, calculate the mapping matrix between them, and solve camera parameters through this matrix through point and picture point on the scaling board,

$z_{\mathrm{<figure-callout\; id="1"\; label="c\; u\; v"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">c</figure-callout>}}\left(\begin{array}{c}u\\ v\end{array}\right)\left(\begin{array}{c}\\ 1\end{array}\right)=\left(\begin{array}{cccc}{f}_x& 0& {\mathrm{<figure-callout\; id="0"\; label="u"\; filenames="HDA00001934594500041.png"\; state="\{\{state\}\}">u</figure-callout>}}_0& 0\\ 0& {\mathrm{<figure-callout\; id="0"\; label="f\; y\; v"\; filenames="HDA00001934594500041.png"\; state="\{\{state\}\}">f</figure-callout>}}_y& v_{}{}_0& 0\\ 0& 0& 1& 0\end{array}\right)\left(\begin{array}{cc}R& T\\ 0^T& 1\end{array}\right)\left(\begin{array}{c}{x}_w\\ y_{\mathrm{<figure-callout\; id="1"\; label="w\; z\; w"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">w</figure-callout>}}& \\ z_w{}_{}\\ 1\end{array}\right)=\left(\begin{array}{cccc}{m}_{11}& m_{12}& m_{13}& m_{14}\\ m_{21}& m_{22}& m_{23}& m_{24}\\ m_{31}& m_{32}& m_{33}& m_{34}\end{array}\right)\left(\begin{array}{c}{x}_w\\ y_w\\ z_w\\ 1\end{array}\right)---\left(1\right)$

Z wherein _cBe the component of certain intersection point in camera coordinate system on the gridiron pattern, f _x, f _y, u ₀, v ₀Inner parameter for video camera; R is a rotation matrix with orthogonality, and T is a translation matrix, and they are external parameters of video camera.x _w, y _w, z _w, 1 is the coordinate of i the point in space; (u, v, 1) is the image coordinate of i point; m _IjThe capable j column element of i for projection matrix M; Formula (1) is decomposed cancellation z _cCan obtain about m _IjLinear equation:

$\left\{\begin{array}{c}{x}_w{m}_{11}+y_w{m}_{12}+z_w{m}_{13}+m_{14}-{\mathrm{ux}}_w-{\mathrm{uy}}_w{m}_{32}-{\mathrm{uz}}_w{m}_{33}={\mathrm{um}}_{34}\\ x_w{m}_{21}+y_w{m}_{22}+z_w{m}_{23}+m_{24}-v{x}_w-v{y}_w{m}_{32}-{\mathrm{vz}}_w{m}_{32}={\mathrm{vm}}_{34}\end{array}\right.---\left(2\right)$

On the calibration piece n known point arranged, obtains 2n linear equation matrix about the Metzler matrix element:

$\left[\begin{array}{cccccccccc}{x}_{w1}& y_{w1}& {\mathrm{<figure-callout\; id="1"\; label="z\; w"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">z</figure-callout>}}_w& 1& 0& 0& 0& -u_1{x}_{w1}& -u_1{y}_{w1}& -u_1{\mathrm{<figure-callout\; id="1"\; label="z\; w"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">z</figure-callout>}}_w\\ 0& 0& 0& 0& x_{w1}& y_{w1}& z_{w1}& -v_1{x}_{w1}& -v_1{y}_{w1}& -v_1{\mathrm{<figure-callout\; id="1"\; label="z\; w"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">z</figure-callout>}}_w\\ & & & & ...& ...& ...& & & \\ & & & & ...& ...& ...& & & \\ x_{\mathrm{wn}}& y_{\mathrm{wn}}& {\mathrm{<figure-callout\; id="1"\; label="z\; wn"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">z</figure-callout>}}_{\mathrm{wn}}& 1& 0& 0& 0& -u_n{x}_{\mathrm{wn}}& -u_n{y}_{\mathrm{wn}}& -u_{\mathrm{<figure-callout\; id="0"\; label="n\; z\; wn"\; filenames="HDA00001934594500041.png"\; state="\{\{state\}\}">n</figure-callout>}}{z}_{\mathrm{wn}}{}_{}\\ 0& 0& 0& 0& x_{\mathrm{wn}}& y_{\mathrm{wn}}& z_{\mathrm{wn}}& -v_n{x}_{\mathrm{wn}}& -v_n{y}_{\mathrm{wn}}& -v_n{z}_{\mathrm{wn}}\end{array}\right]\&times;\left[\begin{array}{c}{m}_{11}\\ m_{12}\\ m_{13}\\ m_{14}\\ m_{21}\\ m_{22}\\ m_{23}\\ m_{24}\\ m_{31}\\ m_{32}\\ m_{33}\end{array}\right]=\left[\begin{array}{c}{u}_1{m}_{34}\\ v_1{m}_{34}\\ ...\\ ...\\ ...\\ \\ ...\\ ...\\ ...\\ u_n{m}_{34}\\ v_n{m}_{34}\end{array}\right]---\left(3\right)$

In formula (2), specify m ₃₄=1, obtain 2n linear equation about other elements of Metzler matrix, the number of unknown element is 11, is designated as 11 dimensional vector m, and formula (3) is write a Chinese character in simplified form:

Km＝U (4)

Wherein, K is left side 2n * 11 matrixes, and m is 11 unknown dimensional vectors, and U is the 2n dimensional vector on the right, and K, U are known vector.Use least square method to obtain as 2n>11 the time above-mentioned linear equation separate for:

m=K ^TK ^-1K ^TU (5)

M vector and m ₃₄=1 constituted the institute find the solution Metzler matrix; Therefore, by 6 above known points in space and their picture point coordinate, can obtain Metzler matrix.

After obtaining Metzler matrix, can divide the whole inside and outside parameter that calculate video camera by relation.

The optical plane calibration process is:

Increase auxiliary camera C during calibration ₂, with video camera C ₁Form the binocular scaling system, the common shooting is positioned at the light belt image on the scaling board.Be based upon video camera C with the world coordinate system initial point this moment ₁The photocentre place, through respectively to C ₁, C ₂Calibration can be confirmed its inner parameter A ₁, A ₂And external parameter B ₁, B ₂Then their projection matrix is respectively M ₁=A ₁B ₁(because world coordinate system is based upon C ₁So the photocentre place is B ₁=I, I are unit matrix)=A ₁, M ₂=A ₂B ₂=A ₂[R ₁₂, T ₁₂] ([R ₁₂, T ₁₂] be video camera C ₁, C ₂Between the position relational matrix).CCD picked-up light belt forms after the RGB image, it is extracted Flame Image Process such as component, gray processing, the computing of difference shadow, binaryzation, refinement, to extract the pixel coordinate of each point on the striation axis.Projection has following form:

$Z_1\left(\begin{array}{c}{u}_1\\ {\mathrm{<figure-callout\; id="1"\; label="v"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">v</figure-callout>}}_1\\ 1\end{array}\right)=M_1\left(\begin{array}{c}{x}_w\\ y_{\mathrm{<figure-callout\; id="1"\; label="w\; z\; w"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">w</figure-callout>}}& \\ z_w{}_{}\\ 1\end{array}\right)=\left(\begin{array}{cccc}{m}_{11}^1& m_{12}^1& m_{13}^1& m_{14}^1\\ m_{21}^1& m_{22}^1& m_{23}^1& m_{24}^1\\ m_{31}^1& m_{32}^1& m_{33}^1& m_{34}^1\end{array}\right)\left(\begin{array}{c}{x}_w\\ y_w\\ z_w\\ 1\end{array}\right)---\left(6\right)$

$Z_2\left(\begin{array}{c}{u}_2\\ {\mathrm{<figure-callout\; id="2"\; label="v"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">v</figure-callout>}}_2\\ 1\end{array}\right)=M_2\left(\begin{array}{c}{x}_w\\ y_{\mathrm{<figure-callout\; id="1"\; label="w\; z\; w"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">w</figure-callout>}}& \\ z_w{}_{}\\ 1\end{array}\right)=\left(\begin{array}{cccc}{m}_{11}^2& m_{12}^2& m_{13}^2& m_{14}^2\\ m_{21}^2& m_{22}^2& m_{23}^2& m_{24}^2\\ m_{31}^2& m_{32}^2& m_{33}^2& m_{34}^2\end{array}\right)\left(\begin{array}{c}{x}_w\\ y_w\\ z_w\\ 1\end{array}\right)---\left(7\right)$

Wherein, Z ₁, Z ₂Be the component of any point P on the scaling board light belt in two camera coordinates, but cancellation during calculating; (u ₁, v ₁), (u ₂, v ₂) be respectively a P at C ₁, C ₂In pixel coordinate.Simultaneous formula (6) and formula (7) obtain:

$\left\{\begin{array}{c}(u_1{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{31}-{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{11})x_w+(u_1{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{32}-{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{12})y_w+(u_1{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{33}-{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{33})z_w={{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{14}-u_1{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{34}\\ (v_1{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{31}-{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{11})x_w+(v_1{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{32}-{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{22})y_w+(v_1{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{33}-{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{23})z_w={{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{24}-v_1{{\mathrm{<figure-callout\; id="1"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^1}_{34}\\ (u_2{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{31}-{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{11})x_w+(u_2{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{32}-{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{12})y_w+(u_2{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{33}-{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{13})z_w={{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{14}-u_2{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{34}\\ (v_2{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{31}-{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{21})x_w+(v_2{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{22}-{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{22})y_w+(v_2{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{33}-{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{23})z_w={{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{24}-v_2{{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}_{34}\end{array}\right.---\left(8\right)$

Thus, can obtain the world coordinates of a P.In like manner, constantly move the locus of scaling board, can obtain (the m > of m on the light belt; 3) world coordinates (x of individual point _Wi, y _Wi, z _Wi) (i=1,2,3 ... m).

The equation on spatial light plane can be represented as follows:

Ax _w+By _w+Cz _w+1=0 (9)

Wherein, A, B, C are 3 components of this planar process vector n.

Because the world coordinates of unique point also satisfies optic plane equations on the light belt, overdetermined equation group of then available m unique point structure, its matrix form is:

$\left[\begin{array}{ccc}{x}_{w1}& y_{w1}& z_{w1}\\ x_{w2}& y_{w2}& {\mathrm{<figure-callout\; id="2"\; label="z\; w"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">z</figure-callout>}}_w\\ .& .& .\\ .& .& .\\ .& .& .\\ x_{\mathrm{wm}}& y_{\mathrm{wm}}& z_{\mathrm{wm}}\end{array}\right]\left[\begin{array}{c}A\\ B\\ C\end{array}\right]=\left[\begin{array}{c}-1\\ -1\\ .\\ .\\ .\\ -1\end{array}\right]---\left(10\right)$

Or be abbreviated as:

GS=L (11)

Wherein, G is the matrix of coefficients on the equality left side, S=[A, B, C] ^T, L=[1,1 ..., 1] ^T, utilize least square method can obtain S=(G then ^TG) ^-1G ^TL and coefficient A, B, C.

Below how explanation obtains the volume coordinate of each pixel of rail profile in the image according to rail profile SPATIAL CALCULATION model, as shown in Figure 3.

Make O _wX _wY _wZ _wBe world coordinate system, O _cX _cY _cZ _cBe camera coordinate system, O _uX _uY _uBe image coordinate system.In native system, world coordinate system is based upon camera coordinates fastens, make both overlap fully, then there are not rotation and translation relation between the two, at this moment

$\left(\begin{array}{cc}R& T\\ 0^{\mathrm{<figure-callout\; id="1"\; label="T"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">T</figure-callout>}}& 1\end{array}\right)=I$

For unit matrix and z is arranged _c=z _w, then

$z_{\mathrm{<figure-callout\; id="1"\; label="c\; u\; v"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">c</figure-callout>}}\left(\begin{array}{c}u\\ v\end{array}\right)\left(\begin{array}{c}\\ 1\end{array}\right)=\left(\begin{array}{cccc}{f}_x& 0& {\mathrm{<figure-callout\; id="0"\; label="u"\; filenames="HDA00001934594500041.png"\; state="\{\{state\}\}">u</figure-callout>}}_0& 0\\ 0& {\mathrm{<figure-callout\; id="0"\; label="f\; y\; v"\; filenames="HDA00001934594500041.png"\; state="\{\{state\}\}">f</figure-callout>}}_y& v_{}{}_0& 0\\ 0& 0& 1& 0\end{array}\right)\left(\begin{array}{c}{x}_w\\ y_w\\ z_w\\ 1\end{array}\right)---\left(12\right)$

This point also in the plane that laser throwed, therefore satisfies the spatial light plane equation simultaneously, and simultaneous formula (12) and formula (9) can obtain:

$\left(\begin{array}{ccc}{f}_x& 0& u_0-u\\ 0& f_y& v_0-v\\ a& b& c\end{array}\right)\left(\begin{array}{c}{x}_w\\ y_w\\ z_w\end{array}\right)=\left(\begin{array}{c}0\\ 0\\ -1\end{array}\right)---\left(13\right)$

Thereby obtain the world coordinates (x of any point on the rail profile outline line _w, y _w, z _w).

Below the method for the web of the rail and rail head lower end unique point coordinate is extracted in explanation.

Web of the rail feature point extraction

Appoint in the web of the rail circular arc behind three-dimensional reconstruction and get 1 X _i(x _Wi, y _Wi, z _Wi), make the space sphere with radius R, note space spherical equation is:

X ^TQX=0 (14)

Wherein x is the coordinate of putting on the sphere, X=[x _w, y _w, z _w, 1] ^T, Q is 4 * 4 symmetric matrix:

$Q=\left[\begin{array}{cccc}1& 0& 0& -{\mathrm{<figure-callout\; id="0"\; label="x\; wi"\; filenames="HDA00001934594500041.png"\; state="\{\{state\}\}">x</figure-callout>}}_{\mathrm{wi}}\\ 0& 1& 0& -{\mathrm{<figure-callout\; id="0"\; label="y\; wi"\; filenames="HDA00001934594500041.png"\; state="\{\{state\}\}">y</figure-callout>}}_{\mathrm{wi}}\\ 0& 0& 1& -z_{\mathrm{wi}}\\ -x_{\mathrm{wi}}& -y_{\mathrm{wi}}& -{\mathrm{<figure-callout\; id="2"\; label="z\; wi\; l"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">z</figure-callout>}}_{\mathrm{wi}}& l^{}{}^2\end{array}\right],$ l ²=x _wi ²+y _wi ²+z _wi ²-R ²。

Because the point on the rail outline line makes x in optical plane _t=x _w, y _t=y _w, then optic plane equations can be write as:

X=Mt (15)

Wherein

$t=\left[\begin{array}{c}{x}_{\mathrm{<figure-callout\; id="1"\; label="t\; y\; t"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">t</figure-callout>}}& \\ y_t{}_{}\\ 1\end{array}\right].$ Can get X by formula (14) and formula (15) ^TQ=X ^Tt ^TM=0,

