CN103913464A - High speed railway track surface defect coupling method based on machine visual inspection - Google Patents

Abstract

The invention discloses a high speed railway track surface defect coupling method based on machine visual inspection, which comprises the following steps: 1) obtaining a rail panoramic image f(x,y); 2)employing a vertical projection method for extracting a rail surface area image f1(x,y) in the rail panoramic image; 3)performing median filter treatment on the rail surface area image to obtain a rail surface area image f2(x,y) through noise removal; 4)performing image preprocessing on the rail surface area image; 5)extracting the defect characteristic information in an image f4(x,y); 6)respectively obtaining the defect characteristic information in a history acquisition image and the defect characteristic information in a real-time acquisition image; and 7)coupling the characteristic information capable of performing real-time defect acquisition with all the characteristic information with stored defect in order and calculating to obtain the coupled defect characteristic information. The method realizes the high speed and high precision coupling for the rail surface defect.

Description

A kind of high speed railway track surface imperfection matching process based on Machine Vision Detection

Technical field

The invention belongs to mechanical vision inspection technology field, particularly a kind of high speed railway track surface imperfection matching process based on Machine Vision Detection.

Background technology

In the traffic transport industry of China, railway is being brought into play huge effect.Along with the development of modern rail technology, the serviceable life of rail is more and more longer, but because the railway network of China is very huge, so the length in serviceable life of rail is indefinite, cannot unify to safeguard and change.When train moves on the defective rail in surface, very easily damage train, the serious train accident that also can cause.In order to guarantee security and the continuity of transportation by railroad, need to carry out regular detection to rail, to there being the rail of damage safeguard and change.Significantly raise speed and heavy haul train start in the situation that at train, understand in time the quality of rail, examination and controlling has just more been necessary especially in real time.

The SPEED VISION detection system of the railroad track surface imperfection based on machine vision technique can detect rail surface imperfection high-speed, high precision, save manpower, realize the intelligent classification identification of rail defect, can improve the suggestion of instructing for railway interests safeguards railroad track timely and effectively.But also have in actual applications many problems, wherein the redundancy of rail view data is exactly a urgent problem.Gather in the SPEED VISION detection system of railroad track surface imperfection the redundancy that there will be data in image process, show certain lengths of rail in once gathering by multi collect, or certain lengths of rail only needs to preserve a in collection repeatedly.This just need to find the rail information of redundancy record their positional information.

In sum, the rail information of None-identified and eliminate redundancy in existing Machine Vision Inspecting System, is badly in need of a kind of matching process of high-speed iron track surface machine vision defects detection, so that eliminate redundancy information.

Summary of the invention

Technical matters to be solved by this invention is to provide a kind of high speed railway track surface imperfection matching process based on Machine Vision Detection, and its object is, finds out the high speed railway track surface imperfection image that is repeated collection and preserves.

A high speed railway track surface imperfection matching process based on Machine Vision Detection, comprises following step:

Step 1: obtain rail panoramic picture f (x, y);

Step 2: adopt vertical projection method to extract rail surf zone image f from rail panoramic picture ₁(x, y);

Step 3: rail surf zone image is carried out to medium filtering processing, obtain the rail surf zone image f that removes noise ₂(x, y);

Step 4: to removing the rail surf zone image f of noise ₂(x, y) carries out successively binaryzation operation and morphology is opened operation, makes f ₂each division region in (x, y) is mutually bonding, obtains image f ₄(x, y);

Step 5: to image f ₄(x, y) carries out projection line by line and adopts blob analytical approach to obtain image f ₄defect characteristic information in (x, y), described defect characteristic information comprises the shape information of defect and the positional information of defect;

Step 6: history collection image and real-time image acquisition are carried out respectively to the operation of step 1-step 5, respectively the defect characteristic information in the historical collection of acquisition image and the defect characteristic information in real-time image acquisition;

Step 7: the characteristic information of Real-time Obtaining defect is mated to calculating with the characteristic information of all storage defect successively, obtain Real-time Obtaining defect and site error Ep and the plesiomorphism degree S of storage defect; Utilize fuzzy control rule according to Ep and S, obtain the defect characteristic information of coupling.

The historical defect characteristic information gathering in image in described step 6, comprises the shape information Dtm (m) of storage defect and the positional information Lt of storage defect _m(LX, LY, RX, RY, XW, YH), m represents m storage defect, and span is 1-M, and M represents storage defect sum;

Defect characteristic information in real-time image acquisition, comprises the shape information Dom (n) of Real-time Obtaining defect and the positional information Lo of Real-time Obtaining defect _n(LX, LY, RX, RY, XW, YH), n represents n Real-time Obtaining defect, and span is 1-N, and N represents the sum of Real-time Obtaining defect;

Wherein, the shape information Am (j) of each defect is according to each defect place image f ₄line number j in (x, y), calculates and obtains according to following formula:

$\mathrm{Am}\left(j\right)=\frac{\underset{y=1}{\overset{\mathrm{Width}}{\mathrm{\Σ}}}{f}_4(j,y)}{255};$

Wherein, j is the integer between [1, Height], and Height is image f ₄the height of (x, y), Width is image f ₄(x, y) wide, LX is the left row coordinate of defect boundary rectangle, LY is the left-hand line coordinate of defect boundary rectangle, RX is the right side row-coordinate of defect boundary rectangle, RY is the right-hand column coordinate of defect boundary rectangle, the width that XW is defect be defect institute across columns, the height that YH is defect is defect institute inter-bank number.

The concrete acquisition process of the defect characteristic information of the coupling in described step 7 is as follows:

1) judge whether Real-time Obtaining defect and storage defect are same defect;

The width Lo (XW) of contrast Real-time Obtaining defect and the width Lt (XW) of storage defect, the height Lo (YH) of contrast Real-time Obtaining defect and the height Lt (YH) of storage defect, if or real-time Obtaining defect is not same defect with storage defect, otherwise, calculate Real-time Obtaining defect and the site error Ep of storage defect, further judge whether Real-time Obtaining defect and storage defect are same defect;

2) calculating Real-time Obtaining defect and the site error Ep of storage defect:

$\mathrm{Ep}=\frac{{\mathrm{Lt}}_m\left(\mathrm{RX}\right)|{\mathrm{Lt}}_m\left(\mathrm{LX}\right)-{\mathrm{Lo}}_n\left(\mathrm{LX}\right)|+{\mathrm{Lt}}_m\left(\mathrm{LX}\right)|{\mathrm{Lt}}_m\left(\mathrm{RX}\right)-{\mathrm{Lo}}_n\left(\mathrm{RX}\right)|}{{2\mathrm{Lt}}_m\left(\mathrm{LX}\right){\mathrm{Lo}}_n\left(\mathrm{RX}\right)}$

If Ep>0.2, Real-time Obtaining defect is not same defect with storage defect, otherwise, calculate Real-time Obtaining defect and the plesiomorphism degree S of storage defect;

3) Real-time Obtaining defect is as follows with the plesiomorphism degree S computation process of storage defect:

Step a: the length of adjusting array Dom (n);

Find peak point Pkt, the Pko of array Dtm (m) and array Dom (n), adjust the length of array Dom (n) take peak point as reference point, wherein, before Dtm (m) array peak point, length is LFt, after peak point, length is LBt, before Dom (n) array peak point, length is LFo, and length is LBo after peak point;

If illustrate that two arrays are reverse, enter step b; Otherwise, if adjust array Dom (n), array Dom (n) peak point previous section is adjusted into LFt length, redundance intercepts, and lacks part and gets Dtm (m) array appropriate section polishing; Array Dom (n) peak point aft section is adjusted into LBt length, and redundance intercepts, and lacks part peek group Dtm (m) appropriate section polishing, enters step c;

Step b: by array Dom (n) upset storage, repeating step a;

Step c: calculating Real-time Obtaining defect and the plesiomorphism degree S of storage defect;

$S=\frac{M\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\mathrm{Dtm}\left(i\right)\mathrm{Dom}\left(i\right)-\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\mathrm{Dtm}\left(i\right)\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\mathrm{Dom}\left(i\right)}{\sqrt{M\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\mathrm{Dtm}\left(i\right)-{\left(\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\mathrm{Dtm}\left(i\right)\right)}^2}\sqrt{M\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\mathrm{Dom}\left(i\right)-{\left(\underset{i=1}{\overset{M}{\mathrm{\Σ}}}\mathrm{Dom}\left(i\right)\right)}^2}};$

4) set membership function, input variable is carried out to Fuzzy Processing;

${\mathrm{\&mu;}}_S\left(S\right)=\left\{\begin{array}{cc}{P}_1& 0.8\&le;S\&le;0.9\\ Q_1& 0.9<S\&le;0.95\\ R_1& 0.95<S\&le;1.0\end{array}\right.,{\mathrm{\&mu;}}_{\mathrm{EP}}\left(\mathrm{EP}\right)=\left\{\begin{array}{cc}{P}_2& 0\&le;\mathrm{EP}\&le;0.1\\ Q_2& 0.1<\mathrm{EP}\&le;0.15\\ R_2& 0.15<\mathrm{EP}\&le;0.2\end{array}\right.$

