CN106053593A - Flaw detection device and flaw part detecting method by flaw detection device - Google Patents
Flaw detection device and flaw part detecting method by flaw detection device Download PDFInfo
- Publication number
- CN106053593A CN106053593A CN201610202357.7A CN201610202357A CN106053593A CN 106053593 A CN106053593 A CN 106053593A CN 201610202357 A CN201610202357 A CN 201610202357A CN 106053593 A CN106053593 A CN 106053593A
- Authority
- CN
- China
- Prior art keywords
- candidate portion
- detection
- hinder
- threshold value
- traumatic part
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N27/00—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means
- G01N27/72—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables
- G01N27/82—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables for investigating the presence of flaws
- G01N27/83—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables for investigating the presence of flaws by investigating stray magnetic fields
- G01N27/84—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables for investigating the presence of flaws by investigating stray magnetic fields by applying magnetic powder or magnetic ink
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N27/00—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means
- G01N27/72—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables
- G01N27/82—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables for investigating the presence of flaws
- G01N27/83—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables for investigating the presence of flaws by investigating stray magnetic fields
- G01N27/87—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables for investigating the presence of flaws by investigating stray magnetic fields using probes
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N27/00—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means
- G01N27/72—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables
- G01N27/74—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables of fluids
- G01N27/76—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables of fluids by investigating susceptibility
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/1717—Systems in which incident light is modified in accordance with the properties of the material investigated with a modulation of one or more physical properties of the sample during the optical investigation, e.g. electro-reflectance
- G01N2021/1727—Magnetomodulation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N2021/1765—Method using an image detector and processing of image signal
- G01N2021/177—Detector of the video camera type
- G01N2021/1776—Colour camera
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8887—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques
Abstract
The invention provides a flaw detection apparatus that prevents detection of a wound portion and an over-detection, thereby improving a detection rate of a wound portion, and a wound detection method using the flaw detection apparatus. A magnetic particle inspection apparatus (1) includes: an image pickup device (16) that captures a surface of an inspection object (10); and a detection device (30) that processes an original image imaged by the imaging device (16) (30) comprises: a first extraction unit (31) for binarizing an original image using a first threshold value to extract a first candidate unit (40); and an inspection region generation unit A second extraction unit (33) for binarizing the inspection region (41) by the second threshold value, and extracting the second detection unit (40) by the second extraction unit (33) 2 injury candidates (42), and an injury determination unit (34) for expanding the second injury candidate unit (42) and detecting the injury unit.
Description
Technical field
The present invention relates to possess camera head that the surface to checked property images and to by
The original image of camera head shooting is processed to the detection device in the traumatic part of detection surface
Failure detector and utilize the traumatic part detection method of failure detector.
Background technology
In the past, by a kind of magnetic-particle test as non-destructive detection method, it is impregnated with spy
Hinder the defect detecting test that test etc. carries out the surface of checked property.In magnetic-particle test,
The surface coating magnetic powder of checked property or the magnetic powder solution containing magnetic powder, and to checked property
Apply magnetic field etc. and make checked property magnetize.Wound due to the surface of flux concentrating to checked property
Portion, so magnetic powder is attracted by this magnetic flux and forms the instruction pattern caused by magnetic powder.Then, pass through
Observe this magnetic powder instruction pattern to check defect.
On the other hand, in being impregnated with crank detection test, first, liquid will be impregnated with and be coated to checked property
Surface, make this be impregnated with liquid and be impregnated at small fault locations such as the crackle of surface opening, pin holes.
It follows that the residue that removal is attached to surface is impregnated with liquid, by developing agent powder coating to surface,
The liquid that is impregnated with being impregnated into fault location is sucked into surface by capillarity.Then, by seeing
Examining this being impregnated with caused by liquid that be impregnated with sucted indicates pattern to check defect.
These magnetic-particle tests, it is impregnated with crank detection test and is carried out by failure detector, institute
State failure detector by automaton, possess the shooting dress that the surface to checked property images
Put and be processed to the original image imaged by camera head detect the inspection in the traumatic part of surface
Survey device.
In patent documentation 1, disclose a kind of magnetic powder inspection device, by image unit to
The magnetic powder pattern that the surface of the examined material of conveyance produces images, from the image inspection of shooting
Look into the surface defect of examined material, it is characterised in that average with bright by the brightness in image
The constant times sum of the standard deviation of degree is set to the threshold value of 2 values, uses this threshold value to enter image
Row 2 value, detects surface defect.
Patent documentation 1: Japanese Unexamined Patent Publication 2011-13007 publication
Summary of the invention
According to the structure of patent documentation 1, even if due to the rough surface of examined material, light source
The flaw detection condition such as brightness disproportionation and the deviation situation of the brightness of defectogram picture there occurs change, also
The threshold value of suitable 2 values can be determined all the time, it is possible to reduce error detection and the detection of exhaustive ground
The defect on surface.
Herein, in magnetic-particle test, the wide wound of the deep wound of the degree of depth, width is imaged bright
Bright.On the other hand, the wound of depth as shallow, the wound of narrow width are imaged the brightest.It is former
Because being, owing to the leakage magnetic flux of the wide wound of the deep wound of the degree of depth, width becomes big, so attracting more
Magnetic powder.Similarly, even if in being impregnated with crank detection test, the wound that the deep wound of the degree of depth, width are wide
Also imaged bright.On the other hand, the most why the wound of depth as shallow, the wound of narrow width are imaged
Bright.This is because the substantial amounts of liquid that is impregnated with is impregnated into the wound that the deep wound of the degree of depth, width are wide.
The size of wound, the degree of depth as the defect being formed at surface are diversified.It addition,
1 continuous print wound, the degree of depth of wound the most linearly extended, width are not constant.That is,
In magnetic-particle test, it is impregnated with in crank detection test, the part of depth as shallow and deep part, width
Wide part is mixed such 1 continuous print wound and is become every for it by shooting with narrow part
Individual part and the different image of lightness.
Therefore, in the case of determining whether wound according to the brightness of image, there are following feelings
Condition: even if 1 continuous print wound, also become desultory wound, be not decided as lacking
The wound fallen into, and over sight (OS) will be hindered.On the other hand, if making it easy to be judged as wound, i.e. lower
As the brightness of benchmark, the most unnecessary small wound, burr, dust etc. are judged as
As the wound of defect, and becoming detection, this crosses detection is the unnecessary wound of detection, dust etc.
State.And, in patent documentation 1, although consider and prevent caused by flaw detection condition
The deviation situation of brightness of image and the error detection that causes, but about due to the degree of depth of wound, width
The error detection caused by deviation of the brightness spent and cause, entirely without consideration.
Therefore, it is an object of the invention to provide a kind of detection preventing traumatic part omit and cross inspection
Survey and improve the failure detector of the verification and measurement ratio in traumatic part and utilize the traumatic part of failure detector to detect
Method.
In order to solve above-mentioned problem, the present invention relates to failure detector, possess: camera head, right
The surface of checked property images；And detection device, to imaged by described camera head
Original image processes, and detects the traumatic part of described surface, it is characterised in that described detection fills
Put and possess: the 1st extracting part, utilize the 1st threshold value that described original image is carried out binary conversion treatment,
Extract the 1st and hinder candidate portion；Inspection area generating unit, to include the described 1st side hindering candidate portion
Formula generates inspection area；2nd extracting part, utilizes the 2nd threshold value that described inspection area is carried out two
Value processes, and candidate portion is hindered in extraction the 2nd；And hinder detection unit, hinder candidate portion to the described 2nd
Carry out expansion process, detect described traumatic part.
