Movatterモバイル変換


[0]ホーム

URL:


CN109470162A - An intelligent detection system and method for micro-hole shape of fuel injector based on machine vision - Google Patents

An intelligent detection system and method for micro-hole shape of fuel injector based on machine vision
Download PDF

Info

Publication number
CN109470162A
CN109470162ACN201811316944.4ACN201811316944ACN109470162ACN 109470162 ACN109470162 ACN 109470162ACN 201811316944 ACN201811316944 ACN 201811316944ACN 109470162 ACN109470162 ACN 109470162A
Authority
CN
China
Prior art keywords
micropore
image
atomizer
electron microscope
control system
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
Application number
CN201811316944.4A
Other languages
Chinese (zh)
Other versions
CN109470162B (en
Inventor
杨淑珍
陈潜
李涵薇
石宵成
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shanghai Polytechnic University
Original Assignee
Shanghai Polytechnic University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Shanghai Polytechnic UniversityfiledCriticalShanghai Polytechnic University
Priority to CN201811316944.4ApriorityCriticalpatent/CN109470162B/en
Publication of CN109470162ApublicationCriticalpatent/CN109470162A/en
Application grantedgrantedCritical
Publication of CN109470162BpublicationCriticalpatent/CN109470162B/en
Activelegal-statusCriticalCurrent
Anticipated expirationlegal-statusCritical

Links

Classifications

Landscapes

Abstract

Translated fromChinese

本发明涉及一种基于机器视觉的喷油嘴微孔孔形智能检测系统及方法。该系统包括电子显微镜、图像处理系统、主控系统和旋转载物台,所述电子显微镜、所述图像处理系统和所述旋转载物台分别连接所述主控系统,所述电子显微镜还与所述图像处理系统连接,用于获得所述喷油嘴微孔的一次观测图像和二次观测图像,所述图像处理系统用于处理所述一次观测图像和所述二次观测图像,并判别所述微孔是否合格和其变形区域,并将结果图像实时显示,所述主控系统用于控制所述电子显微镜的移动和所述旋转载物台的转动。本发明实现了喷油嘴微孔检测的完全自动化、智能化,解决了现有技术中人工检测效率低、劳动强度大、分类质量不均、误判率高等问题。

The invention relates to an intelligent detection system and method for the micro-hole shape of a fuel injection nozzle based on machine vision. The system includes an electron microscope, an image processing system, a main control system and a rotating stage. The electron microscope, the image processing system and the rotating stage are respectively connected to the main control system, and the electron microscope is also connected to the main control system. The image processing system is connected to obtain a primary observation image and a secondary observation image of the micro-hole of the fuel injection nozzle, and the image processing system is used to process the primary observation image and the secondary observation image, and discriminate Whether the micro-hole is qualified and its deformation area, and display the result image in real time, the main control system is used to control the movement of the electron microscope and the rotation of the rotating stage. The invention realizes the complete automation and intelligence of the micro-hole detection of the fuel injection nozzle, and solves the problems of low manual detection efficiency, high labor intensity, uneven classification quality and high misjudgment rate in the prior art.

