CN117630010B - Three-dimensional precision detection method, assembly and system for surface defects of metal plate - Google Patents
Three-dimensional precision detection method, assembly and system for surface defects of metal plate Download PDFInfo
- Publication number
- CN117630010B CN117630010B CN202311593905.XA CN202311593905A CN117630010B CN 117630010 B CN117630010 B CN 117630010B CN 202311593905 A CN202311593905 A CN 202311593905A CN 117630010 B CN117630010 B CN 117630010B
- Authority
- CN
- China
- Prior art keywords
- defect
- dimensional
- depth
- defects
- detection
- 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
- 230000007547 defect Effects 0.000 title claims abstract description 260
- 238000001514 detection method Methods 0.000 title claims abstract description 220
- 229910052751 metal Inorganic materials 0.000 title claims abstract description 43
- 239000002184 metal Substances 0.000 title claims abstract description 43
- 238000005259 measurement Methods 0.000 claims abstract description 39
- 238000000034 method Methods 0.000 claims abstract description 32
- 238000012545 processing Methods 0.000 claims abstract description 22
- 238000007689 inspection Methods 0.000 claims abstract description 9
- 238000013135 deep learning Methods 0.000 claims abstract description 4
- 238000006073 displacement reaction Methods 0.000 claims description 94
- 238000009826 distribution Methods 0.000 claims description 39
- 238000000605 extraction Methods 0.000 claims description 18
- 230000000875 corresponding effect Effects 0.000 claims description 11
- 230000008569 process Effects 0.000 claims description 10
- 230000001276 controlling effect Effects 0.000 claims description 9
- 238000003384 imaging method Methods 0.000 claims description 9
- 230000003595 spectral effect Effects 0.000 claims description 8
- 238000012549 training Methods 0.000 claims description 8
- 230000000007 visual effect Effects 0.000 claims description 8
- 230000015572 biosynthetic process Effects 0.000 claims description 7
- 230000011218 segmentation Effects 0.000 claims description 6
- 230000008859 change Effects 0.000 claims description 4
- 230000000670 limiting effect Effects 0.000 claims description 4
- -1 polytetrafluoroethylene Polymers 0.000 claims description 4
- 229920001343 polytetrafluoroethylene Polymers 0.000 claims description 4
- 239000004810 polytetrafluoroethylene Substances 0.000 claims description 4
- 238000003786 synthesis reaction Methods 0.000 claims description 4
- 238000005520 cutting process Methods 0.000 claims description 3
- 239000000463 material Substances 0.000 claims description 3
- 230000003287 optical effect Effects 0.000 claims description 3
- 238000001303 quality assessment method Methods 0.000 claims description 3
- 230000001360 synchronised effect Effects 0.000 claims description 3
- 238000012544 monitoring process Methods 0.000 claims description 2
- 238000012217 deletion Methods 0.000 abstract description 2
- 230000037430 deletion Effects 0.000 abstract description 2
- 238000005516 engineering process Methods 0.000 description 14
- 238000010586 diagram Methods 0.000 description 8
- 238000004519 manufacturing process Methods 0.000 description 4
- 229910001093 Zr alloy Inorganic materials 0.000 description 3
- 238000013136 deep learning model Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 2
- 238000005286 illumination Methods 0.000 description 2
- 230000002829 reductive effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 239000004020 conductor Substances 0.000 description 1
- 230000008094 contradictory effect Effects 0.000 description 1
- 230000005674 electromagnetic induction Effects 0.000 description 1
- 239000011521 glass Substances 0.000 description 1
- 231100001261 hazardous Toxicity 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 239000007769 metal material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000003758 nuclear fuel Substances 0.000 description 1
- 238000013441 quality evaluation Methods 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 238000012552 review Methods 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 239000011343 solid material Substances 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 238000004381 surface treatment Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012876 topography Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/01—Arrangements or apparatus for facilitating the optical investigation
-
- 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
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/50—Depth or shape recovery
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/01—Arrangements or apparatus for facilitating the optical investigation
- G01N2021/0106—General arrangement of respective parts
- G01N2021/0112—Apparatus in one mechanical, optical or electronic block
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8854—Grading and classifying of flaws
- G01N2021/8861—Determining coordinates of flaws
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8854—Grading and classifying of flaws
- G01N2021/8874—Taking dimensions of defect into account
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8887—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30136—Metal
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Theoretical Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Quality & Reliability (AREA)
- Signal Processing (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The invention discloses a three-dimensional precise detection method, a three-dimensional precise detection component and a three-dimensional precise detection system for surface defects of a metal plate, wherein the detection method provided by the invention realizes defect precision positioning based on a principle of deep learning and a coordinate system, provides a target detection position and a measurement track for deep detection, and solves the problem of defect position deletion during metal plate quality detection; the two-dimensional characteristic size of the defect is calculated by utilizing image processing, the depth characteristic size of the defect is calculated according to the depth information obtained by depth detection, so that the detection of the three-dimensional characteristic size of the defect is completed, meanwhile, the harmful defect on the surface of a detected piece is found, the scanning of the three-dimensional shape of the single harmful defect is completed based on a three-dimensional shape scanning method, the quantitative inspection of the three-dimensional shape of the surface defect is realized, and the inspection efficiency and the inspection reliability of the surface defect of the metal plate are greatly improved.
Description
Technical Field
The invention belongs to the technical field of metal surface quality detection, and particularly relates to a three-dimensional precise detection method, a three-dimensional precise detection assembly and a three-dimensional precise detection system for surface defects of a metal plate.
Background
Sheet metal is one of the common forms of application of metallic materials, and is widely used in various industries. The metal plate element has complex production process, involves a plurality of working procedures, and can possibly cause surface scratch, pit and other surface quality problems in the processes of detection, rolling, surface treatment, production circulation and the like. The presence of surface quality problems, which are the fundamental key elements of a nuclear fuel assembly, can have a great impact on the safety of its use.
The existing metal plate surface defect detection means mainly comprise defects such as manual detection, vortex detection and visual detection. The manual detection is performed by an operator through a microscope, a magnifying glass or a test block comparison, the efficiency is low, the detection precision is low, the dependence on experience and subjective consciousness of the detector is strong, the reliability can be reduced under the long-time working condition, the fine scratches are difficult to judge and easy to misjudge, the quantitative measurement cannot be realized, and the obtained defect information is limited; the eddy current detection utilizes the principle of electromagnetic induction, the detection object is necessarily made of conductive materials, the detection depth and the detection sensitivity are contradictory, and the specific position of the defect is difficult to judge; the metal surface defect detection technology based on machine vision is limited by a surface imaging technology, the detection equipment is limited in obtaining depth information, so that a large number of pseudo defects without surface depth information exist in the detection result, and the false detection rate of defect detection is increased.
Disclosure of Invention
In order to solve the problems of low detection precision, limited obtained defect information, low efficiency and the like in the existing sheet metal surface detection technology, the invention provides a three-dimensional precision detection method, a three-dimensional precision detection assembly and a three-dimensional precision detection system for sheet metal surface defects.
The invention is realized by the following technical scheme:
a three-dimensional precision detection method for surface defects of a metal plate comprises the following steps:
acquiring a two-dimensional plane picture set of the surface of a measured piece;
Detecting the two-dimensional plane picture set by utilizing a defect identification model obtained through pre-training, and identifying the type and position distribution of defects in the picture;
dividing defects from an original image to obtain a sub-defect picture set, extracting the outline of the sub-defect picture set, and calculating to obtain the two-dimensional characteristic size of the defects;
Determining a target position and a measurement track of depth detection according to a contour extraction result by combining a principle of a coordinate system, and controlling a depth detection unit to carry out depth detection according to the target position and the measurement track;
Acquiring depth information acquired by the depth detection unit, and processing according to the depth information to obtain the depth characteristic size of the defect, wherein the depth characteristic size and the two-dimensional characteristic size of the corresponding defect form the three-dimensional characteristic size of the defect;
monitoring the surface quality of the detected piece according to the identified defect types, position distribution and three-dimensional feature sizes, and identifying microscopic defects;
Planning a defect three-dimensional scanning track according to the outline and the position distribution of the microscopic defect, and controlling the depth detection unit to scan the three-dimensional morphology of the microscopic defect according to the scanning track, so that the three-dimensional morphology of the microscopic defect is obtained.
The existing sheet metal surface defect detection technology has the problems of low detection efficiency, low detection precision and the like. The detection method provided by the invention realizes defect precision positioning based on the principle of deep learning and coordinate system, provides a target detection position and a measurement track for deep detection, and solves the problem of defect position deletion during sheet metal quality detection; the two-dimensional characteristic size of the defect is calculated by utilizing image processing, the depth characteristic size of the defect is calculated according to the depth information obtained by depth detection, so that the detection of the three-dimensional characteristic size of the defect is completed, meanwhile, the harmful defect on the surface of a detected piece is found, the scanning of the three-dimensional shape of the single harmful defect is completed based on a three-dimensional shape scanning method, the quantitative inspection of the three-dimensional shape of the surface defect is realized, and the inspection efficiency and the inspection reliability of the surface defect of the metal plate are greatly improved.
In an embodiment, the present invention is a method for obtaining a sub-defect picture set by dividing a defect from an original image, extracting a contour of the sub-defect picture set, and calculating a two-dimensional feature size of the defect, comprising:
Marking defects by using end-to-end rectangular frames and noting defect types;
Dividing the defects from the original image according to the coordinates of the rectangular frame marked with the defects by using a picture cutting method to obtain a sub-defect picture set;
And carrying out contour extraction on the sub-defect picture set, and calculating to obtain the two-dimensional feature size of the defect based on a contour extraction result.
As a preferred embodiment, the method for determining the target position and the measurement track of the depth detection according to the contour extraction result and combining the principle of a coordinate system specifically includes:
Converting the position distribution of the defects into a real space coordinate system according to the contour extraction result to obtain the distribution of the defects in space;
Planning a target position and a measurement track of depth detection according to the distribution of defects in space; the measuring track is a straight line segment, the straight line segment is positioned at the position with the largest gray level change rate of the area where the defect is positioned, the straight line segment is uniformly distributed on two sides of the defect, and two end points of the straight line segment are respectively positioned on the edge contour of the defect.
In a preferred embodiment, the method for obtaining depth information collected by the depth detection unit and obtaining depth feature size of a defect according to the depth information processing specifically includes:
Converting the acquired depth-time curve into a depth-displacement curve to obtain a depth profile of a certain section position of the defect;
The depth feature size of the defect is calculated based on the depth profile.
As a preferred embodiment, the method for planning a defect three-dimensional scanning track according to the microscopic defect outline and the position distribution specifically includes:
Based on a three-dimensional morphology scanning method, establishing an optimal polygon or ellipse for covering the defect;
The coverage area is divided unidirectionally to form a scanning track.
As a preferred embodiment, the optimal polygon or ellipse of the coverage defect of the present invention is specifically: defining a polygon or an ellipse in an image coordinate system to completely cover the position of the defect;
wherein the establishment of the optimal polygon or ellipse should follow the ratio of the scratch area to the coverage area as large as possible.
As a preferred embodiment, the unidirectional division coverage area of the present invention specifically includes:
Establishing a sub-coordinate system in an image coordinate system, and equally dividing the optimal polygon or ellipse along the Y axis of the sub-coordinate system; the segmentation interval is the transverse resolution of the depth detection unit, the segmentation line is parallel to the X axis of the sub-coordinate system and intersects with a polygon or an ellipse to form a multi-parallel line segment; the sub-coordinate system is established specifically as follows: in the polygon coverage area, the Y axis is parallel to the tangential direction of the largest inner angle of the polygon, and the X axis is perpendicular to the Y axis; in the elliptical coverage area, the X axis is parallel to the tangential direction of the maximum point of the elliptical curvature, and the Y axis is perpendicular to the X axis; the optimal polygon or ellipse is located in the first quadrant of the sub-coordinate system and tangent to the two coordinate axes.
In a preferred embodiment, the forming of the scanning track of the present invention specifically refers to forming the scanning track by using the line segment end point nearest to the origin of the sub-coordinate system as a starting point, and sequentially connecting the end points of each line segment in an S shape until the last line segment is connected.
In a second aspect, the present invention provides a three-dimensional precision detection assembly for surface defects of a metal sheet, the assembly comprising:
The deep learning module acquires a two-dimensional plane picture set of the surface of the tested piece, detects the two-dimensional plane picture set by utilizing a defect identification model obtained by training in advance, and identifies the type and position distribution of defects in the picture;
the image processing module is used for dividing the defects from the original image to obtain a sub-defect picture set, extracting the outline of the sub-defect picture set and calculating to obtain the two-dimensional characteristic size of the defects;
The data processing module is used for obtaining depth information acquired by the depth detection unit, processing the depth information according to the acquired depth information to obtain the depth characteristic size of the defect, and forming the three-dimensional characteristic size of the defect by the depth characteristic size and the two-dimensional characteristic size of the corresponding defect;
The surface quality assessment module is used for detecting the surface quality of the detected piece according to the identified defect types, position distribution and three-dimensional characteristic sizes and identifying microscopic defects;
The path planning module is used for determining a target position and a measurement track of depth detection according to a defect contour extraction result and combining a coordinate system principle, and controlling the depth detection unit to carry out depth detection according to the target position and the measurement track; and the path planning module plans a defect three-dimensional scanning track according to the outline and the position distribution of the microscopic defect and controls the depth detection unit to scan the three-dimensional morphology of the microscopic defect according to the scanning track, so that the three-dimensional morphology of the microscopic defect is obtained.
In a third aspect, the present invention provides a three-dimensional precision detection system for surface defects of a metal plate, the system comprising:
the detection component is used as a core for data acquisition, processing and control of the whole system;
the signal control component is used for receiving the control command issued by the detection component so as to drive the mechanical execution component to execute corresponding actions;
The data acquisition component is used for receiving the image data of the surface of the tested piece and the depth information data of the defect and uploading the image data and the depth information data to the detection component;
and the mechanical execution assembly acts according to the movement path planned by the detection assembly to realize corresponding detection.
