CN109584212B - MATLAB-based SLM powder bed powder laying image scratch defect identification method - Google Patents
MATLAB-based SLM powder bed powder laying image scratch defect identification method Download PDFInfo
- Publication number
- CN109584212B CN109584212B CN201811307349.4A CN201811307349A CN109584212B CN 109584212 B CN109584212 B CN 109584212B CN 201811307349 A CN201811307349 A CN 201811307349A CN 109584212 B CN109584212 B CN 109584212B
- Authority
- CN
- China
- Prior art keywords
- image
- matlab
- slm
- detection
- edge
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The invention belongs to the technical field of selective laser melting matching related image detection, and discloses an MATLAB-based SLM powder bed powder laying image scratch defect detection method, which comprises the following steps: collecting a color image of powder spread by an SLM powder bed, introducing the color image into an MATLAB system to serve as a detection image, and performing a series of preprocessing such as graying, gray scale expansion, sharpening filtering, smoothing filtering and the like on each detection image based on the MATLAB system; and continuing to perform image edge detection and Hough transformation by using a canny operator in the MATLAB system, thereby identifying and detecting the final scratch defect and simultaneously giving a position mark. The invention can give full play to the function of the MATLAB system packaging library, efficiently and quickly achieve the aim of automatic identification, is convenient to control and control in the whole process, has high identification rate, and has the advantages of good robustness, good adaptability and the like.
Description
Technical Field
The invention belongs to the technical field of Selective Laser Melting (SLM) matched related image detection, and particularly relates to an MATLAB-based SLM powder bed powder paving image scratch defect detection method which can be well suitable for application occasions and process characteristics of an SLM technology and provides an accurate and rapid identification scheme suitable for the powder paving image scratch defects of the SLM powder bed powder paving image.
Background
Since the invention of 3D printing technology at the end of the 20 th century, 3D printing is rapidly penetrating into various industrial fields. Because a workpiece of Selective Laser Melting (SLM) has the characteristics of small size, high precision, low surface roughness and the like, the SLM has the advantage of being extremely thick in the aspect of manufacturing metal parts with complex structures, so that the SLM has a quite important position in the field of metal additive manufacturing and is applied to more and more fields.
But SLM still has some important technical issues to be optimized in its process. For example, since the SLM is formed by powder laser melting, the existence of powder laying defects in the SLM powder bed directly has a great influence on the performance of the product. In this case, considering that the SLM manufacturing time of the metal part is long, if the powder layering defect state on the SLM powder bed can be quickly identified, it means that the SLM process can be terminated or timely adjusted as quickly as possible, and the error cost is reduced, which is very important for the development of the 3D printing industry. The search finds that the prior art is lack of a scheme for accurately and efficiently identifying the SLM process, particularly the scratch defect of the powder bed powder paving image. Accordingly, there is a need in the art to find a targeted solution to better meet the above technical needs faced in actual production practice.
Disclosure of Invention
Aiming at the defects or the improvement requirements of the prior art, the invention provides an SLM powder bed powder laying image scratch defect detection method based on MATLAB, wherein, by combining the SLM process actual situation and the data characteristics of the powder paving image, an MATLAB system is introduced to detect and identify the scratch defect by replacing human eyes with a robot, and further selects proper image preprocessing, image segmentation and edge detection modes from various algorithms to execute specific operations, accordingly, not only can the functions of an MATLAB system packaging library be fully exerted, and the aim of automatic identification can be efficiently and quickly achieved, the whole process is convenient to control, the recognition rate is high, the method has the characteristic of good robustness, the proper threshold value can be automatically selected aiming at different pictures, therefore, the method is particularly suitable for application occasions needing to carry out high-efficiency and high-precision detection on the linear scratch defects of powder bed powder laying images in the manufacturing process of various SLMs.
In order to achieve the above object, according to the present invention, there is provided a method for detecting a scratch defect of an SLM powder bed powder paving image based on MATLAB, the method comprising the steps of:
(a) image preprocessing step
Collecting color images of powder spread by a plurality of SLM powder beds, introducing the color images into an MATLAB system to serve as detection images, and preprocessing each detection image based on the MATLAB system, wherein the process comprises the following operations: firstly, carrying out binarization and graying processing on a detection image by using an MATLAB system, thereby obtaining a corresponding grayscale image; then, judging an area with concentrated pixels according to the gray distribution histogram, and directly calling an imadjust function in a system toolbox to expand the gray range of the area, thereby obtaining a clearer gray image; then, carrying out sharpening filtering and smoothing filtering on the gray level image in sequence, and then outputting the preprocessed image;
(b) defect acquisition step
Carrying out differential processing on the SLM powder bed powder laying image preprocessed in the step (a) and the current position sectional image of the SLM workpiece, and screening and extracting a long and thin characteristic area of a suspected scratch in the image, so as to obtain information which preliminarily reflects scratch defect distribution;
(c) edge detection and defect identification
Performing image edge detection by using a canny operator in an MATLAB system aiming at the elongated feature region extracted in the step (b), thereby more accurately determining the edge position of a suspected scratch in the image; the process of image edge detection preferably includes the following operations: firstly, setting double thresholds in a threshold selection area of a canny operator, and regarding points lower than a low threshold as non-edge points and regarding points higher than a high threshold as edge points; meanwhile, the point between the edge point and the non-edge point is judged through the connectivity of the edge: if the adjacent edge points are adjacent, the edge points are regarded as edge points; if the points are isolated points, the points are regarded as non-edge points;
and then, continuing to use Hough transformation in the MATLAB system, so that the elongated characteristic region determined as the scratch defect is identified, thereby completing the whole SLM powder bed powder laying image scratch defect identification process.
As a further preference, in step (a), the filter factor is selected automatically, preferably using a laplacian filter in MATLAB system, whereby the corresponding sharpening filtering operation is performed.
As a further preference, in step (a), the filter window is preferably automatically selected using a wiener filter in the MATLAB system, thereby performing the corresponding smoothing filter operation.
As a further preference, in the step (a), the operation of collecting the color image of the SLM powder bed powder is preferably performed in the following manner: keeping the conditions of external factors such as light rays and the like unchanged, and then acquiring the image by using a CDD image acquisition device.
Further preferably, in the step (b), the elongated feature region of the suspected scratch in the extracted image is screened, preferably by using a morphological method, so as to more accurately acquire the region information reflecting the scratch portion.
As a further preference, in step (c), the process of image edge detection preferably includes the following operations: firstly, setting double thresholds in a threshold selection area of a canny operator, and regarding points lower than a low threshold as non-edge points and regarding points higher than a high threshold as edge points; meanwhile, the point between the edge point and the non-edge point is judged through the connectivity of the edge: if the adjacent edge points are adjacent, the edge points are regarded as edge points; if the points are isolated points, the points are regarded as non-edge points.
As a further preference, after the identification and detection of the final scratch defect after the step (c), a corresponding monitoring alarm system is preferably also provided, thereby realizing the real-time early warning of the powder laying defect.
Generally, compared with the prior art, the technical scheme provided by the invention aims at the technical fact that no perfect method is adopted to identify the SLM powder bed scratch defect through a machine, and the MATLAB system is pertinently introduced to realize the automatic identification of the SLM powder bed scratch defect with high efficiency and high precision; particularly, the invention also performs special selection design on the specific operation algorithm of the image preprocessing of the SLM powder bed by combining the self characteristics of the SLM powder bed scratch from the abundant package library functions in the MATLAB system, and performs targeted improvement on the subsequent operations such as image edge detection and the like; more practical test results show that the technical process can not only achieve high recognition rate, but also has good robustness, can select proper threshold values to execute algorithm processing aiming at different SLM powder bed images, can obtain good balance between the efficiency of the whole algorithm and the finally obtained detection precision, and has the advantages of convenience in control and calculation processing.
Drawings
Fig. 1 is a schematic overall process flow diagram of a SLM powder bed powder-spread image scratch defect detection method constructed according to a preferred embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. In addition, the technical features involved in the embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.
Fig. 1 is a schematic overall process flow diagram of a SLM powder bed powder-spread image scratch defect detection method constructed according to a preferred embodiment of the present invention. As shown in fig. 1, the process mainly comprises the following process steps:
step one, an image preprocessing step.
And collecting color images of powder spread by a plurality of SLM powder beds, introducing the color images into an MATLAB system to serve as detection images, and preprocessing each detection image based on the MATLAB system. The key improvement of the invention is that an MATLAB system is introduced to serve as a detection platform of SLM powder bed powder laying images, and multiple functions meeting the requirements of the specific application occasion are realized by virtue of abundant package library functions.