Description

A kind of atomizer micropore hole shape intelligent checking system and method based on machine vision
Technical field
The present invention relates to automatic producing technology fields, more specifically to a kind of atomizer based on machine visionMicropore hole shape intelligent checking system and method.
Background technique
Have multiple small spray orifices on diesel engine atomizer, the processing of spray orifice usually first using electric-spark drilling, thenIt is processed using liquid extrusion and abrasion technique and improves discharge coefficient, improve atomizing effect, improve fuel efficiency.But it is ground by extrudingSpray orifice after mill, it sometimes appear that spray orifice deforms.If the aperture of spray orifice deforms, then will to fuel atomization effect generate compared withIt is big to influence, reduce fuel efficiency.Therefore, to ensure quality, need to detect the orifice deformation of the spray orifice after crush and grindSituation, the hole shape of each spray orifice on nozzle is relatively round, has no notch side as qualification.Existing detection method is, by digital micro-analysisMirror is fixed, and measured workpiece is placed on objective table, then utilizes digital microscope pair to each spray orifice with hand turning of work pieceThese micropores, which amplify, to be imaged on display, directly discriminates whether qualification according to the profile in aperture with human eye.In addition,Since the height in each hole on spherical surface or the conical surface is different, lead to have some holes not at the center of image according to imaging when detectingPrinciple can cause certain barrel distortion so as to cause erroneous judgement.So this artificial detection method low efficiency, large labor intensity,Classification quality is uneven, False Rate is relatively large.
Summary of the invention
For in the prior art to drawbacks described above existing during atomizer micropore pile bore checking, the purpose of the present invention isA kind of atomizer micropore hole shape intelligent checking system and method based on machine vision is provided, which can replace at present artificialAutomatic detection is realized in detection, using machine vision, is realized intelligent distinguishing, is reduced False Rate, improve detection efficiency.For realityThe existing purpose of the present invention, the invention adopts the following technical scheme:
A kind of atomizer micropore hole shape intelligent checking system based on machine vision, the system include electron microscope, figureAs processing system, master control system and rotatable stage for placing the atomizer, at the electron microscope, described imageReason system and the rotatable stage are separately connected the master control system, the electron microscope also with described image processing systemConnection, the electron microscope are used to obtain an observed image R1 and secondary observed image R2 for the atomizer micropore, andObserved image R1 and the secondary observed image R2 are delivered to described image processing system, described image processing systemSystem is for handling an observed image R1 to obtain an observed image center position coordinates (that is, the electron microscope visual fieldCenter position coordinates) (Xa,Ya) and atomizer micropore center position coordinates (Xh,Yh), and by an observed image centre bitSet coordinate (Xa,Ya) and the atomizer micropore center position coordinates (Xh,Yh) it is transmitted to the master control system, and the figureAs processing system is also used to handle the secondary observed image to differentiate whether the atomizer micropore qualified and the atomizerThe deformed region position of micropore, and by result images real-time display, the master control system is for controlling the electron microscopeMovement is so that the atomizer micropore enters the visual field of the electron microscope so that the electron microscope obtains the oil spoutObserved image R1 of mouth micropore, and according to an observed image center position coordinates (Xa,Ya) and the atomizer it is micro-Hole center position coordinates (Xh,Yh) the secondary positioning electron microscope moves along X and Z-direction with by the field of microscopeCenter is adjusted to atomizer micropore center so that the microscope obtains the secondary observed image R2, and describedMaster control system is also used to control the rotatable stage rotation so that next micropore of the atomizer moves to the electronic displayIn microscope fields.
Further, the intelligent checking system further includes positioning mechanism, and the positioning mechanism is for installing the electronicsIt microscope and is connect with the master control system with the electron microscope mobile under the control of the master control system.
Further, the system also includes shell, the rotatable stage and the positioning mechanisms to be mounted on the shellSurface, the master control system are placed on inside the shell.
Further, the side or upper surface of the shell are also equipped with starting detection button.
Further, described image processing system includes following step to the observed image R1 process handledIt is rapid:
(1) by obtaining an observed image size, variable X is seta, YaObtain an observed image centerPosition coordinates (Xa,Ya), Xa=A/2, Ya=B/2, A and B are respectively the length and width of described image;
(2) HSV conversion is carried out to an observed image, is further processed using S image, by adjusting gray scaleValue filters out complete circular hole image within the scope of 0-30, and it is micro- to go out atomizer with select-contours-xld operator extractionThe existing profile in hole carries out preliminary circularity screening, region of the circularity in (0.95-1) range is filtered out, using area_Center operator extraction is to atomizer micropore center position coordinates (Xh,Yh)。
Further, the master control system secondary positioning electron microscope process the following steps are included:
(1) image center location coordinate (X is found outa,Ya) and micropore center position coordinates (Xh,Yh) positional distance △ Xf=Xa-Xh、△Yf=Ya-Yh
(2) the motion profile Tr of electron microscope is determined, specifically: X-axis trackZ axis track:Rotation angle of rotating shaft:Wherein, α is atomizer cone-apex angle, RThe section radius where micropore;
(3) the field of microscope center is adjusted to atomizer micropore center according to the motion profile TrPosition obtains the secondary observed image R2 so as to the microscope.
Further, described image processing system includes following step to the secondary observed image R2 process handledIt is rapid:
(1) decompose operator is used to the secondary observed image R2, converts the channel HSV for triple channel image, selectedThe most clearly s image progress gray value adjustment, gray value magnitude range in three images is taken to extract 0-30 more completeCircular hole;
(2) circular hole region is filled by red, is split using connection operator to picture region, will be disconnectedRegion segmentation obtains isolated area at individual region;Partially exist after segmentation and be not completely filled, notch occurs in marginal portionSituation is filled image, determines that image is a closed figure;Morphological scale-space is carried out to image, after first being corrodedExpansive working is carried out again and obtains image R3, the image border R3 is made to become round and smooth, disconnect narrow interruption and eliminates thin protrusion, is heldIt has gone and image-region has been screened using feature histogram after aforesaid operations, circular hole region is completely covered after the completion of screening,The circle that circularity is not less than 0.75 is only extracted using select-shape-xld operator, the central point of edge pixel is taken more to be sticked onNearly actual micropore profile L1 carries out circularity screening, and judgement of the circularity in (0.9-1) range is qualified, if unqualified, uses edgeIt is fitted split plot design and is fitted positive round L2,
(3) after completing aforesaid operations, filling L1, L2 obtain micro-porous area D1 and fitting circle region D2, then to useDifference operator does region subtraction to D1, D2, identifies deformed region position.
Further, the specific steps of the edge fitting split plot design fitting positive round L2 include: setting huber weighted value,Formula is
Severely deformed part is filtered, finds out former micropore central point, then an xld is fitted to the image R3 Jing Guo preliminary treatmentCircle, that is, positive round L2.
The present invention also provides a kind of oil nozzles using the atomizer micropore hole shape intelligent checking system based on machine visionMicropore hole shape intelligent detecting method, which is characterized in that this method comprises:
S101. tested atomizer is fixed on rotatable stage;
S102. start detection button, digital microscope is moved to the tested spray from origin under the control of master control systemOil nozzle is nearby focused with a micropore of the tested atomizer, is shot to the micropore, is obtained an observed image R1;
S103. image processing system handles observed image of the micropore, obtains an observed image centerPosition coordinates (that is, electron microscope central region position coordinates) (Xa,Ya) and the micropore center position coordinates (Xh,Yh);
S104. master control system is according to an observed image center position coordinates (Xa,Ya) and the micropore centerCoordinate (Xh,Yh) the secondary positioning electron microscope to be to be moved to the micropore center for the electron microscope central regionPosition shoots the micropore again, obtains secondary observed image R2;
S105. described image processing system handles to differentiate whether the micropore closes the secondary observed imageLattice and bore deformation regional location, and by result images real-time display, complete the detection of a micropore;
S106. the rotatable stage rotates under the control of the master control system, the tested atomizer it is mobile so thatNext micropore of the digital microscope and the tested atomizer is focused, and is shot to the micropore, is obtained an observation chartPicture repeats step S103-S106 until all micropores on workpiece are successively detected and finished;
S107. after all micropores all detect on tested atomizer, if all micropores are qualified, the tested sprayOil nozzle is qualified, and otherwise unqualified, described image processing system shows the result;
S108. the digital microscope returns to origin under the control of the master control system, takes out the tested atomizer,Complete the detection of an atomizer.
Further, the intelligent checking system further includes positioning mechanism, and the positioning mechanism is for installing the electronicsIt microscope and is connect with the master control system so as to the electron microscope mobile under the control of the master control system.
Further, the system also includes shell, the rotatable stage and the positioning mechanisms to be mounted on the shellSurface, the master control system are placed on inside the shell.
Further, described image processing system includes following step to the observed image R1 process handledIt is rapid:
(1) by obtaining an observed image size, variable X is seta, YaObtain an observed image centerPosition coordinates (Xa,Ya), Xa=A/2, Ya=B/2,
(2) HSV conversion is carried out to an observed image, is further processed using S image, by adjusting gray scaleValue filters out complete circular hole image within the scope of 0-30, and it is micro- to go out atomizer with select-contours-xld operator extractionThe existing profile in hole carries out preliminary circularity using select-shape-xld operator and screens, filters out circularity in (0.95-1) modelInterior region is enclosed, using area_center operator extraction to atomizer micropore center position coordinates (Xh,Yh)。
Further, the master control system secondary positioning electron microscope process the following steps are included:
(1) image center location coordinate (X is found outa,Ya) and micropore center position coordinates (Xh,Yh) positional distance △ Xf=Xa-Xh、△Yf=Ya-Yh,
(2) the motion profile Tr of electron microscope is determined, specifically: X-axis trackZ axis railMark:Rotation angle of rotating shaft:Wherein, α is the atomizer vertex of a coneAngle, R are section radius where micropore;
(3) the field of microscope center is adjusted to atomizer micropore center according to the motion profile TrPosition obtains the secondary observed image R2 so as to the microscope.
Further, described image processing system includes following step to the secondary observed image R2 process handledIt is rapid:
(1) decompose operator is used to the secondary observed image R2, converts the channel HSV for triple channel image, selectedThe most clearly s image progress gray value adjustment, gray value magnitude range in three images is taken to extract 0-30 more completeCircular hole;
(2) circular hole region is filled by red, is split using connection operator to picture region, will be disconnectedRegion segmentation obtains isolated area at individual region;Partially exist after segmentation and be not completely filled, notch occurs in marginal portionSituation is filled image, determines that image is a closed figure;Morphological scale-space is carried out to image, after first being corrodedExpansive working is carried out again and obtains image R3, the image border R3 is made to become round and smooth, disconnect narrow interruption and eliminates thin protrusion, is heldIt has gone and image-region has been screened using feature histogram after aforesaid operations, circular hole region is completely covered after the completion of screening,The circle that circularity is not less than 0.75 is only extracted using select-shape-xld operator, the central point of edge pixel is taken more to be sticked onNearly actual micropore profile L1 carries out circularity screening, and judgement of the circularity in (0.9-1) range is qualified, if unqualified, uses edgeIt is fitted split plot design and is fitted positive round L2,
(3) after completing aforesaid operations, filling L1, L2 obtain micro-porous area D1 and fitting circle region D2, then to useDifference operator does region subtraction to D1, D2, identifies deformed region position.
Further, the specific steps of the edge fitting split plot design fitting positive round L2 include: setting huber weighted value,Formula is
Severely deformed part is filtered, finds out former micropore central point, then an xld is fitted to the image R3 Jing Guo preliminary treatmentCircle, that is, positive round L2.Edge fitting split plot design fitting positive round is particularly suitable for the feelings that moderate finite deformation occurs for micropore in the present inventionCondition can solve and directly adopt the center that xld fit approach is fitted and the problem of larger offset can actually occur.
Of the invention atomizer micropore hole shape intelligent checking system and method based on machine vision solves artificial detectionThe problems such as workpiece micropore detection efficiency is low, classification quality is uneven.Compared with the prior art, invention achieves below beneficial to effectFruit:
The present invention is based on the micropore hole shape intelligent checking systems of machine vision to obtain atomizer micropore using digital microscopeImage, amplification factor is big, pixel is high, effectively raises the image quality of micropore, solves and requires manually in the prior art notDisconnected rotational workpieces and focusing, the easily generation fatigue of detection for a long time influence detection efficiency, while solving due to detection effectCompletely by eye-observation subjective discrimination gained, the problem that discrimination standard is different, error is relatively high.
The method that the present invention is taken based on the edge fitting segmentation of Huber weight for moderate finite deformation hole, by moderate finite deformationRegion and main outline are distinguished, the center of more accurate determination micropore, solve the generation of moderate finite deformation micropore centerThe problem of relatively large deviation, a possibility that reducing erroneous judgement.
The present invention can also carry out identification mark to the micropore deformed region position after processing, compared with artificial detection, so thatTesting result is more clear, is clear, while can provide preferable analysis of data and foundation to find Deformation Reasons.
The present invention secondary positioning alignment acquisition final process image solves on workpiece one week due to each micropore height notOne, there is some holes not hole shape extra distortion caused by the center of image, so as to cause other problem is judged by accident.
The invention proposes a kind of completely new intelligent solutions for the detection of engine nozzle micropore, are based on machineDevice vision realizes full automation, the intelligence of the detection of workpiece micropore, reduces labor intensity, False Rate and production cost,Improve production efficiency and classification quality.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show belowThere is the attached drawing of technology to do simply to introduce, it is clear that drawings discussed below be only to some embodiments of the present invention, forFor those of ordinary skill in the art, under the preceding topic not made the creative labor, it can also be obtained according to these attached drawings otherAttached drawing.
Fig. 1 is the atomizer micropore hole shape intelligent measurement of the invention based on machine vision according to a specific embodimentThe structural schematic diagram of system.
Fig. 2 is an observed image R1 of the micropore obtained using detection system of the invention.
Fig. 3 is the secondary observed image R2 of the micropore obtained using detection system of the invention.
The micropore profile L1 that Fig. 4 is extracted, it is qualified circular hole that circularity, which is 0.94,.
The positive round L2 obtained after Fig. 5 fitting.
An observed image of severely deformed micropore after the processing that Fig. 6 is obtained using detection system of the invention.
What Fig. 7 was extracted arrives micropore profile L1.
Fig. 8 can accurately determine the center fitting circle D2 according to edge fitting split plot design
Fig. 9 does the deformed region that region subtraction obtains according to D1, D2.
Specific embodiment
In the present invention definition of coordinate refer to Fig. 1, the plane that electron microscope is moved in parallel along the horizontal plane be set as X-axis andPlane where Y-axis, the plane where microscope is set as Z axis along the plane of vertical level moved up and down.It will be of the inventionIntelligent checking system starting before electron microscope position or the intelligent checking system starting after electron microscope do not openThe position coordinates in the electron microscope visual field are set as origin when beginning mobile.
