CN109949291A - A kind of defect inspection method of Cast Aluminum Auto-parts Abroad radioscopic image - Google Patents
A kind of defect inspection method of Cast Aluminum Auto-parts Abroad radioscopic image Download PDFInfo
- Publication number
- CN109949291A CN109949291A CN201910209706.1A CN201910209706A CN109949291A CN 109949291 A CN109949291 A CN 109949291A CN 201910209706 A CN201910209706 A CN 201910209706A CN 109949291 A CN109949291 A CN 109949291A
- Authority
- CN
- China
- Prior art keywords
- defect
- cast aluminum
- profile
- area
- inspection method
- 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.)
- Pending
Links
Abstract
The invention discloses a kind of defect inspection methods of Cast Aluminum Auto-parts Abroad radioscopic image, comprising the following steps: (1) in Cast Aluminum Auto-parts Abroad moving process, acquisition module passes through X ray sensor signal acquisition part picture and is transmitted to image pre-processing module；(2) image pre-processing module pre-processes picture；(3) the pretreated gray scale picture of receiving step (2) carries out morphological image process, and the difference of normal component and defect is searched out to come；(4) Gaussian noise and environment profile are filtered out；(5) shunting that the region searched out in step (3) is realized after the filtering out of step (4) by analyzing and determining to defect profile, is merged to obtain defect profile to controlled imperfections opsition dependent relationship；(6) defect analysis and statistical module count its defect parameters according to step (5) fused defect profile, and generate quality evaluation table；Detection method of the invention can effectively be detected and controlled the quality of product.
Description
Technical field
The present invention relates to casting defect detection technique fields, more particularly, to a kind of lacking for Cast Aluminum Auto-parts Abroad radioscopic image
Fall into detection method.
Background technique
With the development of automotive light weight technology technology, more and more aluminium alloy castings are used as the key components and parts of automobile.
In casting process, some casting flaws are inevitably resulted from, product quality is influenced.Internal defect in cast detection is casting
An important link in production process.Discovery defect can find faulty goods early early, save time and cost.If
Internal flaw is not detected or internal flaw is not detected, these defects may result in critical mechanical component failure.In order to keep away
Exempt from the influence of human-body fatigue, improve detection accuracy, intelligent checking system plays an important role in the production line.
Modern intelligence defects detection is mostly based on vision-based detection, but since visual light can only reach Cast Aluminum Auto-parts Abroad
Surface can not detect the internal flaw of part.And X-ray has penetrability, has to the substance of different densities and different penetrates energy
Power.And in the prior art, there are no the device and method for carrying out the relevant detection and analysis of X-ray to Cast Aluminum Auto-parts Abroad defect.
Summary of the invention
Technical problems to be solved: the object of the present invention is to provide a kind of defects detections of Cast Aluminum Auto-parts Abroad radioscopic image
Method is based on image processing techniques, is identified by the difference of defect and normal configuration in Cast Aluminum Auto-parts Abroad radioscopic image scarce
It falls into, and the defect that will test is counted, Cast Aluminum Auto-parts Abroad criteria of quality evaluation is constructed according to statistical value, to effectively detect
With control product quality.
A kind of technical solution: defect inspection method of Cast Aluminum Auto-parts Abroad radioscopic image, comprising the following steps:
(1) in Cast Aluminum Auto-parts Abroad moving process, acquisition module is transmitted by X ray sensor signal acquisition part picture
To image pre-processing module；
(2) image pre-processing module pre-processes picture；
(3) the pretreated gray scale picture of defect recognition module receiving step (2), and morphological image process is carried out, it will be normal
The difference of component and defect, which searches out, to be come, and labeled as pixel 255(white), remaining is labeled as 0(black)；
(4) it by given threshold, filters out Gaussian noise and environment profile to obtain potential silhouette；
(5) profile after the filtering out of step (4) is analyzed, if profile is not present, continues the signal acquisition of next part；
If profile exists and contour area is greater than the greatest drawback area of setting, defect is excessive not to be available, which should discard；
If contour area is less than the greatest drawback area of setting, these discrete defect profiles are merged by positional relationship；
(6) defect detected to step (4) counts, and generates quality evaluation table according to the defect parameters of statistics, and
Distribute different application scenarios and life cycle；Quality is preferably used for the scene of high-precision high quality demand, and quality is slightly secondary
It can be used for requiring lower scene；If quality is too poor, the aluminum casting is directly discarded.
Preferably, specifically merging standard in the step (5) has:
1. then being merged when defect profile intersects；
2. when defect profile is in same level region or vertical area, and two defect profiles distance be less than threshold X _ min or
Person Y_min, then merge.
Preferably, Gaussian noise and environment profile are filtered out into the method for obtaining potential silhouette in the step (4) are as follows:
It filters out defect area is too small and too big, the too small Gaussian noise for very little of area, ring of the area too greatly outside model
Border profile is not belonging to model internal flaw, to obtain potential silhouette.
Preferably, in the step (2), pretreatment includes local binarization, and local binarization passes through background and feature
Gray difference, background and prospect are distinguished by way of two-value.
Preferably, in the step (2), pretreatment includes median filtering, and median filtering is for keeping the sharp of signal
Variation and elimination impulsive noise.
Preferably, defect parameters include one or more of area, quantity, distribution in the step (6).
Preferably, pretreatment includes histogram equalization in the step (2), and histogram equalization is for enhancing local contrast
Degree reduces background or all too light or too dark phenomenon of prospect.
The utility model has the advantages that the defect detecting system of the Cast Aluminum Auto-parts Abroad radioscopic image of invention, is based on image processing techniques, lead to
The difference of defect and normal configuration is crossed in Cast Aluminum Auto-parts Abroad radioscopic image to identify defect, and the defect that will test is united
Meter constructs Cast Aluminum Auto-parts Abroad criteria of quality evaluation according to statistical value, generates the quality evaluation table of part, true according to evaluation table
Subsequent use and qualification rate are determined, to effectively detect and control product quality.
Detailed description of the invention
Fig. 1 is the flow diagram of the defect inspection method of Cast Aluminum Auto-parts Abroad radioscopic image of the invention.
Specific embodiment
As shown in Figure 1, the device that the defect detecting system of the Cast Aluminum Auto-parts Abroad radioscopic image of the present embodiment uses includes work
Industry X ray sensor, light source and computer.
Computer includes four modules: acquisition module, image pre-processing module, defect recognition module, defect analysis and system
Count module.
The defect inspection method of the Cast Aluminum Auto-parts Abroad radioscopic image of the present embodiment, includes the following steps:
Step 1: when the duty cycle starts, acquiring image information by X ray sensor, and real-time Transmission is to image preprocessing mould
Block；
Step 2: image pre-processing module to picture carry out preliminary treatment, mainly include histogram equalization, local binarization, in
Value filtering.Unnecessary noise is filtered out by these, and keeps brightness of image moderate, prominent features are subsequent analysis and detection
Preferably identification is provided.
Step 3: defect recognition module receives pretreated gray scale picture, carries out Morphological scale-space to picture, will be normal
The difference of component and defect, which searches out, to be come, and labeled as pixel 255(white), remaining is labeled as 0(black).
Step 4: by given threshold, Gaussian noise and environment profile are filtered out；Specifically, defect area is too small
And it is too big filter out, area too it is small mainly very little Gaussian noise, area too greatly be mainly model outside environment profile,
It is not belonging to model internal flaw.To obtain potential silhouette.
Step 5: the profile after filtering out above is analyzed, if profile is not present, the signal for continuing next part is adopted
Collection；If profile exists and contour area is greater than the greatest drawback area of setting, defect is excessive not to be available, which should give up
It abandons；If contour area is less than the greatest drawback area of setting, these discrete defect profiles are merged by positional relationship, are merged
Standard has:
1. then being merged when defect profile intersects；
2. when defect profile is in same level region or vertical area, and two defect profiles distance be less than threshold X _ min or
Person Y_min, then merge；
Step 6: defect analysis and sorting module count the defect detected, and according to the defect parameters (face of statistics
Product, quantity, distribution etc.) quality evaluation table is generated, and distribute different application scenarios and life cycle.Quality is preferably used for
The scene of high-precision high quality demand, quality it is slightly secondary can be used for requiring lower scene.If quality is too poor, the aluminum casting is straight
It connects discarded.
In step 2, histogram equalization reduces background or all too light or too dark phenomenon of prospect for enhancing local contrast；
Local binarization passes through the gray difference of background and feature, and background and prospect are distinguished by way of two-value；Intermediate value filter
Wave is used to keep the sharp change of signal and eliminates impulsive noise.
The above description is only an embodiment of the present invention, is not intended to limit the scope of the invention, all to utilize this hair
Equivalent structure or equivalent flow shift made by bright description is applied directly or indirectly in other relevant technology necks
Domain is included within the scope of the present invention.
Claims (7)
1. a kind of defect inspection method of Cast Aluminum Auto-parts Abroad radioscopic image, it is characterised in that the following steps are included:
(1) in Cast Aluminum Auto-parts Abroad moving process, acquisition module is transmitted by X ray sensor signal acquisition part picture
To image pre-processing module；
(2) image pre-processing module pre-processes picture；
(3) the pretreated gray scale picture of defect recognition module receiving step (2), and morphological image process is carried out, it will be normal
The difference of component and defect, which searches out, to be come, and is labeled as pixel 255, remaining is labeled as 0；
(4) it by given threshold, filters out Gaussian noise and environment profile to obtain potential silhouette；
(5) profile after the filtering out of step (4) is analyzed, if profile is not present, continues the signal acquisition of next part；
If profile exists and contour area is greater than the greatest drawback area of setting, defect is excessive not to be available, which should discard；
If contour area is less than the greatest drawback area of setting, these discrete defect profiles are merged by positional relationship；
(6) defect detected to step (4) counts, and generates quality evaluation table according to the defect parameters of statistics, and
Distribute different application scenarios and life cycle；Quality is preferably used for the scene of high-precision high quality demand, and quality is slightly secondary
It can be used for requiring lower scene；If quality is too poor, the aluminum casting is directly discarded.
2. a kind of defect inspection method of Cast Aluminum Auto-parts Abroad radioscopic image according to claim 1, it is characterised in that described
The step of (5) in specifically merge standard and have:
1. then being merged when defect profile intersects；
2. when defect profile is in same level region or vertical area, and two defect profiles distance be less than threshold X _ min or
Person Y_min, then merge.
3. a kind of defect inspection method of Cast Aluminum Auto-parts Abroad radioscopic image according to claim 1, it is characterised in that described
The step of (4) in Gaussian noise and environment profile filtered out into the method for obtaining potential silhouette are as follows: by defect area it is too small and
Too big filters out, the too small Gaussian noise for very little of area, and environment profile of the area too greatly outside model is not belonging in model
Portion's defect, to obtain potential silhouette.
4. a kind of defect inspection method of Cast Aluminum Auto-parts Abroad radioscopic image according to claim 1, it is characterised in that described
The step of (2) in, pretreatment includes local binarization, local binarization by the gray difference of background and feature, by background and
Prospect is distinguished by way of two-value.
5. a kind of defect inspection method of Cast Aluminum Auto-parts Abroad radioscopic image according to claim 1, it is characterised in that described
The step of (2) in, pretreatment include median filtering, median filtering be used for keep signal sharp change and eliminate impulsive noise.
6. a kind of defect inspection method of Cast Aluminum Auto-parts Abroad radioscopic image according to claim 1, it is characterised in that described
The step of (6) in defect parameters include one or more of area, quantity, distribution.
7. a kind of defect inspection method of Cast Aluminum Auto-parts Abroad radioscopic image according to claim 1, it is characterised in that described
The step of (2) in pretreatment include histogram equalization, histogram equalization reduces background or prospect all for enhancing local contrast
Too light or too dark phenomenon.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910209706.1A CN109949291A (en) | 2019-03-19 | 2019-03-19 | A kind of defect inspection method of Cast Aluminum Auto-parts Abroad radioscopic image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910209706.1A CN109949291A (en) | 2019-03-19 | 2019-03-19 | A kind of defect inspection method of Cast Aluminum Auto-parts Abroad radioscopic image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109949291A true CN109949291A (en) | 2019-06-28 |
Family
ID=67010229
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910209706.1A Pending CN109949291A (en) | 2019-03-19 | 2019-03-19 | A kind of defect inspection method of Cast Aluminum Auto-parts Abroad radioscopic image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109949291A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112819745A (en) * | 2019-10-31 | 2021-05-18 | 合肥美亚光电技术股份有限公司 | Nut kernel center worm-eating defect detection method and device |
| CN113152966A (en) * | 2021-04-21 | 2021-07-23 | 庭院唱库(浙江)科技有限公司 | Garage system |
| CN113516619A (en) * | 2021-04-09 | 2021-10-19 | 重庆大学 | Product surface flaw identification method based on image processing technology |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1401075A (en) * | 2000-02-05 | 2003-03-05 | 伊克斯龙国际X射线有限公司 | Method for automatically detecting casting defects in a test piece |
| CN105976352A (en) * | 2016-04-14 | 2016-09-28 | 北京工业大学 | Weld seam surface detect feature extraction method based on grayscale image morphology |
| CN107808378A (en) * | 2017-11-20 | 2018-03-16 | 浙江大学 | Complicated structure casting latent defect detection method based on vertical co-ordination contour feature |
| CN108760747A (en) * | 2018-04-28 | 2018-11-06 | 浙江大学 | A kind of 3D printing model surface defect visible detection method |
-
2019
- 2019-03-19 CN CN201910209706.1A patent/CN109949291A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1401075A (en) * | 2000-02-05 | 2003-03-05 | 伊克斯龙国际X射线有限公司 | Method for automatically detecting casting defects in a test piece |
| CN105976352A (en) * | 2016-04-14 | 2016-09-28 | 北京工业大学 | Weld seam surface detect feature extraction method based on grayscale image morphology |
| CN107808378A (en) * | 2017-11-20 | 2018-03-16 | 浙江大学 | Complicated structure casting latent defect detection method based on vertical co-ordination contour feature |
| CN108760747A (en) * | 2018-04-28 | 2018-11-06 | 浙江大学 | A kind of 3D printing model surface defect visible detection method |
Non-Patent Citations (1)
| Title |
|---|
| 周健: "基于X射线实时成像的铝合金激光焊接缺陷识别技术研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅰ辑》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112819745A (en) * | 2019-10-31 | 2021-05-18 | 合肥美亚光电技术股份有限公司 | Nut kernel center worm-eating defect detection method and device |
| CN113516619A (en) * | 2021-04-09 | 2021-10-19 | 重庆大学 | Product surface flaw identification method based on image processing technology |
| CN113152966A (en) * | 2021-04-21 | 2021-07-23 | 庭院唱库(浙江)科技有限公司 | Garage system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106053479B (en) | A kind of vision detection system of the workpiece appearance defects based on image procossing | |
| CN106204614B (en) | A kind of workpiece appearance defects detection method based on machine vision | |
| WO2020248439A1 (en) | Crown cap surface defect online inspection method employing image processing | |
| CN109949291A (en) | A kind of defect inspection method of Cast Aluminum Auto-parts Abroad radioscopic image | |
| CN101566582B (en) | Medicine bottle label information online detection system in powder injection production based on mechanical vision | |
| CN109490316A (en) | A kind of surface defects detection algorithm based on machine vision | |
| CN109829876A (en) | Carrier bar on-line detection device of defects and method based on machine vision | |
| CN110211101A (en) | A kind of rail surface defect rapid detection system and method | |
| CN108760747A (en) | A kind of 3D printing model surface defect visible detection method | |
| CN102221559A (en) | Online automatic detection method of fabric defects based on machine vision and device thereof | |
| CN110288584A (en) | Ceramic hot-dip aluminizing detection method of surface flaw and device based on machine vision | |
| CN102175692A (en) | System and method for detecting defects of fabric gray cloth quickly | |
| CN110473184A (en) | A kind of pcb board defect inspection method | |
| CN110108712A (en) | Multifunctional visual sense defect detecting system | |
| CN109507205A (en) | A kind of vision detection system and its detection method | |
| CN102901735B (en) | System for carrying out automatic detections upon workpiece defect, cracking, and deformation by using computer | |
| CN106914428A (en) | A kind of New Algorithm of the steel ball surface defect Differential Detection based on machine vision | |
| CN108776966B (en) | Method and system for detecting magnet appearance defects | |
| Li et al. | The method for glass bottle defects detecting based on machine vision | |
| CN105184792B (en) | A kind of saw blade wear extent On-line Measuring Method | |
| CN106353340A (en) | Surface defect detection method for rod-like high-reflectance part | |
| CN114998205A (en) | Method for detecting foreign matters in bottle in liquid filling process based on optical means | |
| Meng et al. | Rubber hose surface defect detection system based on machine vision | |
| CN110889367A (en) | Deep learning-based kitchen worker wearing standard identification method | |
| CN109626302A (en) | A kind of online test method of beer foreign matter in bottles |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190628 |
|
| RJ01 | Rejection of invention patent application after publication |