CN104006758A - Automatic detection method for pen refill quality - Google Patents
Automatic detection method for pen refill quality Download PDFInfo
- Publication number
- CN104006758A CN104006758A CN201310060629.0A CN201310060629A CN104006758A CN 104006758 A CN104006758 A CN 104006758A CN 201310060629 A CN201310060629 A CN 201310060629A CN 104006758 A CN104006758 A CN 104006758A
- Authority
- CN
- China
- Prior art keywords
- pen core
- data
- stitching
- point
- core quality
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Landscapes
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention relates to an automatic detection method for pen refill quality. The method comprises: (1), acquiring a digit image of a stitch circle; (2), performing scattering Hough transformation on the acquired image: first of all, randomly selecting N points on the whole image, N being taken from 40 to 80, by respectively taking each point as a reference point, drawing a horizontal line and a vertical line, and then performing the scattering Hough transformation on the point on the two lines; the scattering Hough transformation involves transforming a two-dimensional image domain to a parameter domain on the two lines, expressing a circle center coordinate (x, y) and a radius by three parameters c1, c2 and c3, then (x-c1)<2>+(y-c2)<2>=c3<2>; and since the radius is approximately known, updating the accumulation voters of the corresponding parameter values of the c1 and the c2 only according to pixel points, and obtaining the estimation values of the two parameters of the point; (3), integrating the scattering Hough transformation results of the plurality of points to obtain the estimation results of the circle center and the radius; (4), performing edge detection based on evidence accumulation to obtain data of stitch width; and (5), after slide average processing is carried out, making a final judgment about the pen refill quality according to the results.
Description
Technical field
The invention belongs to stationery manufacture and detection field, particularly a kind of automatic testing method of pen core quality.
Background technology
At present, the automatic pen core quality testing of China's system industry is according to belonging to blank, and a large amount of Zhi Bi enterprises adopts the method for artificial quality testing, inefficiency, and in the growing situation of human cost, enterprise is bearing huge financial burden.
Summary of the invention
The object of the invention is to, for realizing the automatic detection of pen core quality, utilize video camera to take the circle stitching that pen core to be measured draws, thereby proposed a kind of automatic testing method of pen core quality.
There is important application in the fields such as image processing that are identified in of Circle in Digital Images profile, and Hough converts a kind of effective solution thinking to us is provided.The collinearity of pixel in the method detected image is a kind of detection method of overall importance.Concentrate while there is noise spot or noise when known data point, it can suppress to disturb or noise well, and it can also fit to many straight lines by the point set of given data simultaneously.But because the ultimate principle of classical hough conversion is that image space is transformed into parameter space, the precision of conversion is not easy to control, in the time that practical problems is higher to the accuracy requirement of detection of straight lines, the computing time that it is required and storage space also increase thereupon, for justifying the more graph outline of this parameter, calculated amount is very large, substantially can not meet the requirement of real-time.For these problems, the present invention proposes a kind of method of loose some Hough conversion.The method is on the basis of Hough conversion, improves for detecting circular contour.Main improvement has: (1) is in the time of calculation of parameter, choose at random portion's spaced point of image, only on the horizontal line by these points, vertical line, carry out Hough conversion, greatly reduced calculated amount, due to the comprehensive computing of horizontal line, vertical line, ensure the accuracy of result; (2) in the situation that radius is approximate known, the three-dimensional search of traditional center of circle (x, y), radius (r) parameter is reduced to centre point (x, the y) two-dimensional search of radius disturbance, has further reduced calculated amount, make algorithm be able to Project Realization.
Edge is the most basic feature of piece image, is that image intensity changes the most obvious part.Rim detection is the first step of image processing often, and the data volume that can reduce graphical analysis also retains the most information of object border structure simultaneously.There is very important effect in pattern-recognition and field of machine vision.At present, the method for conventional rim detection mainly contains following several: (1) gradient method; (2) Second Order Differential Operator; (3) utilize statistic law, wavelet theory, test of hypothesis etc. to carry out rim detection.There is problem in various degree in the practicality of this several method.The present invention, on the Research foundation of classical theory, has adopted a kind of based on the cumulative edge detection algorithm of evidence.
After the discriminative information extracting, the present invention proposes running mean evidence decision method.The method, in the time choosing mean value, by the level and smooth measure such as boundary belt, subband is set, makes the result after average treatment more accurate.
For achieving the above object, the present invention proposes a kind of automatic testing method of pen core quality, and on principle prototype, done application and realized.Described method can comprise following steps:
(1) gather the round digital picture of stitching;
(2) image collecting is done to loose some Hough conversion: first in entire image, choose at random N point, N value is between 40 ~ 80, respectively taking each point as reference point, and picture level, vertical two lines, then, on these two lines, carry out loose some Hough conversion of this point;
Loose described Hough conversion, is, on these two lines, X-Y scheme image field is transformed to parameter field, and central coordinate of circle (x, y) and radius are expressed as to three parameter c 1, c2, c3, (x-c
1)
2+ (y-c
2)
2=c
3 2; Because radius is approximate known, only upgrade c1 according to pixel, the cumulative voting machine of c2 relevant parameter value, obtains the estimated value of two parameters of this point;
(3) loose some Hough transformation results of above-mentioned some points carried out comprehensively, obtaining the estimated result of the center of circle, radius;
(4) taking the center of circle obtained above, radius is as reference, does the data that obtain stitching width based on the cumulative rim detection of evidence, and using these data as the evidence-based that judges pen core quality;
(5) data of above-mentioned stitching width are carried out after running mean processing, make the final judgement of pen core quality according to result, comprising: broken string, monolateral or normal.
Described step (2) also comprises does statistical average to the estimated value of N point, and computation of mean values and variance, reject the estimated value that exceedes 3 times of variances.
In described step (3), after carrying out comprehensively to loose some Hough transformation results, reject obvious off-limits result, finally, obtain the estimated result of the center of circle, radius.
The step of doing the rim detection cumulative based on evidence in described step (4) comprises: first, take the center of circle as reference point, form to extraradial 360 rays, then, described edge detection method carries out on these 360 rays, from the center of circle, judges whether to reach border according to pixel data, if continuous three times eligible; be judged to and reach stitching edge, the difference of outward flange and inward flange is the width of stitching.
In described step (6), running mean processing is stitching width data to be done in the scope of 30 ° on average; taking the 5 ° of scopes in data point left and right as boundary belt; do not participate in summation; the scope of 5 ° ~ 30 ° is subband; participate in summation; form average and, and do normalized, form the data after running mean.
Described step (6) is after running mean is processed, stitching width data is arranged after first running mean, the measured value that the ink marks of rejecting the stitching joint of causing for system thickens, ink marks hangover etc. located, then, remaining data are judged to the evidence of pen core quality uses by being used as.
The rule of the final judgement in described step (6) comprises: stitching has and interrupts being judged to broken string, and level and smooth data variance after average is excessive to be judged to monolaterally, and other are judged to normally.
The monolateral method that is finally judged as in described step (6) is: if the mean value * thickness factor of stitching width data is less than the poor of the maximal value of stitching width data and minimum value, be judged as monolateral.
Described N value is 50.
The invention has the advantages that, the automatic testing method of pen core quality of the present invention, main thought is to utilize video camera to take the circle stitching that pen core to be measured draws, subsequently with the present invention propose loose some Hough mapping algorithm, based on the cumulative edge detection method of evidence and running mean evidence decision method, finally draw the judgement to pen core quality.
Brief description of the drawings
Fig. 1 is the schematic flow sheet of the automatic testing method of pen core quality of the present invention;
Fig. 2 is the stitching photo of the A pen in one embodiment of the invention;
Fig. 3 is the schematic diagram of the stitching width result after the running mean of A pen in the above embodiment of the present invention;
Fig. 4 is that the automatic testing method of pen core quality of the present invention is judged as normal result demonstration;
Fig. 5 is that the automatic testing method of pen core quality of the present invention is judged as monolateral result demonstration;
Fig. 6 is the result demonstration that the automatic testing method of pen core quality of the present invention is judged as broken string.
Embodiment
Below in conjunction with the drawings and specific embodiments, the present invention is described in further detail:
Essence of the present invention is on the streamline of pen core processing, has realized the automatic fast detecting of pen core quality.The method of a kind of automatic pen core quality testing that the present invention proposes has realized application at present on principle prototype.
As shown in Figure 1, described method comprises following steps:
Step 101), on streamline, use the round digital picture of high definition industry camera collection stitching;
Step 102), the image collecting is done to loose some Hough conversion, this step comprises:
(1) first in entire image, choose at random 1 point, taking this point as reference point, picture level, vertical two lines;
(2) on these two lines, carry out Hough conversion, transform to parameter field by X-Y scheme image field, specifically, exactly center of circle x coordinate, center of circle y coordinate and radius are expressed as to three parameter c 1, c2, c3, (x-c
1)
2+ (y-c
2)
2=c
3 2; Because radius is approximate known, only upgrade c1 according to pixel, the cumulative voting machine of c2 relevant parameter value, obtains the estimated value of two parameters, has also completed Hough conversion;
(3) (1) and (2) is repeated N time to (N can be between 40 ~ 80 value, as long as meet required accuracy requirement), obtain independently estimated result the center of circle and radius N time;
Step 103), to step 102) in the estimated result of N time do statistical average, concrete operation comprises:
(1) calculate average and the variance of N estimated result, depart from the estimated value that average exceedes 3 times of variances and think open country value, rejected;
(2) remaining parameter is done to statistical average, obtain the estimated value of three parameters;
Step 104), taking the center of circle that obtains in above-mentioned steps as reference, do the rim detection cumulative based on evidence; Concrete operations are as follows:
(1) taking the center of circle be first reference point, form to extraradial 360 rays,
(2) from the center of circle, judge whether to reach border according to pixel data, if continuous three times eligible; be judged to and reach stitching edge, the difference of outward flange and inward flange is the width of stitching;
(3) edge detection method described in (2) is carried out on these 360 rays.
Step 105), arrange step 104) in the stitching width data that obtains, comprising:
(1) reject illegal value, the measured value that the ink marks of the stitching joint of causing for system thickens, ink marks hangover etc. located is rejected without exception;
(2) remaining data judge that by being used as the evidence of pen core quality uses.
Step 106) by step 105) in stitching width data in the scope of 30 °, do on average, to form final evidence; Concrete operations are as follows:
(1) selected first point, centered by this data point, the scope that left and right is 5 ° is boundary belt, does not participate in computing, the scope of 5 ° ~ 30 ° is subband, do summation operation obtain data and, first data point divided by data and, form the data after average;
(2) data point reach, repeats the operation of (1), until all data points of finishing dealing with;
Step 107) according to the data after running mean, make the final judgement of pen core quality, comprise normal, broken string and monolateral.Stitching has and interrupts being judged to broken string, and level and smooth data variance after average is excessive to be judged to monolaterally, and other are judged to normally.
In the present embodiment, be that the stitching photo of the one-pen to Fig. 2 is processed judgement, Fig. 3 is the stitching width data after running mean, visible, and thickness is maximum is about 4 with minimum differentiation, and what title showed is mean value 7.98.If the thickness factor is set to 0.3, decision gate limit value=7.98*0.3=2.4<4, so be judged as monolateral; Fig. 4 ~ Fig. 6 is sample and judgment result displays normal, monolateral, three kinds of situations of broken string.
It should be noted last that, above embodiment is only unrestricted in order to technical scheme of the present invention to be described.Although the present invention is had been described in detail with reference to embodiment, those of ordinary skill in the art is to be understood that, technical scheme of the present invention is modified or is equal to replacement, do not depart from the spirit and scope of technical solution of the present invention, it all should be encompassed in the middle of claim scope of the present invention.
Claims (9)
1. an automatic testing method for pen core quality, the step of described method comprises:
(1) gather the round digital picture of stitching;
(2) image collecting is done to loose some Hough conversion: first in entire image, choose at random N point, N value is between 40 ~ 80, respectively taking each point as reference point, and picture level, vertical two lines, then, on these two lines, carry out loose some Hough conversion of this point;
Loose described Hough conversion, is, on these two lines, X-Y scheme image field is transformed to parameter field, and central coordinate of circle (x, y) and radius are expressed as to three parameter c 1, c2, c3, (x-c
1)
2+ (y-c
2)
2=c
3 2; Because radius is approximate known, only upgrade c1 according to pixel, the cumulative voting machine of c2 relevant parameter value, obtains the estimated value of two parameters of this point;
(3) loose some Hough transformation results of above-mentioned some points carried out comprehensively, obtaining the estimated result of the center of circle, radius;
(4) taking the center of circle obtained above, radius is as reference, does the data that obtain stitching width based on the cumulative rim detection of evidence, and using these data as the evidence-based that judges pen core quality;
(5) data of above-mentioned stitching width are carried out after running mean processing, make the final judgement of pen core quality according to result, comprising: broken string, monolateral or normal.
2. the automatic testing method of pen core quality according to claim 1, is characterized in that, described step (2) also comprises does statistical average to the estimated value of N point, and computation of mean values and variance, reject the estimated value that exceedes 3 times of variances.
3. the automatic testing method of pen core quality according to claim 1 and 2, is characterized in that, in described step (3), after carrying out comprehensively to loose some Hough transformation results, reject obvious off-limits result, finally, obtain the estimated result of the center of circle, radius.
4. the automatic testing method of pen core quality according to claim 1, it is characterized in that, the step of doing the rim detection cumulative based on evidence in described step (4) comprises: first, take the center of circle as reference point, form to extraradial 360 rays, then, described edge detection method carries out on these 360 rays, from the center of circle, judge whether to reach border according to pixel data, if continuous three times eligible; be judged to and reach stitching edge, the difference of outward flange and inward flange is the width of stitching.
5. the automatic testing method of pen core quality according to claim 1; it is characterized in that; in described step (5), running mean processing is stitching width data to be done in the scope of 30 ° on average; taking the 5 ° of scopes in data point left and right as boundary belt, do not participate in summation, the scope of 5 ° ~ 30 ° is subband; participate in summation; form average and, and do normalized, form the data after running mean.
6. the automatic testing method of pen core quality according to claim 1 or 5, it is characterized in that, described step (5) is after running mean is processed, stitching width data is arranged after first running mean, the measured value that the ink marks of rejecting the stitching joint of causing for system thickens, ink marks hangover etc. located, then, remaining data are judged to the evidence of pen core quality uses by being used as.
7. the automatic testing method of pen core quality according to claim 1, it is characterized in that, the rule of the final judgement in described step (5) comprises: stitching has and interrupts being judged to broken string, and level and smooth data variance after average is excessive to be judged to monolaterally, and other are judged to normally.
8. according to the automatic testing method of the pen core quality described in claim 1 or 7, it is characterized in that, the monolateral method that is finally judged as in described step (5) is: if the mean value * thickness factor of stitching width data is less than the poor of the maximal value of stitching width data and minimum value, be judged as monolateral.
9. the automatic testing method of pen core quality according to claim 1, is characterized in that, described N value is 50.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310060629.0A CN104006758B (en) | 2013-02-26 | 2013-02-26 | A kind of automatic testing method of pen core quality |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310060629.0A CN104006758B (en) | 2013-02-26 | 2013-02-26 | A kind of automatic testing method of pen core quality |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104006758A true CN104006758A (en) | 2014-08-27 |
| CN104006758B CN104006758B (en) | 2017-07-21 |
Family
ID=51367534
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310060629.0A Expired - Fee Related CN104006758B (en) | 2013-02-26 | 2013-02-26 | A kind of automatic testing method of pen core quality |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104006758B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108362708A (en) * | 2018-02-08 | 2018-08-03 | 上海晨光文具股份有限公司 | Pen core scribing line standardized detection system based on machine vision and method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2005283267A (en) * | 2004-03-29 | 2005-10-13 | Suzuki Motor Corp | Through hole measuring device, method, and program for through hole measurement |
| CN101718528A (en) * | 2009-12-10 | 2010-06-02 | 北京科技大学 | Digital image based rapid solving method of circle parameters |
| CN102034101A (en) * | 2010-10-22 | 2011-04-27 | 广东工业大学 | Method for quickly positioning circular mark in PCB visual detection |
-
2013
- 2013-02-26 CN CN201310060629.0A patent/CN104006758B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2005283267A (en) * | 2004-03-29 | 2005-10-13 | Suzuki Motor Corp | Through hole measuring device, method, and program for through hole measurement |
| CN101718528A (en) * | 2009-12-10 | 2010-06-02 | 北京科技大学 | Digital image based rapid solving method of circle parameters |
| CN102034101A (en) * | 2010-10-22 | 2011-04-27 | 广东工业大学 | Method for quickly positioning circular mark in PCB visual detection |
Non-Patent Citations (2)
| Title |
|---|
| 张显全 等: "一种快速的随机Hough变换圆检测算法", 《计算机工程与应用》, 31 December 2008 (2008-12-31), pages 62 - 64 * |
| 李鑫: "基于Hough变换的测量图像质量评价", 《航空兵器》, 31 August 2008 (2008-08-31), pages 7 - 10 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108362708A (en) * | 2018-02-08 | 2018-08-03 | 上海晨光文具股份有限公司 | Pen core scribing line standardized detection system based on machine vision and method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104006758B (en) | 2017-07-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108921176B (en) | Pointer instrument positioning and identifying method based on machine vision | |
| CN106846344B (en) | A kind of image segmentation optimal identification method based on the complete degree in edge | |
| CN102654902B (en) | Contour vector feature-based embedded real-time image matching method | |
| CN104101600B (en) | Cross Section of CC Billet testing of small cracks method | |
| CN103164692B (en) | A kind of special vehicle instrument automatic identification system based on computer vision and method | |
| CN106529559A (en) | Pointer-type circular multi-dashboard real-time reading identification method | |
| CN106952258A (en) | A kind of bottle mouth defect detection method based on gradient orientation histogram | |
| CN107527028A (en) | Target cell recognition methods, device and terminal | |
| CN106339707B (en) | A kind of gauge pointer image-recognizing method based on symmetric characteristics | |
| CN106157310A (en) | The TFT LCD mura defect inspection method being combined with multichannel based on mixed self-adapting Level Set Models | |
| CN101984916A (en) | Blood vessel diameter measuring method based on digital image processing technology | |
| CN102974551A (en) | Machine vision-based method for detecting and sorting polycrystalline silicon solar energy | |
| CN107292310A (en) | A kind of circular pointer dial plate vision positioning and automatic reading method | |
| CN107941808A (en) | 3D printing Forming Quality detecting system and method based on machine vision | |
| CN105574527A (en) | Quick object detection method based on local feature learning | |
| CN110232379A (en) | A kind of vehicle attitude detection method and system | |
| CN108830856B (en) | GA automatic segmentation method based on time series SD-OCT retina image | |
| CN110232682B (en) | Image-based track foreign matter detection method | |
| CN112330598B (en) | Method, device and storage medium for detecting stiff yarn defects on chemical fiber surface | |
| CN103604809A (en) | Pattern cloth flaw online visual inspection method | |
| CN104846054A (en) | Method for automatically detecting moulds in leucorrhea based on morphological characteristics | |
| CN113240623B (en) | Pavement disease detection method and device | |
| CN115100206B (en) | Printing defect identification method for textile with periodic pattern | |
| CN104778707A (en) | Electrolytic capacitor detecting method for improving general Hough transform | |
| CN116563279B (en) | Measuring switch detection method based on computer vision |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170721 Termination date: 20190226 |