Specifically, the image preprocessing process includes the operations of: firstly, carrying out binarization and graying processing on a detection image by using an MATLAB system, thereby obtaining a corresponding grayscale image; then, the area with concentrated pixels is judged according to the gray distribution histogram, and the imadjust function in the system tool box is directly called to expand the gray range, so that a clearer gray image is obtained. In addition, sharpening filtering and smoothing filtering are sequentially performed on the gray level image, and then the preprocessed image is output. As a preferred specific operation manner, a laplacian filter in the MATLAB system may be used to automatically select a filter factor, thereby performing a corresponding sharpening filtering operation; and simultaneously, a wiener filter in the MATLAB system is used for automatically selecting a filter window, so that a corresponding smoothing filter operation is executed.
As one of the key improvements of the present invention, the present invention makes a targeted selection of the specific treatment modes of the gray scale extension, the sharpening filtering and the smoothing filtering in combination with the characteristics and requirements of the application objects. The gray image is expanded by calling the imadjust function in the system toolbox, so that the method is more convenient and faster, more importantly, the gray range of the part of interest can be expanded, and correspondingly, the defect part of the SLM layering powder paving image is more prominent and easier to distinguish. In addition, the specific operation of performing laplacian sharpening filtering and wiener smoothing filtering is performed firstly because the overall noise of the SLM image is not obvious, the original image is blurred by using the smoothing filtering firstly, the defect characteristics are erased, and the edge protruding effect after sharpening is not good, which is not beneficial to the subsequent processing. After sharpening filtering is used firstly, the characteristics of defects become obvious, noise points generated by sharpening filtering are erased after smoothing filtering is carried out, and the comprehensive processing effect is better.
Step two, defect acquisition step
And then, carrying out difference processing on the preprocessed SLM powder bed powder paving image and the current position sectional image of the SLM workpiece, thereby obtaining an image reflecting the SLM powder bed powder paving defect distribution. In the process, the elongated feature region similar to the scratch in the image can be screened and extracted by adopting a morphological method, so that the scratch region can be better acquired.
And step three, edge detection and defect identification.
As another key improvement of the present invention, for the previously extracted elongated feature region, image edge detection is preferably performed using the canny operator in MATLAB system, which is specifically as follows: for example, the OTSU algorithm or other methods may be adopted, first setting a dual threshold in the threshold selection region of the canny operator, and regarding the points lower than the low threshold as non-edge points and the points higher than the high threshold as edge points; meanwhile, the point between the two is judged through the connectivity of the edge: if the adjacent edge points are adjacent, the edge points are regarded as edge points; if the points are isolated points, the points are regarded as non-edge points. In this way, practical tests show that the final edge detection operation can be obtained more comprehensively and accurately, and then the required SLM powder bed powder paving image scratch defect detection result is obtained, so that the linear scratch in the image is preliminarily determined.
The Hough transform in the MATLAB system is then used continuously so that the elongated linear feature regions that are finally determined to be scratches are identified, thereby identifying the detection of the final scratch defect while giving a position marking.
In conclusion, the detection method can better solve the problem that the SLM powder bed scratch defect cannot be identified through a machine in the prior art, and has the advantages of high identification rate, convenience and rapidness in operation, good robustness and the like, so that the method is particularly suitable for application occasions of carrying out high-efficiency and high-precision detection on powder bed powder laying images in various SLM manufacturing processes.
It will be understood by those skilled in the art that the foregoing is only a preferred embodiment of the present invention, and is not intended to limit the invention, and that any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (4)
1. A MATLAB-based SLM powder bed powder laying image scratch defect detection method is characterized by comprising the following steps:
(a) image preprocessing step
Collecting color images of powder spread by a plurality of SLM powder beds, introducing the color images into an MATLAB system to serve as detection images, and preprocessing each detection image based on the MATLAB system, wherein the process comprises the following operations: firstly, carrying out binarization and graying processing on a detection image by using an MATLAB system, thereby obtaining a corresponding grayscale image; then, judging an area with concentrated pixels according to the gray distribution histogram, and directly calling an imadjust function in a system toolbox to expand the gray range of the area, thereby obtaining a clearer gray image; then, carrying out sharpening filtering and smoothing filtering on the gray level image in sequence, and then outputting the preprocessed image;
(b) defect acquisition step
Carrying out differential processing on the SLM powder bed powder laying image preprocessed in the step (a) and the current position sectional image of the SLM workpiece, and screening and extracting a long and thin characteristic area of a suspected scratch in the image, so as to obtain information which preliminarily reflects scratch defect distribution;
(c) edge detection and defect identification
Performing image edge detection by using a canny operator in an MATLAB system aiming at the elongated feature region extracted in the step (b), thereby more accurately determining the edge position of a suspected scratch in the image; the image edge detection process comprises the following operations: firstly, setting double thresholds in a threshold selection area of a canny operator, and regarding points lower than a low threshold as non-edge points and regarding points higher than a high threshold as edge points; meanwhile, the point between the edge point and the non-edge point is judged through the connectivity of the edge: if the adjacent edge points are adjacent, the edge points are regarded as edge points; if the points are isolated points, the points are regarded as non-edge points;
and then, continuing to use Hough transformation in the MATLAB system, so that the elongated characteristic region determined as the scratch defect is identified, thereby completing the whole SLM powder bed powder laying image scratch defect identification process.
2. The SLM powder bed dusting image scratch defect detection method as claimed in claim 1, characterized in that in step (a), a corresponding sharpening filtering operation is performed by automatically selecting a filter factor using a laplacian filter in the MATLAB system.
3. The SLM powder bed powdering image scratch defect detecting method according to claim 1 or 2, characterized in that in step (a), a filter window is automatically selected using a wiener filter in the MATLAB system, whereby a corresponding smoothing filter operation is performed.
4. The SLM powder bed dusting image scratch defect detection method as claimed in any of claims 1-3, characterized in that after step (c), when the final scratch defect is identified and detected, a corresponding monitoring alarm system is further provided, thereby realizing real-time pre-warning of dusting defect.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811307349.4A CN109584212B (en) | 2018-11-05 | 2018-11-05 | MATLAB-based SLM powder bed powder laying image scratch defect identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811307349.4A CN109584212B (en) | 2018-11-05 | 2018-11-05 | MATLAB-based SLM powder bed powder laying image scratch defect identification method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109584212A CN109584212A (en) | 2019-04-05 |
| CN109584212B true CN109584212B (en) | 2022-04-26 |
Family
ID=65921399
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811307349.4A Active CN109584212B (en) | 2018-11-05 | 2018-11-05 | MATLAB-based SLM powder bed powder laying image scratch defect identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109584212B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114199893A (en) * | 2021-12-10 | 2022-03-18 | 北京航空航天大学 | SLM powder laying process defect identification and molten pool state real-time monitoring device and method |
| CN116703898B (en) * | 2023-08-03 | 2023-10-20 | 山东优奭趸泵业科技有限公司 | Quality detection method for end face of precision mechanical bearing |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101865859A (en) * | 2009-04-17 | 2010-10-20 | 华为技术有限公司 | Detection method and device for image scratch |
| CN103175844A (en) * | 2012-03-16 | 2013-06-26 | 沈阳理工大学 | Detection method for scratches and defects on surfaces of metal components |
| CN105388162A (en) * | 2015-10-28 | 2016-03-09 | 镇江苏仪德科技有限公司 | Raw material silicon wafer surface scratch detection method based on machine vision |
| CN105891228A (en) * | 2016-06-07 | 2016-08-24 | 江南工业集团有限公司 | Optical fiber appearance defect detecting and outer diameter measuring device based on machine vision |