Order $C=M^T\mathrm{QM}=\left[\begin{array}{ccc}{(b/a)}^2+1& \mathrm{Bc}/a^2& b/a^2+(b/c)x_{\mathrm{Wi}}-y_{\mathrm{Wi}}\\ \mathrm{Bc}/a^2& {(c/a)}^2+1& c/a^2+(c/a)x_{\mathrm{Wi}}-z_{\mathrm{Wi}}\\ b/a^2+(b/a)x_{\mathrm{Wi}}-y_{\mathrm{Wi}}& c/a^2+(c/a)x_{\mathrm{Wi}}-{\mathrm{<figure-callout\; id="1"\; label="z\; Wi"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">z</figure-callout>}}_{\mathrm{Wi}}& 1/a^2+(2/a)x_{\mathrm{Wi}}+{\mathrm{<figure-callout\; id="2"\; label="l"\; filenames="HDA00001934594500023.png"\; state="\{\{state\}\}">l</figure-callout>}}^2\end{array}\right],$

T is promptly arranged ^TCt=0, C are symmetric matrix, so t is the point on the quafric curve.To get a pair of point be the space sphere of R as radius if in the same circular arc of the web of the rail, appoint, and here R equals the radius of web of the rail arc section, then can obtain two intersection points, shown in (16):

$\left\{\begin{array}{c}{t}^T{C}_{i}t=0\\ t^T{C}_{j}t=0\end{array}\right.---\left(16\right)$

Can know that by detecting principle the near point of two intersection point middle distance video camera photocentres (just apart from the world coordinate system initial point) is the centre point of the arc section of asking, it can calculate acquisition by formula (15) at three-dimensional coordinate figure.In actual engineering,, can obtain out m center of circle X because every section arc profile has m to individual point in the image ₁, X ₂..., X _mCoordinate, the order:

d _k=(x _wo-x _wk) ²+(y _wo-y _wk) ²+(z _wo-z _wk) ²

X wherein _Wo, y _Wo, z _WoBe center of circle optimum point X _oCoordinate components, x _Wk, y _Wk, z _WkBe X _k(k=1,2 ..., coordinate components m), through type (17) can be obtained X _oOptimum solution:

$\left\{\begin{array}{c}\underset{k=1}{\overset{m}{\mathrm{\&Sigma;}}}{d}_k=L\\ \frac{\&PartialD;L}{{\&PartialD;x}_{\mathrm{wo}}}=\frac{\&PartialD;L}{\&PartialD;y_{\mathrm{wo}}}=\frac{\&PartialD;L}{\&PartialD;z_{\mathrm{wo}}}=0\end{array}\right.---\left(17\right)$

Rail head lower end feature point extraction

Because structured light and position for video camera are above the rail side on the detection principle, the real image rail head and the web of the rail exist and significantly cut apart, and therefore can search for the location fast.On the other hand, because can there be the deviation of a pixel in image mostly through micronization processes, so the rail head lower extreme point that obtains of actual search location and 8 adjoint points thereof all might be actual lower extreme points, and through type (13) is asked its { X _B1, X _B2..., X _B9) coordinate, as shown in Figure 4.Ask itself and web of the rail arc section centre point X _oDistance { d _Ob1, d _Ob2..., d _Ob9, the error of note theoretical and actual range is ε _i=| d _Obi-d _Ob|, i=1,2 ..., 9, d wherein _ObBe theoretical.Get

Keep in mind ε _MinThe time pairing X _BiBe X _b, be the rail head lower extreme point of being asked.

Below explanation generates the method for rail profile standard form based on unique point.

Need not three points on same straight line owing to establish space coordinates, after web of the rail arc section unique point and rail head lower extreme point are definite, take up an official post at the rail profile and to get not and X _oAnd X _b1 X of conllinear _r, get vector

With X _oPoint is set up cartesian coordinate system for initial point, and the vector of x ' direction does

Then the vector of z ' direction does $\stackrel{\&RightArrow;}{o^{\&prime;}{z}^{\&prime;}}=\stackrel{\&RightArrow;}{o^{\&prime;}{x}^{\&prime;}}\&times;\stackrel{\&RightArrow;}{X_o{X}_r},$ The vector of y ' direction does

As shown in Figure 5.In case this coordinate system is set up, can know that by the space profiles geometric relationship of standard rail the space profiled outline of whole rail is able to set up, therefore, can in every frame measurement image, generate the standard form of a rail profile dynamically.

Below how explanation contrasts the wide and standard form of rail measuring wheel and obtains wearing valve.

The measurement image that photographs is handled the measurement light belt that obtains the rail profile, calculate to set up according to formula (13) and be positioned at the video camera photocentre rail profile 3 d space coordinate under boundary's coordinate system of conducting oneself in society.Set up new world coordinates; Make its initial point be positioned at the center of rail foot, x " axle overlaps y with the rail base " spool and rail profile central lines; Z " perpendicular to x " y " plane;, therefore can the volume coordinate on the rail profile be mapped to coordinate system o " x " y " z " down, thereby the wearing valve of calculating because the coordinate system o ' x ' y ' z ' that sets up through unique point exists clear and definite rotation and translation relation with o " x " y " z ".