In a preferred embodiment of the invention, a kind of atomizer micropore hole shape based on machine vision of the inventionIntelligent checking system includes electron microscope, image processing system, master control system, rotatable stage, positioning mechanism.Measured workpieceAtomizer is fixed on the rotatable stage;The electron microscope is fixed on the positioning mechanism, in the master controlUnder system control, an observed image R1 of micropore is obtained;Described image processing system is connect with the electron microscope, is used forAn observed image R1 is received, and an observed image R1 is handled, obtains the center of an observed imageCoordinate (that is, electron microscope central region position coordinates) (Xa,Ya) and micropore center position coordinates (Xh,Yh), and will be describedCentral region coordinate (the X of observed image R1a,Ya) and micropore center position coordinates (Xh,Yh) it is transferred to the master control systemSystem;The master control system is connect with image processing system, for receiving the coordinate of image processing system transmission, calculates offset distanceFrom △ Xf、△Yf, determine the motion profile Tr of electron microscope, specifically: X-axis trackZ axis railMark:Rotation angle of rotating shaft:Wherein, α is the atomizer vertex of a coneAngle, R are section radius where micropore, finally carry out secondary positioning to the electron microscope according to motion profile Tr, will be describedMicropore center adjustment is to electron microscope central region position;Then the electron microscope obtains micropore after secondary positioningSecondary observed image R2, and the secondary observed image R2 of the micropore is handled, judges whether tested micropore hole shape qualified, andIdentify deformed region position;The master control system also controls the rotatable stage and next tested atomizer micropore is moved to electricityJudged in sub- field of microscope, is finished until the hole on measured workpiece is successively detected.
In a preferred embodiment of the invention, described image processing system carries out an observed image R1Processing, is handled observed image R1 using the fractional spins based on image definition attribute, process is as follows: passing throughImage size is obtained, variable X is seta,Ya, obtain image R1 center point coordinate (Xa,Ya), Xa=A/2, Ya=B/2, A, B are figurePiece length and width will then obtain the center of micropore, first carry out HSV conversion to image R1, be carried out at next step using S imageReason filters out complete circular hole image by adjusting gray value, with select-contours-xld operator within the scope of 0-30The existing profile of atomizer micropore is extracted, because may be subjected to the influence generation area identification model of the luminance shortage of fringe regionLarger situation is enclosed, preliminary circularity screening is carried out, region of the circularity in (0.75-1) range is filtered out, using area_Center point coordinate (X of the center operator extraction to atomizer microporeh, Yh)。
In a preferred embodiment of the invention, described image processing system observes micropore after the secondary positioningImage R2 is handled, is judged whether qualification specifically includes tested micropore hole shape:, will to image R2 using decompose operatorTriple channel image is converted into the channel HSV, chooses in three images that the most clearly s image carries out gray value adjusting range, experimentMore complete circular hole can be extracted in 0-30 by obtaining gray value magnitude range.Circular hole region is filled by red, is utilizedConnection operator is split picture region, by disconnected region segmentation at individual region, obtains isolated area.PointDiscovery part, which exists, after cutting is not completely filled, and the case where notch occurs in marginal portion, therefore need to be filled to image, determines figureIt seem a closed figure.Morphological scale-space is carried out to image, is first corroded and carries out expansive working afterwards and obtain image R3, make R3Image border becomes round and smooth, disconnects narrow interruption and eliminates thin protrusion.It finds spraying after the experiment of a large amount of real imagesAfter oil nozzle is processed it is possible that edge, which has metallic residue to retain, will affect subsequent identification, after having executed aforesaid operationsImage-region is screened using feature histogram, circular hole region is completely covered after the completion of screening, only relies on gray value tuneThe whole extraction to micropore be it is incomplete, some pictures can be because shooting problem fringe region brightness be lower, if only carrying out ashAngle value will appear the erroneous judgement situation of more profiles when adjusting subsequent extracted profile, accurately to extract former micropore, can be usedSelect-shape-xld operator only extracts the circle that circularity is not less than 0.75.The central point of edge pixel is taken to obtain more close to realityThe micropore profile L1 on border.Circularity screening is carried out, judgement of the circularity in (0.9-1) range is qualified, if unqualified, uses edge fittingSplit plot design is fitted positive round L2, and filling L1, L2 obtain micro-porous area D1 and fitting circle region D2, then to using difference operatorRegion subtraction is done to D1, D2, identifies deformed region position.
In a preferred embodiment of the invention, it is fitted positive round with edge fitting split plot design, by the micropore wheel of extractionWide L1 is compared with positive round, is identified deformed region position, is specifically included: being directlyed adopt when moderate finite deformation occurs for microporeThe center and the problem of larger offset can actually occur that xld fit approach is fitted, the present invention is divided using edge fitting,Huber weighted value is set, and formula is
Severely deformed part can be filtered, convenient for finding out former micropore central point, then it is quasi- to the image R3 Jing Guo preliminary treatmentUnify an xld circle, that is, positive round L2.
In a preferred embodiment of the invention, next tested micropore is moved to electronic display by the master control system controlThe process judged in microscope fields specifically includes: after described image processing system has judged whether current micropore is qualified,Master control system control rotatable stage motor rotates by a certain angle around W axis, and next tested micropore is moved to the electron microscope visual field,Then micropore image R1 is obtained and handled, secondary positioning is carried out, obtains the secondary observed image R2 of micropore after secondary positioning and sentenceWhether other hole shape is qualified.
In a preferred embodiment of the invention, the rotatable stage is specifically used for the control in master control systemUnder, measured workpiece is clamped, and can be by predetermined angular rotational workpieces, so that being evenly distributed in multiple holes of circumference on workpiece successivelyComplete detection.
In a preferred embodiment of the invention, the positioning mechanism includes X-axis and Z axis, and two axis are all made of lead screwEtc. high-precisions drive mechanism, motor use with encoder feedback integrated form step-servo motor, compact-sized, kinematic accuracyHigh, cost performance height.The positioning mechanism is specifically used for: after measuring the rigid incipient stage, being put into and clamp workpiece, according to master control systemThe motor command and parameter of transmission, control X, the rotation of Z axis motor, the mobile electron microscope is from zero-bit (origin) close to workpieceIt (atomizer) and focuses to the 1st micropore;According to the motion profile Tr of the secondary positioning, its X, the operating of Z-direction motor are controlled,The center position coordinates that electron microscope central region is moved to the circular hole complete secondary positioning, to guarantee to obtain for the second timeThe hole shape obtained differentiates that image R2 mesoporous can solve to make hole not at same axis height because of each hole in the center in the visual fieldImage can deviate central region even on the additional bring barrel distortion error of vision periphery;All micropores are equal on workpieceAfter detection, the electron microscope is controlled far from workpiece and returns to zero-bit, in order to take out workpiece, while being next workpieceMeasurement be ready.
In a preferred embodiment of the invention, the digital microscope includes annular LED light source, amplification factor200 times or more, 2,000,000 or more pixel.
In order to make the technical means, the creative features, the aims and the efficiencies achieved by the present invention be readily apparent from understanding, tie belowAttached drawing is closed, the present invention is further explained.
Embodiment 1
Referring to such as Fig. 1, in the example shown in the figure, atomizer micropore hole shape intelligent checking system includes electron microscope1, image processing system 2, master control system 3, rotatable stage 4, positioning mechanism 5 and shell 6.Electron microscope 1 is fixed on localization machineOn structure 5.Rotatable stage 4 and positioning mechanism 5 are fixed on 6 upper surface of shell, and master control system 3 is placed on inside shell 6, image procossing systemSystem 2 can be fixedly connected with shell 6, can also be respectively independent.Positioning mechanism 5, image processing system 2 and rotatable stage 4 connect respectivelyConnect the master control system 3.
Electron microscope 1 is also connect with image processing system 2, described for obtaining the primary observation of the atomizer microporeImage R1 and secondary observed image R2, and an observed image R1 and the secondary observed image R2 are sent at imageReason system 2.
Image processing system 2 receives and handles an observed image R1 to obtain an observed image center seatMark (that is, 1 central region position coordinates of electron microscope) (Xa,Ya) and atomizer micropore center position coordinates (Xh,Yh), and willThe two center position coordinates are transferred to master control system 3.Image processing system 2 also receives and handles the processing secondary observation chartAs R2 to differentiate whether the atomizer micropore is qualified, the deformed region position of the atomizer micropore, and result images are realWhen show,
Master control system 3 is according to the observed image center position coordinates (X receiveda,Ya) and atomizer micropore centre bitSet coordinate (Xh,Yh) the motion profile Tr that determines electron microscope 1, secondary positioning is carried out so that atomizer is micro- to electron microscope 1Hole center is adjusted to 1 central region position of microscope.Master control system 3 also controls the rotation of rotatable stage 4 so that the atomizerNext micropore move in 1 visual field of electron microscope and detected.
In use, tested atomizer is placed on rotatable stage 4, fixed with fixture, Pen-down detection button, master control system3 control clamp atomizers, while controlling positioning mechanism 5 keeps the visual field of electron microscope 1 (former from zero-bit along X-axis and Z axisPoint) start to move, so that digital microscope 1 moves near tested atomizer, and focuses with No. 1 micropore of tested atomizer, it is rightMicropore is shot, and obtains the clearly hole shape image of the micropore, referred to as an observed image R1, and the image transmitting is extremely schemedAs processing system 2;Image processing system 2 handles an observed image R1 of the micropore, obtains in an observed imageHeart position coordinates (Xa,Ya) and micropore center position coordinates (Xh,Yh), and it is delivered to master control system 3;Master control system 3 according to this oneSecondary observed image center position coordinates (Xa,Ya) and the micropore center position coordinates (Xh,Yh) determine the movement of electron microscope 1Then track Tr controls electron microscope 1 along X-axis by positioning mechanism 5 according to motion profile Tr, Z-direction is mobile, will be electricSub- 1 central region of microscope is moved to tested micropore center;Electron microscope 1 shoots micropore again, obtains againMicropore image, referred to as secondary observed image R1, and by the image transmitting to image processing system 2;At image processing system 2It manages secondary observed image R2, differentiates whether tested micropore qualified and bore deformation regional location, and result images R3 is shown in real timeShow, completes the detection of a micropore;Rotatable stage 4 is rotated according to the instruction of master control system 3, is tested atomizer movement, so thatNext micropore enters the visual field of microscope 1, repeats above step until all micropores on atomizer are successively detected and finished;WhenAfter all micropores all detect on tested atomizer, if all micropores are qualified, the workpiece is qualified, otherwise unqualified, shouldAs a result it is shown by image processing system 2;Then, master control system 3 controls positioning mechanism 5 and moves along X-axis and Z-direction, returnsTo dead-center position, finally, taking out workpiece, it is placed into different places according to testing result, the detection of the atomizer in this way finishes.
The acquisition of image can be by external trigger in the present invention, can also be after setup parameter, by 1 automatic collection of microscope.ConsiderMicropore is not or not field of microscope center in when acquiring there may be initial pictures (observed image), so using secondary fixedThe method of position acquires image again, it is ensured that micropore is at field of microscope center.Such as Fig. 2 is shown and is obtained using microscope 1Observed image a R1, Fig. 3 of one micropore of atomizer are by the secondary sight of the micropore obtained after microscope 1 two times positioningAltimetric image R2.
Master control system 3 specifically includes following procedure to the secondary position fixing process of microscope 1:
Variable X is arranged by obtaining an observed image R1 in image processing system 2a, YaObtain image center (Xa,Ya), Xa=A/2, Ya=B/2 then obtains the center of micropore, first carries out HSV conversion to the image, is carried out using S imageIt handles in next step, complete circular hole image is filtered out within the scope of 0-30 by adjusting gray value, with select-Contours-xld operator extraction goes out the existing profile of atomizer micropore, because may be subjected to the shadow of the luminance shortage of fringe regionThe larger situation of generation area identification range is rung, needs to carry out preliminary circularity screening, filters out circularity in (0.95-1) rangeInterior region, using the central point (X of area_center operator extraction to atomizer microporeh,Yh), and it is delivered to control system 3.Control system 3 finds out the positional distance △ X of micropore central point Yu picture (image) central pointf=Xa-Xh, △ Yf=Ya-Yh, and rootDetermine that the motion profile Tr of electron microscope 1 shoots photo, Tr specifically: X to moving lens to micropore center according to the distanceTrack shaftZ axis track:Rotation angle of rotating shaft:Wherein, α is atomizer cone-apex angle, and R is section radius where micropore;Finally according toMotion profile Tr adjusts the microscope (1) central region position to atomizer micropore center so as to described micro-Mirror (1) obtains the secondary observed image R2.
Image processing system 2 handles secondary observed image R2, to judge whether qualification includes tested micropore hole shapeFollowing procedure:
Image processing system 2 uses decompose operator to image R2, converts the channel HSV for triple channel image, choosesThe most clearly s image carries out gray value adjusting range in three images, and experiment obtains gray value magnitude range can mention in 0-30Get more complete circular hole.Circular hole region is filled by red, is split using connection operator to picture region, willDisconnected region segmentation obtains isolated area at individual region.Discovery part, which exists, after segmentation is not completely filled, edge partThe case where separating existing notch, therefore image need to be filled, determine that image is a closed figure.Image is carried out at morphologyReason is first corroded and carries out expansive working afterwards and obtain image R3, so that the image border R3 is become round and smooth, disconnect narrow interruption and eliminationThin protrusion.Discovery is after atomizer is processed it is possible that there is metal in edge after the experiment of a large amount of real imagesResidue, which retains, will affect subsequent identification, be screened using feature histogram to image-region after having executed aforesaid operations, sieveCircular hole region is completely covered after the completion of choosing, and only relying on gray value adjustment is incomplete, some picture meetings to the extraction of microporeBecause it is lower to shoot the brightness of problem fringe region, if only carrying out will appear more profiles when gray value adjustment subsequent extracted profileErroneous judgement situation can be used select-shape-xld operator and only extract circularity not less than 0.75 accurately to extract former microporeCircle.The central point of edge pixel is taken to obtain the micropore profile L1 of more closing to reality.Circularity screening is carried out, circularity is in (0.9-1) judgement in range is qualified.It is illustrated in figure 4 the micropore profile L1 for extracting and obtaining, it is qualified circular hole that circularity, which is 0.94,.
If unqualified, it is fitted positive round L2 with edge fitting split plot design, is specifically included: is straight when for micropore generation moderate finite deformationThe problem of connecing the center being fitted using xld fit approach and larger offset can actually occur, the present invention is quasi- using edgeSplit plot design is closed, setting huber weighted value is configured, formula are as follows:
Severely deformed part can be filtered, convenient for finding out former micropore central point, then it is quasi- to the image R3 Jing Guo preliminary treatmentUnify an xld circle, that is, positive round L2.After completing aforesaid operations, filling L1, L2 obtain micro-porous area D1 and fitting circle regionD2, then region subtraction (21) operation is done to D1, D2 to using difference operator, identify deformed region position.
Fig. 5 show the xld circle positive round L2 being fitted on the basis of Fig. 4.
It is illustrated in figure 6 severely deformed micropore (observed image) after the processing of the acquisition of microscope 1, is unqualified micro-Hole.
It is illustrated in figure 7 the L1 to the unqualified micropore profile of the extraction of image processing system 2.
Fitting circle D2 centre bit can accurately be determined according to edge fitting split plot design by being illustrated in figure 8 image processing system 2It sets.
It is as shown in Figure 9 that the deformed region that region subtraction obtains is done according to D1, D2.Filling L1, L2 obtain micro-porous area D1With fitting circle region D2, then to region subtraction is done to D1, D2 using difference operator, deformed region position is identifiedIt sets.
Basic principles and main features and advantages of the present invention of the invention have been shown and described above.The skill of the industryArt personnel it should be appreciated that the present invention is not limited to the above embodiments, the above embodiments and description only describeThe principle of the present invention, without departing from the spirit and scope of the present invention, various changes and improvements may be made to the invention, theseChanges and improvements all fall within the protetion scope of the claimed invention.The claimed scope of the invention by appended claims andIts equivalent thereof.

Claims (10)

The master control system (3) is used to control the movement of the electron microscope (1) so that described in atomizer micropore entranceThe visual field of electron microscope (1) obtains an observed image of the atomizer micropore, and root so as to the electron microscope (1)According to an observed image center position coordinates (Xa,Ya) and the atomizer micropore center position coordinates (Xh,Yh) secondary fixedThe position electron microscope (1) is moved along X and Z-direction adjusting the microscope (1) central region position to the sprayOil nozzle micropore center obtains the secondary observed image so as to the microscope (1), and the master control system (3) is also used to controlThe rotatable stage (4) rotation is made so that next micropore of the atomizer moves in the electron microscope (1) visual field.
(2) circular hole region is filled by red, is split using connection operator to picture region, by disconnected regionIt is divided into individual region, obtains isolated area, partially exists after segmentation and is not completely filled, the feelings of notch occurs in marginal portionCondition is filled image, determines that image is a closed figure, carries out Morphological scale-space to image, after first being corroded againIt carries out expansive working and obtains image R3, the image border R3 is made to become round and smooth, disconnect narrow interruption and eliminate thin protrusion, executeImage-region is screened using feature histogram after complete aforesaid operations, circular hole region is completely covered after the completion of screening, is madeWith select-shape-xld operator only extract circularity be not less than 0.75 circle, take the central point of edge pixel obtain more close toActual micropore profile L1 carries out circularity screening, and judgement of the circularity in (0.9-1) range is qualified, quasi- with edge if unqualifiedIt closes split plot design and is fitted positive round L2;
CN201811316944.4A2018-11-072018-11-07 An intelligent detection system and method for micro-hole shape of fuel injector based on machine visionActiveCN109470162B (en)

Priority Applications (1)

Application NumberPriority DateFiling DateTitle
CN201811316944.4ACN109470162B (en)2018-11-072018-11-07 An intelligent detection system and method for micro-hole shape of fuel injector based on machine vision

Applications Claiming Priority (1)

Application NumberPriority DateFiling DateTitle
CN201811316944.4ACN109470162B (en)2018-11-072018-11-07 An intelligent detection system and method for micro-hole shape of fuel injector based on machine vision

Publications (2)

Publication NumberPublication Date
CN109470162Atrue CN109470162A (en)2019-03-15
CN109470162B CN109470162B (en)2020-11-10

Family

ID=65672050

Family Applications (1)

Application NumberTitlePriority DateFiling Date
CN201811316944.4AActiveCN109470162B (en)2018-11-072018-11-07 An intelligent detection system and method for micro-hole shape of fuel injector based on machine vision

Country Status (1)

CountryLink
CN (1)CN109470162B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication numberPriority datePublication dateAssigneeTitle
CN111504193A (en)*2020-05-072020-08-07山东菏鲁电子科技有限公司Automatic detection method for data line plug size
CN113175879A (en)*2021-03-252021-07-27潮州三环(集团)股份有限公司Method, device, equipment and medium for detecting T surface of sharp-nose ceramic column

Citations (9)

* Cited by examiner, † Cited by third party
Publication numberPriority datePublication dateAssigneeTitle
CN102519382A (en)*2011-12-222012-06-27山东泰山钢铁集团有限公司Fixed point marking method for deformation detection of cold-rolled steel strip
CN103868479A (en)*2014-03-182014-06-18同济大学Automatic nozzle hole structure parameter measuring method
US20140309795A1 (en)*2013-04-122014-10-16Becton, Dickinson And CompanyAutomated set-up for cell sorting
CN105102923A (en)*2013-01-252015-11-25沃思测量技术股份有限公司 Method and device for determining the geometry of structures by means of computed tomography
CN105844282A (en)*2016-06-142016-08-10上海贝特威自动化科技有限公司Method for detecting defects of fuel injection nozzle O-Ring through line scanning camera
CN106895761A (en)*2017-03-172017-06-27柳州日高滤清器有限责任公司A kind of plastic air intake manifold cubing
CN108257171A (en)*2018-01-092018-07-06江苏科技大学Car radar assembling aperture detection method based on light vision
CN108414358A (en)*2018-05-232018-08-17中国原子能科学研究院A kind of device measuring tensile sample elongation after fracture and the contraction percentage of area
CN209342039U (en)*2018-11-072019-09-03上海第二工业大学 A Microhole Shape Detection System for Fuel Injection Nozzles

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication numberPriority datePublication dateAssigneeTitle
CN102519382A (en)*2011-12-222012-06-27山东泰山钢铁集团有限公司Fixed point marking method for deformation detection of cold-rolled steel strip
CN105102923A (en)*2013-01-252015-11-25沃思测量技术股份有限公司 Method and device for determining the geometry of structures by means of computed tomography
US20140309795A1 (en)*2013-04-122014-10-16Becton, Dickinson And CompanyAutomated set-up for cell sorting
CN103868479A (en)*2014-03-182014-06-18同济大学Automatic nozzle hole structure parameter measuring method
CN105844282A (en)*2016-06-142016-08-10上海贝特威自动化科技有限公司Method for detecting defects of fuel injection nozzle O-Ring through line scanning camera
CN106895761A (en)*2017-03-172017-06-27柳州日高滤清器有限责任公司A kind of plastic air intake manifold cubing
CN108257171A (en)*2018-01-092018-07-06江苏科技大学Car radar assembling aperture detection method based on light vision
CN108414358A (en)*2018-05-232018-08-17中国原子能科学研究院A kind of device measuring tensile sample elongation after fracture and the contraction percentage of area
CN209342039U (en)*2018-11-072019-09-03上海第二工业大学 A Microhole Shape Detection System for Fuel Injection Nozzles

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
刘国阳: "基于机器视觉的微小零件尺寸测量技术研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》*

Cited By (2)

* Cited by examiner, † Cited by third party
Publication numberPriority datePublication dateAssigneeTitle
CN111504193A (en)*2020-05-072020-08-07山东菏鲁电子科技有限公司Automatic detection method for data line plug size
CN113175879A (en)*2021-03-252021-07-27潮州三环(集团)股份有限公司Method, device, equipment and medium for detecting T surface of sharp-nose ceramic column

Also Published As

Publication numberPublication date
CN109470162B (en)2020-11-10

Similar Documents

PublicationPublication DateTitle
CN110539109B (en) A robot automatic welding system and method based on monocular and binocular vision
CN108982546A (en)A kind of intelligent robot gluing quality detecting system and method
CN110142785A (en)A kind of crusing robot visual servo method based on target detection
CN105403568B (en)Nozzle detecting system and method
CN102538672B (en)CMOS (complementary metal-oxide-semiconductor)-machine-vision-based component size measuring system and measurement test method
CN102353349B (en) A detection method of an online detection system for the concentricity of a miniature sound film based on machine vision
CN101701799B (en) Micro-drilling machine vision inspection system and its method
CN106885534B (en)Increasing material manufacturing coaxial powder-feeding nozzle test device and method
CN104851144A (en)Inspection robot image processing system and method with self-adaptive polarizer adjusting function
CN112213320B (en)Biscuit detection device and biscuit detection method based on machine vision
CN109470162A (en) An intelligent detection system and method for micro-hole shape of fuel injector based on machine vision
CN109859192A (en)The detection recognition method of round high reflective workpiece
CN117092113A (en)Camera mould welding quality detection device and system thereof
CN112461846B (en)Workpiece defect detection method and device
CN110160760B (en)Nozzle atomization effect detection method
CN110942059A (en)Equipment data real-time reading method and device based on machine vision recognition
CN109631803A (en)Part angle detects adjusting method
CN108280894A (en)Not parking method for inspecting for power equipment
CN115760893A (en)Single droplet particle size and speed measuring method based on nuclear correlation filtering algorithm
CN109596625A (en)Workpiece, defect detection recognition method in charging tray
Xue et al.Method of electronic component location, grasping and inserting based on machine vision
CN115222698A (en) A system and method for feature recognition of powder bed defects in SLM powder spreading process
CN209342039U (en) A Microhole Shape Detection System for Fuel Injection Nozzles
CN111242979B (en)Method for recognizing and tracking splashing of three-dimensional space in laser welding process
CN109632802A (en)Method for detecting and adjusting front and back surfaces of part

Legal Events

DateCodeTitleDescription
PB01Publication
PB01Publication
SE01Entry into force of request for substantive examination
SE01Entry into force of request for substantive examination
GR01Patent grant
GR01Patent grant

[8]ページ先頭

©2009-2025 Movatter.jp