5) according to fuzzy reasoning table, output fuzzy language variable U, wherein, N ₀expression Real-time Obtaining defect and storage defect are not same defects, P ₀expression Real-time Obtaining defect and storage defect are same defects;

Table 1 fuzzy reasoning table

Numbering	Input variable (S, EP)	Output variable U
			1	(R 1,P 2)	P 0
2	(R 1,Q 2)	P 0
			3	(R 1,R 2)	N 0
4	(Q 1,P 2)	P 0
			5	(Q 1,Q 2)	N 0
6	(Q 1,R 2)	N 0
			7	(P 1,P 2)	N 0
8	(P 1,Q 2)	N 0
			9	(P 1,R 2)	N 0

6) recording output variable is N ₀coupling defect characteristic information.

The specific implementation process of described step 2 is as follows:

1. cumulative gray-scale value by column, produces the gray average array Avg (i) of the each row of rail image f (x, y):

$\mathrm{Avg}\left(i\right)=\frac{\underset{x=1}{\overset{\mathrm{Height}}{\mathrm{\&Sigma;}}}f(x,i)}{\mathrm{Height}},i\&Element;[1,\mathrm{Width}],$ Height is the height of image, and i is image i row;

2. calculate the average Avg_all of Avg (i);

3. Avg (i) array binaryzation, by Avg (i) array one by one with Avg_all comparison, be greater than Avg_all and be made as 1, be less than Avg_all and be made as 0, obtain array Avg_Val (i):

$\mathrm{Avg}\_\mathrm{Val}\left(i\right)=\left\{\begin{array}{cc}1,& \mathrm{Avg}\left(i\right)\&GreaterEqual;\mathrm{Avg}\_\mathrm{all}\\ 0,& \mathrm{Avg}\left(i\right)<\mathrm{Avg}\_\mathrm{all}\end{array}\right.;$

4. in Avg_Val (i) array, finding first continuous ten numbers is all 1 point, the marginal point Begin_Point that this starts as rail surface image; After Avg_Val (i) array Begin_Point, finding first continuous ten numbers is all 0 point, the marginal point End_Point that this starts as rail surface image, the image that intercepting rail image f (x, y) Begin_Point is listed as between End_Point obtains rail surface image f ₁(x, y).

The specific implementation process of described step 3 is to rail surface image f ₁(x, y) carries out the medium filtering operation of a time 3 × 3, removes image noise, obtains the rail surface image f of low noise ₂(x, y).

The specific implementation process of described step 4 is as follows:

1. set global threshold T, to rail surface image f ₂(x, y) carries out binaryzation and obtains bianry image f ₃(x, y):

$f_3(x,y)=\left\{\begin{array}{cc}255,& f_2(x,y)\&GreaterEqual;T\\ 0,& f_2(x,y)<T\end{array}\right.;$

2. use 9 × 9 oval structure element B to bianry image f ₃(x, y) carries out morphology and opens operation and obtain image f ₄(x, y):

In formula, meet " Θ " and represent the corrosion operation in mathematical morphology, symbol represent the closed procedure in mathematical morphology, make the single division region in image obtain mutually bonding.

Beneficial effect

Provided by the inventionly provide a kind of high speed railway track surface imperfection matching process based on Machine Vision Detection, compared with prior art, its outstanding advantage is:

(1) matching speed is fast

The present invention takes into full account the actual conditions of rail surface image, first uses vertical projection method to extract rail surf zone, and the method can obtain rail surf zone quickly and accurately; Then rail area image is carried out to denoising, now to rail region denoising comparison panoramic picture, partial arithmetic amount is saved in denoising; To the bianry image horizontal projection obtaining, can obtain the minimum defect shape information of data volume, for subsequent treatment has reduced computation burden; By the site error of defect and target, defect is carried out to pre-matching, improve the efficiency of follow-up coupling;

(2) accuracy of detection is high

The positional information of the present invention by defect and shape information carry out the coupling of fuzzy algorithm to defect, only have positional information and shape information all to reach when high precision conforms to and be just identified coupling, and precision is high, eliminates much and disturbs; Also retained more details information, therefore rail surface defects detection result is also more accurate simultaneously;

The present invention can realize high speed, the high precision coupling of rail surface imperfection, and easy and existing defect recognition algorithm merges mutually, and realization detects rail surface imperfection efficiently, accurately.

Accompanying drawing explanation

Fig. 1 is overall procedure schematic diagram of the present invention;

Fig. 2 is the two width rail images that same rail defect is taken at different time, wherein, and the rail image of figure (a) for having stored, the rail image that figure (b) is Real-time Collection;

Fig. 3 is the perspective view that figure (a) the rail image in Fig. 2 carries out vertical projection acquisition;

Fig. 4 is for finding rail surface image marginal point schematic diagram in Fig. 3;

Fig. 5 is for obtaining the rail surface image of figure (a) in Fig. 2;

Fig. 6 is the binary map of Fig. 5;

Fig. 7 is the forward feature array comparison diagram of storage defect feature array and Real-time Obtaining defect;

Fig. 8 is the opposite feature array comparison diagram of storage defect feature array and Real-time Obtaining defect.

Embodiment

Below in conjunction with drawings and Examples, the present invention is further illustrated.

Embodiment 1

As shown in Figure 1, be the process flow diagram of a kind of high speed railway track surface imperfection matching process based on Machine Vision Detection of the present invention, concrete steps are as follows:

Step 1: obtain rail panoramic picture f (x, y);

Line scan camera carry ON TRAINS, by the scrambler control collected by camera image on wheel, so that camera trigger rate and train speed match, guarantees that the rail panoramic picture being spliced into has comprised complete rail surface imperfection information.

As shown in Figure 2, the two width rail images that same rail defect is taken at different time, wherein, the rail image of figure (a) for having stored, the rail image that figure (b) is Real-time Collection;

Step 2: adopt vertical projection method to extract rail surf zone image f from rail panoramic picture ₁(x, y);

(1), to the cumulative gray-scale value by column of figure (a) in Fig. 2, produce the gray average array Avg (i) of the each row of rail image f (x, y):

$\mathrm{Avg}\left(i\right)=\frac{\underset{x=1}{\overset{\mathrm{Height}}{\mathrm{\&Sigma;}}}f(x,i)}{\mathrm{Height}},i\&Element;[1,\mathrm{Width}],$ Height is the height of image, and i is image i row;

Fig. 3 is the schematic diagram of the each row gray average of rail image f (x, y) array Avg (i), and wherein horizontal ordinate is the columns i of rail image, and ordinate is the gray average Avg (i) of the each row of image;

(2) the average Avg_all of calculating Avg (i):

$\mathrm{Avg}\_\mathrm{Val}\left(i\right)=\left\{\begin{array}{cc}1,& \mathrm{Avg}\left(i\right)\&GreaterEqual;\mathrm{Avg}\_\mathrm{all}\\ 0,& \mathrm{Avg}\left(i\right)<\mathrm{Avg}\_\mathrm{all}\end{array}\right.;$ Width is the wide of image;

(3) Avg (i) array binaryzation, by Avg (i) array one by one with Avg_all comparison, be greater than Avg_all and be made as 1, be less than Avg_all and be made as 0, obtain array Avg_Val (i): $\mathrm{Avg}\_\mathrm{Val}\left(i\right)=\left\{\begin{array}{cc}1,& \mathrm{Avg}\left(i\right)\&GreaterEqual;\mathrm{Avg}\_\mathrm{all}\\ 0,& \mathrm{Avg}\left(i\right)<\mathrm{Avg}\_\mathrm{all}\end{array}\right.;$ Array Avg_Val (i) as shown in Figure 4, the columns i that wherein horizontal ordinate is image, ordinate is the binary value Avg_Val (i) of the gray average of the each row of image;

(4) in Avg_Val (i) array, finding first continuous ten numbers is all 1 point, the marginal point Begin_Point that this starts as rail surface image; After Avg_Val (i) array Begin_Point, finding first continuous ten numbers is all 0 point, the marginal point End_Point that this starts as rail surface image, the image that intercepting rail image f (x, y) Begin_Point is listed as between End_Point obtains rail surface image f ₁(x, y); As shown in Figure 5;

Step 3: rail surf zone image is carried out to the medium filtering processing of a time 3 × 3, obtain the rail surf zone image f that removes noise ₂(x, y);

Step 4: to removing the rail surf zone image f of noise ₂(x, y) carries out successively binaryzation operation and morphology is opened operation, makes f ₂each division region in (x, y) is mutually bonding, obtains image f ₄(x, y), as shown in Figure 6;

(1) set global threshold T=150, to rail surface image f ₂(x, y) carries out binaryzation and obtains bianry image f ₃(x, y):

$f_3(x,y)=\left\{\begin{array}{cc}255,& f_2(x,y)\&GreaterEqual;T\\ 0,& f_2(x,y)<T\end{array}\right.;$

(2) the oval structure element B of use 9 × 9 is to bianry image f ₃(x, y) carries out morphology and opens operation and obtain image f ₄(x, y):

In formula, meet " Θ " and represent the corrosion operation in mathematical morphology, symbol represent the closed procedure in mathematical morphology, make the single division region in image obtain mutually bonding;

Step 5: to image f ₄(x, y) carries out projection line by line and adopts blob analytical approach to obtain image f ₄defect characteristic information in (x, y), described defect characteristic information comprises the shape information of defect and the positional information of defect;

Step 6: history collection image and real-time image acquisition are carried out respectively to the operation of step 1-step 5, respectively the defect characteristic information in the historical collection of acquisition image and the defect characteristic information in real-time image acquisition;

The described historical defect characteristic information gathering in image, comprises the shape information Dtm (m) of storage defect and the positional information Lt of storage defect _m(LX, LY, RX, RY, XW, YH), m represents m storage defect, and span is 1-M, and M represents storage defect sum;

Defect characteristic information in real-time image acquisition, comprises the shape information Dom (n) of Real-time Obtaining defect and the positional information Lo of Real-time Obtaining defect _n(LX, LY, RX, RY, XW, YH), n represents n Real-time Obtaining defect, and span is 1-N, and N represents the sum of Real-time Obtaining defect;

Wherein, the shape information Am (j) of each defect is according to each defect place image f ₄line number j in (x, y), calculates and obtains according to following formula:

$\mathrm{Am}\left(j\right)=\frac{\underset{y=1}{\overset{\mathrm{Width}}{\mathrm{\&Sigma;}}}{f}_4(j,y)}{255};$

Wherein, j is the integer between [1, Height], and Height is image f ₄the height of (x, y), Width is image f ₄(x, y) wide, LX is the left row coordinate of defect boundary rectangle, LY is the left-hand line coordinate of defect boundary rectangle, RX is the right side row-coordinate of defect boundary rectangle, RY is the right-hand column coordinate of defect boundary rectangle, the width that XW is defect be defect institute across columns, the height that YH is defect is defect institute inter-bank number.

Step 7: the characteristic information of Real-time Obtaining defect is mated to calculating with the characteristic information of all storage defect successively, obtain Real-time Obtaining defect and site error Ep and the plesiomorphism degree S of storage defect; Utilize fuzzy control rule according to Ep and S, obtain the defect characteristic information of coupling;

The concrete acquisition process of the defect characteristic information of coupling is as follows:

1) judge whether Real-time Obtaining defect and storage defect are same defect;

The width Lo (XW) of contrast Real-time Obtaining defect and the width Lt (XW) of storage defect, the height Lo (YH) of contrast Real-time Obtaining defect and the height Lt (YH) of storage defect, if or real-time Obtaining defect is not same defect with storage defect, otherwise, calculate Real-time Obtaining defect and the site error Ep of storage defect, further judge whether Real-time Obtaining defect and storage defect are same defect;

2) calculating Real-time Obtaining defect and the site error Ep of storage defect:

$\mathrm{Ep}=\frac{{\mathrm{Lt}}_m\left(\mathrm{RX}\right)|{\mathrm{Lt}}_m\left(\mathrm{LX}\right)-{\mathrm{Lo}}_n\left(\mathrm{LX}\right)|+{\mathrm{Lt}}_m\left(\mathrm{LX}\right)|{\mathrm{Lt}}_m\left(\mathrm{RX}\right)-{\mathrm{Lo}}_n\left(\mathrm{RX}\right)|}{{2\mathrm{Lt}}_m\left(\mathrm{LX}\right){\mathrm{Lo}}_n\left(\mathrm{RX}\right)}$

If Ep>0.2, Real-time Obtaining defect is not same defect with storage defect, otherwise, calculate Real-time Obtaining defect and the plesiomorphism degree S of storage defect;

3) Real-time Obtaining defect is as follows with the plesiomorphism degree S computation process of storage defect:

Step a: the length of adjusting array Dom (n);

Find peak point Pkt, the Pko of array Dtm (m) and array Dom (n), adjust the length of array Dom (n) take peak point as reference point, wherein, before Dtm (m) array peak point, length is LFt, after peak point, length is LBt, before Dom (n) array peak point, length is LFo, and length is LBo after peak point;

If illustrate that two arrays are reverse, enter step b; Otherwise, if adjust array Dom (n), array Dom (n) peak point previous section is adjusted into LFt length, redundance intercepts, and lacks part and gets Dtm (m) array appropriate section polishing; Array Dom (n) peak point aft section is adjusted into LBt length, and redundance intercepts, and lacks part peek group Dtm (m) appropriate section polishing, enters step c;

Step b: by array Dom (n) upset storage, repeating step a;

Step c: calculating Real-time Obtaining defect and the plesiomorphism degree S of storage defect;

$S=\frac{M\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\mathrm{Dtm}\left(i\right)\mathrm{Dom}\left(i\right)-\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\mathrm{Dtm}\left(i\right)\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\mathrm{Dom}\left(i\right)}{\sqrt{M\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\mathrm{Dtm}\left(i\right)-{\left(\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\mathrm{Dtm}\left(i\right)\right)}^2}\sqrt{M\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\mathrm{Dom}\left(i\right)-{\left(\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\mathrm{Dom}\left(i\right)\right)}^2}};$

As shown in Figure 7, be the forward feature array comparison diagram of storage defect feature array and Real-time Obtaining defect, wherein, solid line is storage defect feature array, dotted line is the forward feature array of Real-time Obtaining defect; Wherein, horizontal ordinate is defect characteristic array sequence, and ordinate is defect characteristic array value;

As shown in Figure 8, be the opposite feature array comparison diagram of storage defect feature array and Real-time Obtaining defect, wherein, solid line is storage defect feature array, dotted line is the opposite feature array of Real-time Obtaining defect; Wherein, horizontal ordinate is defect characteristic array sequence, and ordinate is defect characteristic array value;

4) set membership function, input variable is carried out to Fuzzy Processing;

${\mathrm{\&mu;}}_S\left(S\right)=\left\{\begin{array}{cc}{P}_1& 0.8\&le;S\&le;0.9\\ Q_1& 0.9<S\&le;0.95\\ R_1& 0.95<S\&le;1.0\end{array}\right.,{\mathrm{\&mu;}}_{\mathrm{EP}}\left(\mathrm{EP}\right)=\left\{\begin{array}{cc}{P}_2& 0\&le;\mathrm{EP}\&le;0.1\\ Q_2& 0.1<\mathrm{EP}\&le;0.15\\ R_2& 0.15<\mathrm{EP}\&le;0.2\end{array}\right.$

5) according to fuzzy reasoning table, output fuzzy language variable U, wherein, N ₀expression Real-time Obtaining defect and storage defect are not same defects, P ₀expression Real-time Obtaining defect and storage defect are same defects;

Table 1 fuzzy reasoning table

Numbering	Input variable (S, EP)	Output variable U
			1	(R 1,P 2)	P 0
2	(R 1,Q 2)	P 0
			3	(R 1,R 2)	N 0
4	(Q 1,P 2)	P 0
			5	(Q 1,Q 2)	N 0
6	(Q 1,R 2)	N 0
			7	(P 1,P 2)	N 0
8	(P 1,Q 2)	N 0
			9	(P 1,R 2)	N 0

6) recording output variable is N ₀coupling defect characteristic information.

Two groups of data in Fig. 7, from same defect, obtain under different acquisition conditions.The variation tendency that can see two groups of data is inconsistent, and same similarity is also very low.Analysis reason is known, and this problem is to be caused by the direction gathering.

By Real-time Obtaining defect characteristic array reversion in Fig. 7, obtain the opposite feature array of defect.The opposite feature array that can find out target signature array and defect in Fig. 8, variation tendency is consistent, and similarity is very high.

Comparison diagram 7 and Fig. 8, can find out, same defect is due to the direction difference gathering, and its similarity may be very low, and by after one of them feature array reversion, it is normal that similarity can be recovered.Therefore,, in the time calculating the similarity of two groups of defect characteristic arrays, in once calculating, similarity is very low, after the array of need to reversing, recalculates; While determining whether same defect characteristic, need to carry out twice calculating to similarity, using the high result of calculation of similarity as basis for estimation, could accurately find identical defect.

Claims (6)

1. the high speed railway track surface imperfection matching process based on Machine Vision Detection, is characterized in that, comprises following step:

Step 1: obtain rail panoramic picture f (x, y);

Step 2: adopt vertical projection method to extract rail surf zone image f from rail panoramic picture ₁(x, y);

Step 3: rail surf zone image is carried out to medium filtering processing, obtain the rail surf zone image f that removes noise ₂(x, y);

Step 4: to removing the rail surf zone image f of noise ₂(x, y) carries out successively binaryzation operation and morphology is opened operation, makes f ₂each division region in (x, y) is mutually bonding, obtains image f ₄(x, y);

Step 5: to image f ₄(x, y) carries out projection line by line and adopts blob analytical approach to obtain image f ₄defect characteristic information in (x, y), described defect characteristic information comprises the shape information of defect and the positional information of defect;

Step 6: history collection image and real-time image acquisition are carried out respectively to the operation of step 1-step 5, respectively the defect characteristic information in the historical collection of acquisition image and the defect characteristic information in real-time image acquisition;

Step 7: the characteristic information of Real-time Obtaining defect is mated to calculating with the characteristic information of all storage defect successively, obtain Real-time Obtaining defect and site error Ep and the plesiomorphism degree S of storage defect; Utilize fuzzy control rule according to Ep and S, obtain the defect characteristic information of coupling.

2. the high speed railway track surface imperfection matching process based on Machine Vision Detection according to claim 1, it is characterized in that, the historical defect characteristic information gathering in image in described step 6, comprises the shape information Dtm (m) of storage defect and the positional information Lt of storage defect _m(LX, LY, RX, RY, XW, YH), m represents m storage defect, and span is 1-M, and M represents storage defect sum;

Defect characteristic information in real-time image acquisition, comprises the shape information Dom (n) of Real-time Obtaining defect and the positional information Lo of Real-time Obtaining defect _n(LX, LY, RX, RY, XW, YH), n represents n Real-time Obtaining defect, and span is 1-N, and N represents the sum of Real-time Obtaining defect;

Wherein, the shape information Am (j) of each defect is according to each defect place image f ₄line number j in (x, y), calculates and obtains according to following formula:

$\mathrm{Am}\left(j\right)=\frac{\underset{y=1}{\overset{\mathrm{Width}}{\mathrm{\&Sigma;}}}{f}_4(j,y)}{255};$

Wherein, j is the integer between [1, Height], and Height is image f ₄the height of (x, y), Width is image f ₄(x, y) wide, LX is the left row coordinate of defect boundary rectangle, LY is the left-hand line coordinate of defect boundary rectangle, RX is the right side row-coordinate of defect boundary rectangle, RY is the right-hand column coordinate of defect boundary rectangle, the width that XW is defect be defect institute across columns, the height that YH is defect is defect institute inter-bank number.

3. the high speed railway track surface imperfection matching process based on Machine Vision Detection according to claim 2, is characterized in that, the concrete acquisition process of the defect characteristic information of the coupling in described step 7 is as follows:

1) judge whether Real-time Obtaining defect and storage defect are same defect;

The width Lo (XW) of contrast Real-time Obtaining defect and the width Lt (XW) of storage defect, the height Lo (YH) of contrast Real-time Obtaining defect and the height Lt (YH) of storage defect, if or real-time Obtaining defect is not same defect with storage defect, otherwise, calculate Real-time Obtaining defect and the site error Ep of storage defect, further judge whether Real-time Obtaining defect and storage defect are same defect;

2) calculating Real-time Obtaining defect and the site error Ep of storage defect:

$\mathrm{Ep}=\frac{{\mathrm{Lt}}_m\left(\mathrm{RX}\right)|{\mathrm{Lt}}_m\left(\mathrm{LX}\right)-{\mathrm{Lo}}_n\left(\mathrm{LX}\right)|+{\mathrm{Lt}}_m\left(\mathrm{LX}\right)|{\mathrm{Lt}}_m\left(\mathrm{RX}\right)-{\mathrm{Lo}}_n\left(\mathrm{RX}\right)|}{{2\mathrm{Lt}}_m\left(\mathrm{LX}\right){\mathrm{Lo}}_n\left(\mathrm{RX}\right)}$

If Ep>0.2, Real-time Obtaining defect is not same defect with storage defect, otherwise, calculate Real-time Obtaining defect and the plesiomorphism degree S of storage defect;

3) Real-time Obtaining defect is as follows with the plesiomorphism degree S computation process of storage defect:

Step a: the length of adjusting array Dom (n);

Find peak point Pkt, the Pko of array Dtm (m) and array Dom (n), adjust the length of array Dom (n) take peak point as reference point, wherein, before Dtm (m) array peak point, length is LFt, after peak point, length is LBt, before Dom (n) array peak point, length is LFo, and length is LBo after peak point;

If illustrate that two arrays are reverse, enter step b; Otherwise, if adjust array Dom (n), array Dom (n) peak point previous section is adjusted into LFt length, redundance intercepts, and lacks part and gets Dtm (m) array appropriate section polishing; Array Dom (n) peak point aft section is adjusted into LBt length, and redundance intercepts, and lacks part peek group Dtm (m) appropriate section polishing, enters step c;

Step b: by array Dom (n) upset storage, repeating step a;

Step c: calculating Real-time Obtaining defect and the plesiomorphism degree S of storage defect;

$S=\frac{M\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\mathrm{Dtm}\left(i\right)\mathrm{Dom}\left(i\right)-\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\mathrm{Dtm}\left(i\right)\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\mathrm{Dom}\left(i\right)}{\sqrt{M\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\mathrm{Dtm}\left(i\right)-{\left(\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\mathrm{Dtm}\left(i\right)\right)}^2}\sqrt{M\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\mathrm{Dom}\left(i\right)-{\left(\underset{i=1}{\overset{M}{\mathrm{\&Sigma;}}}\mathrm{Dom}\left(i\right)\right)}^2}};$

4) set membership function, input variable is carried out to Fuzzy Processing;

${\mathrm{\&mu;}}_S\left(S\right)=\left\{\begin{array}{cc}{P}_1& 0.8\&le;S\&le;0.9\\ Q_1& 0.9<S\&le;0.95\\ R_1& 0.95<S\&le;1.0\end{array}\right.,{\mathrm{\&mu;}}_{\mathrm{EP}}\left(\mathrm{EP}\right)=\left\{\begin{array}{cc}{P}_2& 0\&le;\mathrm{EP}\&le;0.1\\ Q_2& 0.1<\mathrm{EP}\&le;0.15\\ R_2& 0.15<\mathrm{EP}\&le;0.2\end{array}\right.$

5) according to fuzzy reasoning table, output fuzzy language variable U, wherein, N ₀expression Real-time Obtaining defect and storage defect are not same defects, P ₀expression Real-time Obtaining defect and storage defect are same defects;

Table 1 fuzzy reasoning table

Numbering Input variable (S, EP) Output variable U 1 (R ₁,P ₂) P ₀ 2 (R ₁,Q ₂) P ₀ 3 (R ₁,R ₂) N ₀ 4 (Q ₁,P ₂) P ₀ 5 (Q ₁,Q ₂) N ₀ 6 (Q ₁,R ₂) N ₀ 7 (P ₁,P ₂) N ₀ 8 (P ₁,Q ₂) N ₀ 9 (P ₁,R ₂) N ₀

6) recording output variable is N ₀coupling defect characteristic information.

4. the high speed railway track surface imperfection matching process based on Machine Vision Detection according to claim 3, is characterized in that, the specific implementation process of described step 2 is as follows:

1. cumulative gray-scale value by column, produces the gray average array Avg (i) of the each row of rail image f (x, y):

$\mathrm{Avg}\left(i\right)=\frac{\underset{x=1}{\overset{\mathrm{Height}}{\mathrm{\&Sigma;}}}f(x,i)}{\mathrm{Height}},i\&Element;[1,\mathrm{Width}],$ Height is the height of image, and i is image i row;

2. calculate the average Avg_all of Avg (i);

3. Avg (i) array binaryzation, by Avg (i) array one by one with Avg_all comparison, be greater than Avg_all and be made as 1, be less than Avg_all and be made as 0, obtain array Avg_Val (i):

$\mathrm{Avg}\_\mathrm{Val}\left(i\right)=\left\{\begin{array}{cc}1,& \mathrm{Avg}\left(i\right)\&GreaterEqual;\mathrm{Avg}\_\mathrm{all}\\ 0,& \mathrm{Avg}\left(i\right)<\mathrm{Avg}\_\mathrm{all}\end{array}\right.;$

4. in Avg_Val (i) array, finding first continuous ten numbers is all 1 point, the marginal point Begin_Point that this starts as rail surface image; After Avg_Val (i) array Begin_Point, finding first continuous ten numbers is all 0 point, the marginal point End_Point that this starts as rail surface image, the image that intercepting rail image f (x, y) Begin_Point is listed as between End_Point obtains rail surface image f ₁(x, y).

5. the high speed railway track surface imperfection matching process based on Machine Vision Detection according to claim 4, is characterized in that, the specific implementation process of described step 3 is to rail surface image f ₁(x, y) carries out the medium filtering operation of a time 3 × 3, removes image noise, obtains the rail surface image f of low noise ₂(x, y).

6. the high speed railway track surface imperfection matching process based on Machine Vision Detection according to claim 5, is characterized in that, the specific implementation process of described step 4 is as follows:

1. set global threshold T, to rail surface image f ₂(x, y) carries out binaryzation and obtains bianry image f ₃(x, y):

$f_3(x,y)=\left\{\begin{array}{cc}255,& f_2(x,y)\&GreaterEqual;T\\ 0,& f_2(x,y)<T\end{array}\right.;$

2. use 9 × 9 oval structure element B to bianry image f ₃(x, y) carries out morphology and opens operation and obtain image f ₄(x, y):

In formula, meet " Θ " and represent the corrosion operation in mathematical morphology, symbol represent the closed procedure in mathematical morphology, make the single division region in image obtain mutually bonding.