CN107576666A - A kind of double light spectrum image-forming rails and fastener method for detecting abnormality - Google Patents
A kind of double light spectrum image-forming rails and fastener method for detecting abnormality Download PDFInfo
- Publication number
- CN107576666A CN107576666A CN201710949482.9A CN201710949482A CN107576666A CN 107576666 A CN107576666 A CN 107576666A CN 201710949482 A CN201710949482 A CN 201710949482A CN 107576666 A CN107576666 A CN 107576666A
- Authority
- CN
- China
- Prior art keywords
- image
- fastener
- rail
- thermal
- area
- 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.)
- Pending
Links
- 238000001228 spectrum Methods 0.000 title claims abstract description 51
- 238000000034 method Methods 0.000 title claims abstract description 43
- 230000005856 abnormality Effects 0.000 title claims abstract description 29
- 238000003384 imaging method Methods 0.000 claims abstract description 72
- 238000001514 detection method Methods 0.000 claims abstract description 59
- 230000003287 optical effect Effects 0.000 claims abstract description 32
- 238000009740 moulding (composite fabrication) Methods 0.000 claims abstract description 19
- 238000005286 illumination Methods 0.000 claims description 12
- 230000002159 abnormal effect Effects 0.000 claims description 11
- 238000001914 filtration Methods 0.000 claims description 8
- 239000011159 matrix material Substances 0.000 claims description 6
- 230000000877 morphologic effect Effects 0.000 claims description 6
- 230000000007 visual effect Effects 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 4
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 230000003628 erosive effect Effects 0.000 claims description 3
- 238000000605 extraction Methods 0.000 claims description 3
- 238000009434 installation Methods 0.000 claims description 3
- 230000007613 environmental effect Effects 0.000 claims description 2
- XEEYBQQBJWHFJM-UHFFFAOYSA-N Iron Chemical compound [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 claims 2
- 230000002452 interceptive effect Effects 0.000 claims 1
- 229910052742 iron Inorganic materials 0.000 claims 1
- 230000013011 mating Effects 0.000 claims 1
- 238000010586 diagram Methods 0.000 description 7
- 208000037656 Respiratory Sounds Diseases 0.000 description 5
- 230000007547 defect Effects 0.000 description 4
- 230000036548 skin texture Effects 0.000 description 3
- 230000015572 biosynthetic process Effects 0.000 description 2
- 235000008331 Pinus X rigitaeda Nutrition 0.000 description 1
- 235000011613 Pinus brutia Nutrition 0.000 description 1
- 241000018646 Pinus brutia Species 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 210000001367 artery Anatomy 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000000701 chemical imaging Methods 0.000 description 1
- 238000010835 comparative analysis Methods 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000001931 thermography Methods 0.000 description 1
- 210000003462 vein Anatomy 0.000 description 1
Abstract
The present invention discloses a kind of double light spectrum image-forming rails and fastener method for detecting abnormality, belongs to railway traffic safety detection technique field, for solving the problems, such as that prior art is difficult to rail and fastener abnormality detection；The present invention forms double optical spectrum imaging devices using visible light camera and thermal camera, double light spectrum image-formings are carried out to rail and fastener area, rail and fastener texture image and thermal-induced imagery are obtained respectively, rail and fastener positioning in thermal-induced imagery are carried out using texture image, compared using background model and carry out fastener missing, loosening, fracture abnormality detection, be used in conjunction with texture image and thermal-induced imagery carries out rail stripping, scratch, ripple mill, crack detection；The present invention can effectively lift rail and fastener abnormality detection efficiency, ensure track traffic safety.
Description
Technical field
The present invention relates to rail traffic safety detection technique field, refers in particular to a kind of double light spectrum image-forming rails and fastener and examines extremely
Survey method.
Background technology
In railway infrastructure detection, for a long time based on artificial and Static Detection, maintenance costs is high in China, and intensity is high,
Security is poor, in recent years China Express Railway fast development, and the automation, real-time to railway detection propose higher requirement.
Existing rail and fastener detection technique is both at home and abroad：Continuous scanning means based on linear array laser, such as German Sick companies；Base
In the Computer Vision Detection device of array image sensor, VIS systems, Germany such as ENSCO companies of the U.S.
The photo-electric rail detection system and Beijing good fortune Sida Company high speed vehicular trajectory diagram of AtlasElectronic companies exploitation
As identifying system etc..But both the shortcomings that, are that detection speed is relatively low, and versatility is not high, it is sometimes desirable to excessive manual intervention, and
And automatic decision fastener whether can not produce loosening before fastener loss and carry out early warning.
Through retrieval, Z.201210192641.2 patent introduces a kind of rail fastener based on infrared thermal imaging and loosens spy at a high speed
Examining system and method：Thermal-induced imagery caused by the contact stress of fastener and rail is obtained using thermal camera, by infrared
Thermal image gray value, judges whether fastener loosens or lack.But this method has the following disadvantages：1) taken the photograph using face battle array is infrared
Camera, collection image data amount is big, fastener area imaging is difficult to align, and the imaging resolution of face battle array thermal camera is low, sets
Standby cost is high；2) directly compared using the temperature value temperature value corresponding with minimum, maximum stress value of infrared image measurement
Detection method, it also there are problems that two in actual applications：First, under different train running speeds, suffered by fastener
Extruding force it is of different sizes, caused heat is also different, and therefore, this method is difficult in adapt to fastener pine under the conditions of friction speed
Dynamic detection demand；Secondly, this method can not detect fastener crackle exception, also abnormal without detection rail.
The content of the invention
In order to solve the above mentioned problem present in existing rail and fastener loosening detection method, the present invention provides a kind of double light
Spectrum imaging rail and fastener method for detecting abnormality.
The technical scheme is that：A kind of double light spectrum image-forming rails and fastener method for detecting abnormality, it is characterised in that：Using
Visible light camera and thermal camera form double optical spectrum imaging devices, and double light spectrum image-formings are carried out to rail and fastener area, point
Not Huo Qu rail and fastener texture image and thermal-induced imagery, rail and fastener in thermal-induced imagery are carried out using texture image and determined
Position, compared using background model and carry out fastener missing, loosening, fracture abnormality detection, be used in conjunction with texture image and infrared chart
As carrying out rail stripping, scratch, ripple mill, crack detection.
Line array video camera or area array camera can be selected in the visible light camera and thermal camera, it is preferable that visible
Light video camera and thermal camera select line array video camera, wherein, it is seen that the imaging resolution of light array camera is not less than
1024 pixels, spectrum sensing range are 400nm~1000nm, and the imaging resolution of infrared line array video camera is not less than 1024 pictures
Element, the temperature control of infrared line array video camera are not less than 0.5 DEG C.
The luminous ray array camera and infrared line array video camera are located at directly over rail, optical axis perpendicular to road surface, into
As section is perpendicular to rail longitudinal direction, shooting visual field covering rail and both sides fastener area.
The line array video camera imaging slice refers to：Imaging region and camera lens in line scan image sensor in line array video camera
The plane that photocentre is formed.
To ensure that the texture image of shooting and infrared chart seem what visual field was alignd, it is seen that light video camera and thermal camera
It is imaged by coaxial light splitting optical path.
The coaxial light splitting optical path refers to：The imaging slice of the optical axis of luminous ray array camera and infrared line array video camera
Vertically, the optical axis of infrared line array video camera is vertical with the imaging slice of luminous ray array camera, it is seen that light array camera
Photocentre to the photocentre of distance and the infrared line array video camera of infrared line array video camera imaging slice to luminous ray array camera into
The distance of picture section is equal, and the imaging angle of two video cameras is identical, optical axis intersects；It is imaged in luminous ray array camera
Section is with the intersection of thermal camera imaging slice, setting a dichroscope；The light splitting surface and luminous ray of dichroscope
The intersection of array camera imaging slice and infrared line array video camera imaging slice is tangent, with luminous ray array camera and infrared ray
The optical axis of array camera is into 45 degree of angles, for incident ray to be separated into visible and near-infrared mixed light, mid-infrared and far-infrared light；When can
When seeing that light array camera is located at the transmitted light path of dichroscope, infrared line array video camera is located at the reflected light path of dichroscope,
From thoroughly visible and near-infrared mixed light, the dichroscope of reflection mid-infrared and far-infrared light；When luminous ray array camera positioned at two to
When the reflected light path of Look mirror, infrared line array video camera are located at the transmitted light path of dichroscope, mixed from reflection is visible with near-infrared
Closing light, the dichroscope for transmiting mid-infrared and far-infrared light.
Preferably, a line source, the projection plane of two line sources are respectively set directly over rail on both sides fastener area
It is coplanar with the imaging slice of luminous ray array camera, provide illumination for luminous ray array camera.
An area source, area source direction of illumination and longitudinal incident, the area source into a angles of rail are arranged above in rail
Imaging region of the irradiation area covering visible light line array video camera on rail, form details in a play not acted out on stage, but told through dialogues with luminous ray array camera and shine
It is bright, it is imaged for rail damage, a angular configurations scope is 10~90 degree.
Two line sources above the fastener and the area source irradiation area above rail be not overlapping, mutually dry to avoid
Disturb.
Preferably, line source and area source are parallel light source.
Preferably, line source and area source select the Single wavelength or narrow-band spectrum LASER Light Source of phase co-wavelength, optical source wavelength
Scope is：400~1000nm, and set in luminous ray array camera front end narrow with line source and area array light source phase co-wavelength
Band optical filter, is disturbed for filtering environmental light.
Preferably, line source selects near-infrared light source with area source, and wave-length coverage is：800~950nm.
To realize to rail during train driving and fastener scanning imagery, a wheel is fixed in train wheel rotating shaft
Encoder is encoded to vehicle wheel rotation angle and produces pulse triggering signal, and pulse triggering signal is used to drive visible ray linear array
Video camera and infrared line array video camera carry out linear array imaging scanning to rail and fastener area, obtain rail and fastener area texture
Image and thermal-induced imagery.
In actual use, respectively above train bottom, left and right rail, respectively set a set of by luminous ray array camera
The double optical spectrum imaging devices formed with infrared line array video camera, imaging is scanned to left and right rail and its both sides fastener area.
Preferably, it is acquisition and the double light spectrum image-forming images of rail length always when train passes through bend, in left and right two
One vehicle wheel rotation encoder is respectively installed, respectively the double optical spectrum imaging devices in the driving left and right sides on the wheel of side.
The inventive method carries out rail surface method for detecting abnormality：
Step 1：According to rail in texture image and thermal-induced imagery image space, be partitioned into rail surface texture image
IwWith thermal-induced imagery Ig, IgSize be h1*w1, h1, w1 span are：50~1000；
Step 2：Establish rail surface thermal-induced imagery background model
Train running speed is divided into n grade L={ l1,...,ln, n span is 1~500, according to train
Speed of service grade n, establish normal rail thermal-induced imagery background model in n group train travelling processesIts
In, background modelIt is by the normal rail surface thermal-induced imagery of the k frames shot under the i-th grade train running speedTrain
Arrive, background modelBy the matrix of 2 h1*w1 dimensionsComposition, wherein,Middle element is that the normal rail surface of k frames is infrared
Thermal image pixel time domain average,It is variance of the normal rail surface thermal-induced imagery pixel of k frames in time domain；
Step 3：Carry out rail surface temperature anomaly detection
Input current train movement velocity v and rail surface thermal-induced imagery
According to train running speed v calculating speed grades lc, from background modelMiddle selection lcThe individual back of the body
Scape model：Background difference operation is performed, obtains background difference image dg：
In formula, dg(x, y) is background difference image dgXth row, y row pixels, x, y are that image is horizontal, ordinate, x value models
Enclose for 1~h1, y 1~w1 of span；
2 width mark image F is set1 g,Image size is h1*w1, and element initial value is 0 in image；
WhenWhen, make F1(x, y)=1；
WhenWhen, make F2(x, y)=1；
Wherein,It is detection threshold value multiplying power factor, span is 0~100；
To image F1 g,Morphologic filtering is carried out respectively and eliminates isolated noise, and finds image F1 g,Middle value is 1
Connected domain；As image F1 gOrIt is middle when the connected domain that area is more than T1 be present, it is abnormal to judge that rail surface is present, wherein T1 is
Detection threshold value, span are 1~1000000000；
Step 4：Rail surface abnormality detection is carried out based on rail texture image
Binary conversion treatment is carried out to rail surface texture image：1 is set to above or equal to threshold value T2 pixel, less than threshold
Value T2 pixel is set to 0, obtains binary imageTo binary imageCarry out morphologic filtering and eliminate picture noise, and
Find imageMiddle value is 1 connected domain；Work as imageIt is middle exist area be more than T2 connected region when, judge rail table
There is exception in face, wherein T2 is detection threshold value, and span is 1~1000000000；
Step 5：Rail surface anomaly classification is carried out based on rail texture binary image and thermal-induced imagery
Operation is merged to the connected domain detected in step 3 and step 4, and calculate merge after connected domain it is maximum external
Rectangle R；Calculate F1 g,The length-width ratio of connected domain, barycenter, major axes orientation, area in middle Zone R domain, constitutive characteristic vector V1；
To pixel extraction histogram in thermal-induced imagery Zone R domain, constitutive characteristic vector V2；Respectively to V1、V2After being normalized,
And with obtaining current thermal-induced imagery train running speed v as rail surface anomaly classification characteristic quantity V'={ v, V1,V2}；So
Afterwards, the extremely positive negative sample of rail surface is gathered, trains sorter model, rail surface abnormal class is judged using sorting technique.
Carrying out the method for fastener abnormality detection is：
Step 1：According to fastener in texture image and thermal-induced imagery imaging region, be partitioned into rail surface texture image
And thermal-induced imagery；
Step 2：Fastener in thermal-induced imagery is carried out using texture image to position：Using image template matching method, detaining
Fastener area-of-interest (Regionofinteresting, ROI) is found in part texture image；Based on the ROI in thermal-induced imagery
Positioning buckling element region, take out fastener area thermal-induced imagery Ik, thermal-induced imagery IkSize be h2*w2, h2, w2 value model
Enclose for：50~1000；
Step 3：Establish fastener thermal-induced imagery background model
Train running speed is divided into n grade L={ l1,...,ln, n span is 1~500, according to train
Speed of service grade n, establish normal fastener thermal-induced imagery background model in n group train travelling processesIts
In, background modelIt is by the normal fastener thermal-induced imagery of the k frames shot under the i-th grade train running speedTraining obtains,
Fastener area thermal-induced imagery size is h2*w2, background modelBy the matrix of 2 h2*w2 dimensionsComposition, wherein,In
Element is average of the normal fastener thermal-induced imagery pixel of k frames in time domain,The normal fastener thermal-induced imagery pixel of k frames when
The variance in domain；
Step 4：Perform background Differential Detection
Input current train movement velocity v and fastener thermal-induced imagery
According to train running speed v calculating speed grades lc, from background modelMiddle selection lcThe individual back of the body
Scape model：Background difference operation is performed, obtains background difference image dk：
In formula, dk(x, y) is background difference image dkXth row, y row pixels, x, y are that image is horizontal, ordinate, x value models
Enclose for 1~h2, y 1~w2 of span；
Three width mark image F is set1 k,Image size is h2*w2, and element initial value is 0 in image；
WhenWhen, make F1 k(x, y)=1；
WhenWhen, make
WhenWhen, make
Wherein, ω1,ω2,ω3It is detection threshold value multiplying power factor, span is 0~100；
Step 5：Perform abnormal fastener detection
To image F1 k,Image erosion, expansive working are carried out respectively, are eliminated in image after isolated noise, are found figure
PictureMiddle value is 1 connected domain；As image F1 kOrIt is middle exist area be more than T3 connected domain when, judge fastener
In the presence of exception, wherein T3 is detection threshold value, and span is 1~1000000000；
Work as imageIt is middle connected region of the area more than T4 to be present and in imageIn be not present area be more than T5 company
Lead to region, in imageIn when area be not present being more than T6 connected region, judge that button is normal；
Work as imageIt is middle connected region of the area more than T5 to be present and in imageIn be not present area be more than T4 company
During logical region, fastener missing is judged；
As image F1 kIt is middle connected region of the area more than T5 to be present and in imageThe middle connection that area be present and be more than T4
During region, judge that fastener loosens；
Work as imageIt is middle connected region of the area more than T8 to be present and in imageThe middle connection that area be present and be more than T4
During region, judge that fastener is cracked, wherein T4, T5, T6, T7, T8 are detection threshold values, and span is 1~1000000000.
Beneficial effect of the present invention：Compared with prior art, particularly with patent Z.201210192641.2 compared with, the present invention
The advantages of be：1) using double spectrum imaging methods, while rail and fastener skin texture images and thermal-induced imagery are obtained, utilized
Texture image carries out detection zone positioning, particularly fastener area to infrared image and positioned, in order to contacting piece region
Infrared image carries out the comparative analysis that becomes more meticulous；2) when carrying out luminous ray array camera scanning imagery to rail, use is oblique
Incident source of parallel light, dark-ground illumination is formed with visible light camera, rail surface ripple mill, crackle, peeling can be highlighted
Equivalent damage, lift rail surface damage check ability；3) area array camera is replaced using linear array thermal camera, 1024 can be used
The infrared line array video camera of pixel, high-resolution thermal-induced imagery is obtained, compared to the face battle array infrared photography of 640*480 resolution ratio
Machine, imaging resolution is higher, cost is low, captured image data amount is few；4) linear array thermal camera imaging plane perpendicular is in rail
And road surface, reference object has same widths size (width on rail direction) in linear array images, and face battle array is infrared
There is transmission projection deformation in video camera, the fastener area size shot at diverse location is different, in order to ensure fastener in face battle array
There is identical size, it is necessary to be precisely controlled video camera and shot directly over fastener, and make imaging in thermal camera shooting image
Control system becomes complicated, and then only needs to use wheel encoder using linear array thermal camera, along rail scanning direction
, this image formation control method has the advantages of simple, reliable；5) split according to fastener in the case where rail extrudes repeatedly, easily
Stress concentration is produced at line, and shows the phenomenon higher than normal fastener temperature, compared to patent Z.201210192641.2 in it is straight
The method judged using thermal-induced imagery gray value is connect, the present invention uses background model comparison method, can not only detected
Whether fastener is lost or is loosened, and can detect to whether there is crackle in fastener；5) present invention uses train running speed
The background model comparison method of classification, rail and fastener abnormality detection reliability can be effectively lifted, fastener is can be especially useful for and splits
Line detects；6) present invention uses coaxial spectroscopic imaging method, it is ensured that double spectrum pictures are alignment.
Brief description of the drawings
Fig. 1 the inventive method schematic diagrams；
The double spectrum face battle array imaging device schematic diagrames of Fig. 2；
Fig. 3 visible ray area array cameras front end sets optical filter schematic diagram；
Fig. 4 light sources illuminate and double spectrum face battle array imaging device scheme of installations
Wherein, (a) is rail cross-sectional direction layout, and (b) is rail longitudinal direction side view；
The normal rail surface dark-ground illumination imaging schematic diagrams of Fig. 5；
Fig. 6 rail surface defect dark-ground illumination imaging schematic diagrams；
The double spectrum linear array imaging device schematic diagrames of Fig. 7；
The double spectrum linear array imagings of Fig. 8 and light source illumination region top view；
The double spectrum linear array imaging device train scheme of installations of Fig. 9；
In figure, 1- rails, 2- fasteners, 3- visible ray area array cameras, the infrared area array cameras of 4-, 5- dichroscopes, 6-
Visible ray, near-infrared and mid and far infrared mixed light, 7- visible rays and near-infrared mixed light, 8- mid-infrared and far-infrared lights, 9- optical filters,
10- directional light area sources, the double spectrum face battle array imaging devices of 11-, 12- rail surface defects, 13- luminous ray array cameras, 14-
Infrared line array video camera, 15 directional light illuminated by surface light source regions, 16- parallel rays light source illumination regions；The double spectrum linear arrays of 17- into
As device imaging region, the double spectrum linear array imaging devices of 18-, 19- vehicle wheel rotation encoders, 20- industrial computers, 21- compartments.
Embodiment
The present invention is described in detail with specific implementation below in conjunction with the accompanying drawings.The present embodiment using technical solution of the present invention as
Premise is implemented, and gives detailed embodiment and specific operating process, but protection scope of the present invention is not limited to down
The embodiment stated.
As shown in figure 1, the present invention uses double optical spectrum imaging devices, double light spectrum image-formings are carried out to rail and fastener area, point
Huo Qu not rail and fastener texture image and thermal-induced imagery；Rail and fastener in thermal-induced imagery are carried out using texture image to determine
Position, compared using background model and carry out fastener missing, loosening, fracture abnormality detection, be used in conjunction with texture image and infrared chart
As carrying out rail stripping, scratch, ripple mill, crack detection.
Embodiment 1
As shown in Fig. 2 form double spectrum faces battle array using 3, infrared area array cameras 4 of a visible ray area array camera
Imaging device 11, wherein, it is seen that smooth surface array camera 3 is vertical with the optical axis of infrared area array camera 4, is imaged in visible smooth surface battle array
Dichroscope 5, the light splitting surface of dichroscope 5 and visible smooth surface battle array camera 3 and infrared area array camera 4 are set on the optical axis of machine
Optical axis respectively into 45 degree of angles；The visible of incidence, near-infrared and mid and far infrared mixed light 6 are separated into visible ray by dichroscope 5
With near-infrared mixed light 7 and mid-infrared and far-infrared light 8, visible ray area array camera 3 and thermal camera 4 are respectively enterd, for obtaining
Double spectrum pictures；Adjust visible ray area array camera 3, infrared area array camera 4 to the light splitting surface of dichroscope 5 away from walk-off angle
Degree, makes the shooting image of visible ray area array camera 3 be overlapped with the infrared shooting image visual field of area array camera 4；Wherein visible smooth surface
The resolution ratio of array camera 3 is 1,000,000 pixels, and the resolution ratio of infrared area array camera 4 is 640*480 pixels；It is based on above-mentioned
Double spectrum face battle array imaging devices 11 that visible ray area array camera 3, infrared area array camera 4 are formed are fixed on the top of rail 1, make
Rail 1 and the region of fastener 2 is completely covered in double light spectrum image-forming visual fields, for obtaining rail 1 and the skin texture images of fastener 2 and infrared
Thermal image.
Embodiment 2
Difference from Example 1 is, sets lighting source on double sides of spectrum face battle array imaging device 11, light source projects arrive
Rail 1 and the region of fastener 2, optical source wavelength 800nm, as shown in figure 3, setting 800nm arrowband in visible light camera front end
Optical filter 9, to eliminate the ambient light interference rail and fastener skin texture images consistent with brightness is obtained.
Embodiment 3
Difference from Example 2 is, as shown in figure 4, setting 3 directional lights on double sides of spectrum face battle array imaging device 11
Area source 10, wherein 2 directional light area sources 10-1,10-2 are located at directly over the region of rail on both sides fastener 2,1 parallel smooth surface
Light source 10-3 oblique incidences are on rail 1, and incident direction and the angle of the longitudinal direction of rail 1 are 20 degree, and 3 light source irradiation areas do not weigh
It is folded.
As shown in figure 5, when parallel area source 10-3 oblique incidence rails 1, formed between visible ray area array camera 3
Dark field illumination system；When rail surface does not occur ripple mill, stripping, crackle equivalent damage, light source 10-3 directional light reaches rail
Behind surface, most light ray energies are projected in a manner of mirror-reflection, cannot be introduced into visible ray area array camera 3, thus in rail area
Domain imaging is very dark；As shown in fig. 6, when ripple mill, stripping, crackle equivalent damage occurs in rail surface, because of rail surface defect 12
Reflection angle is different from track level, and visible ray area array camera formation bright district can be entered by projecting light at defective locations
Domain.Therefore, this dark-ground illumination mode can be utilized, realizes that rail surface defect highlights.
Embodiment 4
Difference from Example 3 is, as shown in fig. 7, replacing visible smooth surface battle array using luminous ray array camera 13
Video camera 3, infrared area array camera 4 is replaced using infrared line array video camera 14, form double spectrum linear array imaging devices 18；Using
Parallel rays light source replaces 2 directional lights area source 10-1,10-2, and illumination is provided for luminous ray array camera 13；Wherein, may be used
The resolution ratio for seeing light array camera 13 is 2048 pixels, line frequency 100KHz, and the resolution ratio of infrared line array video camera 14 is
1024 pixels, line frequency 100KHz；The imaging area of the imaging region of luminous ray array camera 13 and infrared line array video camera 14
Domain overlaps；As shown in figure 8,2 parallel rays light source irradiation areas 16, directional light area source irradiation area 15, respectively positioned at button
Part 2 and the top of rail 1, and the imaging region 17 of double spectrum linear array imaging devices 18 is completely covered.
As shown in fig. 7, the installing wheel rotary encoder 19 on wheel, to vehicle wheel rotation angular coding and produces triggering arteries and veins
Signal is rushed, caused start pulse signal is used to drive luminous ray array camera 13 and infrared line array video camera 14 to rail
Imaging is scanned with fastener area；And the industrial computer 20 by being placed in compartment 21 completes image data acquiring.
Embodiment 5
Difference from Example 4 is, using two vehicle wheel rotation encoders 19 respectively to left and right wheelses rotational angle
Trigger pulse is encoded and is produced, for driving double spectrum linear array imaging device 18-1, the 18-2 in the left and right sides.
Embodiment 6
Rail surface abnormality detection is carried out as follows：
Step 1：According to rail in texture image and thermal-induced imagery image space, be partitioned into rail surface texture image
IwWith thermal-induced imagery Ig, IgSize be h1*w1, h1, w1 span are：50~1000；
Step 2：Establish rail surface thermal-induced imagery background model
Train running speed is divided into n grade L={ l1,...,ln, n span is 1~500, according to train
Speed of service grade n, establish normal rail thermal-induced imagery background model in n group train travelling processesIts
In, background modelIt is by the normal rail surface thermal-induced imagery of the k frames shot under the i-th grade train running speedTrain
Arrive, background modelBy the matrix of 2 h1*w1 dimensionsComposition, wherein,Middle element is that the normal rail surface of k frames is infrared
Thermal image pixel time domain average,It is variance of the normal rail surface thermal-induced imagery pixel of k frames in time domain；
Step 3：Carry out rail surface temperature anomaly detection
Input current train movement velocity v and rail surface thermal-induced imagery
According to train running speed v calculating speed grades lc, from background modelMiddle selection lcThe individual back of the body
Scape model：Background difference operation is performed, obtains background difference image dg：
In formula, dg(x, y) is background difference image dgXth row, y row pixels, x, y are that image is horizontal, ordinate, x value models
Enclose for 1~h1, y 1~w1 of span；
2 width mark image F is set1 g,Image size is h1*w1, and element initial value is 0 in image；
WhenWhen, make F1(x, y)=1；
WhenWhen, make F2(x, y)=1；
Wherein,It is detection threshold value multiplying power factor, span is 0~100；
To image F1 g,Morphologic filtering is carried out respectively and eliminates isolated noise, and finds image F1 g,Middle value is 1
Connected domain；As image F1 gOrIt is middle when the connected domain that area is more than T1 be present, abnormal, the wherein T1 that judges that rail surface is present
It is detection threshold value, span is 1~1000000000；
Step 4：Rail surface abnormality detection is carried out based on rail texture image
Binary conversion treatment is carried out to rail surface texture image：1 is set to above or equal to threshold value T2 pixel, less than threshold
Value T2 pixel is set to 0, obtains binary imageTo binary imageCarry out morphologic filtering and eliminate picture noise, and
Find imageMiddle value is 1 connected domain；Work as imageIt is middle exist area be more than T2 connected region when, judge rail table
There is exception in face, wherein T2 is detection threshold value, and span is 1~1000000000；
Step 5：Rail surface anomaly classification is carried out based on rail texture binary image and thermal-induced imagery
Operation is merged to the connected domain detected in step 3 and step 4, and calculate merge after connected domain it is maximum external
Rectangle R；Calculate F1 g,The length-width ratio of connected domain, barycenter, major axes orientation, area in middle Zone R domain, constitutive characteristic vector V1；
To pixel extraction histogram in thermal-induced imagery Zone R domain, constitutive characteristic vector V2；Respectively to V1、V2After being normalized,
And with obtaining current thermal-induced imagery train running speed v as rail surface anomaly classification characteristic quantity V'={ v, V1,V2}；So
Afterwards, the extremely positive negative sample of rail surface is gathered, trains sorter model, rail surface abnormal class is judged using sorting technique.
