CN101021491A - Automatic decting method and device for wheel hub casting fault based on image understanding - Google Patents

Automatic decting method and device for wheel hub casting fault based on image understanding Download PDF

Info

Publication number
CN101021491A
CN101021491A CN 200610132358 CN200610132358A CN101021491A CN 101021491 A CN101021491 A CN 101021491A CN 200610132358 CN200610132358 CN 200610132358 CN 200610132358 A CN200610132358 A CN 200610132358A CN 101021491 A CN101021491 A CN 101021491A
Authority
CN
China
Prior art keywords
wheel hub
travelling belt
district
workpiece
motor
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
CN 200610132358
Other languages
Chinese (zh)
Other versions
CN100590428C (en
Inventor
黄茜
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.)
South China University of Technology SCUT
Original Assignee
South China University of Technology SCUT
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 South China University of Technology SCUT filed Critical South China University of Technology SCUT
Priority to CN200610132358A priority Critical patent/CN100590428C/en
Publication of CN101021491A publication Critical patent/CN101021491A/en
Application granted granted Critical
Publication of CN100590428C publication Critical patent/CN100590428C/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Analysing Materials By The Use Of Radiation (AREA)

Abstract

The invention discloses an automatic detecting method and device of wheel hub casting defect based on image comprehension. It uses PLC controlling system to control moving and rotating of wheel hub and moving of X light pipe precisely. Parameters of PLC controlling system are set by detecting software of wheel hub casting defect in PC. Utilize X light to shoot images of wheel hub on different positions which are also set by the detecting software. Detecting software adopts detecting algorithm to distinguish wheel hub casting defect automatically. Detecting method contains manual, semi-automatic or automatic detecting program for selecting. The system has high precision and good stability, which improves automatic level of wheel hub detecting and the objectivity of detecting standards greatly.

Description

Wheel hub casting flaw automatic testing method and device based on image understanding
Technical field
The present invention relates to wheel hub casting flaw automatic testing method, the invention still further relates to wheel hub casting flaw automatic detection device based on image understanding based on image understanding.
Background technology
Image understanding is a science of automatic information extraction from image, and the basic fundamental of this subject comprises that image is formed modelling, image is set up and data visualization.For exploitation image understanding based on input picture is explained the artificial intelligence system that makes a policy is an important step (" image understanding ", computer engineering, the 29th volume o. 11th, 23~24,71 pages, in July, 2003).
Wheel hub is wheel rim again, is the important enforcement parts of motor vehicles such as automobile, motorcycle.Vehicle traveling process can produce horizontal and vertical load, follows driving moment, braking moment etc. simultaneously, and wheel hub need bear wherein quite a few load.The fast more requirement to wheel hub of the speed of a motor vehicle is also high more, mainly shows little to the runout requirement in length and breadth, and unbalance amount and moment of inertia also require little, and stiffness and strength requires high, and in a word, it is high that dynamic stability and reliability are wanted.
The structure of wheel hub comprises wheel rim, wheel hub, and spoke three parts, wheel rim is meant the part of being close to tire, wheel hub is meant and is enclosed within the outer field part of axletree that spoke is to be the radial part that is connected to wheel rim by wheel hub.Wheel rim, wheel hub, spoke just began to be integrated by casting technique up to 20 beginnings of the century of 19 end of the centurys.
The wheel hub of motor vehicle has some special requirements in use, at first wants light weight, improves fuel availability, improves acceleration and braking; Next requires good heat dissipation, automobile tyre of running at high speed and ground friction, braking and acceleration etc. produce high temperature, the bad accumulation of heat that causes if dispel the heat, being easy to generate high temperature blows out, braking is weak even malfunctioning etc., the danger that good heat dissipation then can reduce brake system slows down the aging speed of tire, and reduces high temperature and blow out and grow the malfunctioning probability of time brake high temperature; Also can improve vibration damping in addition, make cruising characteristic, corresponding raising driving and ride comfort and security.The wheel hub that aluminium, magnesium alloy are made is all good than steel wheel hub in the performance of above-mentioned each side, therefore, and by being applied in the wheel hub manufacturing industry more and more widely.
Guaranteeing the security of travelling, is the problem that wheel hub at first will be considered in manufacture process.For the quality testing of light-alloy wheel hub, any method all can not be compared with the X ray Non-Destructive Testing.Because detecting in real time, the totally different transversal section cast structure of wheel hub, and complicated geometric configuration, X-ray become the unique selection that guarantees light-alloy wheel hub castability.The inherent vice identification that the image understanding technology is applied to cast article has certain degree of difficulty, mainly be with respect to being for the target of image with casting flaw, change various casting structure and become image background complicated and changeable in the Target Recognition process, and the contour feature that extracts target from the complicated image background fast and accurately is the difficult problem of technical field of image processing always.To being presented at the X-ray wheel hub image on the monitoring screen, prior art is to carry out the identification of defective by human eye, carries out the whether up-to-standard judgement of wheel hub simultaneously.Still unexposed mistake is specifically designed to the image analysis system of wheel hub detection and autocontrol method and the integrated system that cooperates with checkout equipment hardware at present.
Granted publication day is that patent of invention WO 01/23632 A3 " detecting the method for test specimen casting flaw automatically " (Method for automatically detecting casting defects in a test piece) in August 30 calendar year 2001 discloses a kind of method that adopts sequence image tracking and three-dimensional reconstruction technology for detection test specimen casting flaw, the Chinese invention patent CN1140797C on March 3rd, 2004 " internal defect in cast is analyzed recognition device and analysis recognition method thereof automatically " discloses automatic analysis recognition method of a kind of internal defect in cast based on the small echo sorting technique and device, foregoing invention is Foundry Production real-time online internal defect in cast graphical analysis recognition device and analysis recognition method is provided, and overcomes and solved the casting quality detection technique to have prosecutor's eyes fatiguability, poor accuracy, randomness is big, efficient is low, the shortcoming of unreliable grade and problem.But described invention all only is the automatic identification of drawing of rough casting picture, does not relate to the robotization of the whole testing process of cast article that comprises wheel hub as yet.
Granted publication day is that patent of invention WO 2003/085416 A3 " wheel automatic detecting machine structure " (Machinery automatic inspection of wheels) on October 16th, 2003 discloses the mechanism design method that a kind of wheel detects automatically, but the device that this invention only provides wheel to detect does not automatically relate to intelligent identifying system and interactive communication between the connecting gear control system and the mechanism action control method based on the image understanding that are relied in the automatic testing process of wheel.
Summary of the invention
The objective of the invention is provides a kind of wheel hub casting flaw automatic testing method based on image understanding at the product quality detection of wheel hub.
Another object of the present invention provides the pick-up unit of realizing said method.
The technical scheme that realizes the inventive method is as follows: a kind of wheel hub casting flaw automatic testing method based on image understanding, accurately control the rotation of the moving of wheel hub, rotation and X-ray machine with the PLC control system, the parameter of described PLC control system detects software set by the wheel hub casting flaw in the PC, the rotation X-ray machine is taken the image of each position of wheel hub, software set is detected by the wheel hub casting flaw in the PC in described position, and described image is discerned casting flaw automatically by embedded graphical analysis of wheel hub inherent vice and recognizer in the wheel hub casting flaw detection software; Described wheel hub inherent vice graphical analysis recognizer comprises the steps:
(1) from the example storehouse, read the example file parameter;
(2) read in online testing product image, carry out the image pre-service, check whether corresponding typical legend is arranged;
(3), partitioned parameters then is set and detected parameters is set if do not find typical legend; If find typical legend, then from the example storehouse, read in relevant parameters;
(4) press setup parameter partitioned image zone, carry out images match, the output testing result according to existing parameter;
(5) if the wheel hub that is detected belongs to the typical legend that can clearly judge defective, then Flame Image Process parameter and the image with this wheel hub deposits the legend storehouse in, to enlarge the sample storehouse of typical image;
(6) termination routine section;
Described wheel hub casting flaw automatic testing method based on image understanding can be selected automatic mode, semi-automatic pattern and manual mode for use, specifically can detect in the software in wheel hub casting flaw and select; The course of work of described automatic mode is as follows:
(a1) control travelling belt is sent wheel hub into X-ray and is taken the district;
(a2) detect the user and whether send any indication, if any, then finish the automatic inspection line journey, get back to the main interface of PC; If no, then automatically perform next step;
(a3) detect wheel hub automatically and whether entered the shooting district, do not enter and then return step (a2), enter to take to distinguish up to wheel hub and just carry out next step;
(a4) read in camera site and number of times requirement, and initialization current location and number of times;
(a5) judge whether current position and X-ray machine angle mate, as do not match, adjust X-ray machine angle and current location coupling; As having matched, then carry out next step;
(a6) judge that whether the current shooting number of times has reached requirement, as do not have, execution in step (a7), as reach execution in step (a8);
(a7) the control travelling belt rotates wheel hub, continues to take in next position, then execution in step (a9);
(a8) reached requirement as taking number of times, then rotated the angle of X-ray machine, cooperated by travelling belt again and rotate wheel hub, the diverse location of wheel hub can all be photographed;
(a9) when X-ray machine is taken, PC reads in image and detects software by wheel hub casting flaw and carry out analyzing and processing, and detecting the user simultaneously has and manually do not send intervention instruction, if any, finish automatic thread, get back to the main interface of PC;
(a10) wheel hub casting flaw detects software and judges whether detected wheel hub product is qualified automatically, as finds defectively, then product is included into defective class;
(a11) whether all positions of judging wheel hub are all clapped and are over, and as not clapped, judge then whether the number of times counter reaches setting value, if reach setting value, then location counter is added 1, and inferior counter puts 0; If do not reach setting value, then time counter adds 1; Returning step (a6) continues to detect; Finish as taking, then product is included into qualified category, be back to step (a1), carry out the shooting of next wheel hub;
The course of work of described semi-automatic pattern is as follows:
(b1) control travelling belt is sent wheel hub into X-ray and is taken the district;
(b2) detect the user and whether send any indication, if any, then finish the automatic inspection line journey, get back to the main interface of PC; If no, then automatically perform next step;
(b3) detect wheel hub automatically and whether entered the shooting district, do not enter and then return step (b2), enter to take to distinguish up to wheel hub and just carry out next step;
(b4) read in camera site and number of times requirement, and initialization current location and number of times;
(b5) judge whether current position and X-ray machine angle mate, as do not match, adjust X-ray machine angle and current location coupling; As having matched, then carry out next step;
(b6) judge that whether the current shooting number of times has reached requirement, as do not have, execution in step (b7), as reach execution in step (b8);
(b7) the control travelling belt rotates wheel hub, continues to take in next position, then execution in step (b9);
(b8) reached requirement as taking number of times, then rotated the angle of X-ray machine, cooperated by travelling belt again and rotate wheel hub, the diverse location of wheel hub can all be photographed;
(b9) when X-ray machine is taken, PC reads in image and detects software by wheel hub casting flaw and carry out analyzing and processing, and detect the user simultaneously and whether manually send intervention instruction, if any, finish automatic thread, get back to the main interface of PC;
(b10) if do not detect user's intervention instruction, wheel hub casting flaw detects software and judges whether detected wheel hub product is qualified automatically, as finds defectively, and analysis result is shown in main interface, and at this moment main interface is in wait user indicating status;
(b11) user may provide following 4 kinds of instructions:
---1) directly send the indication that stops the wheel hub detection, finish semi-automatic detection thread, and the operation of shut down procedure;
---2) send product is included into indication in the certified products, detected product is included in the certified products; Repeating step (b1) makes next wheel hub enter semi-automatic trace routine to (b11) then;
---3) send product is included into indication in the unacceptable product, detected product is included in the unacceptable product; Repeating step (b1) makes next wheel hub enter semi-automatic trace routine to (b11) then;
---4) instruction of sending " following piece image ", judge whether to have finished whole shooting number of times by PC, if, wait for directly that then the user indicates, return step (b11); If step (b6) is returned in all positions of not having clapped wheel hub, all taken until whole positions of wheel hub and finished, the user sends the instruction of the semi-automatic detection thread of end;
The course of work of described manual mode is as follows:
(c1) control travelling belt is sent wheel hub into X-ray and is taken the district;
(c2) set camera site and number of times, initialization current location and number of times;
(c3) judge whether current position and X-ray machine angle mate, as do not match, adjust X-ray machine angle and current location coupling; As having matched, then carry out next step;
(c4) according to screen display location and shooting count value, judge that whether the current shooting number of times has reached requirement, as do not have, execution in step (c5), as reach execution in step (c6);
(c5) the control travelling belt rotates wheel hub, continues to take in next position, then execution in step (c7);
(c6) reached requirement as taking number of times, then rotated the angle of X-ray machine, cooperated by travelling belt again and rotate wheel hub, the diverse location of wheel hub can all be photographed;
(c7) when X-ray machine is taken, PC reads in image and detects software by wheel hub casting flaw and carry out analyzing and processing, and the result whether product reaches quality requirements is presented on the screen;
(c8) judge that by the user whether detected hub defect surpasses quality standard, as think defective, then is included into defective class by the user with product again;
(c9) wheel hub casting flaw detect all positions that software continues to judge wheel hub whether all bat be over, as not clapped, judge then whether the number of times counter reaches setting value, if reach setting value, then location counter is added 1, and inferior counter puts 0; If do not reach setting value, then time counter adds 1; Returning step (c4) continues to detect; Finish as taking, then product is shown as certified products, and judge finally by the user whether product is qualified, by the manual control product assortment; Program is back to step (c1), waits for the shooting of next wheel hub.
Further preferred embodiment as the inventive method, software in the invention is encrypted design, not with the environment of hardware collaborative work in, the software of copy can not move, be whether wheel hub casting flaw detects PC when detecting software startup and be connected with the PLC hardware control system, if do not detect the PLC hardware control system then close detection software.
The technical scheme that realizes apparatus of the present invention is as follows: based on the wheel hub casting flaw automatic detection device of image understanding, described automatic detection device comprise wheel hub workpiece district to be checked, in and workpiece external form data collection zone, detection zone, unacceptable product mark zone and detect workpiece classification district; Described automatic detection device also need cooperate the wheel hub casting flaw detection software of PC end to use together;
Described wheel hub workpiece district to be checked comprises a motor, a travelling belt and an optoelectronic switch;
Described in and workpiece external form data collection zone comprise plumbous room entrance door, a cylinder, a travelling belt, two motors, switching that is used to control plumbous room entrance door of described motor, one is used to control travelling belt; Also comprise two ccd sensors, one of described ccd sensor is used to discern hub height, and another is used to discern hub diameter, also comprises two optoelectronic switches, one is used for determining that wheel hub has entered this district, and another is used for determining that wheel hub has arrived entrance door place, plumbous room;
Described detection zone comprises two optoelectronic switches, seven motors, two travelling belts, a rotatable X-ray machine, and one of described optoelectronic switch is used to detect the starting point position, and another is used for the control of detection zone; Described motor is respectively and detects travelling belt left side work drive motor, detects the right work drive motor of travelling belt, detects the left and right sides, travelling belt left side motor, detects the left and right sides, travelling belt right side motor, C type arm electric rotating machine, C type arm move up and down motor, starting point positioning conveyor motor; One of described travelling belt is used for workpiece sensing and plays point location, and another is used for workpiece sensing control;
Described unacceptable product mark zone comprises plumbous room outlet portal, two optoelectronic switches, two motors, a travelling belt, whether one of described optoelectronic switch is used for detecting has workpiece to enter the unacceptable product mark zone by detection zone, whether another is used for detecting has workpiece to enter detection workpiece classification district by the unacceptable product mark zone, one in described motor is used for control mark district travelling belt, and another is used for plumbous room outlet portal open and close controlling;
Described detection workpiece classification district comprises: a travelling belt, a motor, an optoelectronic switch;
In the testing process, the control relation of each parts of pick-up unit is as follows:
PC at first detect in and workpiece external form data collection zone have or not workpiece to be checked, if any, start centering district travelling belt motor, trigger two optoelectronic switches in centering district, district to be checked optoelectronic switch work, this moment, workpiece centering district's travelling belt motor and workpiece district to be checked travelling belt motor quit work; After the centering district is used for determining that wheel hub has arrived the optoelectronic switch triggering at entrance door place, plumbous room, the open and close controlling electric motor starting of cylinder and plumbous room entrance door, open plumbous room entrance door, delayed time 20 seconds, treat after wheel hub enters the X-ray check district plumbous room entrance door to be closed, cylinder and plumbous room entrance door open and close controlling motor quit work then; When cylinder starts, start centering district level identification CCD and centering district diameter identification CCD, take the height and the diameter of wheel hub, and send the external physical characteristic picture signal of wheel hub to be checked to model identification and coupling that PC carries out wheel hub;
If detect in and workpiece external form data collection zone do not have workpiece, then directly start centering district level identification CCD and centering district diameter identification CCD, prepare to take by workpiece district to be checked by workpiece district to be checked travelling belt enter into to and the wheel hub of workpiece external form data collection zone; This moment workpiece district to be checked travelling belt motor and workpiece centering district travelling belt motor, and district to be checked optoelectronic switch is all in running order, after workpiece enters workpiece external form data collection zone by district to be checked, trigger the centering district and be used for determining that wheel hub has arrived the optoelectronic switch at entrance door place, plumbous room, start cylinder and plumbous room entrance door open and close controlling motor, open plumbous room entrance door, delayed time 20 seconds, treat after wheel hub enters the X-ray check district plumbous room entrance door to be closed, cylinder quits work then; When cylinder starts, start centering district level identification CCD and centering district diameter identification CCD, take the height and the diameter of wheel hub, and give model identification and the coupling that PC carries out wheel hub the external physical characteristic picture signal of wheel hub to be checked;
Wheel hub has been finished after each action of workpiece external form data collection zone, by centering district travelling belt wheel hub is sent to detection zone; Left and right sides motor is done forward and reverse startup respectively on the left of detecting the left and right sides, travelling belt right side motor and detecting travelling belt, according to the information of hub diameter coupling, clamps the travelling belt centre distance that workpiece transmits about adjustment simultaneously, even left and right sides travelling belt is done spacing in the same way; And start starting point positioning conveyor motor, and detection trigger starting point position optoelectronic switch positions the starting point of wheel hub, and the control wheel hub is to the stroke at X source center; The detection left and right sides, travelling belt right side motor and the left and right sides, detection travelling belt left side motor are just done simultaneously and are started, and adjust the center position of travelling belt, promptly stop after finishing;
When wheel hub is sent to workpiece sensing control zone travelling belt by workpiece sensing starting point positioning conveyor, left and right sides motor is done positive and negative startup once more respectively on the left of detecting the left and right sides, travelling belt right side motor and detecting travelling belt, adjust the spacing of clamping wheel hub, detect the right work drive motor of travelling belt subsequently and detect travelling belt left side work drive motor and just do simultaneously and start, make the detection travelling belt clamp the center that wheel hub is transported to X source, photographing request according to the PC indication, if wheel hub needs rotation, then detect the right work drive motor of travelling belt and do positive and negative startup respectively with detection travelling belt left side work drive motor, promptly two motors are done reverse stepping simultaneously, can clamp angle of wheel hub rotation like this, the anglec of rotation is 0~360 °, and the concrete anglec of rotation detects in the software in wheel hub casting flaw and sets; After finishing the change in location of wheel hub, generally need carry out the shooting of positive and side respectively to wheel hub, at this moment PC will send instruction, the C type arm electric rotating machine that starts X-ray machine is adjusted the angle of X-ray machine, the rotating range of the angle of X-ray machine is 0~60 °, start C type arm then and move up and down motor, adjust the irradiation distance of X-ray, each point of irradiation at X source, all need to adjust the position of wheel hub, at this moment mainly be that wheel hub is rotated, make X-ray can cover each part of wheel hub, repeat above-mentioned action, promptly detect the right work drive motor of travelling belt and do positive and negative startup respectively with detection travelling belt left side work drive motor, two motors are done reverse stepping simultaneously, clamp angle of wheel hub rotation;
In the wheel hub shooting process,, wheel hub is carried out the automatic identification of inherent vice by PC operation detection algorithm.After all positions that need take of wheel hub had all been taken, C type arm electric rotating machine and C type arm moved up and down motor and finish homing action; PC is finished after the automatic identification, starting plumbous room outlet portal open and close controlling motor opens plumbous room outlet portal, to detect the right work drive motor of travelling belt simultaneously and do the forward startup with detection travelling belt left side work drive motor, the wheel hub that has detected is delivered to outlet portal place, plumbous room, detection trigger district optoelectronic switch, start mark zone travelling belt motor simultaneously, triggered mark district optoelectronic switch, plumbous room outlet portal open and close controlling motor is shut plumbous room outlet portal, the work that stops to detect the right work drive motor of travelling belt and detect travelling belt left side work drive motor; In the unacceptable product mark zone unacceptable product is carried out mark, the product that utilizes defective workpiece mark zone travelling belt to have detected is then delivered to and is detected workpiece classification district, triggered mark district optoelectronic switch, notice in and workpiece external form data collection zone carry out the preparation that next wheel hub detects; The wheel hub that startup classification district travelling belt motor utilizes workpiece district to be checked travelling belt to have detected is delivered to the product classification place; Trigger classification district optoelectronic switch, notice classification district's travelling belt motor and mark zone travelling belt motor quit work.
Compared with prior art, the present invention has following technological merit:
(1) the invention provides a kind of automatic control wheel hub and X source relative position and control method and the control device of taking number of times, it makes existing detection system obtain higher lifting, wheel hub is detected realize full process automatization;
(2) communication means and the control program of PC intelligence system provided by the invention and PLC control system are also for accurately each orientation of acquisition wheel hub, the minimum x-ray image of lap have given technique guarantee comprehensively.
(3) automated image analysis recognition methods provided by the invention and system, the user can be freed from heavy visual inspection on the one hand, alleviate working strength of workers, guaranteed the objectivity of quality testing on the other hand, the examination criteria unified by computer utility detects product with identical algorithm, avoided artificial visually examine's subjectivity and randomness;
(4) another characteristics of image analysis processing provided by the invention system are to make more than the rate of accuracy reached to 95% of defect recognition.
System and method of the present invention up to the strong X source of 160KV, provides a kind of technical measures of science for the health care of human body when (5) can make the user utilize the method for Computer Automatic Recognition technology and Long-distance Control suitably to detect away from wheel hub.
Brief description
Fig. 1 is the overall work process flow diagram of expression system of the present invention;
Fig. 2 is a PC intelligence system software general flow chart;
Fig. 3 is the position view of system of the present invention gear train;
Fig. 4 is that the A of Fig. 3 is to synoptic diagram.
Fig. 5 is the control flow chart of autosensing mode underdrive of the present invention mechanism;
The automatic testing process figure of Fig. 6 system of the present invention;
The manual testing process figure of Fig. 7 system of the present invention;
The semi-automatic testing process figure of Fig. 8 system of the present invention.
Fig. 9 is the process flow diagram of automatic analysis of wheel hub inherent vice of the present invention and recognition methods;
Embodiment
The present invention is based on the wheel hub casting flaw automatic testing method and the device of image understanding, adopt the method for software and hardware combining to realize, both can detect surface imperfection, also can detect inherent vice.Software section forms with Fortran or C language compilation on multi-purpose computer.
Below by accompanying drawing method and system of the present invention is described in detail.This should be noted that shape and size in the accompanying drawing only be used for explaining clearer, but not in order to represent strict proportionate relationship.
Fig. 1 is the overall work process flow diagram of apparatus of the present invention.Be provided with the emergency stop button in the device, as needing emergency stop, can cut off the electricity supply after the energized by the emergency stop button.Just often the course of work is as follows:
(1) energized, starting outfit.
(2) the PLC control system is carried out hardware check, and simultaneously the wheel hub casting flaw of PC detects software to carry out initialization be the parameter zero clearing and resets.If break down in arbitrary place, all carry out alarm indication, restart step (1) again after treating manually to fix a breakdown; If do not break down, execution in step (3) then.
(3) check in the lead shield room and have or not workpiece (being wheel hub), if do not have workpiece in the plumbous room, execution in step (4) then; If in the plumbous room workpiece is arranged, then the PLC control system is opened the door 2 that workpiece withdraws from, and startup transmission motor is delivered to district to be detected with workpiece, execution in step (4) after resetting.
(4) open X-ray machine, main interface display appears in PC, enters the operation awaits stage, and the user can select automatic, manual or semi-automatic, and a default operator scheme also can be set in the software.
(5) trace routine of wheel hub reruns, up to receive cease and desist order till.
(6) withdraw from wheel hub casting flaw and detect software systems.
(7) cut off the electricity supply.
Fig. 2 is that PC wheel hub casting flaw detects the software general flow chart.On PC, it is as follows that wheel hub casting flaw of the present invention detects software work:
(1) whether detection hardware exists, and does not exist promptly to stop running software automatically and close software.This is the encryption design of software among the present invention, not with the environment of hardware collaborative work in, the software of copy can not move.After detecting the hardware existence, carry out next step.
(2) transmit hardware check order and initializers system.At this moment following various operations are waited at the PC interface: the image that can read in during park mode in the hard disk carries out various processing, changes corresponding various parameter; Automatically during standby mode, can regulate image parameter, change the parameter of electric machine, can control travelling belt, change the X-ray machine angle, operations such as input and output wheel hub parameter.
(3) press start button after, start operation control automatically.
(4) judge whether to start for the first time, if, open X-ray machine, start travelling belt, wheel hub is sent into X-ray take the district and read in total camera site, always take the parameter of number of times; Initialization current location and shooting number of times parameter.Judge that whether wheel hub enters, as do not enter, continue to wait for,, judge whether current location and X-ray machine angle mate, as do not match, adjust the X-ray machine angle and make it and the current location coupling, carry out X-ray then and take as entering.Clapped when the number of times that needs are taken, then controlled travelling belt and make wheel hub change a camera site,,, continued the diverse location shooting of wheel hub is finished just rotate the angle of X-ray machine if finished predetermined shooting number of times.
(5) computing machine reads in image, stores the image on the internal memory of computing machine, according to the image parameter of having set, carries out Flame Image Process again, the defective in the recognition image, and judge that whether this figure defective surpasses quality standard, as surpass, then this wheel hub is a unacceptable product.
The default setting that wheel hub detects is manual detecting pattern, and whether the systematic parameter in this detecting pattern, each detection position of wheel hub and defective are above standard etc., by artificial judgment.Also can select semi-automatic or autosensing mode, as select semi-automaticly, then detect and finish the back and write down current correlation parameter, for example position and number of times parameter, get back to main interface and wait for some manually operated interventions, for example judge whether defective exceeds standard, whether product is qualified etc.
(6) as selecting automatically, system judges automatically that then whether qualified product is, qualified then automatically product is included into qualified category by control system, otherwise is included into defective class.After a wheel hub detection finished, autotest need not Reparametrization to the wheel hub of follow-up same model, can finish the detection of subsequent product with repeating above-mentioned each detecting pattern automatically.If wheel hub to be checked is with previous to examine hub type different, then before arbitrary in starting manual, semi-automatic or autosensing mode, reset detected parameters, select detecting pattern, detect beginning.
Fig. 3 illustrates integral device and is divided into 5 districts:
1 district is wheel hub workpiece district to be checked;
2 districts be in and workpiece external form data collection zone;
3 districts are detection zone;
4 districts are the unacceptable product mark zone;
5 districts are for detecting workpiece classification district.
The wheel hub conveying control system transmits detected wheel hub by the order by 1 district to 5 district.The course of work is by the driven by motor travelling belt in each district, and workpiece is finally delivered to 5 districts by 1 district.
It below is the note of each code among Fig. 3.
101: workpiece district to be checked travelling belt
102: district to be checked optoelectronic switch
103: workpiece district to be checked travelling belt motor
201: centering district travelling belt
202: centering district optoelectronic switch 1
203: centering district optoelectronic switch 2
204: centering district travelling belt motor
205:208 open and close controlling motor
206: centering district diameter identification CCD
207: centering district level identification CCD
208: plumbous room entrance door
301: workpiece sensing starting point positioning conveyor
302: workpiece sensing control zone travelling belt
303: detect starting point position optoelectronic switch
304: the detection zone optoelectronic switch
305: detect travelling belt left side work drive motor
306: detect the right work drive motor of travelling belt
307: detect the left and right sides, travelling belt left side motor
308: detect the left and right sides, travelling belt right side motor
309:C type arm electric rotating machine
310:C type arm moves up and down motor
311: starting point positioning conveyor motor
The 312:X ray machine
401: defective workpiece mark zone travelling belt
402: mark zone optoelectronic switch 1
403: mark zone optoelectronic switch 2
404: mark zone travelling belt motor
405:406 open and close controlling motor
406: plumbous room outlet portal
501: workpiece district to be checked travelling belt
502: classification district optoelectronic switch
503: classification district travelling belt motor
Fig. 4 be the A of Fig. 3 to synoptic diagram, be used to illustrate the shape of C type arm and the position of motor 305,306,307,307,309 and 310.
Fig. 5 is seen in transmission that workpiece is concrete and action control.
Fig. 5 is the control flow chart of autosensing mode underdrive mechanism.The represented meaning of capitalization before the number in the figure is as follows: the M-motor; The K-optoelectronic switch; The S-CCD sensor.All capitalizations only are used for helping to understand among the figure, and its device is represented by number designation.
At first PC detect in and workpiece external form data collection zone 2 districts have or not workpiece to be checked, if any, start workpiece centering district travelling belt motor 204, after triggering centering district's optoelectronic switch 202 and centering district optoelectronic switch 203,102 work of district to be checked optoelectronic switch, this moment, workpiece centering district's travelling belt motor 204 and workpiece district to be checked travelling belt motor 103 quit work.After centering district optoelectronic switch 203 triggers, cylinder 209 and 208 open and close controlling motors 205 start, and open plumbous room entrance door 208, delay time 20 seconds, treat that wheel hub enters behind 3 districts, X-ray check district and close that cylinder 209 and 208 open and close controlling motors 205 quit work then with 208.When cylinder 209 starts, start centering district level identification CCD 207 and centering district diameter identification CCD206, take the height and the diameter of wheel hub, and give model identification and the coupling that PC carries out wheel hub the external physical characteristic picture signal of wheel hub to be checked.
If detect in and workpiece external form data collection zone 2 districts do not have workpiece, then directly centering district level identification CCD207 and centering district diameter identification CCD206 is started, prepare to take by 1 district, workpiece district to be checked by workpiece district to be checked travelling belt 101 enter into to the wheel hub in workpiece external form data collection zone 2 districts.At this moment workpiece district to be checked travelling belt motor 103 and workpiece centering district travelling belt motor 204, and district to be checked optoelectronic switch 102 is all in running order, after workpiece enters workpiece external form data collection zone 2 districts by 1 district, district to be checked, trigger centering district optoelectronic switch 203, start cylinder 209 and 208 open and close controlling motors 205, open plumbous room entrance door 208, delayed time 20 seconds, treat after wheel hub enters 3 districts, X-ray check district plumbous room entrance door 208 to be closed, cylinder 209 quits work then.When cylinder 209 starts, start centering district level identification CCD 207 and centering district diameter identification CCD206, take the height and the diameter of wheel hub, and give model identification and the coupling that PC carries out wheel hub the external physical characteristic picture signal of wheel hub to be checked.
Wheel hub has been finished after each action in workpiece external form data collection zone 2 districts, by centering district travelling belt 201 wheel hub is sent to detection zone 3 districts.Left and right sides motor 307 is done forward and reverse startup respectively on the left of detecting the left and right sides, travelling belt right side motor 308 and detecting travelling belt, information according to the hub diameter coupling, clamp the travelling belt centre distance that workpiece transmits about adjustment simultaneously, even left and right sides travelling belt is done spacing in the same way.And start starting point positioning conveyor motor 311, and detection trigger starting point position optoelectronic switch 303 positions the starting point of wheel hub, and the control wheel hub is to the stroke at X source center.The detection left and right sides, travelling belt right side motor 308 and the left and right sides, detection travelling belt left side motor 307 are just done simultaneously and are started, and adjust the center position of travelling belt, promptly stop after finishing.
When wheel hub is sent to workpiece sensing control zone travelling belt 302 by workpiece sensing starting point positioning conveyor 301, left and right sides motor 307 is done positive and negative startup once more respectively on the left of detecting the left and right sides, travelling belt right side motor 308 and detecting travelling belt, adjust the spacing of clamping wheel hub, detect the right work drive motor 306 of travelling belt subsequently and detect travelling belt left side work drive motor 305 and just do simultaneously and start, make the detection travelling belt clamp the center that wheel hub is transported to X source, photographing request according to the PC indication, if wheel hub needs rotation, then detect the right work drive motor 306 of travelling belt and do positive and negative startup respectively with detection travelling belt left side work drive motor 305, promptly two motors are done reverse stepping simultaneously, can clamp angle of wheel hub rotation like this, rotation angle range is 0~360 °, the concrete anglec of rotation detects in the software in wheel hub casting flaw and sets for example 30 ° or 45 °.After finishing the change in location of wheel hub, generally need carry out the shooting of positive and side respectively to wheel hub, at this moment PC will send instruction, the C type arm electric rotating machine 309 that starts X-ray machine 312 is adjusted the angle of X-ray, the rotating range of the angle of X-ray machine is 0~60 °, for example 30 °, starts C type arm then and moves up and down motor 310, adjust the irradiation distance of X-ray, for example X source is from workpiece 25cm.Same each point of irradiation at X source, all need to adjust the position of wheel hub, at this moment mainly be that wheel hub is rotated, make X-ray can cover each part of wheel hub, at this moment, repeat above-mentioned action, promptly detect the right work drive motor 306 of travelling belt and do positive and negative startup respectively with detection travelling belt left side work drive motor 305, two motors are done reverse stepping simultaneously, clamp angle of wheel hub rotation.
After all positions that need take of wheel hub had all been taken, PC operation detection algorithm carried out the automatic identification of inherent vice to wheel hub, and at this moment, C type arm electric rotating machine 309 and C type arm move up and down motor 310 and finish homing action.
PC is finished after the automatic identification, starting 406 open and close controlling motors 405 opens plumbous room outlet portal 406, to detect the right work drive motor 306 of travelling belt simultaneously and do the forward startup with detection travelling belt left side work drive motor 305, the wheel hub that has detected is delivered to outlet portal 406 places, plumbous room, detection trigger district optoelectronic switch 304, start mark zone travelling belt motor 404 simultaneously, triggered mark district optoelectronic switch 403,406 open and close controlling motors 405 are shut plumbous room outlet portal 406, the work that stops to detect the right work drive motor 306 of travelling belt and detect travelling belt left side work drive motor 305.4 districts carry out mark with unacceptable product in the unacceptable product mark zone, the product that utilizes defective workpiece mark zone travelling belt 401 to have detected is then delivered to and is detected 5 districts, workpiece classification district, triggered mark district optoelectronic switch 402, notice in and workpiece external form data collection zone 2 districts carry out the preparation of next wheel hub detection.The wheel hub that startup classification district travelling belt motor 503 utilizes workpiece district to be checked travelling belt 501 to have detected is delivered to the product classification place.Trigger classification district optoelectronic switch 502, notice classification district's travelling belt motor 503 and mark zone travelling belt motor 404 quit work.
The automatic testing process figure of Fig. 6 system of the present invention.As shown in Figure 1, selected automatic mode when detecting on the software interface in PC wheel hub casting flaw, whole X-ray wheel hub detection system then begins to enter full-automatic detected state.The implementation of full-automatic detected state is as follows:
(a1) control travelling belt is sent wheel hub into X-ray and is taken the district;
(a2) detect the user and whether send any indication, if any, then finish the automatic inspection line journey, get back to the main interface of PC; If no, then automatically perform next step;
(a3) detect wheel hub automatically and whether entered the shooting district, do not enter and then return step (a2), enter to take to distinguish up to wheel hub and just carry out next step;
(a4) read in camera site and number of times requirement, and initialization current location and number of times;
(a5) judge whether current position and X-ray machine angle mate, as do not match, adjust X-ray machine angle and current location coupling; As having matched, then carry out next step;
(a6) judge that whether the current shooting number of times has reached requirement, as do not have, execution in step (a7), as reach execution in step (a8);
(a7) the control travelling belt rotates wheel hub, continues to take in next position, then execution in step (a9);
(a8) reached requirement as taking number of times, then rotated the angle of X-ray machine, cooperated by travelling belt again and rotate wheel hub, the diverse location of wheel hub can all be photographed;
(a9) when X-ray machine is taken, PC reads in image and detects software by wheel hub casting flaw and carry out analyzing and processing, and detecting the user simultaneously has and manually do not send intervention instruction, if any, finish automatic thread, get back to the main interface of PC;
(a10) wheel hub casting flaw detects software and judges whether detected wheel hub product is qualified automatically, as finds defectively, then product is included into defective class;
(a11) whether all positions of judging wheel hub are all clapped and are over, and as not clapped, judge then whether the number of times counter reaches setting value, if reach setting value, then location counter is added 1, and inferior counter puts 0; If do not reach setting value, then time counter adds 1; Returning step (a6) continues to detect; Finish as taking, then product is included into qualified category, be back to step (a1), carry out the shooting of next wheel hub.
The manual testing process figure of Fig. 7 system of the present invention.Selected manual mode on the PC system interface, it is manual detected state that whole X-ray wheel hub detection system then begins to enter manual intervention.Manually the implementation of detected state is as follows:
(c1) control travelling belt is sent wheel hub into X-ray and is taken the district;
(c2) set camera site and number of times, initialization current location and number of times;
(c3) judge whether current position and X-ray machine angle mate, as do not match, adjust X-ray machine angle and current location coupling; As having matched, then carry out next step;
(c4) according to screen display location and shooting count value, judge that whether the current shooting number of times has reached requirement, as do not have, execution in step (c5), as reach execution in step (c6);
(c5) the control travelling belt rotates wheel hub, continues to take in next position, then execution in step (c7);
(c6) reached requirement as taking number of times, then rotated the angle of X-ray machine, cooperated by travelling belt again and rotate wheel hub, the diverse location of wheel hub can all be photographed;
(c7) when X-ray machine is taken, PC reads in image and detects software by wheel hub casting flaw and carry out analyzing and processing, and the result whether product reaches quality requirements is presented on the screen;
(c8) judge that by the user whether detected hub defect surpasses quality standard, as think defective, then is included into defective class by the user with product again;
(c9) wheel hub casting flaw detect all positions that software continues to judge wheel hub whether all bat be over, as not clapped, judge then whether the number of times counter reaches setting value, if reach setting value, then location counter is added 1, and inferior counter puts 0; If do not reach setting value, then time counter adds 1; Returning step (c4) continues to detect; Finish as taking, then product is shown as certified products, and judge finally by the user whether product is qualified, by the manual control product assortment; Program is back to step (c1), waits for the shooting of next wheel hub.
Fig. 8 is the semi-automatic testing process figure of system of the present invention.Selected semi-automatic pattern on the PC system interface, it is semi-automatic detected state that whole X-ray wheel hub detection system then begins the entering part manual intervention.The implementation of semi-automatic detected state is as follows:
(b1) control travelling belt is sent wheel hub into X-ray and is taken the district;
(b2) detect the user and whether send any indication, if any, then finish the automatic inspection line journey, get back to the main interface of PC; If no, then automatically perform next step;
(b3) detect wheel hub automatically and whether entered the shooting district, do not enter and then return step (b2), enter to take to distinguish up to wheel hub and just carry out next step;
(b4) read in camera site and number of times requirement, and initialization current location and number of times;
(b5) judge whether current position and X-ray machine angle mate, as do not match, adjust X-ray machine angle and current location coupling; As having matched, then carry out next step;
(b6) judge that whether the current shooting number of times has reached requirement, as do not have, execution in step (b7), as reach execution in step (b8);
(b7) the control travelling belt rotates wheel hub, continues to take in next position, then execution in step (b9);
(b8) reached requirement as taking number of times, then rotated the angle of X-ray machine, cooperated by travelling belt again and rotate wheel hub, the diverse location of wheel hub can all be photographed;
(b9) when X-ray machine is taken, PC reads in image and detects software by wheel hub casting flaw and carry out analyzing and processing, and detect the user simultaneously and whether manually send intervention instruction, if any, finish automatic thread, get back to the main interface of PC;
(b10) if do not detect user's intervention instruction, wheel hub casting flaw detects software and judges whether detected wheel hub product is qualified automatically, as finds defectively, and analysis result is shown in main interface, and at this moment main interface is in wait user indicating status;
(b11) user may provide following 4 kinds of instructions:
---1) directly send the indication that stops the wheel hub detection, finish semi-automatic detection thread, and the operation of shut down procedure;
---2) send product is included into indication in the certified products, detected product is included in the certified products; Repeating step (b1) makes next wheel hub enter semi-automatic trace routine to (b11) then;
---3) send product is included into indication in the unacceptable product, detected product is included in the unacceptable product; Repeating step (b1) makes next wheel hub enter semi-automatic trace routine to (b11) then;
---4) instruction of sending " following piece image ", judge whether to have finished whole shooting number of times by PC, if, wait for directly that then the user indicates, return step (b11); If step (b6) is returned in all positions of not having clapped wheel hub, all taken until whole positions of wheel hub and finished, the user sends the instruction of the semi-automatic detection thread of end.
Fig. 9 is the process flow diagram that embodies the automatic analysis of wheel hub inherent vice of the present invention and recognition methods and system handles method.Because the wheel hub quality testing is main according to the wheel hub in the storer that pre-deposits PC various taxonomic defficiency images that may occur and the numerical characteristic storehouse of being set up, without the human eye analysis, directly the product image of online acquisition is handled, so that automatic analysis system can be finished pattern match and identification by relative simple numerical computing with computing machine.Program is read the example file parameter earlier from the example storehouse, read in online testing product image again, carries out the image pre-service, checks whether corresponding typical legend is arranged.If find typical legend, then from the example storehouse, read in relevant parameters, press setup parameter partitioned image zone, detect the output testing result according to existing parameter.If by clearly judging the figure of defective with typical legend matching detection, promptly be can be by the characteristic parameter of typical legend defective, for example defect area, area girth compare etc., differentiate clearly the whether qualified defective patterns of the product of examining, then deposit its parameter and image in the legend storehouse, to enlarge the sample storehouse of typical image.If not, then direct termination routine.At the legend examination phase,, partitioned parameters then is set and detected parameters is set if do not find typical legend.

Claims (3)

1, a kind of wheel hub casting flaw automatic testing method based on image understanding, it is characterized in that: accurately control the rotation of the moving of wheel hub, rotation and X-ray machine with the PLC control system, the parameter of described PLC control system detects software set by the wheel hub casting flaw in the PC, the rotation X-ray machine is taken the image of each position of wheel hub, software set is detected by the wheel hub casting flaw in the PC in described position, and described image is discerned casting flaw automatically by embedded graphical analysis of wheel hub inherent vice and recognizer in the wheel hub casting flaw detection software; Described wheel hub inherent vice graphical analysis recognizer comprises the steps:
(1) from the example storehouse, reads the example file parameter;
(2) read in online testing product image, carry out the image pre-service, check whether corresponding typical legend is arranged;
(3), partitioned parameters then is set and detected parameters is set if do not find typical legend; If find typical legend, then from the example storehouse, read in relevant parameters;
(4) press setup parameter partitioned image zone, carry out images match, the output testing result according to existing parameter;
(5) if the wheel hub that is detected belongs to the typical legend that can clearly judge defective, then Flame Image Process parameter and the image with this wheel hub deposits the legend storehouse in, to enlarge the sample storehouse of typical image;
(6) termination routine section;
Described wheel hub casting flaw automatic testing method based on image understanding can be selected automatic mode, semi-automatic pattern and manual mode for use, specifically can detect in the software in wheel hub casting flaw and select; The course of work of described automatic mode is as follows:
(a1) control travelling belt is sent wheel hub into X-ray and is taken the district;
(a2) detect the user and whether send any indication, if any, then finish the automatic inspection line journey, get back to the main interface of PC; If no, then automatically perform next step;
(a3) detect wheel hub automatically and whether entered the shooting district, do not enter and then return step (a2), enter to take to distinguish up to wheel hub and just carry out next step;
(a4) read in camera site and number of times requirement, and initialization current location and number of times;
(a5) judge whether current position and X-ray machine angle mate, as do not match, adjust X-ray machine angle and current location coupling; As having matched, then carry out next step;
(a6) judge that whether the current shooting number of times has reached requirement, as do not have, execution in step (a7), as reach execution in step (a8);
(a7) the control travelling belt rotates wheel hub, continues to take in next position, then execution in step (a9);
(a8) reached requirement as taking number of times, then rotated the angle of X-ray machine, cooperated by travelling belt again and rotate wheel hub, the diverse location of wheel hub can all be photographed;
(a9) when X-ray machine is taken, PC reads in image and detects software by wheel hub casting flaw and carry out analyzing and processing, and detecting the user simultaneously has and manually do not send intervention instruction, if any, finish automatic thread, get back to the main interface of PC;
(a10) wheel hub casting flaw detects software and judges whether detected wheel hub product is qualified automatically, as finds defectively, then product is included into defective class;
(a11) whether all positions of judging wheel hub are all clapped and are over, and as not clapped, judge then whether the number of times counter reaches setting value, if reach setting value, then location counter is added 1, and inferior counter puts 0; If do not reach setting value, then time counter adds 1; Returning step (a6) continues to detect; Finish as taking, then product is included into qualified category, be back to step (a1), carry out the shooting of next wheel hub;
The course of work of described semi-automatic pattern is as follows:
(b1) control travelling belt is sent wheel hub into X-ray and is taken the district;
(b2) detect the user and whether send any indication, if any, then finish the automatic inspection line journey, get back to the main interface of PC; If no, then automatically perform next step;
(b3) detect wheel hub automatically and whether entered the shooting district, do not enter and then return step (b2), enter to take to distinguish up to wheel hub and just carry out next step;
(b4) read in camera site and number of times requirement, and initialization current location and number of times;
(b5) judge whether current position and X-ray machine angle mate, as do not match, adjust X-ray machine angle and current location coupling; As having matched, then carry out next step;
(b6) judge that whether the current shooting number of times has reached requirement, as do not have, execution in step (b7), as reach execution in step (b8);
(b7) the control travelling belt rotates wheel hub, continues to take in next position, then execution in step (b9);
(b8) reached requirement as taking number of times, then rotated the angle of X-ray machine, cooperated by travelling belt again and rotate wheel hub, the diverse location of wheel hub can all be photographed;
(b9) when X-ray machine is taken, PC reads in image and detects software by wheel hub casting flaw and carry out analyzing and processing, and detect the user simultaneously and whether manually send intervention instruction, if any, finish automatic thread, get back to the main interface of PC;
(b10) if do not detect user's intervention instruction, wheel hub casting flaw detects software and judges whether detected wheel hub product is qualified automatically, as finds defectively, and analysis result is shown in main interface, and at this moment main interface is in wait user indicating status;
(b11) user may provide following 4 kinds of instructions:
---1) directly send the indication that stops the wheel hub detection, finish semi-automatic detection thread, and the operation of shut down procedure;
---2) send product is included into indication in the certified products, detected product is included in the certified products; Repeating step (b1) makes next wheel hub enter semi-automatic trace routine to (b11) then;
---3) send product is included into indication in the unacceptable product, detected product is included in the unacceptable product; Repeating step (b1) makes next wheel hub enter semi-automatic trace routine to (b11) then;
---4) instruction of sending " following piece image ", judge whether to have finished whole shooting number of times by PC, if, wait for directly that then the user indicates, return step (b11); If step (b6) is returned in all positions of not having clapped wheel hub, all taken until whole positions of wheel hub and finished, the user sends the instruction of the semi-automatic detection thread of end;
The course of work of described manual mode is as follows:
(c1) control travelling belt is sent wheel hub into X-ray and is taken the district;
(c2) set camera site and number of times, initialization current location and number of times;
(c3) judge whether current position and X-ray machine angle mate, as do not match, adjust X-ray machine angle and current location coupling; As having matched, then carry out next step;
(c4) according to screen display location and shooting count value, judge that whether the current shooting number of times has reached requirement, as do not have, execution in step (c5), as reach execution in step (c6);
(c5) the control travelling belt rotates wheel hub, continues to take in next position, then execution in step (c7);
(c6) reached requirement as taking number of times, then rotated the angle of X-ray machine, cooperated by travelling belt again and rotate wheel hub, the diverse location of wheel hub can all be photographed;
(c7) when X-ray machine is taken, PC reads in image and detects software by wheel hub casting flaw and carry out analyzing and processing, and the result whether product reaches quality requirements is presented on the screen;
(c8) judge that by the user whether detected hub defect surpasses quality standard, as think defective, then is included into defective class by the user with product again;
(c9) wheel hub casting flaw detect all positions that software continues to judge wheel hub whether all bat be over, as not clapped, judge then whether the number of times counter reaches setting value, if reach setting value, then location counter is added 1, and inferior counter puts 0; If do not reach setting value, then time counter adds 1; Returning step (c4) continues to detect; Finish as taking, then product is shown as certified products, and judge finally by the user whether product is qualified, by the manual control product assortment; Program is back to step (c1), waits for the shooting of next wheel hub.
2, the wheel hub casting flaw automatic testing method based on image understanding according to claim 1, it is characterized in that: detect PC when described wheel hub casting flaw detects software startup and whether be connected, if do not detect the PLC hardware control system then close detection software with the PLC hardware control system.
3, a kind of wheel hub casting flaw automatic detection device based on image understanding of realizing claim 1 or 2 described methods is characterized in that: described automatic detection device comprise wheel hub workpiece district to be checked, in and workpiece external form data collection zone, detection zone, unacceptable product mark zone and detect workpiece classification district; Described automatic detection device also need cooperate the wheel hub casting flaw detection software of PC end to use together;
Described wheel hub workpiece district to be checked comprises a motor, a travelling belt and an optoelectronic switch;
Described in and workpiece external form data collection zone comprise plumbous room entrance door, a cylinder, a travelling belt, two motors, switching that is used to control plumbous room entrance door of described motor, one is used to control travelling belt; Also comprise two ccd sensors, one of described ccd sensor is used to discern hub height, and another is used to discern hub diameter, also comprises two optoelectronic switches, one is used for determining that wheel hub has entered this district, and another is used for determining that wheel hub has arrived entrance door place, plumbous room;
Described detection zone comprises two optoelectronic switches, seven motors, two travelling belts, a rotatable X-ray machine, and one of described optoelectronic switch is used to detect the starting point position, and another is used for the control of detection zone; Described motor is respectively and detects travelling belt left side work drive motor, detects the right work drive motor of travelling belt, detects the left and right sides, travelling belt left side motor, detects the left and right sides, travelling belt right side motor, C type arm electric rotating machine, C type arm move up and down motor, starting point positioning conveyor motor; One of described travelling belt is used for workpiece sensing and plays point location, and another is used for workpiece sensing control;
Described unacceptable product mark zone comprises plumbous room outlet portal, two optoelectronic switches, two motors, a travelling belt, whether one of described optoelectronic switch is used for detecting has workpiece to enter the unacceptable product mark zone by detection zone, whether another is used for detecting has workpiece to enter detection workpiece classification district by the unacceptable product mark zone, one in described motor is used for control mark district travelling belt, and another is used for plumbous room outlet portal open and close controlling;
Described detection workpiece classification district comprises: a travelling belt, a motor, an optoelectronic switch;
In the testing process, the control relation of each parts of pick-up unit is as follows:
PC at first detect in and workpiece external form data collection zone have or not workpiece to be checked, if any, start centering district travelling belt motor, trigger two optoelectronic switches in centering district, district to be checked optoelectronic switch work, this moment, workpiece centering district's travelling belt motor and workpiece district to be checked travelling belt motor quit work; After the centering district is used for determining that wheel hub has arrived the optoelectronic switch triggering at entrance door place, plumbous room, the open and close controlling electric motor starting of cylinder and plumbous room entrance door, open plumbous room entrance door, delayed time 20 seconds, treat after wheel hub enters the X-ray check district plumbous room entrance door to be closed, cylinder and plumbous room entrance door open and close controlling motor quit work then; When cylinder starts, start centering district level identification CCD and centering district diameter identification CCD, take the height and the diameter of wheel hub, and send the external physical characteristic picture signal of wheel hub to be checked to model identification and coupling that PC carries out wheel hub;
If detect in and workpiece external form data collection zone do not have workpiece, then directly start centering district level identification CCD and centering district diameter identification CCD, prepare to take by workpiece district to be checked by workpiece district to be checked travelling belt enter into to and the wheel hub of workpiece external form data collection zone; This moment workpiece district to be checked travelling belt motor and workpiece centering district travelling belt motor, and district to be checked optoelectronic switch is all in running order, after workpiece enters workpiece external form data collection zone by district to be checked, trigger the centering district and be used for determining that wheel hub has arrived the optoelectronic switch at entrance door place, plumbous room, start cylinder and plumbous room entrance door open and close controlling motor, open plumbous room entrance door, delayed time 20 seconds, treat after wheel hub enters the X-ray check district plumbous room entrance door to be closed, cylinder quits work then; When cylinder starts, start centering district level identification CCD and centering district diameter identification CCD, take the height and the diameter of wheel hub, and give model identification and the coupling that PC carries out wheel hub the external physical characteristic picture signal of wheel hub to be checked;
Wheel hub has been finished after each action of workpiece external form data collection zone, by centering district travelling belt wheel hub is sent to detection zone; Left and right sides motor is done forward and reverse startup respectively on the left of detecting the left and right sides, travelling belt right side motor and detecting travelling belt, according to the information of hub diameter coupling, clamps the travelling belt centre distance that workpiece transmits about adjustment simultaneously, even left and right sides travelling belt is done spacing in the same way; And start starting point positioning conveyor motor, and detection trigger starting point position optoelectronic switch positions the starting point of wheel hub, and the control wheel hub is to the stroke at X source center; The detection left and right sides, travelling belt right side motor and the left and right sides, detection travelling belt left side motor are just done simultaneously and are started, and adjust the center position of travelling belt, promptly stop after finishing;
When wheel hub is sent to workpiece sensing control zone travelling belt by workpiece sensing starting point positioning conveyor, left and right sides motor is done positive and negative startup once more respectively on the left of detecting the left and right sides, travelling belt right side motor and detecting travelling belt, adjust the spacing of clamping wheel hub, detect the right work drive motor of travelling belt subsequently and detect travelling belt left side work drive motor and just do simultaneously and start, make the detection travelling belt clamp the center that wheel hub is transported to X source, photographing request according to the PC indication, if wheel hub needs rotation, then detect the right work drive motor of travelling belt and do positive and negative startup respectively with detection travelling belt left side work drive motor, promptly two motors are done reverse stepping simultaneously, can clamp angle of wheel hub rotation like this, the anglec of rotation is 0~360 °, and the concrete anglec of rotation detects in the software in wheel hub casting flaw and sets; After finishing the change in location of wheel hub, generally need carry out the shooting of positive and side respectively to wheel hub, at this moment PC will send instruction, the C type arm electric rotating machine that starts X-ray machine is adjusted the angle of X-ray machine, the rotating range of the angle of X-ray machine is 0~60 °, start C type arm then and move up and down motor, adjust the irradiation distance of X-ray, each point of irradiation at X source, all need to adjust the position of wheel hub, at this moment mainly be that wheel hub is rotated, make X-ray can cover each part of wheel hub, repeat above-mentioned action, promptly detect the right work drive motor of travelling belt and do positive and negative startup respectively with detection travelling belt left side work drive motor, two motors are done reverse stepping simultaneously, clamp angle of wheel hub rotation;
In the wheel hub shooting process, by PC operation detection algorithm, wheel hub is carried out the automatic identification of inherent vice, after all positions that need take of wheel hub had all been taken, C type arm electric rotating machine and C type arm moved up and down motor and finish homing action;
PC is finished after the automatic identification, starting plumbous room outlet portal open and close controlling motor opens plumbous room outlet portal, to detect the right work drive motor of travelling belt simultaneously and do the forward startup with detection travelling belt left side work drive motor, the wheel hub that has detected is delivered to outlet portal place, plumbous room, detection trigger district optoelectronic switch, start mark zone travelling belt motor simultaneously, triggered mark district optoelectronic switch, plumbous room outlet portal open and close controlling motor is shut plumbous room outlet portal, the work that stops to detect the right work drive motor of travelling belt and detect travelling belt left side work drive motor; In the unacceptable product mark zone unacceptable product is carried out mark, the product that utilizes defective workpiece mark zone travelling belt to have detected is then delivered to and is detected workpiece classification district, triggered mark district optoelectronic switch, notice in and workpiece external form data collection zone carry out the preparation that next wheel hub detects; The wheel hub that startup classification district travelling belt motor utilizes workpiece district to be checked travelling belt to have detected is delivered to the product classification place; Trigger classification district optoelectronic switch, notice classification district's travelling belt motor and mark zone travelling belt motor quit work.
CN200610132358A 2006-12-28 2006-12-28 Automatic decting method and device for wheel hub casting fault based on image understanding Expired - Fee Related CN100590428C (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN200610132358A CN100590428C (en) 2006-12-28 2006-12-28 Automatic decting method and device for wheel hub casting fault based on image understanding

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN200610132358A CN100590428C (en) 2006-12-28 2006-12-28 Automatic decting method and device for wheel hub casting fault based on image understanding

Publications (2)

Publication Number Publication Date
CN101021491A true CN101021491A (en) 2007-08-22
CN100590428C CN100590428C (en) 2010-02-17

Family

ID=38709325

Family Applications (1)

Application Number Title Priority Date Filing Date
CN200610132358A Expired - Fee Related CN100590428C (en) 2006-12-28 2006-12-28 Automatic decting method and device for wheel hub casting fault based on image understanding

Country Status (1)

Country Link
CN (1) CN100590428C (en)

Cited By (25)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101539533B (en) * 2009-03-11 2011-06-22 华南理工大学 Automatic detecting device and method for battery internal defect
CN102506776A (en) * 2011-11-23 2012-06-20 丹东奥龙射线仪器有限公司 X ray hub automatic detecting device and control method thereof
CN102589441A (en) * 2012-01-11 2012-07-18 华中科技大学 Fully-automatic nondestructive measurement system and measurement method for phenotype parameters of potted rice
CN102928444A (en) * 2012-11-06 2013-02-13 信阳同合车轮有限公司 X-ray flaw detection continuous detection and real-time imaging device
CN103090790A (en) * 2012-12-21 2013-05-08 宁波赛恩斯智能科技有限公司 Automatic identification device and automatic identification method for hub
CN103604461A (en) * 2013-12-05 2014-02-26 丹东奥龙射线仪器集团有限公司 X-ray piston automatic detecting device
CN104237261A (en) * 2014-09-25 2014-12-24 陈丽华 Positioning device for X-ray nondestructive flaw detection of wheel hub
CN104923759A (en) * 2015-07-17 2015-09-23 昆山莱捷有色金属有限公司 Product printing device
CN105004740A (en) * 2015-08-14 2015-10-28 丹东奥龙射线仪器集团有限公司 X-ray multi-station detection device adopting automatic closed-loop control
CN105095581A (en) * 2015-07-22 2015-11-25 华南理工大学 Generation method for defect images in casting shrinkage
CN105118044A (en) * 2015-06-16 2015-12-02 华南理工大学 Method for automatically detecting defects of wheel-shaped cast product
CN106124534A (en) * 2016-09-17 2016-11-16 天津立中汽车铝合金配件有限公司 A kind of x-ray examination equipment of hub for vehicle wheel
CN106423894A (en) * 2016-09-20 2017-02-22 图灵视控(北京)科技有限公司 Automatic wheel hub appearance sorting device, sorting method and sorting system based on machine vision
CN106643566A (en) * 2016-09-30 2017-05-10 华南理工大学 Method of automatically measuring initial angle of spokes of wheel hub
CN106725556A (en) * 2017-01-18 2017-05-31 中国科学院深圳先进技术研究院 X-ray machine, X-ray machine pivot method of adjustment and device
CN108132262A (en) * 2016-11-30 2018-06-08 住友化学株式会社 The manufacturing method of flaw detection apparatus, defect detecting method and membrane winding body
CN108153229A (en) * 2017-12-31 2018-06-12 芜湖哈特机器人产业技术研究院有限公司 Product automatic identification alignment system and method
CN109766391A (en) * 2019-01-21 2019-05-17 武汉裕量信息科技有限公司 Detection system, detection method and computer-readable medium
CN111610205A (en) * 2020-06-12 2020-09-01 上海鸢安智能科技有限公司 X-ray image defect detection device for metal parts
CN113421262A (en) * 2021-08-23 2021-09-21 深圳市信润富联数字科技有限公司 Hub defect detection method and device, electronic equipment and storage medium
CN113470018A (en) * 2021-09-01 2021-10-01 深圳市信润富联数字科技有限公司 Hub defect identification method, electronic device, device and readable storage medium
US11322795B2 (en) 2016-11-30 2022-05-03 Sumitomo Chemical Company, Limited Method for producing separator roll
CN114755178A (en) * 2022-03-03 2022-07-15 燕山大学 Automatic detection device for hub surface defects
CN115494813A (en) * 2022-11-16 2022-12-20 优铸科技(北京)有限公司 Domain control method and domain controller for casting work part
CN117347394A (en) * 2023-11-10 2024-01-05 重庆日联科技有限公司 Method for improving detection efficiency of hub nondestructive X-ray detection system

Cited By (36)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101539533B (en) * 2009-03-11 2011-06-22 华南理工大学 Automatic detecting device and method for battery internal defect
CN102506776A (en) * 2011-11-23 2012-06-20 丹东奥龙射线仪器有限公司 X ray hub automatic detecting device and control method thereof
CN102589441A (en) * 2012-01-11 2012-07-18 华中科技大学 Fully-automatic nondestructive measurement system and measurement method for phenotype parameters of potted rice
CN102928444A (en) * 2012-11-06 2013-02-13 信阳同合车轮有限公司 X-ray flaw detection continuous detection and real-time imaging device
CN103090790B (en) * 2012-12-21 2016-05-04 宁波赛恩斯智能科技有限公司 Wheel hub automatic identifying method
CN103090790A (en) * 2012-12-21 2013-05-08 宁波赛恩斯智能科技有限公司 Automatic identification device and automatic identification method for hub
CN103604461A (en) * 2013-12-05 2014-02-26 丹东奥龙射线仪器集团有限公司 X-ray piston automatic detecting device
CN104237261A (en) * 2014-09-25 2014-12-24 陈丽华 Positioning device for X-ray nondestructive flaw detection of wheel hub
CN105118044A (en) * 2015-06-16 2015-12-02 华南理工大学 Method for automatically detecting defects of wheel-shaped cast product
US10803573B2 (en) 2015-06-16 2020-10-13 South China University Of Technology Method for automated detection of defects in cast wheel products
CN105118044B (en) * 2015-06-16 2017-11-07 华南理工大学 A kind of wheel shape cast article defect automatic testing method
CN104923759A (en) * 2015-07-17 2015-09-23 昆山莱捷有色金属有限公司 Product printing device
CN105095581A (en) * 2015-07-22 2015-11-25 华南理工大学 Generation method for defect images in casting shrinkage
CN105095581B (en) * 2015-07-22 2018-09-14 华南理工大学 A kind of casting shrinking hole defect image generation method
CN105004740A (en) * 2015-08-14 2015-10-28 丹东奥龙射线仪器集团有限公司 X-ray multi-station detection device adopting automatic closed-loop control
CN105004740B (en) * 2015-08-14 2017-09-12 丹东奥龙射线仪器集团有限公司 X-ray closed-loop automatic control Optimum Design device
CN106124534A (en) * 2016-09-17 2016-11-16 天津立中汽车铝合金配件有限公司 A kind of x-ray examination equipment of hub for vehicle wheel
CN106423894A (en) * 2016-09-20 2017-02-22 图灵视控(北京)科技有限公司 Automatic wheel hub appearance sorting device, sorting method and sorting system based on machine vision
CN106423894B (en) * 2016-09-20 2019-01-15 苏州香农智能科技有限公司 Wheel hub appearance automatic sorting device, method for sorting and sorting system based on machine vision
CN106643566A (en) * 2016-09-30 2017-05-10 华南理工大学 Method of automatically measuring initial angle of spokes of wheel hub
CN106643566B (en) * 2016-09-30 2018-06-19 华南理工大学 A kind of method of automatic measurement wheel hub spoke initial angle
CN108132262A (en) * 2016-11-30 2018-06-08 住友化学株式会社 The manufacturing method of flaw detection apparatus, defect detecting method and membrane winding body
US11322795B2 (en) 2016-11-30 2022-05-03 Sumitomo Chemical Company, Limited Method for producing separator roll
CN106725556B (en) * 2017-01-18 2023-08-08 中国科学院深圳先进技术研究院 X-ray machine, X-ray machine rotation center adjusting method and device
CN106725556A (en) * 2017-01-18 2017-05-31 中国科学院深圳先进技术研究院 X-ray machine, X-ray machine pivot method of adjustment and device
CN108153229A (en) * 2017-12-31 2018-06-12 芜湖哈特机器人产业技术研究院有限公司 Product automatic identification alignment system and method
CN109766391A (en) * 2019-01-21 2019-05-17 武汉裕量信息科技有限公司 Detection system, detection method and computer-readable medium
CN109766391B (en) * 2019-01-21 2024-01-23 武汉易万科技有限公司 Detection system, detection method, and computer-readable medium
CN111610205A (en) * 2020-06-12 2020-09-01 上海鸢安智能科技有限公司 X-ray image defect detection device for metal parts
CN113421262A (en) * 2021-08-23 2021-09-21 深圳市信润富联数字科技有限公司 Hub defect detection method and device, electronic equipment and storage medium
CN113470018B (en) * 2021-09-01 2021-12-21 深圳市信润富联数字科技有限公司 Hub defect identification method, electronic device, device and readable storage medium
CN113470018A (en) * 2021-09-01 2021-10-01 深圳市信润富联数字科技有限公司 Hub defect identification method, electronic device, device and readable storage medium
CN114755178A (en) * 2022-03-03 2022-07-15 燕山大学 Automatic detection device for hub surface defects
CN114755178B (en) * 2022-03-03 2024-02-02 燕山大学 Automatic detection device for surface defects of hub
CN115494813A (en) * 2022-11-16 2022-12-20 优铸科技(北京)有限公司 Domain control method and domain controller for casting work part
CN117347394A (en) * 2023-11-10 2024-01-05 重庆日联科技有限公司 Method for improving detection efficiency of hub nondestructive X-ray detection system

Also Published As

Publication number Publication date
CN100590428C (en) 2010-02-17

Similar Documents

Publication Publication Date Title
CN100590428C (en) Automatic decting method and device for wheel hub casting fault based on image understanding
CN201034948Y (en) Nave casting flaw automatic detection device based on image comprehention
CN105817430B (en) Product inspection method based on machine vision
CN105548185A (en) Automobile wheel hub screw hole recognition method based on machine vision and shielding method and system
CN106652551A (en) Parking stall detection method and device
CN104063873A (en) Shaft sleeve part surface defect on-line detection method based on compressed sensing
CN101509766B (en) On-line detecting method for spring end plane angle by machine vision system
CN107273802B (en) Method and device for detecting fault of brake shoe drill rod ring of railway train
CN101532926A (en) On-line vision detecting system for automatic impact specimen processing device and image processing method thereof
CN103528523A (en) Thread detection method and system based on three-dimensional modeling
CN102928431A (en) Device for automatically grading pearls on line according to size and shape on basis of monocular multi-view machine vision
CN112669262B (en) Motor axle vibration abnormity detection and prediction system and method
JP2020042794A (en) Self-driving vehicle visual sensing method, self-driving vehicle visual sensing device, control instrument and computer readable storage medium
CN106706656A (en) Machine vision-based zipper detection device and method
CN101762241A (en) Tyre detecting device and detecting method based on line laser and single camera
CN106353336A (en) Lens coating automatic detection system
CN106546185A (en) A kind of profile quality determining method based on Machine Vision Detection
CN204440490U (en) A kind of Model For The Bush-axle Type Parts surface imperfection on-line measuring device based on compressed sensing
CN207081666U (en) A kind of zipper detecting device based on machine vision
CN103575483B (en) The detection method of hand brake valve on-line performance Detecting data control system
CN113189005A (en) Portable surface defect integrated detection device and surface defect automatic detection method
WO2021030322A1 (en) System and method of object detection using ai deep learning models
CN106643566B (en) A kind of method of automatic measurement wheel hub spoke initial angle
CN203479499U (en) Detection equipment for headlamp assembly of automobile
CN116008301A (en) Method for detecting welding quality of vehicle body plate

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
C17 Cessation of patent right
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20100217

Termination date: 20121228