Embodiment:

It is detected object that present embodiment adopts standard 60 rails; Camera adopts the CCD industrial camera of German AVT-FC125 1934 interfaces; Resolution is 1280 * 960 pixels, in practice, can start windowing function, improves Flame Image Process and data transmission capabilities simultaneously not reducing image resolution ratio.Laser output wavelength is the red planar structure light of 635nm, and power is 40mW.In the laboratory, carry out the calibration of camera parameter and optic plane equations parameter earlier, confirmed the space geometry relation of camera and optical plane.60 standard rail section profiles are taken; As shown in Figure 6; Through obtaining its space curve behind the three-dimensional reconstruction, calculate the theoretical profile spatial value of relevant position through obtaining of unique point, table 1 has been listed the measured value and the template theory value that dynamically generates of rail head profile parameters.See that from error the maximum error of x component is 0.0509mm, the maximum error of y component is 0.1846mm, and the maximum error of z component is 0.1531mm.Because detection system requires optical plane to overlap with rail profile fully in theory, the actual detected system is difficult to guarantee this condition, therefore has certain error.Error also relates to calibration error and non-sub-pix error of Flame Image Process of camera parameter and light-plane parameters etc.Above global error can satisfy the measurement requirement of abrasion less than 0.16mm.

The comparison of table 1 rail head of rail measurement point and theoretical value

The detection system of accomplishing calibration is installed on the small-sized railcar of experiment, and small-sized railcar satisfies railway security of operation standard fully, can run on the circuit that standard gauge is 1435mm highest running speed 45km/h.In the experiment same location of rail is measured, Fig. 7 a and Fig. 7 b are the rail profile that measures and the curve comparison diagram of nominal contour, and table 2 is 10 resulting measurement of wear values of duplicate measurements.Measurement result repeatability is high, and standard deviation has good stable property in 0.07mm.

Table 2 measurement of wear result

The present invention proposes to carry out the rail web of the rail and rail head lower end feature point extraction based on real image, dynamically generates the rail profile template under the world coordinate system according to unique point, and then compares the measuring method of calculating the acquisition wearing valve.This method only need confirm to measure two unique points of rail can set up measurement coordinate system according to the geometric relationship of standard rail profile; And can obtain standard form fast, accurately; Do not need to measure the analysis of aliging of profile and reference design profile; Broken traditional problem, improved the accuracy of measurement of wear greatly based on static template coupling difficulty; Reduce the operand of image analysis processing and had good stable property.

More than show and described ultimate principle of the present invention, principal character and advantage of the present invention.The technician of the industry should understand; The present invention is not restricted to the described embodiments; That describes in the foregoing description and the instructions just explains principle of the present invention; The present invention also has various changes and modifications under the prerequisite that does not break away from spirit and scope of the invention, and these variations and improvement all fall in the scope of the invention that requires protection.The present invention requires protection domain to be defined by appending claims and equivalent thereof.

Claims (8)

1. the measurement of rail wear based method based on dynamic template is characterized in that, may further comprise the steps:

1) at two one steel rail inside fronts ccd video camera and fan-shaped LASER Light Source are installed respectively;

2) demarcate each ccd video camera and light-plane parameters;

3) obtain the volume coordinate of each pixel of rail profile in the image according to rail profile SPATIAL CALCULATION model;

4) extract the web of the rail and rail head lower end unique point coordinate;

5) generate rail profile standard form based on unique point;

6) contrast rail measuring wheel exterior feature and standard form obtain wearing valve.

2. measurement of rail wear based method as claimed in claim 1; It is characterized in that; Every side rail is provided with a ccd video camera and a fan-shaped LASER Light Source in the step 1), all is positioned at the rail inside front, and the optical plane that sends of fan-shaped LASER Light Source and rail is vertical vertical.

3. measurement of rail wear based method as claimed in claim 1 is characterized in that step 2) in to demarcate the detailed process of each ccd video camera parameter following,

Video camera is fixed on the plane, takes the gridiron pattern scaling board image that is positioned at diverse location more than three, mate, calculate the mapping matrix between them through point and picture point on the scaling board:

$z_c\left(\begin{array}{c}u\\ v\\ 1\end{array}\right)=\left(\begin{array}{cccc}{f}_x& 0& u_0& 0\\ 0& f_y& v_0& 0\\ 0& 0& 1& 0\end{array}\right)\left(\begin{array}{cc}R& T\\ 0^T& 1\end{array}\right)\left(\begin{array}{c}{x}_w\\ y_w\\ z_w\\ 1\end{array}\right)=\left(\begin{array}{cccc}{m}_{11}& m_{12}& m_{13}& m_{14}\\ m_{21}& m_{22}& m_{23}& m_{24}\\ m_{31}& m_{32}& m_{33}& m_{34}\end{array}\right)\left(\begin{array}{c}{x}_w\\ y_w\\ z_w\\ 1\end{array}\right),$

And solve camera parameters through this matrix; Z wherein _cBe the component of certain intersection point in camera coordinate system on the gridiron pattern, f _x, f _y, u ₀, v ₀Inner parameter for video camera; R is a rotation matrix with orthogonality, and T is a translation matrix, and they are external parameters of video camera; x _w, y _w, z _w, 1 is the coordinate of i the point in space; (u, v, 1) is the image coordinate of i point; m _IjThe capable j column element of i for projection matrix M; Decompose cancellation z _cObtain about m _IjLinear equation:

$\left\{\begin{array}{c}{x}_w{m}_{11}+y_w{m}_{12}+z_w{m}_{13}+m_{14}-{\mathrm{ux}}_w-{\mathrm{uy}}_w{m}_{32}-{\mathrm{uz}}_w{m}_{33}={\mathrm{um}}_{34}\\ x_w{m}_{21}+y_w{m}_{22}+z_w{m}_{23}+m_{24}-v{x}_w-v{y}_w{m}_{32}-{\mathrm{vz}}_w{m}_{32}={\mathrm{vm}}_{34}\end{array}\right.,$

On the calibration piece n known point arranged, obtains 2n linear equation matrix about the Metzler matrix element:

$\left[\begin{array}{cccccccccc}{x}_{w1}& y_{w1}& z_{w1}& 1& 0& 0& 0& -u_1{x}_{w1}& -u_1{y}_{w1}& -u_1{z}_{w1}\\ 0& 0& 0& 0& x_{w1}& y_{w1}& z_{w1}& -v_1{x}_{w1}& -v_1{y}_{w1}& -v_1{z}_{w1}\\ & & & & ...& ...& ...& & & \\ & & & & ...& ...& ...& & & \\ x_{\mathrm{wn}}& y_{\mathrm{wn}}& z_{\mathrm{wn}}& 1& 0& 0& 0& -u_n{x}_{\mathrm{wn}}& -u_n{y}_{\mathrm{wn}}& -u_n{z}_{\mathrm{wn}}\\ 0& 0& 0& 0& x_{\mathrm{wn}}& y_{\mathrm{wn}}& z_{\mathrm{wn}}& -v_n{x}_{\mathrm{wn}}& -v_n{y}_{\mathrm{wn}}& -v_n{z}_{\mathrm{wn}}\end{array}\right]\&times;\left[\begin{array}{c}{m}_{11}\\ m_{12}\\ m_{13}\\ m_{14}\\ m_{21}\\ m_{22}\\ m_{23}\\ m_{24}\\ m_{31}\\ m_{32}\\ m_{33}\end{array}\right]=\left[\begin{array}{c}{u}_1{m}_{34}\\ v_1{m}_{34}\\ ...\\ ...\\ ...\\ \\ ...\\ ...\\ ...\\ u_n{m}_{34}\\ v_n{m}_{34}\end{array}\right],$

Specify m ₃₄=1, obtain 2n linear equation about other elements of Metzler matrix, the number of unknown element is 11, is designated as 11 dimensional vector m, writes a Chinese character in simplified form: Km=U, K are left side 2n * 11 matrixes; M is 11 unknown dimensional vectors; U is the 2n dimensional vector on the right; K, U are known vector; Obtain > with least square method as 2n; 11 o'clock above-mentioned linear equations separate for:

m＝K ^TK ^-1K ^TU，

M vector and m ₃₄=1 constituted the institute find the solution Metzler matrix; By 6 above known points in space and their picture point coordinate, obtain Metzler matrix;

After obtaining Metzler matrix, divide the whole inside and outside parameter that calculate video camera by relation.

4. measurement of rail wear based method as claimed in claim 1 is characterized in that step 2) in to demarcate the detailed process of light-plane parameters following:

Increase auxiliary camera C during calibration ₂, with video camera C ₁Form the binocular scaling system, the common shooting is positioned at the light belt image on the scaling board, and be based upon video camera C with the world coordinate system initial point this moment ₁The photocentre place, through respectively to C ₁, C ₂Its inner parameter A is confirmed in calibration ₁, A ₂And external parameter B ₁, B ₂, their projection matrix is respectively M ₁=A ₁B ₁=A ₁, M ₂=A ₂B ₂=A ₂[R ₁₂, T ₁₂], CCD picked-up light belt forms after the RGB image, it is extracted Flame Image Process such as component, gray processing, the computing of difference shadow, binaryzation, refinement, and to extract the pixel coordinate of each point on the striation axis, projection has following form:

$Z_1\left(\begin{array}{c}{u}_1\\ v_1\\ 1\end{array}\right)=M_1\left(\begin{array}{c}{x}_w\\ y_w\\ z_w\\ 1\end{array}\right)=\left(\begin{array}{cccc}{m}_{11}^1& m_{12}^1& m_{13}^1& m_{14}^1\\ m_{21}^1& m_{22}^1& m_{23}^1& m_{24}^1\\ m_{31}^1& m_{32}^1& m_{33}^1& m_{34}^1\end{array}\right)\left(\begin{array}{c}{x}_w\\ y_w\\ z_w\\ 1\end{array}\right),$

$Z_2\left(\begin{array}{c}{u}_2\\ v_2\\ 1\end{array}\right)=M_2\left(\begin{array}{c}{x}_w\\ y_w\\ z_w\\ 1\end{array}\right)=\left(\begin{array}{cccc}{m}_{11}^2& m_{12}^2& m_{13}^2& m_{14}^2\\ m_{21}^2& m_{22}^2& m_{23}^2& m_{24}^2\\ m_{31}^2& m_{32}^2& m_{33}^2& m_{34}^2\end{array}\right)\left(\begin{array}{c}{x}_w\\ y_w\\ z_w\\ 1\end{array}\right),$

Wherein, Z ₁, Z ₂Be the component of some P in two camera coordinates on the scaling board light belt; (u ₁, v ₁), (u ₂, v ₂) be respectively a P at C ₁, C ₂In pixel coordinate; Above two formulas of simultaneous obtain:

$\left\{\begin{array}{c}(u_1{m^1}_{31}-{m^1}_{11})x_w+(u_1{m^1}_{32}-{m^1}_{12})y_w+(u_1{m^1}_{33}-{m^1}_{33})z_w={m^1}_{14}-u_1{m^1}_{34}\\ (v_1{m^1}_{31}-{m^1}_{11})x_w+(v_1{m^1}_{32}-{m^1}_{22})y_w+(v_1{m^1}_{33}-{m^1}_{23})z_w={m^1}_{24}-v_1{m^1}_{34}\\ (u_2{m^2}_{31}-{m^2}_{11})x_w+(u_2{m^2}_{32}-{m^2}_{12})y_w+(u_2{m^2}_{33}-{m^2}_{13})z_w={m^2}_{14}-u_2{m^2}_{34}\\ (v_2{m^2}_{31}-{m^2}_{21})x_w+(v_2{m^2}_{22}-{m^2}_{22})y_w+(v_2{m^2}_{33}-{m^2}_{23})z_w={m^2}_{24}-v_2{m^2}_{34}\end{array}\right.,$

Thus, obtain the world coordinates of a P; Constantly move the locus of scaling board, obtain (the m > of m on the light belt; 3) world coordinates (x of individual point _Wi, y _Wi, z _Wi) (i=1,2,3 ... m);

The The Representation Equation on spatial light plane is following:

Ax _w+By _w+Cz _w+1=0，

Wherein, A, B, C are 3 components of this planar process vector n;

With overdetermined equation group of m unique point structure, its matrix form is:

$\left[\begin{array}{ccc}{x}_{w1}& y_{w1}& z_{w1}\\ x_{w2}& y_{w2}& z_{w2}\\ .& .& .\\ .& .& .\\ .& .& .\\ x_{\mathrm{wm}}& y_{\mathrm{wm}}& z_{\mathrm{wm}}\end{array}\right]\left[\begin{array}{c}A\\ B\\ C\end{array}\right]=\left[\begin{array}{c}-1\\ -1\\ .\\ .\\ .\\ -1\end{array}\right],$

Or be abbreviated as GS=L, wherein, G is the matrix of coefficients on the equality left side, S=[A, B, C] ^T, L=[1,1 ..., 1] ^T, utilize least square method to obtain S=(G then ^TG) ^-1G ^TL and coefficient A, B, C.

5. measurement of rail wear based method as claimed in claim 1 is characterized in that, it is following to obtain the detailed process of the volume coordinate of each pixel of rail profile in the image according to rail profile SPATIAL CALCULATION model in the step 3):

World coordinate system is based upon camera coordinates fastens, make both overlap fully, then do not have rotation and translation relation between the two, at this moment,

$\left(\begin{array}{cc}R& T\\ 0^T& 1\end{array}\right)=I$

Be unit matrix, and z is arranged _c=z _w, then

$z_c\left(\begin{array}{c}u\\ v\\ 1\end{array}\right)=\left(\begin{array}{cccc}{f}_x& 0& u_0& 0\\ 0& f_y& v_0& 0\\ 0& 0& 1& 0\end{array}\right)\left(\begin{array}{c}{x}_w\\ y_w\\ z_w\\ 1\end{array}\right);$

The p point also in the plane that laser throwed, satisfies optic plane equations simultaneously:

Ax _w+By _w+Cz _w+1=0，

In the formula, A, B, C are the optical plane coefficient; Simultaneous obtains:

$\left(\begin{array}{ccc}{f}_x& 0& u_0-u\\ 0& f_y& v_0-v\\ a& b& c\end{array}\right)\left(\begin{array}{c}{x}_w\\ y_w\\ z_w\end{array}\right)=\left(\begin{array}{c}0\\ 0\\ -1\end{array}\right),$

Obtain the world coordinates (x of any point on the rail profile outline line _w, y _w, z _w).

6. measurement of rail wear based method as claimed in claim 1 is characterized in that, the detailed process of extracting the web of the rail and rail head lower end unique point coordinate in the step 4) is following,

(1) web of the rail feature point extraction:

Appoint in the web of the rail circular arc behind three-dimensional reconstruction and get 1 X _i(x _Wi, y _Wi, z _Wi), make the space sphere with radius R, note space spherical equation is:

X ^TQX=0，

Wherein x is the coordinate of putting on the sphere, X=[x _w, y _w, z _w, 1] ^T, Q is 4 * 4 symmetric matrix:

$Q=\left[\begin{array}{cccc}1& 0& 0& -x_{\mathrm{wi}}\\ 0& 1& 0& -y_{\mathrm{wi}}\\ 0& 0& 1& -z_{\mathrm{wi}}\\ -x_{\mathrm{wi}}& -y_{\mathrm{wi}}& -z_{\mathrm{wi}}& l^2\end{array}\right],$ l ²=x _wi ²+y _wi ²+z _wi ²-R ²；

Make x _t=x _w, y _t=y _w, optic plane equations is write as:

X＝Mt，

Wherein, $M=\left[\begin{array}{ccc}-b/a& -c/a& -1/a\\ 1& 0& 0\\ 0& 1& 0\\ 0& 0& 1\end{array}\right],$ $t=\left[\begin{array}{c}{x}_t\\ y_t\\ 1\end{array}\right],$ Obtain X ^TQX=t ^TM ^TQMt=0, order:

$C=M^T\mathrm{QM}=\left[\begin{array}{ccc}{(b/a)}^2+1& \mathrm{bc}/a^2& b/a^2+(b/c)x_{\mathrm{wi}}-y_{\mathrm{wi}}\\ \mathrm{bc}/a^2& {(c/a)}^2+1& c/a^2+(c/a)x_{\mathrm{wi}}-z_{\mathrm{wi}}\\ b/a^2+(b/a)x_{\mathrm{wi}}-y_{\mathrm{wi}}& c/a^2+(c/a)x_{\mathrm{wi}}-z_{\mathrm{wi}}& 1/a^2+(2/a)x_{\mathrm{wi}}+l^2\end{array}\right],$

T is promptly arranged ^TCt=0, C are symmetric matrix, and t is the point on the quafric curve; Appoint in the same circular arc of the web of the rail that to get a pair of point be the space sphere of R as radius, here R equals the radius of web of the rail arc section, obtains two intersection points:

$\left\{\begin{array}{c}{t}^T{C}_{i}t=0\\ t^T{C}_{j}t=0\end{array}\right.;$

Two near points of intersection point middle distance video camera photocentre are the centre point of the arc section of asking, and it can calculate acquisition at three-dimensional coordinate figure; Every section arc profile has m to individual point in the image, obtains out m center of circle X ₁, X ₂..., X _mCoordinate, the order:

d _k=(x _wo-x _wk) ²+(y _wo-y _wk) ²+(z _wo-z _wk) ²

X wherein _Wo, y _Wo, z _WoBe center of circle optimum point X _oCoordinate components, x _Wk, y _Wk, z _WkBe X _k(k=1,2 ..., coordinate components m) is obtained X through following formula _oOptimum solution:

$\left\{\begin{array}{c}\underset{k=1}{\overset{m}{\mathrm{\&Sigma;}}}{d}_k=L\\ \frac{\&PartialD;L}{{\&PartialD;x}_{\mathrm{wo}}}=\frac{\&PartialD;L}{\&PartialD;y_{\mathrm{wo}}}=\frac{\&PartialD;L}{\&PartialD;z_{\mathrm{wo}}}=0\end{array}\right.;$

(2) rail head lower end feature point extraction:

Ask rail head lower end unique point { X _B1, X _B2..., X _B9With web of the rail arc section centre point X _oDistance

The error of note theoretical and actual range is ε _i=| d _Obi-d _Ob|, i=1,2 ..., 9, d wherein _ObBe theoretical; Get ε _Min=min{ ε ₁, ε ₂..., ε ₉, keep in mind ε _MinThe time pairing X _BiBe X _b, be the rail head lower extreme point of being asked.

7. measurement of rail wear based method as claimed in claim 1 is characterized in that, generates rail profile standard form based on unique point in the step 5), and its detailed process is following:

After web of the rail arc section unique point and rail head lower extreme point are confirmed, take up an official post at the rail profile and to get not and X _oAnd X _b1 X of conllinear _r, get vector

With X _oPoint is set up cartesian coordinate system for initial point, and the vector of x ' direction does $\stackrel{\&RightArrow;}{o^{\&prime;}{x}^{\&prime;}}=\stackrel{\&RightArrow;}{X_o{X}_b}=X_b-X_o,$ The vector of z ' direction does $\stackrel{\&RightArrow;}{o^{\&prime;}{z}^{\&prime;}}=\stackrel{\&RightArrow;}{o^{\&prime;}{x}^{\&prime;}}\&times;\stackrel{\&RightArrow;}{X_o{X}_r},$ The vector of y ' direction does

Set up the space profiled outline of rail, in every frame measurement image, generate the standard form of a rail profile.

8. measurement of rail wear based method as claimed in claim 1 is characterized in that, contrast rail measuring wheel exterior feature and standard form obtain wearing valve in the step 6), and its detailed process is following:

The measurement image that photographs is handled the measurement light belt that obtains the rail profile, calculate to set up and be positioned at the video camera photocentre rail profile 3 d space coordinate under boundary's coordinate system of conducting oneself in society; Set up new world coordinates, make its initial point be positioned at the center of rail foot, " axle overlaps with the rail base x; y " Axle and rail profile central lines; " plane is mapped to coordinate system o " x " y " z " down with the volume coordinate on the rail profile to z " perpendicular to x " y, calculates wearing valve.

Images