Embodiment 7
Fastener abnormality detection is carried out using following methods：
Step 1：According to fastener in texture image and thermal-induced imagery imaging region, be partitioned into rail surface texture image
And thermal-induced imagery；
Step 2：Fastener in thermal-induced imagery is carried out using texture image to position：Using image template matching method, detaining
Fastener area-of-interest (Regionofinteresting, ROI) is found in part texture image；Based on the ROI in thermal-induced imagery
Positioning buckling element region, take out fastener area thermal-induced imagery Ik, thermal-induced imagery IkSize be h2*w2, h2, w2 value model
Enclose for：50~1000；
Step 3：Establish fastener thermal-induced imagery background model
Train running speed is divided into n grade L={ l1,...,ln, n span is 1~500, according to train
Speed of service grade n, establish normal fastener thermal-induced imagery background model in n group train travelling processesIts
In, background modelIt is by the normal fastener thermal-induced imagery of the k frames shot under the i-th grade train running speedTraining obtains,
Fastener area thermal-induced imagery size is h2*w2, background modelBy the matrix of 2 h2*w2 dimensionsComposition, wherein,
Middle element is average of the normal fastener thermal-induced imagery pixel of k frames in time domain,It is that the normal fastener thermal-induced imagery pixel of k frames exists
The variance of time domain；
Step 4：Perform background Differential Detection
Input current train movement velocity v and fastener thermal-induced imagery
According to train running speed v calculating speed grades lc, from background modelMiddle selection lcThe individual back of the body
Scape model：Background difference operation is performed, obtains background difference image dk：
In formula, dk(x, y) is background difference image dkXth row, y row pixels, x, y are that image is horizontal, ordinate, x value models
Enclose for 1~h2, y 1~w2 of span；
Three width mark image F is set1 k,Image size is h2*w2, and element initial value is 0 in image；
WhenWhen, make F1 k(x, y)=1；
WhenWhen, make
WhenWhen, make
Wherein, ω1,ω2,ω3It is detection threshold value multiplying power factor, span is 0~100；
Step 5：Perform abnormal fastener detection
To image F1 k,Image erosion, expansive working are carried out respectively, are eliminated in image after isolated noise, are found figure
As F1 k,Middle value is 1 connected domain；As image F1 kOrIt is middle exist area be more than T3 connected domain when, judge fastener
In the presence of exception, wherein T3 is detection threshold value, and span is 1~1000000000；
Work as imageIt is middle connected region of the area more than T4 to be present and in image F1 kIn be not present area be more than T5 company
Lead to region, in imageIn when area be not present being more than T6 connected region, judge that button is normal；
Work as imageIt is middle connected region of the area more than T5 to be present and in imageIn be not present area be more than T4 company
During logical region, fastener missing is judged；
Work as imageIt is middle connected region of the area more than T5 to be present and in imageThe middle connection that area be present and be more than T4
During region, judge that fastener loosens；
Work as imageIt is middle connected region of the area more than T8 to be present and in imageThe middle connection that area be present and be more than T4
During region, judge that fastener is cracked, wherein T4, T5, T6, T7, T8 are detection threshold values, and span is 1~1000000000.
Although present disclosure is discussed in detail by above preferred embodiment, but it should be appreciated that above-mentioned
Description should not be considered as limitative to the invention.After those skilled in the art have read the above, for the present invention
A variety of modifications and substitutions all will be apparent.Therefore, protection scope of the present invention should be limited by appended claim
It is fixed.
Claims (8)
1. a kind of double light spectrum image-forming rails and fastener method for detecting abnormality, it is characterized in that：Using visible light camera and infrared take the photograph
Camera forms double optical spectrum imaging devices, carries out double light spectrum image-formings to rail and fastener area, obtains rail and fastener texture respectively
Image and thermal-induced imagery, rail and fastener in thermal-induced imagery are carried out using texture image and positioned, is compared using background model
Carry out fastener missing, loosening, fracture abnormality detection, be used in conjunction with texture image and thermal-induced imagery carry out rail stripping, scratch,
Ripple mill, crack detection.
2. double light spectrum image-forming rails according to claim 1 and fastener method for detecting abnormality, it is characterized in that：Visible image capturing
Line array video camera or area array camera can be selected in machine and thermal camera；
Preferably, it is seen that light video camera and thermal camera select line array video camera, wherein, it is seen that the imaging of light array camera
Resolution ratio is not less than 1024 pixels, and spectrum sensing range is 400nm~1000nm, and the imaging resolution of infrared line array video camera is not
Less than 1024 pixels, the temperature control of infrared line array video camera is not less than 0.5 DEG C；
The luminous ray array camera and infrared line array video camera are located at directly over rail, optical axis is cut perpendicular to road surface, imaging
Piece is perpendicular to rail longitudinal direction, shooting visual field covering rail and both sides fastener area；
The line array video camera imaging slice refers to：Imaging region and camera lens photocentre in line scan image sensor in line array video camera
The plane of composition.
3. double light spectrum image-forming rails and fastener method for detecting abnormality according to claim 1 and 2, it is characterized in that：To ensure
It seems what visual field was alignd to shoot texture image and infrared chart, it is seen that light video camera passes through coaxial light splitting optical path with thermal camera
Imaging；The coaxial light splitting optical path refers to：The imaging slice of the optical axis of luminous ray array camera and infrared line array video camera hangs down
Directly, the optical axis of infrared line array video camera is vertical with the imaging slice of luminous ray array camera, it is seen that the light of light array camera
The heart is imaged to the distance of infrared line array video camera imaging slice and the photocentre of infrared line array video camera to luminous ray array camera
The distance of section is equal, and the imaging angle of two video cameras is identical, optical axis intersects；Cut in the imaging of luminous ray array camera
Piece is with the intersection of thermal camera imaging slice, setting a dichroscope；The light splitting surface of dichroscope and visible ray linear array
Video camera imaging cut into slices with the intersection of infrared line array video camera imaging slice it is tangent, with luminous ray array camera and infrared linear array
The optical axis of video camera is into 45 degree of angles, for incident ray to be separated into visible and near-infrared mixed light, mid-infrared and far-infrared light；When visible
When light array camera is located at the transmitted light path of dichroscope, infrared line array video camera is located at the reflected light path of dichroscope, choosing
With thoroughly visible and near-infrared mixed light, the dichroscope of reflection mid-infrared and far-infrared light；When luminous ray array camera is located at dichroic
When the reflected light path of mirror, infrared line array video camera are located at the transmitted light path of dichroscope, mixed from reflection is visible with near-infrared
Light, the dichroscope for transmiting mid-infrared and far-infrared light.
4. double light spectrum image-forming rails and fastener method for detecting abnormality according to claims 1 to 3, it is characterized in that：In rail
One line source is respectively set directly over both sides fastener areas, the projection planes of two line sources and luminous ray array camera into
It is coplanar as cutting into slices, provide illumination for luminous ray array camera；An area source, area source irradiation side are arranged above in rail
To with rail longitudinal direction into a angles incident, imaging region of the area source irradiation area covering visible light line array video camera on rail,
Dark-ground illumination is formed with luminous ray array camera, is imaged for rail damage, a angular configurations scope is 10~90 degree；It is described
Two line sources above fastener and the area source irradiation area above rail be not overlapping, to avoid interfering；
Preferably, line source and area source are parallel light source；
Preferably, line source and area source select the Single wavelength or narrow-band spectrum LASER Light Source of phase co-wavelength, optical source wavelength scope
For：400~1000nm, and set in luminous ray array camera front end and filtered with the arrowband of line source and area array light source phase co-wavelength
Mating plate, disturbed for filtering environmental light.
5. double light spectrum image-forming rails according to claim 4 and fastener method for detecting abnormality, it is characterized in that：Line source and face
Light source selects near-infrared light source, and wave-length coverage is：800~950nm, and set and light source ripple in luminous ray array camera front end
Long identical near infrared light narrow band pass filter.
6. double light spectrum image-forming rails according to claim 2 and fastener method for detecting abnormality, it is characterized in that：To realize to row
Rail and fastener scanning imagery during car traveling, fix a wheel encoder to vehicle wheel rotation angle in train wheel rotating shaft
Degree is encoded and produces pulse triggering signal, and pulse triggering signal is used to drive luminous ray array camera and infrared linear array to take the photograph
Camera carries out linear array imaging scanning to rail and fastener area, obtains rail and fastener area texture image and thermal-induced imagery；
In actual use, respectively above train bottom, left and right rail, respectively set a set of by luminous ray array camera and infrared ray
Double optical spectrum imaging devices that array camera is formed, imaging is scanned to left and right rail and its both sides fastener area；
Preferably, it is acquisition and the double light spectrum image-forming images of rail length always when train passes through bend, in left and right sides car
One vehicle wheel rotation encoder of each installation on wheel, the double optical spectrum imaging devices in the left and right sides are driven respectively.
7. double light spectrum image-forming rails and fastener method for detecting abnormality according to claims 1 to 6, it is characterized in that：Carry out iron
Track surface method for detecting abnormality is：
Step 1：According to rail in texture image and thermal-induced imagery image space, be partitioned into rail surface texture image IwWith
Thermal-induced imagery Ig, IgSize be h1*w1, h1, w1 span are：50~1000；
Step 2：Establish rail surface thermal-induced imagery background model
Train running speed is divided into n grade L={ l1,...,ln, n span is 1~500, according to train operation
Speed class n, establish normal rail thermal-induced imagery background model in n group train travelling processesWherein,
Background modelIt is by the normal rail surface thermal-induced imagery of the k frames shot under the i-th grade train running speedTraining obtains,
Background modelBy the matrix of 2 h1*w1 dimensionsComposition, wherein,Middle element is the normal rail surface infrared chart of k frames
As pixel is in the average of time domain,It is variance of the normal rail surface thermal-induced imagery pixel of k frames in time domain；
Step 3：Carry out rail surface temperature anomaly detection
Input current train movement velocity v and rail surface thermal-induced imagery
According to train running speed v calculating speed grades lc, from background modelMiddle selection lcIndividual background mould
Type：Background difference operation is performed, obtains background difference image dg：
<mrow>
<msup>
<mi>d</mi>
<mi>g</mi>
</msup>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<msubsup>
<mi>I</mi>
<mi>c</mi>
<mi>g</mi>
</msubsup>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<msubsup>
<mi>&mu;</mi>
<msub>
<mi>l</mi>
<mi>c</mi>
</msub>
<mi>g</mi>
</msubsup>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
</mrow>
In formula, dg(x, y) is background difference image dgXth row, y row pixels, x, y are that image is horizontal, ordinate, and x spans are
1~h1, y 1~w1 of span；
2 width mark image F is set1 g,Image size is h1*w1, and element initial value is 0 in image；
WhenWhen, make F1(x, y)=1；
WhenWhen, make F2(x, y)=1；
Wherein,It is detection threshold value multiplying power factor, span is 0~100；
To image F1 g,Morphologic filtering is carried out respectively and eliminates isolated noise, and finds image F1 g,Middle value is 1 connection
Domain；As image F1 gOrIt is middle when the connected domain that area is more than T1 be present, it is abnormal to judge that rail surface is present, wherein T1 is detection
Threshold value, span are 1~1000000000；
Step 4：Rail surface abnormality detection is carried out based on rail texture image
Binary conversion treatment is carried out to rail surface texture image：1 is set to above or equal to threshold value T2 pixel, less than threshold value T2
Pixel be set to 0, obtain binary imageTo binary imageCarry out morphologic filtering and eliminate picture noise, and find
ImageMiddle value is 1 connected domain；Work as imageIt is middle exist area be more than T2 connected region when, judge rail surface deposit
In exception, wherein T2 is detection threshold value, and span is 1~1000000000；
Step 5：Rail surface anomaly classification is carried out based on rail texture binary image and thermal-induced imagery
Operation is merged to the connected domain detected in step 3 and step 4, and calculates the maximum boundary rectangle of connected domain after merging
R；Calculate F1 g,The length-width ratio of connected domain, barycenter, major axes orientation, area in middle Zone R domain, constitutive characteristic vector V1；To red
Pixel extraction histogram in outer thermal image Zone R domain, constitutive characteristic vector V2；Respectively to V1、V2After being normalized, and with
Current thermal-induced imagery train running speed v is obtained as rail surface anomaly classification characteristic quantity V'={ v, V1,V2}；Then, adopt
Collect the extremely positive negative sample of rail surface, train sorter model, rail surface abnormal class is judged using sorting technique.
8. double light spectrum image-forming rails and fastener method for detecting abnormality according to claims 1 to 6, it is characterized in that：Detained
The method of part abnormality detection is：
Step 1：According to fastener in texture image and thermal-induced imagery imaging region, be partitioned into rail surface texture image and red
Outer thermal image；
Step 2：Fastener in thermal-induced imagery is carried out using texture image to position：Using image template matching method, in fastener line
Fastener area-of-interest (Region ofinteresting, ROI) is found in reason image；Determined based on the ROI in thermal-induced imagery
Position fastener area, takes out fastener area thermal-induced imagery Ik, thermal-induced imagery IkSize be h2*w2, h2, w2 span
For：50~1000；
Step 3：Establish fastener thermal-induced imagery background model
Train running speed is divided into n grade L={ l1,...,ln, n span is 1~500, according to train operation
Speed class n, establish normal fastener thermal-induced imagery background model in n group train travelling processesWherein,
Background modelIt is by the normal fastener thermal-induced imagery of the k frames shot under the i-th grade train running speedTraining obtains, fastener
Region thermal-induced imagery size is h2*w2, background modelBy the matrix of 2 h2*w2 dimensionsComposition, wherein,Middle member
Element is average of the normal fastener thermal-induced imagery pixel of k frames in time domain,It is the normal fastener thermal-induced imagery pixel of k frames in time domain
Variance；
Step 4：Perform background Differential Detection
Input current train movement velocity v and fastener thermal-induced imagery
According to train running speed v calculating speed grades lc, from background modelMiddle selection lcIndividual background mould
Type：Background difference operation is performed, obtains background difference image dk：
<mrow>
<msup>
<mi>d</mi>
<mi>k</mi>
</msup>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<msubsup>
<mi>I</mi>
<mi>c</mi>
<mi>k</mi>
</msubsup>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<msubsup>
<mi>&mu;</mi>
<msub>
<mi>l</mi>
<mi>c</mi>
</msub>
<mi>k</mi>
</msubsup>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>2</mn>
<mo>)</mo>
</mrow>
</mrow>
In formula, dk(x, y) is background difference image dkXth row, y row pixels, x, y are that image is horizontal, ordinate, and x spans are
1~h2, y 1~w2 of span；
Three width mark image F is set1 k,Image size is h2*w2, and element initial value is 0 in image；
WhenWhen, make F1 k(x, y)=1；
WhenWhen, make
WhenWhen, make
Wherein, ω1,ω2,ω3It is detection threshold value multiplying power factor, span is 0~100；
Step 5：Perform abnormal fastener detection
To image F1 k,Image erosion, expansive working are carried out respectively, are eliminated in image after isolated noise, are found image F1 k,Middle value is 1 connected domain；As image F1 kOrIt is middle exist area be more than T3 connected domain when, judge fastener exist
Abnormal, wherein T3 is detection threshold value, and span is 1~1000000000；
Work as imageIt is middle connected region of the area more than T4 to be present and in image F1 kIn be not present area be more than T5 connected region
Domain, in imageIn when area be not present being more than T6 connected region, judge that button is normal；
As image F1 kIt is middle connected region of the area more than T5 to be present and in imageIn be not present area be more than T4 connected region
When, judge fastener missing；
As image F1 kIt is middle connected region of the area more than T5 to be present and in imageThe middle connected region that area be present and be more than T4
When, judge that fastener loosens；
Work as imageIt is middle connected region of the area more than T8 to be present and in imageThe middle connected region that area be present and be more than T4
When, judge that fastener is cracked, wherein T4, T5, T6, T7, T8 are detection threshold values, and span is 1~1000000000.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710949482.9A CN107576666A (en) | 2017-10-13 | 2017-10-13 | A kind of double light spectrum image-forming rails and fastener method for detecting abnormality |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710949482.9A CN107576666A (en) | 2017-10-13 | 2017-10-13 | A kind of double light spectrum image-forming rails and fastener method for detecting abnormality |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107576666A true CN107576666A (en) | 2018-01-12 |
Family
ID=61037026
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710949482.9A Pending CN107576666A (en) | 2017-10-13 | 2017-10-13 | A kind of double light spectrum image-forming rails and fastener method for detecting abnormality |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107576666A (en) |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108318263A (en) * | 2018-01-30 | 2018-07-24 | 中道恒通（北京）科技有限公司 | A kind of locomotive vibration condition monitoring system |
| CN108344738A (en) * | 2018-01-22 | 2018-07-31 | 翰飞骏德（北京）医疗科技有限公司 | Imaging method and its device for hydroxyapatite |
| CN108596968A (en) * | 2018-04-02 | 2018-09-28 | 成都精工华耀科技有限公司 | A kind of taking count of crosstie method based on track 3D depth images |
| CN108846824A (en) * | 2018-04-02 | 2018-11-20 | 成都精工华耀科技有限公司 | A kind of positioning of linear array scan image sleeper and method of counting based on gradient projection |
| CN109187560A (en) * | 2018-10-12 | 2019-01-11 | 成都精工华耀科技有限公司 | A kind of low-power consumption track two dimension and three-dimensional fusion imaging system |
| CN109283188A (en) * | 2018-10-12 | 2019-01-29 | 成都精工华耀科技有限公司 | A kind of track visualization inspection two dimension and three-dimensional integratedization imaging system |
| CN109360236A (en) * | 2018-10-26 | 2019-02-19 | 郑州轻工业学院 | A kind of floral disc of sunflower dimension measurement method |
| CN109636789A (en) * | 2018-12-13 | 2019-04-16 | 银河水滴科技（北京）有限公司 | The part of rail clip based on image segmentation lacks detection method and device |
| CN109916912A (en) * | 2019-04-01 | 2019-06-21 | 北京通运腾达科技发展有限公司 | A kind of railway rail clip Defect inspection method and system |
| CN110567397A (en) * | 2018-06-05 | 2019-12-13 | 成都精工华耀科技有限公司 | Fastener spring tongue separation detection method |
| CN110567680A (en) * | 2018-06-05 | 2019-12-13 | 成都精工华耀科技有限公司 | Track fastener looseness detection method based on angle comparison |
| CN110634121A (en) * | 2018-06-05 | 2019-12-31 | 成都精工华耀科技有限公司 | Track fastener loosening detection method based on texture and depth images |
| CN110632121A (en) * | 2018-06-05 | 2019-12-31 | 波音公司 | Dynamic high-speed non-destructive evaluation for railcar axles |
| CN110961289A (en) * | 2019-12-09 | 2020-04-07 | 国网智能科技股份有限公司 | Transformer substation insulator anti-pollution flashover coating spraying tool and spraying method |
| CN111062920A (en) * | 2019-12-13 | 2020-04-24 | 北京百度网讯科技有限公司 | Method and device for generating semiconductor detection report |
| CN111256841A (en) * | 2020-03-26 | 2020-06-09 | 深圳市永达电子信息股份有限公司 | Track state detection method and detection system thereof |
| CN115063416A (en) * | 2022-08-08 | 2022-09-16 | 苏州立创致恒电子科技有限公司 | Rail fastener state detection method and system |
| JP7165745B2 (en) | 2018-09-10 | 2022-11-04 | メル・メック・ソチエタ・ペル・アツィオーニ | Equipment for detecting defects in railway equipment |
| CN116109638A (en) * | 2023-04-13 | 2023-05-12 | 中铁四局集团有限公司 | Rail break detection method and system |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6556945B1 (en) * | 1997-05-26 | 2003-04-29 | Stn Atlas Elektronik Gmbh | Measurement of grooves and long waves on rails with a longitudinal streak of light |
| CN101142462A (en) * | 2004-06-30 | 2008-03-12 | 乔治敦铁轨设备公司 | System and method for inspecting railroad track |
| CN102507587A (en) * | 2011-09-20 | 2012-06-20 | 株洲南车时代电气股份有限公司 | Perambulated inspection system and method |
| CN102721700A (en) * | 2012-06-12 | 2012-10-10 | 上海交通大学 | System and method for high-speed detection of loosening of railway fasteners on basis of infrared thermal imaging |
| CN103534581A (en) * | 2011-03-21 | 2014-01-22 | 费德罗-莫格尔公司 | Multi-spectral imaging system and method of surface inspection therewith |
| CN205262989U (en) * | 2015-12-16 | 2016-05-25 | 深圳大学 | Rail connector defect detecting device |
-
2017
- 2017-10-13 CN CN201710949482.9A patent/CN107576666A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6556945B1 (en) * | 1997-05-26 | 2003-04-29 | Stn Atlas Elektronik Gmbh | Measurement of grooves and long waves on rails with a longitudinal streak of light |
| CN101142462A (en) * | 2004-06-30 | 2008-03-12 | 乔治敦铁轨设备公司 | System and method for inspecting railroad track |
| CN103534581A (en) * | 2011-03-21 | 2014-01-22 | 费德罗-莫格尔公司 | Multi-spectral imaging system and method of surface inspection therewith |
| CN102507587A (en) * | 2011-09-20 | 2012-06-20 | 株洲南车时代电气股份有限公司 | Perambulated inspection system and method |
| CN102721700A (en) * | 2012-06-12 | 2012-10-10 | 上海交通大学 | System and method for high-speed detection of loosening of railway fasteners on basis of infrared thermal imaging |
| CN205262989U (en) * | 2015-12-16 | 2016-05-25 | 深圳大学 | Rail connector defect detecting device |
Cited By (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108344738A (en) * | 2018-01-22 | 2018-07-31 | 翰飞骏德（北京）医疗科技有限公司 | Imaging method and its device for hydroxyapatite |
| CN108318263A (en) * | 2018-01-30 | 2018-07-24 | 中道恒通（北京）科技有限公司 | A kind of locomotive vibration condition monitoring system |
| CN108596968A (en) * | 2018-04-02 | 2018-09-28 | 成都精工华耀科技有限公司 | A kind of taking count of crosstie method based on track 3D depth images |
| CN108846824A (en) * | 2018-04-02 | 2018-11-20 | 成都精工华耀科技有限公司 | A kind of positioning of linear array scan image sleeper and method of counting based on gradient projection |
| CN108846824B (en) * | 2018-04-02 | 2022-05-17 | 成都精工华耀科技有限公司 | Linear array scanning image sleeper positioning and counting method based on gradient projection |
| CN108596968B (en) * | 2018-04-02 | 2022-04-19 | 成都精工华耀科技有限公司 | Sleeper counting method based on track 3D depth image |
| CN110632121A (en) * | 2018-06-05 | 2019-12-31 | 波音公司 | Dynamic high-speed non-destructive evaluation for railcar axles |
| CN110567680B (en) * | 2018-06-05 | 2022-03-29 | 成都精工华耀科技有限公司 | Track fastener looseness detection method based on angle comparison |
| CN110567397A (en) * | 2018-06-05 | 2019-12-13 | 成都精工华耀科技有限公司 | Fastener spring tongue separation detection method |
| CN110567680A (en) * | 2018-06-05 | 2019-12-13 | 成都精工华耀科技有限公司 | Track fastener looseness detection method based on angle comparison |
| CN110634121A (en) * | 2018-06-05 | 2019-12-31 | 成都精工华耀科技有限公司 | Track fastener loosening detection method based on texture and depth images |
| JP7165745B2 (en) | 2018-09-10 | 2022-11-04 | メル・メック・ソチエタ・ペル・アツィオーニ | Equipment for detecting defects in railway equipment |
| CN109187560A (en) * | 2018-10-12 | 2019-01-11 | 成都精工华耀科技有限公司 | A kind of low-power consumption track two dimension and three-dimensional fusion imaging system |
| CN109283188A (en) * | 2018-10-12 | 2019-01-29 | 成都精工华耀科技有限公司 | A kind of track visualization inspection two dimension and three-dimensional integratedization imaging system |
| CN109360236B (en) * | 2018-10-26 | 2022-05-17 | 郑州轻工业学院 | Sunflower disc size measurement method |
| CN109360236A (en) * | 2018-10-26 | 2019-02-19 | 郑州轻工业学院 | A kind of floral disc of sunflower dimension measurement method |
| CN109636789A (en) * | 2018-12-13 | 2019-04-16 | 银河水滴科技（北京）有限公司 | The part of rail clip based on image segmentation lacks detection method and device |
| CN109916912A (en) * | 2019-04-01 | 2019-06-21 | 北京通运腾达科技发展有限公司 | A kind of railway rail clip Defect inspection method and system |
| CN110961289B (en) * | 2019-12-09 | 2021-06-29 | 国网智能科技股份有限公司 | Transformer substation insulator anti-pollution flashover coating spraying tool and spraying method |
| CN110961289A (en) * | 2019-12-09 | 2020-04-07 | 国网智能科技股份有限公司 | Transformer substation insulator anti-pollution flashover coating spraying tool and spraying method |
| CN111062920A (en) * | 2019-12-13 | 2020-04-24 | 北京百度网讯科技有限公司 | Method and device for generating semiconductor detection report |
| CN111062920B (en) * | 2019-12-13 | 2023-06-20 | 北京百度网讯科技有限公司 | Method and device for generating semiconductor detection report |
| CN111256841A (en) * | 2020-03-26 | 2020-06-09 | 深圳市永达电子信息股份有限公司 | Track state detection method and detection system thereof |
| CN115063416A (en) * | 2022-08-08 | 2022-09-16 | 苏州立创致恒电子科技有限公司 | Rail fastener state detection method and system |
| CN115063416B (en) * | 2022-08-08 | 2022-11-04 | 苏州立创致恒电子科技有限公司 | Rail fastener state detection method and system |
| CN116109638A (en) * | 2023-04-13 | 2023-05-12 | 中铁四局集团有限公司 | Rail break detection method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107576666A (en) | A kind of double light spectrum image-forming rails and fastener method for detecting abnormality | |
| CN207908384U (en) | A kind of double light spectrum image-forming rails and fastener abnormal detector | |
| CN101639452B (en) | 3D detection method for rail surface defects | |
| CN106053479B (en) | A kind of vision detection system of the workpiece appearance defects based on image procossing | |
| CN104608799B (en) | Based on information fusion technology Railway wheelset tread damage on-line checking and recognition methodss | |
| US8050486B2 (en) | System and method for identifying a feature of a workpiece | |
| CN104732235B (en) | A kind of vehicle checking method for eliminating the reflective interference of road at night time | |
| CN102495069B (en) | Method for detecting defects of chain belts of zipper on basis of digital image processing | |
| CN102221553B (en) | Structured light-based high-speed detection system and method for railway fastener | |
| CN102175692A (en) | System and method for detecting defects of fabric gray cloth quickly | |
| WO2015055060A1 (en) | Online detecting method for continuous casting slab surface quality | |
| CN105158257A (en) | Sliding plate measurement method and device | |
| CN103837087B (en) | Pantograph automatic testing method based on active shape model | |
| CN103303337B (en) | Based on rail fastener adaptive location method and the high-speed inspection system of image sequence | |
| CN1843822A (en) | Dynamic detection system of train wheel pair and detection method thereof | |
| CN107576667A (en) | A kind of railway rail clip abnormality detection system based on linear array thermal camera | |
| CN104729426A (en) | Automatic angle iron online detecting system and method based on machine vision | |
| CN108759670A (en) | A kind of contact line abrasion device for dynamically detecting based on non-contact detection technology | |
| CN101699273A (en) | Auxiliary detection device and method of image processing for on-line flaw detection of rails | |
| CN104483320A (en) | Digitized defect detection device and detection method of industrial denitration catalyst | |
| CN102288614A (en) | Method for detecting pantograph crack fault based on curvelet domain moving parallel window | |
| CN103010258A (en) | System and method for detecting cracks of fasteners of high-speed rails and subways | |
| CN202548069U (en) | On-line detection device for surface defects of high temperature continuous casting billet | |
| US7116814B2 (en) | Image-based container defects detector | |
| CN113673614B (en) | Metro tunnel foreign matter intrusion detection device and method based on machine vision |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 610021 Longquanyi Jingkai District, Chengdu City, Sichuan Province, 309 Dingfeng Power Port, 12 buildings, 7 units 501 Applicant after: CHENGDU JINGGONG HUAYAO TECHNOLOGY Co.,Ltd. Address before: 610021 Longquanyi Jingkai District, Chengdu City, Sichuan Province, 309 Dingfeng Power Port, 12 buildings, 7 units 501 Applicant before: CHENGDU JINGGONG HUAYAO MACHINERY MANUFACTURING CO.,LTD. |
|
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180112 |