| EP3095591A1 (en) * | 2015-05-19 | 2016-11-23 | MTU Aero Engines GmbH | Method and device for detecting at least sections of a contour of a layer of an object obtainable by additive processing |
| CN107153067A (en) * | 2017-05-30 | 2017-09-12 | 镇江苏仪德科技有限公司 | A kind of surface defects of parts detection method based on MATLAB |
| CN107402220A (en) * | 2017-07-01 | 2017-11-28 | 华中科技大学 | A kind of selective laser fusing shaping powdering quality vision online test method and system |
| CN108169236A (en) * | 2016-12-07 | 2018-06-15 | 广州映博智能科技有限公司 | A kind of cracks of metal surface detection method of view-based access control model |
| CN108665458A (en) * | 2018-05-17 | 2018-10-16 | 杭州智谷精工有限公司 | Transparent body surface defect is extracted and recognition methods |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017172611A1 (en) * | 2016-03-28 | 2017-10-05 | General Dynamics Mission Systems, Inc. | System and methods for automatic solar panel recognition and defect detection using infrared imaging |
-
2018
- 2018-11-05 CN CN201811307349.4A patent/CN109584212B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101865859A (en) * | 2009-04-17 | 2010-10-20 | 华为技术有限公司 | Detection method and device for image scratch |
| CN103175844A (en) * | 2012-03-16 | 2013-06-26 | 沈阳理工大学 | Detection method for scratches and defects on surfaces of metal components |
| EP3095591A1 (en) * | 2015-05-19 | 2016-11-23 | MTU Aero Engines GmbH | Method and device for detecting at least sections of a contour of a layer of an object obtainable by additive processing |
| CN105388162A (en) * | 2015-10-28 | 2016-03-09 | 镇江苏仪德科技有限公司 | Raw material silicon wafer surface scratch detection method based on machine vision |
| CN105891228A (en) * | 2016-06-07 | 2016-08-24 | 江南工业集团有限公司 | Optical fiber appearance defect detecting and outer diameter measuring device based on machine vision |
| CN108169236A (en) * | 2016-12-07 | 2018-06-15 | 广州映博智能科技有限公司 | A kind of cracks of metal surface detection method of view-based access control model |
| CN107153067A (en) * | 2017-05-30 | 2017-09-12 | 镇江苏仪德科技有限公司 | A kind of surface defects of parts detection method based on MATLAB |
| CN107402220A (en) * | 2017-07-01 | 2017-11-28 | 华中科技大学 | A kind of selective laser fusing shaping powdering quality vision online test method and system |
| CN108665458A (en) * | 2018-05-17 | 2018-10-16 | 杭州智谷精工有限公司 | Transparent body surface defect is extracted and recognition methods |
Non-Patent Citations (2)
| Title |
|---|
| "Image processing based fault detection approach for rail surface";Orhan Yaman 等;《2015 23nd Signal Processing and Communications Applications Conference (SIU)》;20150622;全文 * |
| "基于机器视觉的磁瓦表面缺陷检测技术研究";胡环星;《中国优秀硕士学位论文全文数据库-信息科技辑》;20160315;第2016年卷(第3期);I138-5966 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109584212A (en) | 2019-04-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109685760B (en) | MATLAB-based SLM powder bed powder laying image convex hull depression defect detection method | |
| CN111292305B (en) | Improved YOLO-V3 metal processing surface defect detection method | |
| CN111310558B (en) | Intelligent pavement disease extraction method based on deep learning and image processing method | |
| CN116721106B (en) | Profile flaw visual detection method based on image processing | |
| CN109490316B (en) | Surface defect detection algorithm based on machine vision | |
| CN114359270B (en) | Computer vision-based automobile engine oil way copper sleeve defect detection method | |
| CN106920219B (en) | Article defect detection method, image processing system and computer readable recording medium | |
| CN106780486B (en) | Steel plate surface defect image extraction method | |
| CN109682839B (en) | Online detection method for surface defects of metal arc-shaped workpiece | |
| CN108765402B (en) | Non-woven fabric defect detection and classification method | |
| CN116777907A (en) | Sheet metal part quality detection method | |
| CN113298757A (en) | Metal surface defect detection method based on U-NET convolutional neural network | |
| CN106780526A (en) | A kind of ferrite wafer alligatoring recognition methods | |
| CN111815555A (en) | Metal additive manufacturing image detection method and device combining anti-neural network with local binary | |
| CN115861320B (en) | Intelligent detection method for automobile part machining information | |
| CN106780464A (en) | A kind of fabric defect detection method based on improvement Threshold segmentation | |
| CN113034488B (en) | Visual inspection method for ink-jet printed matter | |
| CN110070523B (en) | Foreign matter detection method for bottle bottom | |
| CN111539927B (en) | Detection method of automobile plastic assembly fastening buckle missing detection device | |
| Daniel et al. | Automatic road distress detection and analysis | |
| CN109584212B (en) | MATLAB-based SLM powder bed powder laying image scratch defect identification method | |
| JP2020532122A (en) | Defect detection for transparent or translucent wafers | |
| CN115100191A (en) | Metal casting defect identification method based on industrial detection | |
| CN111160451A (en) | Flexible material detection method and storage medium thereof | |
| CN109544513A (en) | A kind of steel pipe end surface defect extraction knowledge method for distinguishing |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |