Utility model content
To solve the deficiencies in the prior art, the purpose of this utility model is to provide a kind of online vision-based detection systems of connecting hole
System.
In order to achieve the above objectives, the utility model adopts the following technical solutions:
A kind of online vision detection system of connecting hole, including lathe, workpiece to be measured, Vision imaging system and computer, machine
Bed includes main shaft one, main shaft two and workbench, and workbench is equipped with the clamping device for clamping workpiece to be measured, and main shaft one connects
Cutter is connect, main shaft one connects Vision imaging system for processing workpiece to be measured, main shaft two, and main shaft two is for driving visual imaging
System detection workpiece to be measured, main shaft one and main shaft two are all connected with Machine-Tool Control computer;
Vision imaging system includes knife handle, camera support, industrial digital camera and camera lens;Main shaft two connects phase by knife handle
Machine support, camera support are all connected with electricity for fixed industrial digital camera and camera lens, Machine-Tool Control computer and Vision imaging system
Brain.
As the prioritization scheme for being originally utility model, a kind of online vision detection system of connecting hole above-mentioned, camera support
Including clamping part interconnected and camera fixed part, clamping part connects knife handle, camera fixed part connect industrial digital camera and
Camera lens, industrial digital camera connect computer.
As the prioritization scheme for being originally utility model, a kind of online vision detection system of connecting hole above-mentioned, visual imaging
System further includes light source, and light source is connected to camera lens.
As the prioritization scheme for being originally utility model, a kind of online vision detection system of connecting hole above-mentioned, light source is ring
Shape light source, light source are arranged concentrically with camera lens.
As the prioritization scheme for being originally utility model, a kind of online vision detection system of connecting hole above-mentioned, light source is LED
Light source.
As the prioritization scheme for being originally utility model, a kind of online vision detection system of connecting hole above-mentioned, clamping device
It is any one in mechanical gripping means, hydraulic clamp, starting clamping device or electromagnetic clamp device.
As the prioritization scheme for being originally utility model, a kind of online vision detection system of connecting hole above-mentioned, clamping part
Section is round.
As the prioritization scheme for being originally utility model, a kind of online vision detection system of connecting hole above-mentioned, industrial digital
Camera is CCD camera.
The utility model it is achieved the utility model has the advantages that
The vision detection system that 1. measurement accuracy is reliable and stable, measurement efficiency is high, non-contact, the live adaptability of manufacture is good,
The problems such as solving hand-made hole Operational preparation period length, large labor intensity, low efficiency, not high hole position accuracy.
2. only need to carry out a coordinate system parameters after cutter and Vision imaging system clamping and appraise and decide, avoid repeating to demarcate
Verification;
3. completing processing and detection under the clamped one time of part, the error of repeated clamping generation is avoided;
4. playing processing quality evaluation model based on Element Design establishment of coordinate system, processing quality essential elements of evaluation includes: hole location
Centre deviation, X are to deviation, Y-direction deviation, Z-direction deviation, aperture value, ovality;
5. setting up the connecting hole vision detection system of complete set, include detection path planning program and on-line checking
System;
6. veritified by specific part, compared with three coordinate measurement results pair, vision-based detection precision is higher.
Specific embodiment
The utility model is further described with reference to the accompanying drawing.Following embodiment is only used for clearly illustrating this
The technical solution of utility model, and cannot be used as a limitation the limitation protection scope of the utility model.
Present embodiment discloses a kind of online vision detection systems of connecting hole, including lathe, work to be measured as shown in Figure 1:
Part 4, Vision imaging system 6 and computer, lathe include main shaft 1, main shaft 22 and workbench, and workbench is equipped with for clamping
The clamping device 3 of workpiece 4 to be measured, main shaft 1 connect cutter, and main shaft 1 is connected for processing workpiece 4 to be measured, main shaft 22
Vision imaging system 6, for main shaft 22 for driving Vision imaging system 6 to detect workpiece 4 to be measured, main shaft 1 and main shaft 22 are equal
Connecting machine tool controls computer.
It is as shown in Figure 2: specifically, the Vision imaging system 6 of the present embodiment includes knife handle 61, camera support 62, industrial number
Word camera 63 and camera lens 64;Main shaft 22 connects camera support 62 by knife handle 61, and camera support 62 is for fixing industrial digital phase
Machine 63 and camera lens 64, Machine-Tool Control computer and Vision imaging system 6 are all connected with computer.
The camera support 62 of the present embodiment includes clamping part 621 interconnected and camera fixed part 622, clamping part 621
Knife handle 61 is connected, camera fixed part 622 connects industrial digital camera 63 and camera lens 64, and industrial digital camera 63 connects computer.
In order to promote the image-capturing resolution of industrial digital camera 63, the Vision imaging system 6 of the present embodiment further includes light source
65, light source 65 is connected to camera lens 64, the preferred annular light source of light source 65, and light source 65 and is preferably arranged concentrically with camera lens 64.
The preferred power consumption of light source 65 is lower, the higher LED light source of brightness.
Mechanical gripping means, hydraulic clamp, starting clamping device or electromagnetism can be selected in the clamping device 3 of the present embodiment
Any one in clamping device only needs that the workbench that workpiece 4 to be measured is fixed on to lathe can be reached.
The section of clamping part 621 is circle, and circle can cooperate cutter holder in the prior art to use, such as drill bit cutter holder,
Milling cutter cutter holder etc., the section of retaining part are circles, without specially making new cutter holder, are able to ascend visual imaging in this way
The versatility of system 6.
The preferred CCD camera of industrial digital camera 63 of the present embodiment.
It is as shown in Figure 3: when measurement, to input processing program in Machine-Tool Control computer first, then pass through cutter in main shaft 1
Workpieces processing, the process are not different with parts machining in the prior art.After completing the process, workbench will be completed the process
Workpiece to be measured 4 be moved to main shaft 22, in this process, clamping device 3 does not unclamp.The detection positioning and demarcating time can be saved,
The efficiency and detection accuracy for improving production and detection complete processing and detection under the clamped one time of part, avoid repetition
The error that clamping generates.
Start to detect path planning at this time, detection path planning part major function is the processing using workpiece 4 to be measured
Path and three-dimensional digital-to-analogue generate vision-based detection path planning file, and machining path therein and three-dimensional digital-to-analogue are carried out in main shaft 1
It has been input to Machine-Tool Control computer before processing, has belonged to the conventional means of the prior art.
Detection path planning file in contain connecting hole to be measured theoretical information (hole location, hole to and aperture) and detection it is suitable
Sequence information.On-line checking part is first carried out system parameter calibration operation, determines industrial digital camera in Vision imaging system 6
63 imaging parameters and industrial digital camera 63 are equivalent in one 1 process of main shaft with respect to the pose parameter of 22 end of main shaft
Aim at tool operation.
To guarantee that Vision imaging system 6 can collect each connecting hole image data to be measured, detection path planning text is utilized
Part and system calibrating parameter calculate the corresponding numerical control of machine tools information of each shooting pose of Vision imaging system 6;Then figure is executed
As automatic collection operation, the image of all tested connecting holes is acquired;Finally by being handled acquired image and analyzed,
Obtain connecting hole processing quality evaluation result.
Fig. 4 is the system communication schematic diagram of image automatic collection module.The purpose of this module design is to utilize numerical control code
Controlling lathe drives visual sensor to reach predetermined pose, then automatic collection image.Wherein lathe drives Vision imaging system 6
It reaches predetermined pose and drives cutter progress machining principle identical with main shaft 1.
Specific communicating step are as follows:
(1) it is established using computer where On-machine Test system software as server end, using Machine-Tool Control computer as client
TCP/IP communication connection (prior art)
(2) after numerically-controlled machine tool executes numerical control code arrival designated position, computer, which is sent, reaches command word to server end, together
When lathe can execute the delay operation several seconds;
(3) during lathe delay, control industrial digital camera 63 is to current tested after server end computer receives command word
Connecting hole is taken pictures, and server end computer, which returns to shooting to client computer, after taking pictures terminates command word;
(4) subsequent code is executed after lathe delay, reaches next position, repeat step (2), (3) operation, until
EP (end of program), all detection image acquisitions are completed.
System calibrating: as shown in Fig. 5 (a), wherein Cm is machine tool chief axis coordinate system to coordinate system relationship in the present embodiment system,
Cc indicates camera coordinates system, and Cp is calibration article coordinate system, and Tc-p is 63 coordinate system of industrial digital camera to demarcating article coordinate system
Rigid body translation matrix (4 × 4 homogeneous matrix, rear same), Tp-m is calibration object (being equivalent to workpiece 4 to be measured) coordinate system to lathe
The rigid body translation matrix of two coordinate system of main shaft, Tc-m are the changes of 63 coordinate system of industrial digital camera to 22 coordinate system of machine tool chief axis
Change matrix.The purpose of system calibrating is: the parameter (intrinsic parameter) of Vision imaging system 6 is obtained, to realize the hole under physical size
Position, Pore Diameter Detection;The transformation relation between 63 coordinate system of industrial digital camera and machine tool chief axis coordinate system is obtained, to guarantee camera
The image of tested connecting hole is shot by planning pose.
The imaging parameters of industrial digital camera 63 are demarcated: 63 parameter calibration algorithm of industrial digital camera is flat using Zhang Shi classics
Face standardization, because calibration object needs clamping centering, this system on lathe specially to devise calibration object as shown in Fig. 5 (b),
Calibration object surface by specially treated to facilitate the clamping centering on lathe, and array circular hole be then used for industrial digital camera 63 at
As parameter calibration.It is similar with circular hole feature to be detected to demarcate object circular hole, belongs to marginal surface feature, rather than by illumination poor contrast
Abnormity at edge feature, therefore conventional scaling board image processing method can not adapt to the uncalibrated image of this system.For this purpose, this
Text proposes a kind of local ring belt edge extraction algorithm with Robust Estimation.
As shown in fig. 6, this algorithm flow is as follows: (1) source images after gaussian filtering being carried out binary conversion treatment, analysis two
The pre-determined bit result in the circular hole information acquisition calibration hole in value image;(2) using calibration hole predetermined bit information in source images number
The annular area-of-interest (ROI) that connection bore edges are established on carries out side to the image in ROI using Canny operator [5]
Edge extraction operation obtains initial edge;(3) initial edge profile point is subjected to Robust Estimation screening, then by screening point through Asia
Pixel localization process obtains profile point final position;
Final profile point is subjected to ellipse fitting, completes the positioning of elliptical center.1.2 cameras and machine tool chief axis relationship mark
Determine the transformation relation in Fig. 5 (a) between camera coordinates system and machine tool chief axis coordinate system and is also referred to as trick relationship.In order to guarantee phase function
The image for collecting tested gaging hole needs to control machine tool chief axis and drives camera motion to predetermined position, and machine tool motion information
It calculates and depends on accurate trick relationship.Every uncalibrated image can obtain phase in 63 imaging parameters calibration algorithm of industrial digital camera
The Tc-p information answered, and formal matter is demarcated first in lathe centering, lathe reading when according to every uncalibrated image of shooting can obtain
The accurate information of Tp-m is obtained, the trick relationship Tc-m for needing to solve can be calculated according to formula (1).
Tc-m=Tp-mTc-p formula (1)
Theoretically an image is achieved with trick relationship, but to inhibit possible trick relationship to calculate deviation, right herein
The trick relationship that all images generate carries out Robust Estimation, rejects there are the trick relational matrix of large error, is then averaged
Value.
Vision-based detection numerical control code generates: Vision imaging system 6 can be regarded as a Cutting tool installation manner to lathe knife handle
On, but since the axis of Vision imaging system 6 forms coaxial pass with machine tool chief axis coordinate system Z axis without image of Buddha tool axis is the same
System, therefore the detection system of this paper can not directly use numerical control code when connecting hole machined, and need to recalculate vision inspection
Survey numerical control code.In the case where the axis of Vision imaging system 6 and the law vector to gaging hole are aligned, Vision imaging system 6 is clapped
The image imaging effect in hole to be measured taken the photograph is best, also most beneficial for image procossing, the herein referred to as positive bit image of measured hole.To protect
Blur-free imaging is demonstrate,proved, camera needs and is kept fixed distance d to gaging hole, which is obtained by camera calibration step.Note is to gaging hole hole axle
Vector homogeneous coordinates are Vh=ijk0 TT, are Vc to homogeneous coordinates of the gaging hole hole axle vector under 6 coordinate system of Vision imaging system
=00 10 TT is Ph Ph=Xh Yh Zh 1 to hole heart position homogeneous coordinates of the gaging hole under Element Design coordinate system
TT is 1 TT of Pc=00d to homogeneous coordinates of the gaging hole hole location coordinate under camera coordinates system.Note machine tool chief axis coordinate system arrives
Element Design coordinate system is converted to Tm-h, and entire coordinate converts mathematical expression form are as follows:
Vh=Tm-hTc-mVc formula (2)
Ph=Tm-hTc-mPc formula (3)
In formula: Tm-h matrix is the transformation matrix for needing to solve, and different lathe configurations lead to the component motion solved not
Together.The solution for AC Double swing head lathe, A yaw C turntable lathe and oblique 45 ° of B yaw C turntable lathe has been developed herein
Calculate program.By taking A yaw C turntable lathe as an example, the numerical control of machine tools component for needing to solve is (X, Y, Z, C, A), then Tm-h can be indicated
Are as follows:
Tm h=TXYZTRCTRA T m h=TXYZTRCTRA T m h=TXYZTRCTRA formula (4)
In formula: TRA is the rotational transformation matrix around A axis, and TRC is the rotational transformation matrix around C axis, and TXYZ is translation transformation matrix.
In conjunction with formula (2) and (4), numerical control component (C, A) is solved first, then solves number using formula (3) and (4)
It controls component (X, Y, Z), obtains the motion information of lathe cnc mechanism when camera shoots each connecting hole.Motion information is with numerical control generation
The form of code provides, as shown in Figure 7.
The evaluation of connecting hole processing quality: the evaluation of connecting hole processing quality is in order to obtain to the hole location of gaging hole, Pore Diameter Detection
As a result.The detection image of acquisition is solved using the Accurate translation relationship between each coordinate system in the system obtained
Out each to the theoretical image-forming information of gaging hole, using information interested area division (ROI) on the image, can reduce at image
It manages workload, improve processing speed.For the detection image in each hole, process flow and scaling board circular hole process flow class
Seemingly, it but considers the actually detected environment of connecting hole, is carried out in detection image processing using Deriche operator substitution Canny operator
Edge extracting.Deriche operator can effectively cut down interference of the surface texture to edge extracting of part, guarantee that edge extracting is steady
Qualitative and accuracy.The measured hole ellipse information that image procossing terminates to obtain is and to connect hole machined in two-dimensional image plane
Quality evaluation target is to obtain the processing quality evaluation result under Element Design coordinate system, therefore establish herein as shown in Figure 8
Evaluation model.
Evaluate connecting hole processing quality error using two-dimensional image information: tested connecting hole is curved surface normal direction hole, measured hole
The smaller plane hole that can be regarded as being located in facet in aperture, while camera is normotopia shooting to gaging hole, therefore connecting hole
Facet and the plane of delineation are parallel relations where edge, using this preset parameter of shooting distance, obtain connecting hole in camera
Practical hole location and aperture parameters under coordinate system;Numerical control code calculate step in, shooting each to gaging hole phase machine coordinate system with
The transformation relation of part coordinate system is determining and known, the hole using the transformation relation by measured hole under camera coordinates system
Position, aperture parameters are transformed under part coordinate system, compare measured hole theory hole location and aperture parameters, so that it may obtain measured hole
Processing quality evaluation result.
Software function module brief introduction: this system is based on 2013 platform of Visual Studio, is opened with assembly module thinking
Hair, Fig. 9 are system software module exploded view, be divided into path planning (Figure 10 (a) be software detection path planning preview interface) and
On-line checking (Figure 10 (b) is on-line checking startup interface) two parts.It, will for convenience of the exploitation, maintenance and update of system software
System is independent program by the exploitation of functions of modules difference.By systems parameters document and customized between system module and module
The data file of format realizes information interchange and data transmitting.Operation is detected every time when system detection to be engineered thinking management
Data detect a corresponding detection engineering catalogue every time to save all file datas relevant to this detection.2 systems are fast
Fast scaling method quickly and easily completes proving operation for convenience of operator, and system parameter calibration program (see Figure 10 (c)) utilizes
Image automatic collection mechanism realizes a kind of quick calibrating method in conjunction with image processing module and system parameter computing module: (1)
Calibration numerical control code used in acquisition uncalibrated image is first woven before acquisition uncalibrated image;(2) image is executed using calibration code
Automatic collection operation, acquires all uncalibrated images;(3) it opens camera parameter and demarcates thread, be automatically performed all parameter marks of system
It is fixed.When 3 parallel processing image mechanism systems detect, image automatic collection program can remain operational from the background as stand-alone program,
Acquisition image data is simultaneously saved in detection image file.Online image processing program has an individual thread to be responsible for prison in real time
Amount of images situation of change in detection image file is controlled, once detect that new image input is switched on sub thread to image
Carry out processing analysis, the processing quality evaluation result that processing terminate can in main interface thread real-time update.Figure 10 (d) is online
The working interface of image processing program, program displays source images and processing result image, synchronized update in experiment result list
All testing result.
Testing result and discussion: for the detection accuracy of verification vision detection system, to airframe zero shown in Figure 11
Part has carried out repeatable accuracy verifying, and is compared with the testing result of three coordinate measuring machine (CMM), as a result such as Figure 12,13.
With the verifying of three coordinate measuring engine measurement Comparative result to the aviation part point for having 118 connecting holes shown in Figure 11
It is not measured using VIHAPIS system and three coordinate measuring machine, (three is counted to respective deviation average and standard deviation
The average value and standard value of measurement of coordinates machine testing are respectively 0.0745mm and 0.0299mm, and VIHAPIS system detection is averaged
Value and standard value are respectively 0.0421mm and 0.0223mm), and distribution comparison is carried out to the hole location deviation of two methods, it ties
Fruit is as shown in figure 13.By the duplicate measurements precision of analysis system, and compare three coordinate measuring engine measurement result, it can be seen that this
(repetition measurement deviation is not more than the connecting hole On-machine Test system VIHAPIS accuracy with higher of text research and development
0.01mm), it compares deviation with three coordinate measuring engine measurement result and is respectively less than 0.1mm.Since this paper detection method is in machine original
Position detection, without carrying and clamping centering again, therefore in scene property, stability and the measurement efficiency (detection in every hole of measurement
Time has comprehensive advantage compared with three-coordinate measuring method less than 5s) etc..
The above is only the preferred embodiment of the utility model, it is noted that for the common skill of the art
For art personnel, without deviating from the technical principle of the utility model, several improvement and deformations can also be made, these change
It also should be regarded as the protection scope of the utility model into deformation.