As a preferred embodiment, the mechanical actuating assembly of the present invention comprises:
the space positioning unit comprises a bearing table and a positioning angle;
The three-dimensional movement unit is arranged on the bearing table, a gantry structure is adopted to carry out two-dimensional movement on the surface of the detected piece in a horizontal plane to drive the vision acquisition unit to image the surface of the detected piece, the depth detection unit is moved to a target position, and the depth detection unit is ensured to be positioned at a safe position and a working distance in the vertical direction through movement in the vertical direction;
The visual acquisition unit is used for imaging the surface of the measured piece and acquiring image information of the surface of the measured piece;
and the depth detection unit moves according to the planned movement path and acquires depth information of the defect in real time in the movement process.
As an optimal implementation mode, the bearing table is placed on a horizontal plane, and the positioning angle comprises four right-angle blocks which are respectively arranged at detection limit positions [ Xmax, ymax ], [ Xmin, ymax ], [ Xmax, ymin ], [ Xmin, ymin ] of the detection system on the horizontal plane and are used for limiting the position of a detected piece on the bearing table.
As a preferred embodiment, the right angle block of the present invention is made of polytetrafluoroethylene material.
As a preferred embodiment, the three-dimensional motion unit of the present invention includes an X-direction motor displacement axis, a Y-direction motor displacement axis, a Z-direction motor displacement axis, and a gantry;
The X-direction motor displacement shaft consists of two shafts which are arranged in parallel, and the two shafts have the same stroke, are aligned at two ends and are arranged on the bearing table in parallel to realize transverse synchronous movement;
the bottom end of the portal frame is arranged on a bearing table of the X-direction motor displacement shaft;
The Y-direction motor displacement shaft side is arranged at the top end of the portal frame to realize longitudinal movement;
The Z-direction motor displacement shaft is vertically arranged on the bearing table of the Y-direction motor displacement shaft, so that vertical movement is realized.
As a preferred embodiment, the vision acquisition unit of the present invention includes an industrial camera, a fixed focus lens, and a coaxial light source;
the visual acquisition unit is arranged on a bearing table of the Y-direction motor displacement shaft; the industrial camera is connected with the fixed focus lens through a C interface, the coaxial light source is arranged right below the fixed focus lens, and the center of the light emitting surface of the coaxial light source coincides with the center of the optical axis of the industrial camera.
As a preferred embodiment, the depth detection unit of the present invention comprises a micro-motion displacement platform and a spectral confocal sensor;
The depth detection unit is arranged on a bearing table of the Z-direction motor displacement shaft; the micro-motion displacement platform consists of an X-direction micro-motion displacement shaft and a Y-direction micro-motion displacement shaft, wherein the X-direction micro-motion displacement shaft and the Y-direction micro-motion displacement shaft are vertically arranged, the X-direction micro-motion displacement shaft and the X-direction motor displacement shaft are arranged in parallel, and the Y-direction micro-motion displacement shaft and the Y-direction motor displacement shaft are arranged in parallel; the spectral confocal sensor is arranged on the micro-motion displacement platform through a clamp, and the X-direction micro-motion displacement shaft and the Y-direction micro-motion displacement shaft realize measuring tracks of different angles in a plane through speed vector synthesis.
Compared with the prior art, the invention has the following advantages and beneficial effects:
1. The invention images the surface defect of the measured piece, utilizes a depth learning method to identify and position the defect, provides more accurate target detection position and measurement track for depth detection, utilizes an image processing technology to calculate and obtain the two-dimensional characteristic size of the defect, controls a depth detection unit to carry out depth detection according to the target detection position and the measurement track, utilizes the depth information obtained by detection to calculate and obtain the depth characteristic size, and completes the measurement of the three-dimensional characteristic size of the defect; the three-dimensional morphology scanning of the harmful defects is realized based on the defect three-dimensional morphology method, so that the quantitative detection of the three-dimensional morphology of the surface defects of the metal plate is realized, and the detection efficiency and the detection reliability of the surface defects of the metal plate are improved;
2. The technology provided by the invention can be popularized and applied to the quality inspection requirements of various metal plate surface defects, such as plate production and manufacturing, process products and final products, and the technology relates to the detection and control of the surface quality of a flat plate, and has a relatively high market application value.
Drawings
The accompanying drawings, which are included to provide a further understanding of embodiments of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the principles of the application. In the drawings:
fig. 1 is a flowchart of a detection method according to an embodiment of the present invention.
FIG. 2 is a schematic block diagram of a detection assembly according to an embodiment of the present invention.
Fig. 3 is a schematic diagram of a detection system architecture according to an embodiment of the invention.
Fig. 4 is a schematic diagram of the operation of the coaxial light source according to an embodiment of the present invention.
FIG. 5 is a flowchart illustrating the operation of a detection system according to an embodiment of the present invention.
Fig. 6 is a schematic diagram of path planning of a detecting unit according to an embodiment of the present invention.
FIG. 7 is a schematic diagram of comparing an optimal polygon/ellipse with an optimal bounding rectangle according to an embodiment of the present invention.
Fig. 8 is a schematic diagram of sub-coordinate system establishment according to an embodiment of the present invention.
FIG. 9 is a schematic diagram of a three-dimensional topography scan trajectory planning for defects according to an embodiment of the present invention.
FIG. 10 is a three-dimensional morphology detection chart of defects of a zirconium alloy plate in an embodiment of the invention.
Reference numerals and corresponding part names:
The device comprises a 1-detection component, a 2-signal control component, a 3-data acquisition component, a 4-mechanical execution component, a first 5-X-direction motor displacement shaft, a second 6-X-direction displacement shaft, a first 7-Y-direction displacement shaft, a second 8-Z-direction displacement shaft, a 9-industrial camera, a 10-fixed focus lens, a 11-coaxial light source, a 12-X-direction micro-motion displacement shaft, a 13-Y-direction micro-motion displacement shaft, a 14-spectral confocal sensor, a 15-bearing table, a 16-positioning angle, a 17-measured piece, a 18-portal frame, a 19-LED light source, a 20-lens and a 21-spectroscope.
Detailed Description
For the purpose of making apparent the objects, technical solutions and advantages of the present invention, the present invention will be further described in detail with reference to the following examples and the accompanying drawings, wherein the exemplary embodiments of the present invention and the descriptions thereof are for illustrating the present invention only and are not to be construed as limiting the present invention.
Example 1
On the surface of the metal plate in the defect detection technique: the manual detection technology has the problems of low efficiency, low detection precision, strong dependence on experience and subjective consciousness of detection personnel, incapability of guaranteeing reliability, easiness in misjudgment, incapability of realizing quantitative measurement and the like; however, the eddy current inspection technique is difficult to determine the specific location of the defect; the detection technology based on machine vision is limited by the surface imaging technology, and the acquisition of depth information is limited, so that a large number of pseudo defects are caused, and the false detection rate is increased. In view of this, the present embodiment provides a three-dimensional precision detection method for surface defects of a metal plate.
As shown in fig. 1, the detection method provided in this embodiment specifically includes the following steps:
Step 1, acquiring a two-dimensional plane picture set of the surface of a measured piece.
And 2, detecting the acquired two-dimensional plane picture set by utilizing a defect identification model obtained through pre-training, and identifying the type and position distribution of defects in the picture.
And step 3, dividing the defects from the original image to obtain a sub-defect picture set, extracting the outline of the sub-defect picture set, and calculating to obtain the two-dimensional feature size of the defects.
And 4, determining a target position and a measurement track of the depth detection according to the contour extraction result and combining a principle of a coordinate system, and controlling a depth detection unit to carry out the depth detection according to the target position and the measurement track.
And step 5, acquiring depth information acquired by the depth detection unit, and processing according to the acquired depth information to obtain the depth characteristic size of the defect.
And 6, detecting the surface quality of the detected piece according to the identified defect types, position distribution and three-dimensional feature sizes, and identifying microscopic defects.
And 7, planning a three-dimensional scanning track of the defect according to the outline and the position distribution of the microscopic defect, and controlling the depth detection unit to scan the three-dimensional morphology of the microscopic defect according to the scanning track, so as to obtain the three-dimensional morphology of the microscopic defect.
According to the detection method provided by the embodiment, firstly, the surface defects of the detected piece are fully detected, namely, the three-dimensional characteristic dimensions (namely, the two-dimensional characteristic dimensions and the depth characteristic dimensions) of the surface defects of the detected piece are obtained, the accurate positioning of the defects is realized specifically based on the principle of depth learning and a coordinate system, reliable target detection positions and measurement tracks are provided for the depth detection, the problem that the defect positions are missing during the quality detection of the metal sheet is solved, meanwhile, the two-dimensional characteristic dimensions of the defects are calculated by utilizing an image processing technology, the depth information is detected according to the defect positions and the measurement tracks which are accurately positioned, the depth characteristic dimensions are obtained, the three-dimensional characteristic dimension detection of the defects is completed, meanwhile, the three-dimensional appearance of single defects is obtained, the reasons for causing the defects are analyzed, and the reliability and the detection efficiency of the defect detection are improved.
Further, a training data set is constructed by utilizing the collected two-dimensional plane picture sets on the surfaces of various metal plates, and the deep learning model is trained by utilizing the training data set to obtain a defect identification model for identifying the types and the position distribution of defects. It should be noted that, the deep learning model herein adopts the existing deep learning model architecture, and will not be repeated here.
Further, marking defects by using an end-to-end rectangular frame and noting defect types; and dividing the defects from the original image according to the coordinates of the rectangular frame by using a picture cutting method to obtain a sub-defect picture set, extracting the outline of the sub-defect picture set, and counting the number of pixel points in the outline area and the outline edge to obtain the two-dimensional characteristic size value of the defects.
Further, according to the contour extraction result, the position distribution of the defects is converted into a real space coordinate system to obtain the distribution of the defects in space, and the target position and the measurement track of depth detection are planned. Specifically, the measuring track is a straight line segment, which is positioned at the place where the gray level change rate of the area where the defect is positioned is the largest, and is uniformly distributed on two sides of the defect, and two end points of the straight line segment are respectively positioned on the edge contour of the defect.
Further, according to the target position and the measurement track of the depth detection, the depth detection unit is controlled to move to the target position, and the depth detection unit is controlled to move according to the planned measurement track according to the principle of speed vector synthesis, and meanwhile depth data are collected; and calculating the synthetic motion speed, and converting the acquired depth-time curve into a depth-displacement curve by utilizing a time-displacement conversion relation to obtain a depth profile of a certain section position of the defect. Therefore, the position point information of the target can be determined through the displacement distance, and then the depth information of the position point can be further known through the displacement depth function, so that the three-dimensional feature size detection of the defect is completed.
Further, according to the identified defect types, position distribution and three-dimensional feature sizes, the surface quality of the detected piece is evaluated by using a surface quality evaluation technology, and the harmful defects on the surface of the detected piece are found out.
Further, based on a three-dimensional morphology scanning method, an optimal polygon/ellipse covering the defect is established, the covering area is divided in one direction, a scanning track is formed, scanning of the three-dimensional morphology of the single defect is completed, and the obtained three-dimensional discrete points are fitted into a three-dimensional curved surface, so that the three-dimensional morphology of the single defect is obtained.
Specifically, the optimal polygon/ellipse covering the defect refers to that the polygon/ellipse is customized in the image coordinate system so as to completely cover the position of the defect. For defects with directionality such as scratches, the polygon is selected for covering, and the principle of polygon establishment should follow that the number of sides is as small as possible, and the polygon can be a convex polygon/a concave polygon. Elliptical coverage is selected for defects such as pits. The establishment of the optimal polygon/ellipse should follow the ratio of scratch area to coverage area as large as possible.
The unidirectional division coverage area specifically includes: and establishing a sub-coordinate system in the image coordinate system, equally dividing the optimal polygon/ellipse along the Y axis of the sub-coordinate system, wherein the dividing distance is the transverse resolution of the depth detection unit, and the dividing line is parallel to the X axis of the sub-coordinate system and intersects the polygon/ellipse to form a multi-parallel line segment. The established sub-coordinate system specifically means that in the polygonal coverage area, the Y axis is parallel to the tangential direction of the largest inner angle, and the X axis is perpendicular to the Y axis; in the elliptical coverage area, the X axis is parallel to the tangential direction of the point of maximum curvature, and the Y axis is perpendicular to the X axis; the optimal polygons/ellipses are all positioned in the first quadrant of the sub-coordinate system and tangent to the two coordinate axes.
The formation of the scanning track specifically refers to that the end point of a line segment closest to the origin of the sub-coordinate system is taken as a starting point, and the end points of all the line segments are sequentially connected in an S shape until the last line segment is connected, so that the scanning track is formed.
Based on the same technical concept, the embodiment also provides a three-dimensional precision detection assembly for surface defects of a metal plate, as shown in fig. 2, the detection assembly comprises:
The depth learning module acquires a two-dimensional plane picture set of the surface of the tested piece, detects the acquired two-dimensional plane picture set by utilizing a defect identification model obtained through pre-training, and identifies the type and the position distribution of defects in the picture.
And the image processing module is used for dividing the defects from the original image to obtain a sub-defect picture set, extracting the outline of the sub-defect picture set and calculating to obtain the two-dimensional feature size of the defects.
And the data processing module is used for obtaining the depth information acquired by the depth detection unit and processing the depth information according to the acquired depth information to obtain the depth characteristic size of the defect.
And the surface quality assessment module is used for detecting the surface quality of the detected piece according to the identified defect types, the position distribution and the three-dimensional characteristic size and determining microscopic defects.
The path planning module is used for determining a target position and a measurement track of the depth detection according to the contour extraction result and combining a principle of a coordinate system, and controlling the depth detection unit to carry out the depth detection according to the target position and the measurement track; the path planning module plans a three-dimensional scanning track of the defect according to the outline and the position distribution of the microscopic defect, and controls the depth detection unit to scan the three-dimensional morphology of the microscopic defect according to the scanning track, so that the three-dimensional morphology of the microscopic defect is obtained.
Based on the above-mentioned three-dimensional accuracy detection component for surface defects of metal plates, this embodiment also provides a three-dimensional accuracy detection system for surface defects of metal plates, as shown in fig. 3, the detection system mainly includes: the three-dimensional precision detection assembly 1 serving as a data acquisition processing and control core can adopt computer equipment with the functions of calculation, processing, storage and the like for the detection assembly 1; a signal control assembly 2; a data acquisition component 3 and a mechanical execution component 4.
The signal control component 2 is configured to receive a control command issued by the detection component 1 to drive the mechanical execution component 4 to execute a corresponding action, the data acquisition component 3 is configured to receive related image data and upload the related image data to the detection component 1, and the mechanical execution component 3 is configured to perform image acquisition and depth information detection of a measured object under the control of the detection component 1. It should be noted that, the signal control component 2 and the data acquisition component 3 may be conventional signal control devices and data acquisition devices, which are not described herein in detail.
The mechanical execution assembly 4 comprises a spatial positioning unit, a three-dimensional motion unit, a vision acquisition unit and a depth detection unit.
The spatial positioning unit comprises a carrier 15 and a positioning angle 16. The bearing table 15 is placed on a horizontal plane and used for bearing and installing other units of the mechanical execution assembly, the positioning angle 16 comprises four right-angle blocks, the right-angle blocks are respectively installed at detection limit positions [ Xmax, ymax ], [ Xmin, ymin ] of the detection system in the X, Y direction and used for limiting the position of the detected piece 17 on the bearing table, and the detected piece is required to be placed in a region limited by the positioning angle; the polytetrafluoroethylene is the material with the lowest friction coefficient in the solid material, so that secondary damage to the measured piece in the detection process can be avoided, and therefore, in the embodiment, the right-angle block is preferably made of the polytetrafluoroethylene. The coordinate system of the detection system specifically comprises: the X-Y plane is parallel to the plane of the carrying table 15, and the direction perpendicular to the X-Y plane is the Z direction.
The three-dimensional movement unit adopts a gantry structure, so that horizontal plane two-dimensional movement and vertical movement are realized, the horizontal plane two-dimensional movement drives the vision acquisition unit to image the surface of the measured piece on one hand, the depth detection unit is moved to a target detection position on the other hand, and meanwhile, the depth detection unit is ensured to be in a safe position and a working distance in the vertical direction through the vertical movement; the three-dimensional motion unit specifically comprises a precise linear motor displacement shaft (X-direction displacement shafts 5 and 6; Y-direction displacement shaft 7;Z-direction displacement shaft 8) and a portal frame 18, and is arranged on a bearing table 15; the X-direction displacement shaft is composed of two shafts (namely 5 and 6) which are arranged in parallel, and the two shafts have the same stroke, are aligned at two ends and are arranged on the bearing table 15 in parallel, so that synchronous movement in the X direction is realized. The bottom end of the portal frame 18 is mounted on a bearing table of the X-direction displacement shaft. The Y-direction displacement shaft 7 is arranged at the top end of the portal frame 18, so that the Y-direction movement can be realized. The Z-direction displacement shaft 8 is vertically arranged on the bearing table of the Y-direction displacement shaft to realize the motion in the Z direction.
The visual acquisition unit comprises an industrial camera 9, a fixed focus lens 10 and a coaxial light source 11, and is arranged on a bearing table of the Y-direction displacement shaft 7; the industrial camera 9 is connected with the fixed focus lens 10 through a C interface, the coaxial light source 11 is arranged right below the fixed focus lens 10, and the center of the light emitting surface of the coaxial light source is coincident with the center of the optical axis of the industrial camera 9. The coaxial light source 11 provides more uniform illumination, overcomes the interference of reflection of light on the surface of a metal object, and improves the imaging quality of the metal surface. Specifically, the coaxial light source 11 may be composed of an LED light source 19, a lens 20 and a spectroscope 21, where the working principle is as shown in fig. 4, the LED light source is a high-brightness light source, and is changed into parallel light after passing through the lens, and part of the light is reflected to the surface of the measured piece through the spectroscope placed at 45 degrees, so as to realize illumination of the measured piece, and part of the light reflected from the measured piece enters the lens through the spectroscope, so as to realize imaging of an object. The light rays emitted by the coaxial light source are irradiated on the surface of the measured piece in parallel and perpendicular, so that the defect of the surface of an object can be highlighted, the reflection of the surface of the object is avoided, and the imaging quality of the metal surface is improved.
The depth detection unit comprises a micro-motion displacement platform (X-direction micro-motion displacement shaft 12; Y-direction micro-motion displacement shaft 13) and a spectral confocal sensor 14, and is arranged on the bearing table of the Z-direction displacement shaft 8. Wherein, the X-direction micro-motion displacement shaft 12 is installed vertically with the Y-direction micro-motion displacement shaft 13, the X-direction micro-motion displacement shaft 12 is arranged in parallel with the X-direction displacement shaft, the Y-direction micro-motion displacement shaft 13 is arranged in parallel with the Y-direction displacement shaft 7, and the X-direction micro-motion displacement shaft 12 and the Y-direction micro-motion displacement shaft 13 are installed on the bearing platform of the Z-direction displacement shaft 8 through structural members. The spectral confocal sensor 14 is cylindrical and mounted to the micro-motion displacement stage by a fixture. In the initial state, the centers of the X-direction micro-motion displacement shaft 12 and the Y-direction micro-motion displacement shaft 13 are coincident, and the motion in the range of 0-360 degrees can be realized through the combination of the displacement direction and the speed.
Specifically, the longitudinal resolution of the spectral confocal sensor 14 can reach the nanometer level, the working distance, that is, the distance between the front end of the measuring head and the surface of the measured piece is smaller and generally not more than 10cm, in order to ensure the safety of the sensor measuring head, when the spectral confocal sensor 14 does not work, the sensor measuring head is lifted to a safety position by utilizing the Z-direction displacement shaft 8, and when scratch depth information is measured, the working distance is reduced.
As shown in fig. 5, the working process of the detection system provided in this embodiment specifically includes: when the system is initialized, the three-dimensional motion unit returns to the system zero point, namely, each motor displacement shaft moves to the mechanical zero point to be the system origin. Placing the tested piece. And (3) taking the positioning angle closest to the mechanical zero position as an initial position in the field of view of the camera, and moving the three-dimensional motion unit to a detection zero point, namely, moving each motor displacement shaft to the initial position of the system as a detection origin. Under the control of the detection components, the signal control components drive the three-dimensional movement units to automatically move at intervals of the imaging units, and the vision acquisition units are controlled to carry out static acquisition on the surface images of the detected piece in the stay period of 0.1s for each detection unit so as to obtain a picture set of the surface of the detected piece; transmitting the picture set to a detection assembly through a data acquisition assembly for processing to obtain the type, position distribution and two-dimensional characteristic size of the defect, and converting the defect into a real space coordinate system according to the position distribution of the defect in the picture to provide a target position and a measurement track for depth detection; the signal control assembly drives the three-dimensional movement unit to enable the sensor in the depth detection unit to move to a target position, enables the micro-motion displacement platform in the depth detection unit to be in an initial position, utilizes speed vector synthesis to move according to a provided measurement track, collects depth data in the movement process to obtain depth information of the defect, returns the micro-motion displacement platform to the initial position again after the collection is completed, and stores the obtained depth information in one-to-one correspondence with the two-dimensional feature size, so that the three-dimensional feature size of the defect is obtained; the three-dimensional movement unit drives the depth detection unit to sequentially measure according to the measurement track until the depth information of all defects is measured, so that the full detection of the surface defects of the measured piece is completed, and the types, the position distribution and the three-dimensional feature sizes of the surface defects in the two-dimensional plane are obtained. And evaluating the surface quality of the tested piece according to the type, the position distribution and the three-dimensional characteristic size of the surface defect in the two-dimensional plane, and marking the harmful scratch. Finding out a defect contour and a defect position corresponding to the hazardous scratch, customizing an optimal coverage polygon/ellipse in an image coordinate system, planning a three-dimensional scanning track, and obtaining a scanning starting point and a scanning track; based on the principle of a coordinate system, a moving track of the micro-motion displacement platform is formed, a three-dimensional motion unit is driven to enable an origin of the micro-motion displacement platform to be aligned with a scanning starting point, the micro-motion displacement platform starts to move according to the moving track, a sensor acquires depth data in real time in the motion process, three-dimensional discrete data of a coverage area are finally obtained, and three-dimensional discrete points are fitted to obtain the three-dimensional shape of the defect. And (3) finishing the individual detection of all the harmful scratches according to the process, returning the three-dimensional motion unit to the zero point of the system, removing the measured piece, and finishing the measurement. The surface image file and defect depth data of the measured piece are stored in a local file, and a report can be derived from the measurement result, so that the measured piece is convenient to review.
The motion path planning of the mechanical execution assembly 4 in the system proposed in this embodiment mainly includes a visual detection unit path planning and a depth detection unit path planning. The visual detection unit is used for imaging the surface of the detected piece, and the surface of the detected piece is imaged in a small area of rows, columns and faces through path planning, as shown in fig. 6, so that a detection unit is formed. In the measuring process, the surface of each measured piece is divided into a plurality of detection units, and in the same detection area, the reciprocating motion of the detection equipment can be avoided by reducing the number of the detection units, so that the measuring time is shortened, and the measuring efficiency is improved. The depth detection unit is used for measuring the depth information of the defects, optimizing the detection sequence among the defects through path planning, and improving the detection efficiency. Specifically, in each detection unit, the position of the three-dimensional motion unit is taken as the detection origin of the current detection unit when the 1 st photo is taken, and the three-dimensional motion unit drives the vision acquisition unit to sequentially image the detected piece according to the path planning from 1 to 2 to 3 to … to n until the last detection unit n stops. The position of each picture in the coordinate system can be known under the movement of the displacement axis. And transmitting the photo imaged by each detection unit to a detection assembly for processing to obtain the distribution of defects in the detection unit and the set of target detection positions. The resulting set of target detection positions is a set of absolute position coordinates at the detection origin of the detection unit. The moving depth detection unit starts to measure from the defects identified in the nth picture, and sequentially measures different defects in one picture according to preset priorities. And measuring the defects identified in the n-1 th picture in reverse order after the defect measurement in the n-1 th picture is completed until the defect measurement in the 1 st picture is completed.
As shown in fig. 7, for a stripe defect with directionality such as a scratch, an area where the defect is located is covered with an optimal circumscribed polygon; and for defects such as pits, the area where the defect is located is covered by adopting an optimal ellipse. The optimal polygonal/elliptical coverage method has a higher area ratio than the circumscribed rectangle. Therefore, when a scanning path is planned, the depth information acquisition of a non-defect area is effectively avoided, and the efficiency of three-dimensional morphology acquisition is improved.
As shown in fig. 8, the maximum internal angle in the optimal polygon is α, the dotted line is a tangent line of the angle α, the Y axis is parallel to the direction of the dotted line, and the X axis is perpendicular to the Y; the maximum curvature point A in the optimal ellipse is the tangent line of the point A, the X axis is parallel to the direction of the dotted line, and the Y axis is perpendicular to the X axis. The optimal polygon/ellipse is located in the first quadrant of the sub-coordinate system and tangent to the two coordinate axes.
As shown in fig. 9, the optimal polygon/ellipse is divided at equal intervals along the Y-axis in the sub-coordinate system, and the dividing line is parallel to the X-axis and intersects with the optimal polygon/ellipse to form multiple parallel line segments. And (3) connecting the parallel line segments in an S-shaped sequence by taking the line segment end point nearest to the origin of the sub-coordinate system as a scanning start point S to form a three-dimensional morphology scanning track, wherein the three-dimensional morphology detection diagram of the zirconium alloy plate defect is shown in FIG. 10.
The detection technology provided by the embodiment can realize automatic detection of the surface defects of the zirconium alloy plate and three-dimensional morphology precision measurement, and the main technical indexes of the detection technology comprise: (1) two-dimensional image recognition of surface micro defects: resolution 10 μm, recognition rate 1s (10 mm. Times.10 mm); (2) three-dimensional appearance detection of surface micro defects: longitudinal resolution 0.1 μm, longitudinal precision 2 μm, and longitudinal measurement range 300 μm; the lateral resolution is 15 μm.
The foregoing description of the embodiments has been provided for the purpose of illustrating the general principles of the invention, and is not meant to limit the scope of the invention, but to limit the invention to the particular embodiments, and any modifications, equivalents, improvements, etc. that fall within the spirit and principles of the invention are intended to be included within the scope of the invention.
Claims (11)
1. The three-dimensional precision detection method for the surface defects of the metal plate is characterized by comprising the following steps of:
acquiring a two-dimensional plane picture set of the surface of a measured piece;
Detecting the two-dimensional plane picture set by utilizing a defect identification model obtained through pre-training, and identifying the type and position distribution of defects in the picture;
dividing defects from an original image to obtain a sub-defect picture set, extracting the outline of the sub-defect picture set, and calculating to obtain the two-dimensional characteristic size of the defects;
Determining a target position and a measurement track of depth detection according to a contour extraction result by combining a principle of a coordinate system, and controlling a depth detection unit to carry out depth detection according to the target position and the measurement track;
Acquiring depth information acquired by the depth detection unit, and processing according to the depth information to obtain the depth characteristic size of the defect, wherein the depth characteristic size and the two-dimensional characteristic size of the corresponding defect form the three-dimensional characteristic size of the defect;
monitoring the surface quality of the detected piece according to the identified defect types, position distribution and three-dimensional feature sizes, and identifying microscopic defects;
Planning a defect three-dimensional scanning track according to the outline and the position distribution of the microscopic defect, and controlling the depth detection unit to scan the three-dimensional morphology of the microscopic defect according to the scanning track so as to obtain the three-dimensional morphology of the microscopic defect;
According to the contour extraction result, determining the target position and the measurement track of the depth detection by combining a principle of a coordinate system, wherein the method specifically comprises the following steps:
Converting the position distribution of the defects into a real space coordinate system according to the contour extraction result to obtain the distribution of the defects in space;
Planning a target position and a measurement track of depth detection according to the distribution of defects in space; the measuring track is a straight line segment, is positioned at the position with the largest gray level change rate of the area where the defect is positioned, is uniformly distributed on two sides of the defect, and two end points of the straight line segment are respectively positioned on the edge contour of the defect;
planning a defect three-dimensional scanning track according to the microscopic defect outline and the position distribution, and specifically comprising the following steps:
Based on a three-dimensional morphology scanning method, establishing an optimal polygon or ellipse for covering the defect;
Unidirectional dividing the coverage area to form a scanning track;
The optimal polygon or ellipse covering the defect is specifically: defining a polygon or an ellipse in an image coordinate system to completely cover the position of the defect;
The establishment of the optimal polygon or ellipse should follow the occupation ratio of the scratch area in the coverage area as large as possible;
The unidirectional division coverage area specifically comprises:
establishing a sub-coordinate system in an image coordinate system, and equally dividing the optimal polygon or ellipse along the Y axis of the sub-coordinate system; the segmentation interval is the transverse resolution of the depth detection unit, the segmentation line is parallel to the X axis of the sub-coordinate system and intersects with a polygon or an ellipse to form a multi-parallel line segment; the sub-coordinate system is established specifically as follows: in the polygon coverage area, the Y axis is parallel to the tangential direction of the largest inner angle of the polygon, and the X axis is perpendicular to the Y axis; in the elliptical coverage area, the X axis is parallel to the tangential direction of the maximum point of the elliptical curvature, and the Y axis is perpendicular to the X axis; the optimal polygon or ellipse is positioned in the first quadrant of the sub-coordinate system and tangent to the two coordinate axes;
the formation of the scanning track specifically refers to that the end point of a line segment closest to the origin of the sub-coordinate system is taken as a starting point, and the end points of all the line segments are sequentially connected in an S shape until the last line segment is connected, so that the scanning track is formed.
2. The method for three-dimensional precision detection of surface defects of a metal plate according to claim 1, wherein the method is characterized in that defects are segmented from original pictures to obtain sub-defect picture sets, contour extraction is performed on the sub-defect picture sets, and two-dimensional feature sizes of the defects are calculated, and specifically comprises the following steps:
Marking defects by using end-to-end rectangular frames and noting defect types;
Dividing the defects from the original image according to the coordinates of the rectangular frame marked with the defects by using a picture cutting method to obtain a sub-defect picture set;
And carrying out contour extraction on the sub-defect picture set, and calculating to obtain the two-dimensional feature size of the defect based on a contour extraction result.
3. The three-dimensional precision detection method for the surface defects of the metal sheet according to claim 1, wherein depth information acquired by the depth detection unit is acquired, and depth feature sizes of the defects are obtained according to the depth information, specifically comprising:
Converting the acquired depth-time curve into a depth-displacement curve to obtain a depth profile of a certain section position of the defect;
The depth feature size of the defect is calculated based on the depth profile.
4. A three-dimensional precision inspection assembly for surface defects of sheet metal, the assembly comprising:
The deep learning module acquires a two-dimensional plane picture set of the surface of the tested piece, detects the two-dimensional plane picture set by utilizing a defect identification model obtained by training in advance, and identifies the type and position distribution of defects in the picture;
the image processing module is used for dividing the defects from the original image to obtain a sub-defect picture set, extracting the outline of the sub-defect picture set and calculating to obtain the two-dimensional characteristic size of the defects;
The data processing module is used for obtaining depth information acquired by the depth detection unit, processing the depth information according to the acquired depth information to obtain the depth characteristic size of the defect, and forming the three-dimensional characteristic size of the defect by the depth characteristic size and the two-dimensional characteristic size of the corresponding defect;
The surface quality assessment module is used for detecting the surface quality of the detected piece according to the identified defect types, position distribution and three-dimensional characteristic sizes and identifying microscopic defects;
The path planning module is used for determining a target position and a measurement track of depth detection according to a defect contour extraction result and combining a coordinate system principle, and controlling the depth detection unit to carry out depth detection according to the target position and the measurement track; the path planning module plans a defect three-dimensional scanning track according to the outline and the position distribution of the microscopic defect and controls the depth detection unit to scan the three-dimensional morphology of the microscopic defect according to the scanning track, so that the three-dimensional morphology of the microscopic defect is obtained;
According to the contour extraction result, determining the target position and the measurement track of the depth detection by combining a principle of a coordinate system, wherein the method specifically comprises the following steps:
Converting the position distribution of the defects into a real space coordinate system according to the contour extraction result to obtain the distribution of the defects in space;
Planning a target position and a measurement track of depth detection according to the distribution of defects in space; the measuring track is a straight line segment, is positioned at the position with the largest gray level change rate of the area where the defect is positioned, is uniformly distributed on two sides of the defect, and two end points of the straight line segment are respectively positioned on the edge contour of the defect;
planning a defect three-dimensional scanning track according to the microscopic defect outline and the position distribution, and specifically comprising the following steps:
Based on a three-dimensional morphology scanning method, establishing an optimal polygon or ellipse for covering the defect;
Unidirectional dividing the coverage area to form a scanning track;
The optimal polygon or ellipse covering the defect is specifically: defining a polygon or an ellipse in an image coordinate system to completely cover the position of the defect;
The establishment of the optimal polygon or ellipse should follow the occupation ratio of the scratch area in the coverage area as large as possible;
The unidirectional division coverage area specifically comprises:
establishing a sub-coordinate system in an image coordinate system, and equally dividing the optimal polygon or ellipse along the Y axis of the sub-coordinate system; the segmentation interval is the transverse resolution of the depth detection unit, the segmentation line is parallel to the X axis of the sub-coordinate system and intersects with a polygon or an ellipse to form a multi-parallel line segment; the sub-coordinate system is established specifically as follows: in the polygon coverage area, the Y axis is parallel to the tangential direction of the largest inner angle of the polygon, and the X axis is perpendicular to the Y axis; in the elliptical coverage area, the X axis is parallel to the tangential direction of the maximum point of the elliptical curvature, and the Y axis is perpendicular to the X axis; the optimal polygon or ellipse is positioned in the first quadrant of the sub-coordinate system and tangent to the two coordinate axes;
the formation of the scanning track specifically refers to that the end point of a line segment closest to the origin of the sub-coordinate system is taken as a starting point, and the end points of all the line segments are sequentially connected in an S shape until the last line segment is connected, so that the scanning track is formed.
5. A three-dimensional precision detection system for surface defects of a metal sheet, the system comprising:
The detection assembly of claim 4, which serves as a core for data acquisition, processing and control of the overall system;
the signal control component is used for receiving the control command issued by the detection component so as to drive the mechanical execution component to execute corresponding actions;
The data acquisition component is used for receiving the image data of the surface of the tested piece and the depth information data of the defect and uploading the image data and the depth information data to the detection component;
and the mechanical execution assembly acts according to the movement path planned by the detection assembly to realize corresponding detection.
6. The three-dimensional precision detection system for surface defects of sheet metal according to claim 5, wherein said mechanical actuator assembly comprises:
the space positioning unit comprises a bearing table and a positioning angle;
The three-dimensional movement unit is arranged on the bearing table, a gantry structure is adopted to carry out two-dimensional movement on the surface of the detected piece in a horizontal plane to drive the vision acquisition unit to image the surface of the detected piece, the depth detection unit is moved to a target position, and the depth detection unit is ensured to be positioned at a safe position and a working distance in the vertical direction through movement in the vertical direction;
The visual acquisition unit is used for imaging the surface of the measured piece and acquiring image information of the surface of the measured piece;
and the depth detection unit moves according to the planned movement path and acquires depth information of the defect in real time in the movement process.
7. The three-dimensional precision detection system for surface defects of a metal plate according to claim 6, wherein the bearing table is placed on a horizontal plane, and the positioning angle comprises four right-angle blocks which are respectively installed at detection limit positions [ Xmax, ymax ], [ Xmin, ymin ] of the detection system on the horizontal plane and are used for limiting the position of a detected piece on the bearing table.
8. The three-dimensional precision detection system for surface defects of a metal plate according to claim 7, wherein the right-angle block is made of polytetrafluoroethylene materials.
9. The three-dimensional precision detection system for surface defects of a metal plate according to claim 6, wherein the three-dimensional motion unit comprises an X-direction motor displacement shaft, a Y-direction motor displacement shaft, a Z-direction motor displacement shaft and a portal frame;
The X-direction motor displacement shaft consists of two shafts which are arranged in parallel, and the two shafts have the same stroke, are aligned at two ends and are arranged on the bearing table in parallel to realize transverse synchronous movement;
the bottom end of the portal frame is arranged on a bearing table of the X-direction motor displacement shaft;
The Y-direction motor displacement shaft side is arranged at the top end of the portal frame to realize longitudinal movement;
The Z-direction motor displacement shaft is vertically arranged on the bearing table of the Y-direction motor displacement shaft, so that vertical movement is realized.
10. The three-dimensional precision detection system for surface defects of a metal plate according to claim 9, wherein the vision acquisition unit comprises an industrial camera, a fixed focus lens and a coaxial light source;
the visual acquisition unit is arranged on a bearing table of the Y-direction motor displacement shaft; the industrial camera is connected with the fixed focus lens through a C interface, the coaxial light source is arranged right below the fixed focus lens, and the center of the light emitting surface of the coaxial light source coincides with the center of the optical axis of the industrial camera.
11. The three-dimensional precision detection system for surface defects of a metal plate according to claim 9, wherein the depth detection unit comprises a micro-motion displacement platform and a spectral confocal sensor;
The depth detection unit is arranged on a bearing table of the Z-direction motor displacement shaft; the micro-motion displacement platform consists of an X-direction micro-motion displacement shaft and a Y-direction micro-motion displacement shaft, wherein the X-direction micro-motion displacement shaft and the Y-direction micro-motion displacement shaft are vertically arranged, the X-direction micro-motion displacement shaft and the X-direction motor displacement shaft are arranged in parallel, and the Y-direction micro-motion displacement shaft and the Y-direction motor displacement shaft are arranged in parallel; the spectral confocal sensor is arranged on the micro-motion displacement platform through a clamp, and the X-direction micro-motion displacement shaft and the Y-direction micro-motion displacement shaft realize measuring tracks of different angles in a plane through speed vector synthesis.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311593905.XA CN117630010B (en) | 2023-11-27 | 2023-11-27 | Three-dimensional precision detection method, assembly and system for surface defects of metal plate |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311593905.XA CN117630010B (en) | 2023-11-27 | 2023-11-27 | Three-dimensional precision detection method, assembly and system for surface defects of metal plate |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117630010A CN117630010A (en) | 2024-03-01 |
| CN117630010B true CN117630010B (en) | 2024-10-22 |
Family
ID=90033228
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311593905.XA Active CN117630010B (en) | 2023-11-27 | 2023-11-27 | Three-dimensional precision detection method, assembly and system for surface defects of metal plate |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117630010B (en) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103868929A (en) * | 2013-11-29 | 2014-06-18 | 中科华核电技术研究院有限公司 | Three-dimensional detection method for defects of sealing surface |
Family Cites Families (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104792793B (en) * | 2015-04-28 | 2019-09-10 | 刘凯 | Optical defect detection method and system |
| CN107255673A (en) * | 2017-05-05 | 2017-10-17 | 西安交通大学 | High temperature blade internal flaw three dimensional lossless detection method based on ultrasonic phase array |
| CN111044522B (en) * | 2019-12-14 | 2022-03-11 | 中国科学院深圳先进技术研究院 | Defect detection method and device and terminal equipment |
| CN113970560B (en) * | 2021-10-28 | 2023-05-30 | 中国科学院光电技术研究所 | Defect three-dimensional detection method based on multi-sensor fusion |
| CN114820474A (en) * | 2022-04-11 | 2022-07-29 | 南京拓控信息科技股份有限公司 | Train wheel defect detection method based on three-dimensional information |
| CN116429768A (en) * | 2023-03-27 | 2023-07-14 | 超音速人工智能科技股份有限公司 | Sealing nail welding quality detection method, system, equipment and storage medium |
| CN116148277B (en) * | 2023-04-19 | 2023-07-04 | 武汉精一微仪器有限公司 | Three-dimensional detection method, device and equipment for defects of transparent body and storage medium |
| CN116358842B (en) * | 2023-06-02 | 2023-08-01 | 中国科学院长春光学精密机械与物理研究所 | Method and device for detecting surface defects of large-caliber optical element based on mechanical arm |
| CN117036259A (en) * | 2023-07-27 | 2023-11-10 | 苏州虹尔德金属制品有限公司 | Metal plate surface defect detection method based on deep learning |
-
2023
- 2023-11-27 CN CN202311593905.XA patent/CN117630010B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103868929A (en) * | 2013-11-29 | 2014-06-18 | 中科华核电技术研究院有限公司 | Three-dimensional detection method for defects of sealing surface |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117630010A (en) | 2024-03-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108801914B (en) | Method and system for detecting forming defects of multi-groove profile plate | |
| CN108332689B (en) | Optical measurement system and method for detecting surface roughness and surface damage | |
| CN107356608B (en) | Rapid dark field detection method for surface microdefects of large-caliber fused quartz optical element | |
| CN108956645A (en) | A kind of the optical mirror slip defect detecting device and method of more vision systems | |
| US20200150055A1 (en) | Automatic detection device and method for detecting surface detects of large caliber cambered optical element | |
| TWI426228B (en) | Method and apparatus of shape measurement of specular reflective surface | |
| CN106248686A (en) | Glass surface defects based on machine vision detection device and method | |
| KR101343277B1 (en) | Surface inspection apparatus | |
| CN109211110A (en) | A kind of flat-type workpiece size detection system and method based on machine vision | |
| CN110853018B (en) | Computer vision-based vibration table fatigue crack online detection system and detection method | |
| CN101004389A (en) | Method for detecting 3D defects on surface of belt material | |
| CN105783723A (en) | Machine vision-based precise die surface processing precision detection device and method | |
| CN113134683A (en) | Laser marking method and device based on machine learning | |
| CN108489394A (en) | A kind of large-scale sheet metal works almost T-stable automatic detection device and method | |
| Yang et al. | An automatic aperture detection system for LED cup based on machine vision | |
| CN113219622A (en) | Objective lens focusing method, device and system for panel defect detection | |
| CN110873718A (en) | Steel plate surface defect detection system and method based on machine vision | |
| CN113610083B (en) | Character recognition and character engraving depth detection system and method for vehicle VIN code | |
| CN117630010B (en) | Three-dimensional precision detection method, assembly and system for surface defects of metal plate | |
| CN208042989U (en) | A kind of large-scale sheet metal works almost T-stable automatic detection device | |
| CN109829897A (en) | A kind of gear burr detection method and gear high-precision vision measuring system | |
| CN211452221U (en) | Universal part size tester | |
| CN111583241B (en) | Mobile detection method and device for regularity of ultra-large area honeycomb products | |
| CN211697564U (en) | Optical filter detection device | |
| CN209624417U (en) | Touch screen defect detecting device based on the confocal camera of line |
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 |