And then, the failure detector of the present invention is characterised by, described 2nd threshold value is less than described the
1 threshold value.
And then, the failure detector of the present invention is characterised by, described inspection area is to include by making
The rectangle that the described 1st minimum rectangle hindering candidate portion expands the width of regulation and obtains surrounds
Region.
And then, the failure detector of the present invention is characterised by, described expansion process is to make described
2 hinder the process that candidate portion expands to respective long axis direction.
And then, the present invention relates to the use of the traumatic part detection method of failure detector, examine in described traumatic part
In survey method, the surface of checked property is imaged by camera head, and detection device is to by described
The original image of camera head shooting processes, and detects the traumatic part of described surface, described traumatic part
Detection method is characterised by, described detection device utilizes the 1st threshold value to carry out described original image
Binary conversion treatment, candidate portion is hindered in extraction the 1st, raw to include in the way of the described 1st hinders candidate portion
Become inspection area, utilize the 2nd threshold value that described inspection area carries out binary conversion treatment, extraction the
2 hinder candidate portion, hinder candidate portion to the described 2nd and carry out expansion process, detect described traumatic part.
And then, the traumatic part detection method utilizing failure detector of the present invention is characterised by, described
2nd threshold value is less than described 1st threshold value.
And then, the traumatic part detection method utilizing failure detector of the present invention is characterised by, described
Inspection area is by making to include that the described 1st minimum rectangle hindering candidate portion expands the width of regulation
The region that the rectangle spent and obtain surrounds.
And then, the traumatic part detection method utilizing failure detector of the present invention is characterised by, described
Expansion process is to make the described 2nd to hinder the process that candidate portion expands to respective long axis direction.
According to the present invention, failure detector possesses: camera head, carries out the surface of checked property
Shooting；And detection device, described camera head the original image imaged is processed, inspection
Surveying the traumatic part of described surface, wherein, described detection device possesses: the 1st extracting part, utilizes
1st threshold value carries out binary conversion treatment to described original image, and candidate portion is hindered in extraction the 1st；Test zone
Territory generating unit, generates inspection area in the way of the described 1st hinders candidate portion including；2nd extraction
Portion, utilizes the 2nd threshold value that described inspection area carries out binary conversion treatment, and candidate is hindered in extraction the 2nd
Portion；And hinder detection unit, hinder candidate portion to the described 2nd and carry out expansion process, detect described wound
Portion, it is possible to provide the detection preventing traumatic part to omit and cross detection and improve the inspection in traumatic part
The failure detector of survey rate.
And then, it is less than the structure of described 1st threshold value according to described 2nd threshold value, using the teaching of the invention it is possible to provide anti-
Only the detection in traumatic part is omitted and is crossed detection and improve the failure detector of the verification and measurement ratio in traumatic part.
And then, it is by making to include that the described 1st hinders the minimum of candidate portion according to described inspection area
The structure in the region that the rectangle that rectangle expands the width of regulation and obtains surrounds, it is possible to prevent inspection
Survey the increase of the operand of device.
And then, it is to make the described 2nd to hinder candidate portion to respective major axis side according to described expansion process
Structure to the process expanded, it is possible to prevent the detection in traumatic part from omitting with higher accuracy.
According to the present invention, in the traumatic part detection method utilizing failure detector, camera head is to quilt
Check that the surface of thing images, described camera head the original image imaged is entered by detection device
Row processes, and detects the traumatic part of described surface, and wherein, described detection device utilizes the 1st threshold value
Described original image carries out binary conversion treatment, and extraction the 1st hinders candidate portion, to include the described 1st
The mode hindering candidate portion generates inspection area, utilizes the 2nd threshold value that described inspection area is carried out two
Value processes, and candidate portion is hindered in extraction the 2nd, hinders candidate portion to the described 2nd and carries out expansion process,
Detect described traumatic part, it is possible to provide the detection preventing traumatic part to omit and cross detection and improve
The traumatic part detection method that utilizes failure detector of the verification and measurement ratio in traumatic part.
And then, the method being less than described 1st threshold value according to described 2nd threshold value, using the teaching of the invention it is possible to provide anti-
Only the detection in traumatic part is omitted and is crossed detection and improve the utilization flaw detection dress of the verification and measurement ratio in traumatic part
The traumatic part detection method put.
And then, it is by making to include that the described 1st hinders the minimum of candidate portion according to described inspection area
The method in the region that the rectangle that rectangle expands the width of regulation and obtains surrounds, it is possible to prevent inspection
Survey the increase of the operand of device.
And then, it is to make the described 2nd to hinder candidate portion to respective major axis side according to described expansion process
Method to the process expanded, it is possible to prevent the detection in traumatic part from omitting with higher accuracy.
Accompanying drawing explanation
Fig. 1 is the magnetic powder inspection of an example of the failure detector being shown as present embodiment
The schematic diagram of the structure of device.
Fig. 2 is the block diagram of the control system of magnetic powder inspection device.
Fig. 3 is the flow chart for illustrating to detect an example of the detection action of device.
Fig. 4 is the example illustrating the image having been carried out binary conversion treatment by the 1st extracting part
Skeleton diagram.
Fig. 5 is the skeleton diagram of the image illustrating the extract the 1st example hindering candidate portion.
Fig. 6 is the skeleton diagram of the image of the example illustrating generated inspection area.
Fig. 7 is the example illustrating the image having been carried out binary conversion treatment by the 2nd extracting part
Skeleton diagram.
Fig. 8 is the skeleton diagram of the image illustrating the extract the 2nd example hindering candidate portion.
Fig. 9 is to illustrate to hinder the image that candidate portion has carried out an example of expansion process to the 2nd
Skeleton diagram.
Symbol description
1: magnetic powder inspection device (failure detector)；10: checked property；16: camera head；
30: detection device；31: the 1 extracting part；32: inspection area generating unit；33: the 2 take out
Take portion；34: hinder detection unit；40: the 1 hinder candidate portion；41: inspection area；42: the 2
Hinder candidate portion.
Detailed description of the invention
Hereinafter, referring to the drawings, narration in detail is for implementing the optimal mode of the present invention.Fig. 1
It it is the knot of the magnetic powder inspection device 1 of an example of the failure detector being shown as present embodiment
The schematic diagram of structure.It addition, the arrow in Fig. 1 represents the conveyance direction of checked property 10, from
Right side in Fig. 1 transports checked property 10 towards left side.
As it is shown in figure 1, magnetic powder inspection device 1 is to use magnetic powder to detect example by automatically controlling
The traumatic part as defect of the surface of the checked properties 10 such as the steel of prism-shaped such as long chi
Failure detector.It addition, magnetic powder inspection device 1 possesses carrying device 11, magnetic powder dissemination apparatus
12, magnetizing assembly 13, air-blast device 14, ultraviolet crack inspection lamp 15, camera head 16 and
Labelling (marking) device 17 etc..It addition, magnetic powder inspection device 1 is also equipped with not scheming herein
The controller C that shows, detection device 30 etc..
Carrying device 11 transports checked property 10.Carrying device 11 is to be made up of multiple rollers etc.
Roller conveyer, be configured to desired speed transport checked property 10.It addition, at figure
In 1, transport from the right side residing for magnetic powder dissemination apparatus 12 to the left side residing for labelling apparatus 17
Checked property 10.It addition, carrying device 11 possesses the most not shown conveying distance measures dress
Put 18.Conveying distance measurement apparatus 18 measures the conveying distance of checked property 10.As conveyance
Distance-measuring device 18, it is possible to use is by the displacement of the rotation of the roller measuring carrying device 11
The determinator that rotary encoder etc. are constituted.It addition, the structure of conveying distance measurement apparatus 18
It is not particularly limited, it is possible to use utilize the determinator of cordless, such as laser surface
Velometer, it is also possible to be the structure being combined with them.It addition, the knot of carrying device 11
Structure is also not particularly limited, such as, can also be the belt conveyor being made up of the band etc. not having joint
Deng.
Magnetic powder dissemination apparatus 12 is the dress on the surface that magnetic particle inspection liquid spreads to checked property 10
Put, the upstream side being configured in conveyance direction.Magnetic powder dissemination apparatus 12 by not shown tank,
Pump, nozzle etc. are constituted.The magnetic particle inspection liquid being contained in tank is sent to by pump nozzle,
Magnetic particle inspection liquid is sprayed from nozzle.Magnetic particle inspection liquid is the solution containing magnetic powder, the table of this magnetic powder
Face is coated with by fluorophor.Magnetic powder dissemination apparatus 12 is configured to examine the magnetic powder of desired amount
Look into liquid and spread to the surface of checked property 10 continuously.It addition, the knot of magnetic powder dissemination apparatus 12
Structure is not particularly limited.
Magnetizing assembly 13 is the device that checked property 10 applies magnetic field, is configured at magnetic adjacently
The downstream of powder dissemination apparatus 12.Magnetizing assembly 13 includes being opposed to be configured at conveyance direction
Upstream side and two through coils 19,20 in downstream and be in two through coils 19,
Two interpolar coils 21,22 etc. of arow configuration between 20 and in conveyance direction.Through
Coil 19,20 is formed as circular, and checked property 10 is with through above-mentioned through coil 19,20
The mode at center transport.On the other hand, interpolar coil 21,22 is formed as U-shaped, quilt
Check that thing 10 transports in the way of by the space of above-mentioned interpolar coil 21,22.It addition, magnetic
Gasifying device 13 possesses this two interpolar coils 21,22 such that it is able to two through coils 19,
Uniform rotating excitation field is produced between 20.
More specifically, by flowing through in through coil 19,20 and interpolar coil 21,22
Electric current, thus in through coil 19,20, the conveyance direction of checked property 10 generates
Magnetic field, in interpolar coil 21,22, with the checked property 10 as its direction, space
Magnetic field is generated on the direction that conveyance direction is orthogonal.And, when in through coil 19,20 and pole
Between time coil 21,22 flows through the alternating current of phase shifting 90 degree, by by through
The direction (conveyance direction) in the magnetic field that coil 19,20 generates and raw by interpolar coil 21,22
The direction (direction orthogonal with conveyance direction) in magnetic field that becomes and in the plane that formed, generate with
The rotating excitation field that certain magnetic field intensity rotates.
It addition, the structure of magnetizing assembly 13 is not particularly limited, it is possible to be suitably designed through
Coil 19,20, the quantity etc. of interpolar coil 21,22.Such as, magnetizing assembly 13 can also
It is in addition to that there is the structure also beyond interpolar coil 21,22 with multiple interpolar coil.It addition,
Magnetizing assembly 13 can also be made up of 1 through coil 19 and 1 interpolar coil 21.
Air-blast device 14 towards checked property 10 to the direction blows air contrary with gravity.Drum
Wind apparatus 14 is between two through coils 19,20, is respectively arranged at through coil 19
And between interpolar coil 21, between two interpolar coils 21,22 and interpolar coil 22 with
Between through coil 20.Can be regulated at checked property 10 by this air-blast device 14
The flow velocity of the magnetic particle inspection liquid that surface is flow through.It addition, the structure of air-blast device 14 is the most especially
Limit.
Ultraviolet crack inspection lamp 15 between two through coils 19,20 to checked property 10
The magnetic particle inspection liquid irradiation ultraviolet radiation on surface.In order to avoid by magnetizing assembly 13 generate strong
The impact of rotating excitation field, is arranged in ultraviolet crack inspection lamp 15 and leaves regulation from checked property 10
The position of distance, such as 600mm～about 2000mm.It addition, ultraviolet crack inspection lamp 15
Structure be not particularly limited, can also be such as the structure having implemented magnetic shield.
The camera head 16 checked property 10 to having been irradiated ultraviolet by ultraviolet crack inspection lamp 15
Surface image.Camera head 16 is configured to the surface for checked property 10
Image from vertical direction.It addition, in order to avoid the strong rotation generated by magnetizing assembly 13
Turn the impact in magnetic field, camera head 16 is arranged in from checked property 10 leave regulation distance,
The position of such as about 600mm～2000mm, and implemented magnetic shield.It addition, shooting
As long as device 16 can be to the checked property 10 having been irradiated ultraviolet by ultraviolet crack inspection lamp 15
Surface carry out imaging, its shooting direction do not limit.But, with high accuracy inspection
Survey traumatic part viewpoint in, camera head 16 be preferably configured as the surface for checked property 10 from
Vertical direction images.As camera head 16, it is possible to use region video camera or line to take the photograph
Camera, and be not particularly limited.For instance, it is possible to use CCD (Charge Coupled
Device, charge-coupled image sensor) video camera.
The surface of the labelling apparatus 17 checked property 10 to being detected by aftermentioned detection device 30
The traumatic part as defect at place carries out labelling that can be visual.As labelling apparatus 17, it is possible to make
With such as labelling rifle, this labelling rifle is by using air pressure to spray ink, thus is marked.
It addition, the structure of labelling apparatus 17 is not particularly limited.
Detection device 30 is by reading in the picture signal (original image) imaged by camera head 16
And the process that original image specified, thus detect the traumatic part on the surface of checked property 10,
About its structure and detection method, aftermentioned.
It follows that the control system of the magnetic powder inspection device 1 of explanation present embodiment.Fig. 2
It it is the block diagram of the control system of magnetic powder inspection device 1.Magnetic powder inspection device 1 possesses controller C,
It is configured to control carrying device 11, conveying distance measurement apparatus by this controller C
18, magnetic powder dissemination apparatus 12, magnetizing assembly 13, air-blast device 14, ultraviolet crack inspection lamp 15,
Camera head 16, labelling apparatus 17 etc., detect the table of checked property 10 by automatically controlling
The traumatic part in face.
Controller C is configured to by reading in various setting values, utilizing various sensor to obtain
The input signals such as detected value also export control signal, thus control magnetic powder inspection device 1 and possessed
The action of various devices.Controller C is by the process dress carrying out calculation process and control process
The main storage means etc. putting, preserving data is constituted.Controller C is such as to possess as processing dress
The CPU (Central Processing Unit, CPU) that puts, as primary storage
The ROM (Read Only Memory, read only memory) of device, RAM (Random
Access Memory, random access memory), intervalometer, input circuit, output circuit,
The microcomputer of power circuit etc..In main storage means, preserve for performing this enforcement
The control program of the action of mode, various data etc..It addition, these various programs, data etc.
Can also be following manner: be stored in the storage device being disposed separately with controller C, by controlling
Device C processed reads.
Controller C has been electrically connected carrying device 11, conveying distance measurement apparatus 18, magnetic powder
Dissemination apparatus 12, magnetizing assembly 13, air-blast device 14, ultraviolet crack inspection lamp 15, shooting dress
Put 16, labelling apparatus 17 and detection device 30.It addition, controller C has been electrically connected figure
Various sensors etc. beyond structure shown in 2.
Detection device 30 is in the same manner as controller C, by carrying out calculation process and control process
Processing means, the main storage means etc. that preserves data is constituted, and is such as to possess CPU, main memory
The microcomputer of storage device, intervalometer, input circuit, output circuit, power circuit etc..
Detection device 30 has the 1st extracting part 31.1st extracting part 31 is by such as program structure
Become.Aftermentioned about detailed content, the 1st extracting part 31 is configured to utilize the 1st predetermined
Threshold value carries out binary conversion treatment to the picture signal (original image) imaged by camera head 16,
Extract the 1st and hinder candidate portion.It addition, by the original image that imaged by camera head 16 via control
Device C is input to detect device 30.
Detection device 30 also has inspection area generating unit 32.Inspection area generating unit 32 is also
It is made up of such as program.Aftermentioned about detailed content, inspection area generating unit 32 be configured to
Inspection area is generated including the 1st mode hindering candidate portion extracted by the 1st extracting part 31.
Detection device 30 also has the 2nd extracting part 33.2nd extracting part 33 is also by such as journey
Sequence is constituted.Aftermentioned about detailed content, the 2nd extracting part 33 is configured to utilize and predetermines
2nd threshold value inspection area to being generated by inspection area generating unit 32 carries out binary conversion treatment,
Extract the 2nd and hinder candidate portion.
Detection device 30 also has hinders detection unit 34.Hinder detection unit 34 also by such as program structure
Become.Aftermentioned about detailed content, hinder detection unit 34 and be configured to extract by the 2nd extracting part 33
The 2nd gone out is hindered candidate portion and is carried out expansion process, detects traumatic part.It is then detected that device 30 structure
Become and will be delivered to controller C by the position data hindering the traumatic part that detection unit 34 detects.
It follows that the action of explanation magnetic powder inspection device 1.Magnetic powder inspection device 1 is configured to logical
Cross carrying device 11 and checked property 10 is transported to successively each device, detect checked property 10
The traumatic part on surface.First checked property 10 is transported to magnetic powder dissemination apparatus 12, at surface quilt
Spread magnetic particle inspection liquid.On surface, the checked property 10 of distributed magnetic particle inspection liquid is transported to
By in the rotating excitation field region that magnetizing assembly 13 is formed.
Now, in the case of the surface of checked property 10 exists traumatic part, produce due to this wound
Portion and the stray field that causes, the magnetic powder that magnetic particle inspection liquid is comprised is attracted by this stray field.Herein,
The magnetic field rotating formed by magnetizing assembly 13, so the direction (shape no matter traumatic part extends
Shape) how, all producing stray field, magnetic powder attracted to traumatic part.Then, magnetic powder set is to wound
Portion, thus the magnetic powder instruction pattern caused due to traumatic part is formed on the surface of checked property 10.
It addition, suitably regulate flow through on the surface of checked property 10 by air-blast device 14
The flow velocity of magnetic particle inspection liquid.In the viewpoint making magnetic powder instruction pattern be stably formed, magnetic powder
Check that the flow velocity of liquid is preferably 5～100mm/s.If the velocity ratio 100mm/s of magnetic particle inspection liquid
Hurry up, then the stray field that magnetic powder becomes to be difficult to be caused due to traumatic part attracts, magnetic powder instruction pattern
Become unstable.On the other hand, if the velocity ratio 5mm/s of magnetic particle inspection liquid is slow, then stable
Magnetic powder instruction pattern formation need the substantial amounts of time, thus relate to check efficiency decline.
About the magnetic powder instruction pattern so formed, by being irradiated purple by ultraviolet crack inspection lamp 15
Outside line, so that the phosphor excitation on the surface of cladding magnetic powder, is carried out by camera head 16
Shooting.Deliver to the original image imaged by this camera head 16 detect device 30.Detection device
The process that 30 pairs of these original images specify, the traumatic part in detection original image, by its position data
Deliver to controller C.Then, controller C surveys according to position data and the conveying distance in this traumatic part
The measurement data of amount device 18, controls the action of labelling apparatus 17, to checked property 10
The traumatic part on surface is marked.
It follows that explanation utilizes the detection method in the traumatic part of detection device 30.Fig. 3 be for
Illustrate to detect the flow chart of an example of the detection action of device 30.Detection device 30 via
Controller C is taken into the picture signal (original image) (step S1) imaged by camera head 16.
Herein, the original image of shooting is made up of the pixel such as indulging 960 × horizontal 1280.It addition, will be by
The moment of camera head 16 shooting associates with original image.
It follows that detection device 30 is by the 1st extracting part 31, for the original image being taken into,
As pre-treatment, carry out employing various wave filter emphasize process (step S2).Herein,
Process as long as emphasize to process as the traumatic part emphasizing in original image, e.g. LUT
(Look Up Table, look-up table) conversion process, expansion/shrink process, shade (shading)
Process, it is also possible to take the process these various process combined.
Then, the 1st extracting part 31 utilizes the 1st threshold value to enter the image having carried out emphasizing to process
Row binary conversion treatment (step S3).Fig. 4 represents that being shown through the 1st extracting part 31 is carried out
The skeleton diagram of one example of the image of binary conversion treatment.It addition, the 1st threshold value is to preset
Value, be stored in detection device 30 not shown main storage means.
It addition, the 1st extracting part 31 indicates in the image carrying out binary conversion treatment
(labeling) (step S4) is processed.Herein, sign processes and refers to, at binary image
In give identical label for continuous print pixel and carry out the process of region division.
Then, the 1st extracting part 31, for having been assigned each region of label, calculates such as
Area, length, longitudinal width, horizontal width etc., be used as decision content, these is counted
The decision content calculated compares with the 1st determinating reference value pre-saved, it is determined whether be
1 hinders candidate portion, and extraction hinders corresponding region, candidate portion (step S5) with the 1st.
Herein, the 1st hinders candidate portion and represents the region being assumed to be traumatic part.It addition, the 1st extracting part
31 for having been assigned all regions of label, the calculating carrying out decision content, the judgement calculated
The comparison of value and the 1st determinating reference value and whether be the 1st judgement hindering candidate portion.It addition,
Fig. 5 represents the skeleton diagram of the image illustrating the extract the 1st example hindering candidate portion 40.
It addition, Fig. 5 is as decision content, the length in region to be extracted the 1st hinder candidate portion 40
State, is to extract 4 the 1st states hindering candidate portion 40a, 40b, 40c, 40d.
It addition, decision content is not limited to above-mentioned example, be can the value of feature of comparison domain,
Suitably it is set.It addition, whether be that the 1st judgement hindering candidate portion 40 can be according to 1
Plant decision content to carry out, it is also possible to carry out according to multiple decision content.It addition, according to multiple
Decision content and determine whether the 1st hinder candidate portion 40 in the case of, at these multiple decision contents
In complete to judge according at least to a kind of decision content.For example, it is also possible to be 3 kinds of judgements
By this regional determination for being in the case of in value, at least two decision content meets the 1st determinating reference
1 hinders structure as candidate portion 40.
It addition, the 1st extracting part 31 is not limited to step S5 is repeated for each region
Structure, it is also possible to be for all regions calculate decision content, carry out then for each region
The decision content calculated and the comparison of the 1st determinating reference value and be whether the 1st to hinder candidate portion
Judgement as structure.
It follows that detection device 30 is by inspection area generating unit 32, for by the 1st extraction
What portion 31 extracted each 1st hinders candidate portion 40, generates and includes respective inspection area 41 (step
Rapid S6).It addition, Fig. 6 represents the image of an example of the inspection area 41 illustrating generation
Skeleton diagram.It addition, Fig. 6 be generate for 4 the 1st hinder candidate portion 40a, 40b, 40c,
The state of inspection area 41a, 41b, 41c, 41d of 40d.Inspection area 41 is by vertical
To and the region that surrounds of the rectangle that transversely extends.And, inspection area 41 is by making it
Area is that minimum rectangle obtains at the width longitudinally and expanding regulation the most respectively
Rectangle surround region.Herein, the shape of inspection area 41, size etc. do not limit,
It is to include the 1st region hindering candidate portion 40.Such as, inspection area 41 can also be by
The region that rhombus, trapezoidal, round, oval etc. surrounds.It addition, inspection area 41 can also be
The region tilted, such as, can also be to hinder the 1st to extend on the long axis direction in candidate portion 40
Rectangle etc..
It follows that detection device 30 is right with inspection area 41 by the 2nd extracting part 33 extraction
The original image (step S7) answered.
Then, the 2nd extracting part 33 is in the same manner as the 1st extracting part 31, to having extracted out and inspection
The original image of region 41 correspondence and the image that obtains is emphasized processing (step S8).And then,
2nd extracting part 33 utilizes the 2nd threshold value that this image having carried out emphasizing to process is carried out binaryzation
Process (step S9).Fig. 7 represents and illustrates and carried out binary conversion treatment by the 2nd extracting part 33
The skeleton diagram of an example of image.It addition, the 2nd threshold value is value set in advance, preserve
Not shown main storage means in detection device 30.And, the 2nd threshold value is less than the 1st
Threshold value.Herein, in the figure 7, compared with Fig. 4, many less regions are defined.Separately
Outward, hindering region corresponding for candidate portion 40b with the 1st and hindering candidate portion 40d with the 1st in Fig. 6
Corresponding region links, and is formed as 1 region.This is because, the 2nd threshold value is less than the 1st threshold
Value, is the state also including the less region of brightness.
Herein, there is the big region of brightness and bright randomly in traumatic part according to its degree of depth, width
Imaged under the state spending little region.But it is possible to by utilizing the 2nd threshold value to carry out two
Step S9 that value processes and the region that makes brightness less is also contained in the aftermentioned 2nd and hinders candidate portion,
It is prevented from the region suitable with traumatic part and becomes desultory wound.
It addition, the 2nd extracting part 33 is in the same manner as the 1st extracting part 31, carrying out binaryzation
In the image processed, carry out sign process (step S10).Then, the 2nd extracting part 33
For having been assigned each region of label, calculate such as area, length, longitudinal width,
Horizontal width etc., is used as decision content, decision content that these are calculated with pre-save
2nd determinating reference value compares, it is determined whether be the 2nd to hinder candidate portion, extraction and the 2nd wound
Corresponding region, candidate portion (step S11).
Herein, the 2nd hinders candidate portion in the same manner as the 1st hinders candidate portion, represents and is assumed to be traumatic part
Region.It addition, the 2nd determinating reference value in the 2nd extracting part 33 with in the 1st extracting part 31
The 1st determinating reference value different.2nd determinating reference value is set to be difficult to than the 1st determinating reference value
The state in extraction region.In addition, it is difficult to extraction state representation such as extract bigger region or
The state in longer region.On the other hand, it is easy to the state representation of extraction such as extracts less
Region or the state in shorter region.It addition, such as, about the length in region, the 2nd judges
The scope of reference value is the scope of the narrow range than the 1st determinating reference value, and is set to the 1st judgement
The scope that the scope of reference value is comprised.
It addition, the 2nd extracting part 33 is for having been assigned all regions of label, judge
The calculating of value, the decision content calculated and the comparison of the 2nd determinating reference value and whether be
2 judgements hindering candidate portion.It addition, Fig. 8 represents illustrates that the extract the 2nd hinders candidate portion 42
The skeleton diagram of the image of one example.It addition, Fig. 8 is as decision content using the length in region
Extract the 2nd state hindering candidate portion 42, be to extract two the 2nd to hinder candidate portion 42a, 42b
State.It addition, in fig. 8, do not extract in Fig. 5 to hinder candidate portion 40c with the 1st corresponding
Region.This is because, more difficult than the 1st determinating reference value owing to the 2nd determinating reference value is set to
To extract the state in region, it is determined that for hindering region corresponding for candidate portion 40c with the 1st be not
It is assumed to be the region in traumatic part.
It addition, decision content is not limited to above-mentioned example, be can the value of feature of comparison domain,
Suitably it is set.It addition, whether be that the 2nd judgement hindering candidate portion 42 can be according to 1
Plant decision content to carry out, it is also possible to carry out according to multiple decision content.It addition, according to multiple
Decision content determine whether the 2nd hinder candidate portion 42 in the case of, in these multiple decision contents
Complete to judge according at least to a kind of decision content.For example, it is also possible to be at 3 kinds of decision contents
Middle at least two decision content meets this regional determination in the case of determinating reference for being that the 2nd wound is waited
Structure as benefit portion 42.
It addition, the 2nd determinating reference value is set to be difficult to extract the shape in region than the 1st determinating reference value
State, is not limited to said structure.Such as, the 2nd determinating reference value and the 1st determinating reference value
Can also be identical, in this case, the 2nd extracting part 33 be configured to according to ratio the 1st
The decision content of the kind of extracting part more than 31 and determine whether the 2nd and hinder candidate portion.
It addition, the 2nd extracting part 33 is in the same manner as the 1st extracting part 31, it is not limited to for each
Region is repeated the structure of step S11, it is also possible to be to calculate decision content for all regions,
Decision content and the comparison of determinating reference value, the Yi Jishi calculated is carried out then for each region
No is structure as the 1st judgement hindering candidate portion.
It follows that detection device 30 is by hindering detection unit 34, hinder 42 points of candidate portion to the 2nd
Do not carry out expansion process (step S12).Herein, expansion process refers to, makes the 2nd to hinder candidate
The process that portion 42 expands.It addition, hinder detection unit 34 to hinder candidate portion 42 for each 2nd, meter
Can be regarded as the direction for the major axis in each profile, make the 2nd to hinder candidate portion 42 to this long axis direction
Expand the amount of regulation.That is, the expansion process hindering detection unit 34 is to make the 2nd candidate portion 42 elongated
Such process.It addition, Fig. 9 represents illustrates that hindering candidate portion 42 to the 2nd has been carried out at expansion
The skeleton diagram of the image of one example of reason.It addition, Fig. 9 is to two the 2nd wounds in Fig. 8
Candidate portion 42a, 42b have carried out the state of expansion process.It addition, the 2nd hinders candidate portion 42a
Hinder candidate portion 42b with the 2nd to link owing to expanding respectively, define 1 region 43.
It addition, hinder detection unit 34 in the same manner as the 1st extracting part the 31 and the 2nd extracting part 33,
Carry out expansion process hindering candidate portion 42 to the 2nd and in the image that obtains, carried out marked position
Reason (step S13).Then, hinder detection unit 34 for having been assigned each region of label,
Calculate such as area, length, longitudinal width, horizontal width etc., be used as decision content,
The decision content these calculated compares with the determinating reference value of hindering pre-saved, it is determined that be
No for traumatic part, it is traumatic part (step S14) by region detection corresponding with traumatic part.
Herein, the hinder in determinating reference value and the 2nd extracting part 33 the 2nd in detection unit 34 is hindered
Determinating reference value is different.Hinder determinating reference value to be set to be difficult to extract region than the 2nd determinating reference value
State.Such as, about the length in region, the scope hindering determinating reference value is to judge than the 2nd
The scope of the narrow range of reference value, and the scope that the scope being set to the 2nd determinating reference value is comprised.
It addition, hinder detection unit 34 for having been assigned all regions of label, carry out decision content
Calculating, the decision content that calculates with hinder the comparison of determinating reference value and whether be traumatic part
Judge.Such as, hinder detection unit 34 and will hinder candidate portion 42a by the 2nd after expanding in fig .9
1 region 43 hindering candidate portion 42b link with the 2nd and formed is detected as 1 traumatic part.So
After, the position data in the traumatic part detected is sent to controller C by detection device 30.
It addition, decision content is not limited to above-mentioned example, be can the value of feature of comparison domain,
Suitably it is set.It addition, be whether that the judgement in traumatic part can be entered according to a kind of decision content
OK, it is also possible to carry out according to multiple decision content.It addition, according to the judgement of multiple decision content be
No for traumatic part in the case of, complete according at least to a kind of decision content in these multiple decision contents
Judge.For example, it is also possible to be that at least two decision content meets judgement in 3 kinds of decision contents
By this regional determination for being structure as traumatic part in the case of benchmark.
It addition, hinder determinating reference value to be set to be difficult to extract the state in region than the 2nd determinating reference value
, it is not limited to said structure.Such as, hinder determinating reference value also may be used with the 2nd determinating reference value
To be identical, in this case, hinder detection unit 34 to be configured to according to than the 2nd extraction
The decision content of the kind in portion more than 33 and determine whether traumatic part.
It addition, hinder detection unit 34 in the same manner as the 1st extracting part the 31, the 2nd extracting part 33,
It is not limited to be repeated the structure of step S14 for each region, it is also possible to be for all districts
Territory calculates decision content, carries out decision content and the determinating reference value calculated then for each region
Comparison and whether be structure as the 1st judgement hindering candidate portion.
It addition, detection device 30 can also be configured to hinder candidate portion 40 by extracting the 1st and obtain
To image, extract generated inspection area 41, the 2nd hinder candidate portion 42 and obtain
Image, hinder the 2nd candidate portion 42 carried out expansion process and obtain image, wound detected
The data such as the image in portion, the decision content calculated suitably are saved in main storage portion.
It addition, detection device 30 can also be configured to will carry out in step s 2 emphasizing to process
Image be saved in main storage portion, in the step s 7 extract out and carrying out corresponding to inspection area 41
This emphasizes the image processed, and utilizes the 2nd threshold value to inspection area 41 is corresponding carries out with this
Emphasize that the image processed carries out binary conversion treatment.By being made such structure, it is possible to omit
It is emphasized step S8 processed, it is possible to reduce the operand of detection device 30.
So, in the present embodiment, detection device 30, by the 1st extracting part 31, utilizes
1st threshold value carries out binary conversion treatment to original image, and candidate portion 40 is hindered in extraction the 1st, by checking
Area generation portion 32, generates inspection area 41, by the 2nd extracting part 33, utilizes the 2nd threshold
Value carries out binary conversion treatment to inspection area 41, and extraction the 2nd is hindered candidate portion 42, sentenced by wound
Determine portion 34, hinder candidate portion 42 to the 2nd and carry out expansion process, detect traumatic part.According to the method,
The detection preventing traumatic part can be provided to omit and cross detection and improve the verification and measurement ratio in traumatic part
Utilize the traumatic part detection method of failure detector.
Herein, the 2nd threshold value in step S9 is less than the 1st threshold value in step S3.That is,
The step S9 pixel that also extraction brightness is lower compared with step S3.It addition, step S9 is pin
Process to the inspection area 41 as the region limited.And, detection device 30 is first
Extraction hinders candidate portion 40 as the 1st of the region being assumed to be traumatic part, for including that the 1st wound is waited
In the region of the restriction in benefit portion 40, again extraction as be assumed to be traumatic part region, include
Candidate portion 42 is hindered in the 2nd of the part that brightness is lower.Therefore, detection device 30 is not for artwork
What all region extractions of picture included the low part of brightness hinders candidate portion, it is possible to prevent micro-
Little wound, burr, dust etc. are detected as traumatic part (crossing detection).It addition, step S9 is to make
The binary conversion treatment of inspection area 41 in the region for limiting, so preventing from detecting device 30
Operand increases.
Hinder candidate portion 42 to the 2nd carry out expansion process to detect traumatic part it addition, hinder detection unit 34,
It is possible to prevent 1 continuous print wound is judged to it is desultory wound rather than traumatic part, anti-
Only the detection in traumatic part is omitted.It addition, hinder detection unit 34 to from employing brightness ratio the 1st threshold value
What the binary conversion treatment of the 2nd little threshold value extracted the 2nd hinders candidate portion 42 and carries out expansion process,
It is possible to the detection more efficiently preventing from traumatic part is omitted.
It addition, be to make the 2nd to hinder candidate portion 42 to length by hindering the expansion process that detection unit 34 carries out
Direction of principal axis expand process, i.e. make to being assumed to be it in the case of being 1 continuous print traumatic part
The direction in the direction extended carries out the process expanded.Therefore, hinder detection unit 34 to prevent more reliably
Only it is judged to that the region suitable with traumatic part is desultory wound rather than traumatic part, it is possible to higher
Accuracy prevent the detection in traumatic part from omitting.
It addition, detection device 30 by suitably set the 1st threshold value, the 2nd threshold value, the 1st
Determinating reference value, the 2nd determinating reference value, hinder determinating reference value such that it is able to suitably change
It is detected as the size in the traumatic part in traumatic part, it is possible to suitably regulate the inspection precision in traumatic part, usability
Well.
As described above, in the present embodiment, at camera head 16 to checked property 10
Surface image, detection the device 30 original image to being imaged by camera head 16 carry out
Reason, in the traumatic part detection method of the magnetic powder inspection device 1 in the traumatic part of detection surface, detection dress
Putting 30 utilizes the 1st threshold value that original image carries out binary conversion treatment, and candidate portion 40 is hindered in extraction the 1st,
Generate inspection area 41 in the way of 1st hinders candidate portion 40 including, utilize the 2nd threshold value to inspection
Looking into region 41 and carry out binary conversion treatment, extraction the 2nd is hindered candidate portion 42, is hindered candidate portion to the 2nd
42 carry out expansion process, detect traumatic part.
Therefore, according to present embodiment, using the teaching of the invention it is possible to provide prevent the detection in traumatic part from omitting and crossing inspection
Survey and improve the traumatic part detection method utilizing failure detector of the verification and measurement ratio in traumatic part.
It addition, a kind of magnetic powder inspection device 1, possess: camera head 16, to checked property
The surface of 10 images；And detection device 30, to the artwork imaged by camera head 16
As processing, the traumatic part of detection surface, wherein, described detection device 30 possesses: the 1st
Extracting part 31, utilizes the 1st threshold value that original image carries out binary conversion treatment, and candidate is hindered in extraction the 1st
Portion 40；Inspection area generating unit 32, generates inspection in the way of the 1st hinders candidate portion 40 including
Region 41；2nd extracting part 33, utilizes the 2nd threshold value to carry out inspection area 41 at binaryzation
Reason, candidate portion 42 is hindered in extraction the 2nd；And hinder detection unit 34, hinder candidate portion 42 to the 2nd and enter
Row expansion process, detects traumatic part.
It addition, according to present embodiment, using the teaching of the invention it is possible to provide prevent the detection in traumatic part from omitting and crossing inspection
Survey and improve the magnetic powder inspection device 1 of the verification and measurement ratio in traumatic part.
It addition, the magnetic powder inspection device 1 as failure detector is not limited to said structure.Such as,
Controller C can also be integrally formed with detection device 30.I.e., it is also possible to be controller C
Possess the 1st extracting part 31, inspection area generating unit the 32, the 2nd extracting part 33 and wound to judge
The structure in portion 34.By being made such structure, magnetic powder inspection device 1 is simplified, real
Existing miniaturization.It addition, detection device 30 can also be the structure possessing monitor, described prison
Visual organ make to extract the 1st hinder candidate portion 40 and obtain image, extract generated inspection
Region 41, the 2nd hinder candidate portion 42 and obtain image, hinder candidate portion 42 to the 2nd and carry out
Expansion process and the numbers such as the image obtained, the image in traumatic part detected, the decision content that calculates
According to display.The testing result in traumatic part can be suitably confirmed, usability by such structure
Well.
It addition, magnetic powder inspection device 1 can also be other that be also equipped with being connected with controller C
The structure of controller.Other controller in the same manner as controller C, by carry out calculation process and
Processing means, the main storage means etc. of preservation data that control processes are constituted, and are such as to possess
CPU, main storage means, intervalometer, input circuit, output circuit, power circuit etc. micro-
Type computer.Controller C by extract the 1st hinder candidate portion 40 and obtain image, extraction
Go out generated inspection area 41, the 2nd hinder candidate portion 42 and obtain image, to the 2nd wound
Candidate portion 42 has carried out expansion process and the image obtained, the image in traumatic part detected, meter
The data such as the decision content calculated deliver to other controller.On the other hand, other controller will be from control
These data that device C processed sends here count with the size of checked property 10 pre-saved, material etc.
Its main storage means it is saved according to associating.
By being made such structure, it is possible to make checked property 10 by other controller
The mapping data in traumatic part, it is possible to easily carry out the production management of checked property 10.It addition,
Controller C can also be configured to and receive the size of checked property 10, material from other controller
Etc. data, make the mapping data in the traumatic part of checked property 10.
Industrial applicability
It addition, the failure detector of the present invention is not limited to magnetic powder inspection device 1, for example, it is also possible to
Be use be impregnated with liquid and is detected a flaw in the traumatic part on the surface of checked property be impregnated with failure detector,
Can be applied to possess camera head that the surface to checked property images and to by imaging
The original image of device shooting is processed to all of the detection device in the traumatic part of detection surface
Failure detector.
Claims (8)
1. a failure detector, possesses:
Camera head, images the surface of checked property；And
Detection device, processes the original image imaged by described camera head, and detection is described
The traumatic part of surface,
It is characterized in that,
Described detection device possesses:
1st extracting part, utilizes the 1st threshold value that described original image is carried out binary conversion treatment, extraction
1st hinders candidate portion；
Inspection area generating unit, generates test zone in the way of the described 1st hinders candidate portion including
Territory；
2nd extracting part, utilizes the 2nd threshold value that described inspection area is carried out binary conversion treatment, takes out
Take the 2nd and hinder candidate portion；And
Hinder detection unit, hinder candidate portion to the described 2nd and carry out expansion process, detect described traumatic part.
Failure detector the most according to claim 1, it is characterised in that
Described 2nd threshold value is less than described 1st threshold value.
3. according to the failure detector described in claims 1 or 2, it is characterised in that
Described inspection area is by making the rectangle including the described 1st minimum hindering candidate portion expand
The width of regulation and region that the rectangle that obtains surrounds.
4., according to the failure detector described in any one in claims 1 to 3, its feature exists
In,
Described expansion process makes the described 2nd to hinder candidate portion to expand to respective long axis direction
Process.
5. utilize a traumatic part detection method for failure detector,
In the detection method of described traumatic part,
The surface of checked property is imaged by camera head,
The original image imaged by described camera head is processed by detection device, detects described table
Traumatic part at face,
Described traumatic part detection method is characterised by,
Described detection device utilizes the 1st threshold value that described original image is carried out binary conversion treatment, extraction
1st hinders candidate portion, generates inspection area including in the way of the described 1st hinders candidate portion, utilizes
2nd threshold value carries out binary conversion treatment to described inspection area, and extraction the 2nd hinders candidate portion, to institute
State the 2nd to hinder candidate portion and carry out expansion process, detect described traumatic part.
The traumatic part detection method utilizing failure detector the most according to claim 5, its feature
It is,
Described 2nd threshold value is less than described 1st threshold value.
7. according to the traumatic part detection method utilizing failure detector described in claim 5 or 6,
It is characterized in that,
Described inspection area is by making the rectangle including the described 1st minimum hindering candidate portion expand
The width of regulation and region that the rectangle that obtains surrounds.
8. according to the traumatic part utilizing failure detector described in any one in claim 5 to 7
Detection method, it is characterised in that
Described expansion process makes the described 2nd to hinder candidate portion to expand to respective long axis direction
Process.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2015-076333 | 2015-04-02 | ||
| JP2015076333A JP6596188B2 (en) | 2015-04-02 | 2015-04-02 | Flaw detector and flaw detection method using flaw detector |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106053593A true CN106053593A (en) | 2016-10-26 |
| CN106053593B CN106053593B (en) | 2021-05-11 |
Family
ID=57173507
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610202357.7A Active CN106053593B (en) | 2015-04-02 | 2016-04-01 | Flaw detection device and flaw detection method using flaw detection device |
Country Status (3)
| Country | Link |
|---|---|
| JP (1) | JP6596188B2 (en) |
| KR (1) | KR102518378B1 (en) |
| CN (1) | CN106053593B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109799240A (en) * | 2019-02-01 | 2019-05-24 | 信阳舜宇光学有限公司 | A kind of detection method of surface flaw |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP6697302B2 (en) * | 2016-03-25 | 2020-05-20 | マークテック株式会社 | Flaw detection device and defect detection method using flaw detection device |
| CN110892255B (en) * | 2017-07-07 | 2024-03-01 | 佳能株式会社 | Image processing apparatus, method and storage medium for detecting defect from image |
| CN108844965A (en) * | 2018-07-04 | 2018-11-20 | 武汉黎赛科技有限责任公司 | Defect image acquires display system, method, apparatus, computer and storage medium |
| CN111562273A (en) * | 2020-06-05 | 2020-08-21 | 大连工业大学 | Hyperspectrum-based fish water jet descaling slight damage visualization method |
| CN115082431B (en) * | 2022-07-20 | 2023-01-06 | 惠州威尔高电子有限公司 | PCB surface defect detection method |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0882616A (en) * | 1994-09-13 | 1996-03-26 | Sumitomo Metal Ind Ltd | Method and apparatus for inspecting flaw with magnetic particle |
| JPH10304196A (en) * | 1997-04-30 | 1998-11-13 | Nippon Telegr & Teleph Corp <Ntt> | Device and method for binarizing gradation image information |
| JP2000002523A (en) * | 1998-06-18 | 2000-01-07 | Ohbayashi Corp | Crack measuring method for structure surface |
| JP2003168114A (en) * | 2001-12-04 | 2003-06-13 | Olympus Optical Co Ltd | Defect sorting device |
| CN101464418A (en) * | 2007-12-18 | 2009-06-24 | 大日本网屏制造株式会社 | Flaw detection method and apparatus |
| CN101484818A (en) * | 2006-06-14 | 2009-07-15 | 浜松光子学株式会社 | Semiconductor defect analysis device, defect analysis method, and defect analysis program |
| JP2011013007A (en) * | 2009-06-30 | 2011-01-20 | Kobe Steel Ltd | Magnetic particle flaw inspection apparatus |
| CN104062354A (en) * | 2013-03-18 | 2014-09-24 | 宝山钢铁股份有限公司 | Steel pipe magnetic powder inspection fluorescent image detection apparatus and detection method |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP4009409B2 (en) | 1999-10-29 | 2007-11-14 | 株式会社日立製作所 | Pattern defect inspection method and apparatus |
| JP2001330567A (en) | 2000-05-19 | 2001-11-30 | Shimadzu Corp | X-ray inspection device |
| JP2008082860A (en) | 2006-09-27 | 2008-04-10 | Ntn Corp | Crack inspection method and its inspection apparatus |
| JP5315417B2 (en) * | 2009-11-18 | 2013-10-16 | 本田技研工業株式会社 | Surface inspection apparatus and surface inspection method |
| KR101492336B1 (en) | 2013-09-24 | 2015-02-11 | 서울여자대학교 산학협력단 | Automatic detection system and the method of internal and surface cracks of metal products |
-
2015
- 2015-04-02 JP JP2015076333A patent/JP6596188B2/en active Active
-
2016
- 2016-03-21 KR KR1020160033396A patent/KR102518378B1/en active IP Right Grant
- 2016-04-01 CN CN201610202357.7A patent/CN106053593B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0882616A (en) * | 1994-09-13 | 1996-03-26 | Sumitomo Metal Ind Ltd | Method and apparatus for inspecting flaw with magnetic particle |
| JPH10304196A (en) * | 1997-04-30 | 1998-11-13 | Nippon Telegr & Teleph Corp <Ntt> | Device and method for binarizing gradation image information |
| JP2000002523A (en) * | 1998-06-18 | 2000-01-07 | Ohbayashi Corp | Crack measuring method for structure surface |
| JP2003168114A (en) * | 2001-12-04 | 2003-06-13 | Olympus Optical Co Ltd | Defect sorting device |
| CN101484818A (en) * | 2006-06-14 | 2009-07-15 | 浜松光子学株式会社 | Semiconductor defect analysis device, defect analysis method, and defect analysis program |
| CN101464418A (en) * | 2007-12-18 | 2009-06-24 | 大日本网屏制造株式会社 | Flaw detection method and apparatus |
| JP2011013007A (en) * | 2009-06-30 | 2011-01-20 | Kobe Steel Ltd | Magnetic particle flaw inspection apparatus |
| CN104062354A (en) * | 2013-03-18 | 2014-09-24 | 宝山钢铁股份有限公司 | Steel pipe magnetic powder inspection fluorescent image detection apparatus and detection method |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109799240A (en) * | 2019-02-01 | 2019-05-24 | 信阳舜宇光学有限公司 | A kind of detection method of surface flaw |
Also Published As
| Publication number | Publication date |
|---|---|
| KR20160118932A (en) | 2016-10-12 |
| JP6596188B2 (en) | 2019-10-23 |
| KR102518378B1 (en) | 2023-04-04 |
| CN106053593B (en) | 2021-05-11 |
| JP2016197033A (en) | 2016-11-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106053593A (en) | Flaw detection device and flaw part detecting method by flaw detection device | |
| KR102342791B1 (en) | A flaw detection device, and a method for detecting defects by the flaw detection device | |
| Benedek et al. | Solder paste scooping detection by multilevel visual inspection of printed circuit boards | |
| BR102017026255A2 (en) | method of inspecting a steel strip | |
| WO2002065107A3 (en) | Web inspection system | |
| JP2016099195A (en) | Method and apparatus for measuring flow rate of particle image | |
| EP3904868A1 (en) | Learned model generation method, learned model, surface defect inspection method, steel manufacturing method, pass/fail determination method, grade determination method, surface defect determination program, pass/fail determination program, determination system, and steel manufacturing equipment | |
| JP2011013007A (en) | Magnetic particle flaw inspection apparatus | |
| JP2016156647A (en) | Inspection device using electromagnetic wave | |
| CN107709977A (en) | Surface defect detection apparatus and detection method of surface flaw | |
| CN102331430A (en) | Irregularly shaped meat foreign matter detection system | |
| JP5354187B2 (en) | Traveling material surface quality judging device and surface quality judging method | |
| CN107464336A (en) | A kind of colourless phosphorescence printing figure line quality online test method and device | |
| JP7295568B2 (en) | Inspection device and program | |
| TW201504591A (en) | Hole inspection device | |
| US10062155B2 (en) | Apparatus and method for detecting defect of image having periodic pattern | |
| CN109632810A (en) | Display panel crack detecting method and system | |
| JP2016217989A (en) | Defect inspection device and defect inspection method | |
| WO2022065110A1 (en) | X-ray inspection device and x-ray inspection method | |
| JP5889426B1 (en) | Web inspection device, web inspection method, web inspection program | |
| JP2015059854A (en) | Defect inspection method and defect inspection device | |
| JP2005172649A (en) | Apparatus for inspecting defect | |
| Sarkar et al. | On online counting of cigarette in packets—an image processing approach | |
| JP2005031952A (en) | Image processing inspection method and image processing inspection device | |
| Ma et al. | Research and develop on PCB defect intelligent visual inspection robot |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |