CN104236498B - Method for measuring end size of switch rail - Google Patents

Method for measuring end size of switch rail Download PDF

Info

Publication number
CN104236498B
CN104236498B CN201410520138.4A CN201410520138A CN104236498B CN 104236498 B CN104236498 B CN 104236498B CN 201410520138 A CN201410520138 A CN 201410520138A CN 104236498 B CN104236498 B CN 104236498B
Authority
CN
China
Prior art keywords
rail part
face
rail
edge
image
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201410520138.4A
Other languages
Chinese (zh)
Other versions
CN104236498A (en
Inventor
刘洪柱
窦燕
郑玉倩
范安然
戴坡
詹春艳
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
China Railway Shanhaiguan Bridge Group Co Ltd
Original Assignee
China Railway Shanhaiguan Bridge Group Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by China Railway Shanhaiguan Bridge Group Co Ltd filed Critical China Railway Shanhaiguan Bridge Group Co Ltd
Priority to CN201410520138.4A priority Critical patent/CN104236498B/en
Publication of CN104236498A publication Critical patent/CN104236498A/en
Application granted granted Critical
Publication of CN104236498B publication Critical patent/CN104236498B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Abstract

The invention relates to a method for measuring end size of a switch rail, overcomes the defect that indication error of a tape measure manually used for measurement, reading errors caused by tension changes during measuring and errors caused by temperature changes finally cause a final measurement error, aims to directly measure the size of a related part of each end face of a rail member in rail workshops, to achieve noncontact measurement and to finish measurement on the site. The method has the advantages that the method is flexible and convenient, measurement errors are reduced, measurement results are accurate, measuring precision and measuring speed are greatly increased, and production efficiency is greatly improved accordingly.

Description

A kind of measuring method of railroad turnout steel rail part face size
Technical field
The invention belongs to measuring method technical field, it is related to a kind of measuring method of railroad turnout steel rail part face size.
Background technology
Track switch is the crucial track equipment that railway realizes that train turns line operation, and its technical merit has embodied a concentrated reflection of a country The manufacture level of railroad track, switch manufacture is one of important technical specification with the high accuracy of assembling with high ride, track switch The relative dimensions accuracy of manufacture of rail part is impact one of track switch high accuracy and key factor of high ride.Railroad turnout steel rail at present The face size measurement of part is all generally manually by the way of ribbon measurement, but the error of indication of ribbon itself, survey The error that error in reading that during amount, tension variations cause, temperature change are brought is likely to cause final measurement error, precision Low, measurement efficiency is low.
Content of the invention
The present invention, in order to overcome the defect of prior art, devises a kind of measuring method of railroad turnout steel rail part face size, Achieve non-contact measurement, overcome the defect that when manually adopting ribbon measurement, measurement error is big, precision is low, efficiency is low, Reduce measurement error, improve certainty of measurement and measurement efficiency.
The concrete technical scheme that the present invention is taken is:
A kind of measuring method of railroad turnout steel rail part face size it is important to: described method comprises the following steps:
A, prepare rail part and calibrating block, it is identical that calibrating block is that a size with rail part end face, shape are dug in centre Hole scaling board, be located at the middle position of scaling board by cutouts, both sides web of the rail position has chequered with black and white chessboard figure, In addition, also respectively having one piece of chessboard figure in this scaling board left and right sides, two pieces of chessboard figures become fixed angle with scaling board, and three is one Overall;
B, carry out industrial camera demarcation: demarcated using Zhang Zhengyou plane reference method, shoot the image of calibrating block, to mark Determine block image carry out extract angle point operation, extract all of angle point it is established that image coordinate system, industrial camera coordinate system with And the corresponding relation between three coordinate systems of world coordinate system, obtain the inside and outside parameter of industrial camera, you can set up pixel and sit Mark is as follows with the relation between the point coordinates of corresponding space:
z u v 1 = 1 / dx tan α / dy u 0 0 1 / dy v 0 0 0 1 f 0 0 0 0 f 0 0 0 0 1 0 x y z 1 = f x γ u 0 0 0 f y v 0 0 0 0 1 0 r t 0 t 1 x w y w z w 1
Wherein, coordinate in industrial camera coordinate system for (x, y, z) representation space point p, α denotation coordination axle out of plumb is by mistake Difference, fx=f/dx, fy=f/dy is the equivalent focal length being represented with pixel wide and length, and γ represents the nonopiate of image coordinate axle Property;
C, rail part is positioned over ground or is placed on platform so that the flange of rail contact ground of rail part or platform Place, remove the debris around rail part, calibrating block is close to rail part end face, the height of two industrial cameras of adjustment and angle Degree, makes two industrial cameras and rail part end face be in sustained height, and allow two industrial cameras just to rail part end face it is desirable to The optical axis of two industrial cameras is parallel, and the optical axis of two industrial cameras is perpendicular to rail part end face;
D, open lamphouse light source, side illuminates rail part end face so that light projects end profile completely, to remove the moon Shadow, adjusts two industrial cameras, makes rail part end face fall into the center of screen picture, and takes each screen of two industrial cameras View, the image of collection rail part end face now;
E, control device first carry out median filter process to the image of step d gained, then carry out binaryzation, using being based on The edge detection algorithm of canny obtains the edge of image;
Gestalt rule in f, foundation topology consciousness theory, the profile of rail part end face should be continuous, close, completely , if the parameterized form of a line edge is b=v (t)={ v1(t1),v2(t2),...,vn(tn)},0≤ti≤1,1≤i≤ If n. v (0)=v (1), b are closed contours;If this closed contour just contains the end face of rail part, b It is exactly that we are required;But often contour of object is made up of a series of edges, these edges are often discontinuous, here A kind of method that two edges are coupled together composition continuous boundary: a line edge of the image obtained by hypothesis step e is proposed There are two end points p1(x, y) and p2(x, y), and have both direction it is stipulated that the direction gradient angle of end points is the side of this two end points To,
In formula, r (p1)y、r(p1)xRepresent point p respectively1Projection in y direction and x direction;r(p2)y、r(p2)xRepresent respectively Point p2In the projection in y direction and x direction, if the direction of a line edge end points and a certain bar edge in its surrounding edge The angular separation of one end points is minimum, then couple together this two end points straight lines and be merged into a new edge, Detect eventually one meet seriality, closure, the edge of integrity be exactly rail part end face closed contour;
G, the data of the closed contour obtaining in step f is fitted to geometric primitive, using least square fitting principle Lai Fitting a straight line, extracts n discrete point on graph line edge feature, asks for fitting a straight line y=kx+b of these points, by step This n discrete point coordinates (x that f obtainsi, yi) (i=1,2 ..., n) meet linear equation or as close from straight line as possible, ask down The minima of face expression formula:
s = σ i = 1 n [ y i - ( kx i + b ) ] 2
The matching of circle is to carry out even the squared-distance of n point on profile to fitting circle plus sue for peace so as to minimize:
s = σ i = 1 n ( ( x i - a ) 2 + ( y i - b ) 2 - r ) 2
Wherein, (a, b) is central coordinate of circle, and r is the radius of circle, so image outline segmentation can be in line and curve, By to the detection of these shapes and identification, according between the pixel coordinate set up in step b and corresponding space point coordinates Relation, obtains the size that we want;
H, control device are by the standard size of the size of step g gained and the same model rail part being saved in data base Compare, if measured size is in the margin of tolerance of permission, can conclude that end surface shape and the chi of this rail part Very little is qualified, otherwise points out underproof position, for being processed further providing foundation.
The margin of tolerance allowing in step h is: rail head width ± 0.5mm;Web of the rail thickness+0.75mm, -0.5mm;Flange of rail width Degree+1.0mm, -2.0mm;Rail height+0.8mm.
The invention has the beneficial effects as follows: provide a kind of measuring method of railroad turnout steel rail part face size, overcome artificial Error in reading that when error of indication due to ribbon itself that exists when being measured using ribbon, measurement, tension variations cause, The error that temperature change is brought causes the defect of final measurement error, can be directly realized by rail part in rail part workshop The measurement of end face region of interest size is it is achieved that non-contact measurement is so that measurement work at the scene can be flexibly square to complete Just, reduce measurement error, measurement result accurately, will greatly improve the accuracy and speed of measurement, thus greatly improving production effect Rate.
Brief description
Fig. 1 is the structural representation of measurement apparatus in the present invention.
The position relationship schematic diagram of calibrating block, industrial camera and rail part when Fig. 2 measures for the present invention.
In accompanying drawing, 1 represents base, and 2 represent industrial camera, and 3 represent bracing frame, and 4 represent leading screw, and 5 represent motor, 6 representatives Reductor, 7 represent slide block, and 8 represent retention bar, and 9 represent positioning vertical pole, and 10 represent all-directional tripod head, and 11 represent shaft coupling, and 12 Represent pulley, 13 represent control station, 14 represent support bar, 15 represent Turning-over plate chair, 16 represent rail part, and 17 represent calibrating block.
Specific embodiment
The present invention will be further described below in conjunction with the accompanying drawings:
As shown in figure 1, a kind of railroad turnout steel rail part face size measurement apparatus, including 1, two industrial cameras 2 of base, control Device, arrangement for adjusting height, two industrial cameras 2 and arrangement for adjusting height are all connected with control device, arrangement for adjusting height edge Vertical direction is arranged on base 1, two industrial cameras 2 be symmetricly set in the horizontal direction arrangement for adjusting height both sides and with Arrangement for adjusting height forms lifting cooperation, and two industrial cameras 2 form lifting cooperation by arrangement for adjusting height and base 1.
Arrangement for adjusting height includes the bracing frame 3 of door type structure, leading screw 4, motor 5, reductor 6, fixed support, motor 5 Output shaft be connected with reductor 6 by shaft coupling 11, the bottom of bracing frame 3 is fixedly connected with base 1, and leading screw 4 is along vertical side To the inside being arranged on bracing frame 3 door type structure, motor 5 is arranged on the top of bracing frame 3 and is connected with control device, leading screw 4 Top break-through bracing frame 3 be fixedly connected with the output shaft of reductor 6, the bottom of leading screw 4 and base 1 form and are rotatably assorted, silk Slide block 7 is set with thick stick 4, fixed support includes retention bar 8, the two ends of retention bar 8 are respectively arranged with a positioning vertical pole 9, two positioning vertical poles 9 are symmetrical arranged and are all vertically arranged with retention bar 8, and the end of two positioning vertical poles 9 is all fixed with entirely Orientation head 10, industrial camera 2 is hinged with all-directional tripod head 10, and retention bar 8 is fixedly connected with slide block 7, industrial camera 2 by Slide block 7 and leading screw 4 cooperation form lifting cooperation with base 1.During measurement, motor 5 drives reductor 6 to rotate by shaft coupling 11, Reductor 6 drives leading screw 4 to rotate, and so that slide block 7 is risen or fallen on leading screw 4, and slide block 7 drives two industry by fixed support Camera 2 rises or falls, and when two industrial cameras 2 reach required height, motor 5 stops operating, and then utilizes omnibearing cloud Platform 10 adjusts the angle of industrial camera 2, and structure is simple, easy to operate.
As to further improvement of the present invention, in the bottom of base 1, multiple pulleys 12 are set so that this device is being carried out Convenient laborsaving when moving horizontally.
As to further improvement of the present invention, set up the control station 13 for placing control device, the back of the body of control station 13 Face is fixedly connected with bracing frame 3, and the bottom of control station 13 is fixed with support bar 14, the upper end of the bottom of support bar 14 and base 1 It is fixedly connected, control device is arranged on control station 13, as an entirety, facilitate operator to observe during measurement and control.
As to further improvement of the present invention, set up the Turning-over plate chair 15 matching with control station 13, Turning-over plate chair 15 Bottom be fixedly connected with base 1, operator can be sitting in and be operated on Turning-over plate chair 15, convenient comfortable.
A kind of measuring method of use railroad turnout steel rail part face size measurement apparatus it is important to: described method include with Lower step:
A, prepare rail part 16 and calibrating block 17, calibrating block 17 be centre dig a size with rail part 16 end face, The scaling board in shape identical hole, is located at the middle position of scaling board by cutouts, and both sides web of the rail position has chequered with black and white Chessboard figure, in addition, also respectively having one piece of chessboard figure in this scaling board left and right sides, two pieces of chessboard figures become fixed angles with scaling board Degree, three is an entirety;
B, carry out industrial camera demarcation: industrial camera 2 demarcate be exactly set up industrial camera 2 image pixel coordinates with corresponding Relation between the point coordinates of space, carries out extracting angle point operation to the calibrating block image of the different angles collecting, extracts institute Some angle points, the corresponding pass between the image coordinate system of foundation, industrial camera 2 coordinate system and three coordinate systems of world coordinate system System, tries to achieve the inside and outside parameter of industrial camera 2, the accuracy of the space point coordinates of industrial camera 2 parameter determination of demarcation, final shadow Ring the accuracy to dimensional measurement, the present invention uses the scaling method based on two-dimensional planar template of Zhang Zhengyou proposition, should Method requires no knowledge about concrete orientation and the displacement information of plane template movement, and plane template makes simply, compared to it His method is more flexible, due to the particularity of the calibrating block 17 that the present invention adopts, only needs to gather a chessboard figure during demarcation, and And industrial camera 2 is fixed, so only needing to demarcate once, (zhengyou after system is demarcated zhang.a flexible new technique for camera calibration.ieee transactions on Pattern analysis and machine intelligence, vol.22, no.11, november 2000), you can build The relation erecting between pixel coordinate and corresponding space point coordinates is as follows:
z u v 1 = 1 / dx tan α / dy u 0 0 1 / dy v 0 0 0 1 f 0 0 0 0 f 0 0 0 0 1 0 x y z 1 = f x γ u 0 0 0 f y v 0 0 0 0 1 0 r t 0 t 1 x w y w z w 1
Wherein, coordinate in industrial camera 2 coordinate system for (x, y, z) representation space point p, α denotation coordination axle out of plumb Error, fx=f/dx, fy=f/dy is the equivalent focal length being represented with pixel wide and length, and γ represents the anon-normal of image coordinate axle The property handed over;
C, rail part 16 is positioned over ground or be placed on platform so that the flange of rail contact ground of rail part 16 or Platform is placed, and removes the debris around rail part 16, and calibrating block 17 is close to rail part 16 end face, adjusts two industrial cameras 2 height and angle, make two industrial cameras 2 and rail part 16 end face be in sustained height, and make two industrial cameras 2 just right , it is desirable to the optical axis of two industrial cameras 2 is parallel, the optical axis of two industrial cameras 2 is perpendicular to rail part 16 end for rail part 16 end face Face, as shown in Figure 2;
D, open lamphouse light source, side illuminates rail part 16 end face so that light projects end profile completely, to remove the moon Shadow, adjusts two industrial cameras 2, makes rail part 16 end face fall into the center of screen picture, and it is each to take two industrial cameras 2 Screen view, the image of collection rail part 16 end face now, the particularity of the calibrating block 17 due to adopting, week can be masked The debris enclosing, prominent rail part 16 end face, we only need two industrial cameras 2 respectively according to a sheet photo, but environmental factorss Impact is inevitable, when gathering image more gather several width, general three width, select the work that the preferable width of effect carries out the later stage Make;
E, control device first carry out median filter process to the image of step d gained, then carry out binaryzation, using being based on The edge detection algorithm of canny obtains the edge of image, particularly as follows:
Medium filtering is by a kind of nonlinear digital filter technology of tukey proposition first, is frequently used for removing image Or the noise in other signals, traditional medium filtering is general to adopt the sliding window containing odd number point, with each in window The gray value to replace specified point for the intermediate value of point gray value, the mathematical description of medium filtering is: sets s as pixel (x0,y0) neighbour Domain is gathered, and (x, y) is the element of s, and the gray value of point (x, y) is represented with h (x, y), and | s | is the number of element in s set, to point (x0,y0) smooth be expressed as
h ′ ( x 0 , y 0 ) = [ sort ( x , y ) &element; s h ( x , y ) ] | s | + 1 2
Wherein, sort is expressed as h (x, y) is ranked up, and this is the principle of standard medium filtering, and standard medium filtering needs Carry out substantial amounts of sequence work, amount of calculation is very big, the performance simultaneously removing impulsive noise is affected relatively by filter window size Greatly, there is certain contradiction in terms of suppression picture noise and protection details, therefore present invention uses a kind of improved intermediate value Filtering algorithm, i.e. the filtering algorithm based on extreme value and direction, this algorithm first will according to the extreme point whether certain point is its neighborhood Noise image is divided into suspicious noise spot and signal point set, then constructs 8 direction gradient moulds in neighborhood to suspicious noise spot Plate determines whether this point is real noise spot, finally carries out medium filtering to noise spot, result shows this algorithm denoising effect Substantially, simultaneously Edge preservation very well (quoted from Dou Yan, Kong Lingfu. a kind of filtering algorithm [j] based on extreme value and direction. computer Engineering and application .2007,43 (28): 41-43.);
Then binaryzation is carried out to image, obtain the edge of image, canny edge detection operator using edge detection algorithm Being j.canny proposed in 1986, had that signal to noise ratio is big, accuracy of detection is high and amount of calculation compared with other edge detection operators Little advantage, is widely used;According to the definition of canny operator, the intensity of center edge is the maximum on gradient direction, point Do not judge whether each has maximum in the neighborhood on its gradient direction, if so, then this point is marginal point;Canny side Edge detection algorithm is the maximum finding shade of gray in topography, test problems is converted into the maximum finding a function Process, and detect strong and weak two edges respectively with two threshold values, therefore the method noise resisting ability is strong, canny rim detection Algorithm specifically includes: Gaussian smoothing, gradient calculation, non-maxima suppression and double-threshold comparison four are most of;Traditional canny side Edge detection algorithm uses Gaussian smoothing, and edge also can be entered while filtering noise by Gaussian filter as high fdrequency components Row is smooth, may lose marginal information in non-maxima suppression afterwards;Carry out edge using fixing high-low threshold value to carry Take, lack the adaptivity to different images, and the setting of high-low threshold is completely dependent on manually obtaining, automaticity is low, right In smooth, we adopt aforementioned improved median filtering algorithm, so can retain the edge letter of needs Breath, then calculates the amplitude that have passed through smooth rear gradient of image and gray scale and direction, carries out the finite difference formulations of single order local derviation, width It is worth for m (x, y), gradient direction is h (x, y), and expression formula is:
m ( x , y ) = f x ( x , y ) 2 + f y ( x , y ) 2 H (x, y)=arctan (fy(x,y)/fx(x,y))
Wherein, fx(x, y) and fy(x,y)
The neighborhood being respectively in the x and y direction is poor;Then carry out non-maxima suppression, finally select ostu as threshold value The method chosen, its maximum feature is its adaptivity is based on the grey level histogram of image, takes target and background Variance maximum as segmentation threshold, because one-dimensional ostu algorithm is only able to find single threshold value, made by we by the threshold value obtaining For high threshold, this two values are realized double-threshold comparison, base as the high-low threshold value in canny algorithm 2/5 as Low threshold by it Sub-pixel edge detection in least square fitting can make measurement result reach higher precision;
F, the complete outline of rail member 16 end face of being tapped using Gestalt Rule Extraction, Gestalt psychology conclusion is numerous, We mainly utilize its conclusion related to image procossing,
First, seriality: refer to a kind of special consciousness to lines for the mankind, people easily will be very short for some distances or mutual Close part is combined into an entirety;
Second, integrity: refer to the mankind can tend to be allowed to conjunction as far as possible by incoherent figure jaggy thus Become an entirety;
3rd, similarity: refer to people and easily similar part (such as color is close) is combined into an entirety;
The characteristic such as in addition also have propinquity, altogether (straight/bent) linear, closure, collimation and Curvature varying minimum is it is believed that people Class visual system tends to perceive them as an entirety, according to gestalt above rule in topological consciousness theory, With multiscale analysis technology, above detected profile is organized into groups, thus obtaining objects in images closed contour, tool Body is as follows:
Gestalt rule in foundation topology consciousness theory, the profile of rail part 16 end face should be continuous, close, completely , if the parameterized form of a line edge is b=v (t)={ v1(t1),v2(t2),...,vn(tn)},0≤ti≤1,1≤i≤ If n. v (0)=v (1), b are closed contours;If this closed contour just contains the end face of rail part 16, Then b is exactly that we are required;But often contour of object is made up of a series of edges, these edges are often discontinuous, A kind of method that two edges are coupled together composition continuous boundary: of the image obtained by hypothesis step e is proposed here There are two end points p at edge1(x, y) and p2(x, y), and have both direction it is stipulated that the direction gradient angle of end points is this two end points Direction,
In formula, r (p1)y、r(p1)xRepresent point p respectively1Projection in y direction and x direction;r(p2)y、r(p2)xRepresent respectively Point p2In the projection in y direction and x direction, if the direction of a line edge end points and a certain bar edge in its surrounding edge The angular separation of one end points is minimum, then couple together this two end points straight lines and be merged into a new edge, Detect eventually one meet seriality, closure, the edge of integrity be exactly rail part 16 end face closed contour;
G, the data of the closed contour obtaining in step f is fitted to geometric primitive, using least square fitting principle Lai Fitting a straight line, extracts n discrete point on graph line edge feature, asks for fitting a straight line y=kx+b of these points, by step This n discrete point coordinates (x that f obtainsi, yi) (i=1,2 ..., n) meet linear equation or as close from straight line as possible, ask down The minima of face expression formula:
s = σ i = 1 n [ y i - ( kx i + b ) ] 2
The matching of circle is to carry out even the squared-distance of n point on profile to fitting circle plus sue for peace so as to minimize:
s = σ i = 1 n ( ( x i - a ) 2 + ( y i - b ) 2 - r ) 2
Wherein, (a, b) is central coordinate of circle, and r is the radius of circle, so image outline segmentation can be in line and curve, By to the detection of these shapes and identification, according between the pixel coordinate set up in step b and corresponding space point coordinates Relation, obtains the size that we want;
H, control device are by the gauge of the size of step g gained and the same model rail part 16 being saved in data base Very little compare, if measured size is in the margin of tolerance of permission, can conclude that the end surface shape of this rail part 16 And size is qualified, otherwise point out underproof position, for be processed further being given according to it is allowed to the margin of tolerance refer to: Rail head width ± 0.5mm;Web of the rail thickness+0.75mm, -0.5mm;Flange of rail width+1.0mm, -2.0mm;Rail height+0.8mm.
Measuring method in the present invention overcomes the showing due to ribbon itself existing when manually measuring using ribbon The error that when value error, measurement, the error in reading that causes of tension variations, temperature change are brought causes lacking of final measurement error Fall into, the measurement of rail part end face region of interest size can be directly realized by rail part 16 workshop it is achieved that contactless So that measurement work can reduce measurement error to complete, flexibly and easily at the scene, measurement result is accurate, Ke Yi great for measurement The big accuracy and speed improving measurement, thus greatly improve production efficiency.

Claims (2)

1. a kind of measuring method of railroad turnout steel rail part face size it is characterised in that: described method comprises the following steps:
A, prepare rail part (16) and calibrating block (17), calibrating block (17) be centre dig one big with rail part (16) end face The little, scaling board in shape identical hole, is located at the middle position of scaling board by cutouts, and both sides web of the rail position has black and white phase Between chessboard figure, in addition, also respectively having one piece of chessboard figure in this scaling board left and right sides, two pieces of chessboard figures become fixed angles with scaling board Degree, three is an entirety;
B, carry out industrial camera demarcation: demarcated using Zhang Zhengyou plane reference method, shoot the image of calibrating block (17), to mark The image determining block (17) carries out extracting angle point operation, extracts all of angle point it is established that image coordinate system, industrial camera (2) Corresponding relation between coordinate system and three coordinate systems of world coordinate system, obtains the inside and outside parameter of industrial camera (2), you can build The relation erecting between pixel coordinate and corresponding space point coordinates is as follows:
z u v 1 = 1 / dx tan α / dy u 0 0 1 / dy v 0 0 0 1 f 0 0 0 0 f 0 0 0 0 1 0 x y z 1 = f x γ u 0 0 0 f y v 0 0 0 0 1 0 r t 0 t 1 x w y w z w 1
Wherein, coordinate in industrial camera (2) coordinate system for (x, y, z) representation space point p, α denotation coordination axle out of plumb is by mistake Difference, fx=f/dx, fy=f/dy is the equivalent focal length being represented with pixel wide and length, and γ represents the nonopiate of image coordinate axle Property;
C, rail part (16) is positioned over ground or be placed on platform so that the flange of rail contact ground of rail part (16) or Platform is placed, and removes the debris around rail part (16), and calibrating block (17) is close to rail part (16) end face, adjusts two works The height of industry camera (2) and angle, make two industrial cameras (2) and rail part (16) end face be in sustained height, and allow two Industrial camera (2) just to rail part (16) end face it is desirable to the optical axis of two industrial cameras (2) is parallel, two industrial cameras (2) Optical axis is perpendicular to rail part (16) end face;
D, open lamphouse light source, side illuminates rail part (16) end face so that light projects end profile completely, to remove the moon Shadow, adjusts two industrial cameras (2), makes rail part (16) end face fall into the center of screen picture, and takes two industrial phases The each screen view of machine (2), the image of collection rail part (16) end face now;
E, control device first carry out median filter process to the image of step d gained, then carry out binaryzation, using being based on The edge detection algorithm of canny obtains the edge of image;
Gestalt rule in f, foundation topology consciousness theory, the profile of rail part (16) end face should be continuous, close, completely , if the parameterized form of a line edge is b=v (t)={ v1(t1),v2(t2),...,vn(tn)},0≤ti≤1,1≤i≤ If n. v (0)=v (1), b are closed contours;If this closed contour just contains the end of rail part (16) Face, then b is exactly that we are required;But often contour of object is made up of a series of edges, these edges are often discontinuous , a kind of method that two edges are coupled together composition continuous boundary: of the image obtained by hypothesis step e is proposed here There are two end points p at bar edge1(x, y) and p2(x, y), and have both direction it is stipulated that the direction gradient angle of end points is this two ends The direction of point,
In formula, r (p1)y、r(p1)xRepresent point p respectively1Projection in y direction and x direction;r(p2)y、r(p2)xRepresent point p respectively2 In the projection in y direction and x direction, if one of the direction of a line edge end points and a certain bar edge in its surrounding edge The angular separation of end points is minimum, then couple together this two end points straight lines and be merged into a new edge, final inspection Measure one meet seriality, closure, the edge of integrity be exactly rail part (16) end face closed contour;
G, the data of the closed contour obtaining in step f is fitted to geometric primitive, using least square fitting principle come matching Straight line, extracts n discrete point on graph line edge feature, asks for fitting a straight line y=kx+b of these points, is obtained by step f This n discrete point coordinates (x arrivingi, yi) (i=1,2 ..., n) meet linear equation or as close from straight line as possible, seek following table Reach the minima of formula:
s = σ i = 1 n [ y i - ( kx i + b ) ] 2
The matching of circle is to carry out even the squared-distance of n point on profile to fitting circle plus sue for peace so as to minimize:
s = σ i = 1 n ( ( x i - a ) 2 + ( y i - b ) 2 - r ) 2
Wherein, (a, b) is central coordinate of circle, and r is the radius of circle, so image outline segmentation can be in line and curve, pass through Detection to these shapes and identification, according to the relation between the pixel coordinate set up in step b and corresponding space point coordinates, Obtain the size that we want;
H, control device are by the standard size of the size of step g gained and same model rail part (16) being saved in data base Compare, if measured size is in the margin of tolerance of permission, can conclude that the end surface shape of this rail part (16) And size is qualified, otherwise point out underproof position, for being processed further providing foundation.
2. a kind of railroad turnout steel rail part face size according to claim 1 measuring method it is characterised in that: in step h The margin of tolerance allowing is: rail head width ± 0.5mm;Web of the rail thickness+0.75mm, -0.5mm;Flange of rail width+1.0mm ,- 2.0mm;Rail height+0.8mm.
CN201410520138.4A 2014-09-30 2014-09-30 Method for measuring end size of switch rail Active CN104236498B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410520138.4A CN104236498B (en) 2014-09-30 2014-09-30 Method for measuring end size of switch rail

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410520138.4A CN104236498B (en) 2014-09-30 2014-09-30 Method for measuring end size of switch rail

Publications (2)

Publication Number Publication Date
CN104236498A CN104236498A (en) 2014-12-24
CN104236498B true CN104236498B (en) 2017-01-18

Family

ID=52225048

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410520138.4A Active CN104236498B (en) 2014-09-30 2014-09-30 Method for measuring end size of switch rail

Country Status (1)

Country Link
CN (1) CN104236498B (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104778379B (en) * 2015-05-07 2017-07-07 攀钢集团攀枝花钢钒有限公司 The parametrization tolerance decision method of the heavy rail head breadth and coronal plumpness
CN104880156A (en) * 2015-06-05 2015-09-02 贵州航天精工制造有限公司 Method for measuring large R value by using CAD technology
CN108240793A (en) * 2018-01-26 2018-07-03 广东美的智能机器人有限公司 Dimension of object measuring method, device and system
CN108489383A (en) * 2018-03-01 2018-09-04 北京科技大学 A kind of measuring device and method of H-type cross dimensions

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101314932A (en) * 2008-07-16 2008-12-03 张小虎 Camera shooting measuring method for track geometric parameter
CN101576375A (en) * 2009-05-21 2009-11-11 北京航空航天大学 Fast processing method of laser vision image of steel rail wear
CN101871768A (en) * 2010-05-12 2010-10-27 中国科学院自动化研究所 Automatic acquisition system of baggage appearance information based on monocular vision and method thereof
CN104048606A (en) * 2014-06-25 2014-09-17 燕山大学 Machine vision non-contact automatic precise measuring device for size of railway turnout steel rail part

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101314932A (en) * 2008-07-16 2008-12-03 张小虎 Camera shooting measuring method for track geometric parameter
CN101576375A (en) * 2009-05-21 2009-11-11 北京航空航天大学 Fast processing method of laser vision image of steel rail wear
CN101871768A (en) * 2010-05-12 2010-10-27 中国科学院自动化研究所 Automatic acquisition system of baggage appearance information based on monocular vision and method thereof
CN104048606A (en) * 2014-06-25 2014-09-17 燕山大学 Machine vision non-contact automatic precise measuring device for size of railway turnout steel rail part

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
基于双目视觉的钢轨磨耗测量系统关键技术的研究与实现;王玉柱;《中国优秀硕士学位论文全文数据库信息科技辑》;20091115(第11期);第I140-480页 *
基于空间和物体的视觉注意计算方法及实验研究;窦燕;《中国博士学位论文全文数据库信息科技辑》;20100815(第8期);第I138-18页 *
相机标定与三维重建技术研究;李云翔;《中国优秀硕士学位论文全文数据库信息科技辑》;20091015(第10期);第I138-558页 *

Also Published As

Publication number Publication date
CN104236498A (en) 2014-12-24

Similar Documents

Publication Publication Date Title
US20200364849A1 (en) Method and device for automatically drawing structural cracks and precisely measuring widths thereof
CN104236498B (en) Method for measuring end size of switch rail
CN104897062B (en) A kind of vision measuring method and device of part antarafacial parallel hole shape position error
CN102032875B (en) Image-processing-based cable sheath thickness measuring method
CN102297660B (en) Measuring method of shield tunnel lining segment seam open width and apparatus thereof
CN104359403A (en) Plane part size measurement method based on sub-pixel edge algorithm
CN105160652A (en) Handset casing testing apparatus and method based on computer vision
CN104899609B (en) A kind of mechanical meter recognition methods based on image registration
CN104050446A (en) Meter pointer image identification method based on pointer width character
CN106017350A (en) Machine-vision-based rapid detection device and detection method for medium and small module gears
CN102519400B (en) Large slenderness ratio shaft part straightness error detection method based on machine vision
CN109060821A (en) Tunnel defect detection method and tunnel defect detection device based on laser detection
CN105956942A (en) Detection method for quality of electric power pipe gallery reinforcing steel bar mesh based on machine vision and detection device
CN104268853A (en) Infrared image and visible image registering method
CN104700395A (en) Method and system for detecting appearance crack of structure
CN110929710B (en) Method and system for automatically identifying meter pointer reading based on vision
CN108256394A (en) A kind of method for tracking target based on profile gradients
CN105222727A (en) The measuring method of linear array CCD camera imaging plane and the worktable depth of parallelism and system
CN109738460A (en) A kind of evaluation detection system for civil engineering structure visual defects
CN107392929A (en) A kind of intelligent target detection and dimension measurement method based on human vision model
Zhang et al. Automatic road-marking detection and measurement from laser-scanning 3D profile data
CN107341781A (en) Based on the SAR image correcting methods for improving the matching of phase equalization characteristic vector base map
CN104390584B (en) Binocular vision laser calibration measurement apparatus and measuring method
CN108725511A (en) The real time position modification method of rail corrugation string measuring point
CN109507198B (en) Mask detection system and method based on fast Fourier transform and linear Gaussian

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant