CN113218960B - Method for detecting defects of ruby bearing with complex surface - Google Patents
Method for detecting defects of ruby bearing with complex surface Download PDFInfo
- Publication number
- CN113218960B CN113218960B CN202110393320.8A CN202110393320A CN113218960B CN 113218960 B CN113218960 B CN 113218960B CN 202110393320 A CN202110393320 A CN 202110393320A CN 113218960 B CN113218960 B CN 113218960B
- Authority
- CN
- China
- Prior art keywords
- ruby
- image
- area
- ring
- circle
- 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
- 229910001750 ruby Inorganic materials 0.000 title claims abstract description 136
- 239000010979 ruby Substances 0.000 title claims abstract description 135
- 230000007547 defect Effects 0.000 title claims abstract description 103
- 238000000034 method Methods 0.000 title claims abstract description 41
- 238000001514 detection method Methods 0.000 claims abstract description 85
- 238000011068 loading method Methods 0.000 claims abstract description 32
- 238000007781 pre-processing Methods 0.000 claims abstract description 16
- 238000012216 screening Methods 0.000 claims description 15
- 238000001914 filtration Methods 0.000 claims description 14
- 238000004891 communication Methods 0.000 claims description 8
- 230000005484 gravity Effects 0.000 claims description 6
- 230000011218 segmentation Effects 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 238000001035 drying Methods 0.000 claims description 3
- 238000000605 extraction Methods 0.000 claims description 3
- 239000003550 marker Substances 0.000 claims description 3
- 230000003287 optical effect Effects 0.000 claims description 2
- 239000004575 stone Substances 0.000 description 3
- 238000011049 filling Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000012535 impurity Substances 0.000 description 1
- 238000005498 polishing 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
- G01N21/95—Investigating the presence of flaws or contamination characterised by the material or shape of the object to be examined
- G01N21/9515—Objects of complex shape, e.g. examined with use of a surface follower device
-
- 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
- G01N21/84—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
Landscapes
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a method for detecting defects of a ruby bearing with a complex surface, which comprises the following steps: s1, automatic detection and initialization: finishing feeding, and rotating the disc to a specified front detection station; s2, acquiring a first image: the method comprises the steps that a front camera, a 2-fold coaxial telecentric lens and a white point light source acquire an ruby bearing image which is a first coaxial light image; s3, the white point light source 1 is turned off, the ruby is motionless, and the annular combined light of the 0-degree blue annular side light and the back annular 90-degree blue backlight is lightened simultaneously; s4, acquiring a second image: the method comprises the steps that a front camera, a 2-fold coaxial telecentric lens, 0-degree blue ring-shaped side light and back ring-shaped 90-degree blue backlight ring-shaped combined light are combined to obtain a ruby bearing image, and the ruby bearing image is a first ring-shaped light image; s5, loading a first coaxial light image for image preprocessing; s6, performing first coaxial light image defect detection, judging whether a defect result exists, and if so, ending.
Description
The patent application is filed on the date of 2018, 12 and 27 and has the application number of 201811614523.X, and the name of a divisional application of a device and a method for detecting defects of a complex surface ruby bearing.
Technical Field
The invention belongs to the technical field of ruby bearing detection, and particularly relates to a method for detecting defects of a ruby bearing with a complex surface.
Background
Ruby bearings are timepiece function stone elements mounted on timepiece hubs, and are also suitable for use on instruments and various devices. The functional stone element is used for improving the friction stability of the contact surface of the timepiece element and reducing the wear of the timepiece stone. The ruby bearing has the specification that the outer diameter is between 0.7 and 3mm, and is characterized in that a groove surface exists or not, the outer diameter is chamfered, and the surface stepwise change process of hole-chamfer-curved surface-plane-chamfer exists. The surface conditions are complex. In the processing process, the defects of external collapse, kong Beng, fracture, assembly collapse, scratch and the like are easily caused.
At present, for defect detection of ruby bearings, defects are mostly selected visually under manual high-power microscope, and the problems of low efficiency and poor stability exist.
Disclosure of Invention
The invention aims to provide a method for detecting defects of a ruby bearing with a complex surface.
In order to solve the technical problems, the invention adopts the following technical scheme:
The front detection station of the device for detecting the defects of the complex surface ruby bearing comprises a front detection camera, a 2-time coaxial telecentric lens, a white point light source, a red filter, 0-degree blue ring-shaped side light and 90-degree blue ring-shaped backlight; a transparent detection disc; the back detection station comprises a back 90-degree blue ring-shaped backlight, a back red filter, a back 0-degree blue ring-shaped side light, a back 2-time zone coaxial telecentric lens and a back detection camera, wherein:
The transparent detection disc is used for bearing the ruby bearing, and sequentially rotating the ruby bearing to the positions of the front camera and the back camera for detection; the front camera, the 2 times of the coaxial telecentric lens and the white point light source are used for collecting a first coaxial light image of the ruby bearing, calculating the image coordinates and the radius of each ring of the ruby bearing, and detecting defects of the plane and the outline; the front camera, the 2-fold coaxial telecentric lens, the 0-degree blue ring-shaped side light and the 90-degree blue ring-shaped backlight are used for collecting a first annular light image of the ruby bearing and detecting defects on the outer diameter, the groove diameter, the aperture and the curved surface; the red filter is used for filtering ambient stray light; the back camera, the back 2 times of the back camera with the coaxial telecentric lens and the back white point light source are used for collecting a second coaxial light image and detecting defects of the working surface; the back camera, the back 2 times of the coaxial telecentric lens, the back 0-degree blue annular side light and the back 90-degree blue annular backlight are combined together to collect a second annular light image for working face outer diameter chamfering and working hole position defect detection, and the method is characterized by comprising the following steps:
S1, automatic detection and initialization: finishing feeding, and rotating the disc to a specified front detection station;
S2, acquiring a first image: the method comprises the steps that a front camera, a 2-fold coaxial telecentric lens and a white point light source acquire an ruby bearing image which is a first coaxial light image;
S3, the white point light source 1 is turned off, the ruby is motionless, and the annular combined light of the 0-degree blue annular side light and the back annular 90-degree blue backlight is lightened simultaneously;
S4, acquiring a second image: the method comprises the steps that a front camera, a 2-fold coaxial telecentric lens, 0-degree blue ring-shaped side light and back ring-shaped 90-degree blue backlight ring-shaped combined light are combined to obtain a ruby bearing image, and the ruby bearing image is a first ring-shaped light image;
S5, loading a first coaxial light image for image preprocessing: obtaining a ruby area, judging the position of the ruby in the image, calculating position information, calculating sample information, and storing the sample information including the circle center and the radius of each ring into a database; obtaining which surface of the ruby faces upwards, and storing a surface-up information zone bit; dividing the region;
S6, performing first coaxial light image defect detection, judging whether a defect result exists or not, and if yes, ending;
S7, otherwise, loading a first annular light image, and carrying out image preprocessing: loading the center coordinates and the radius of each circular ring in the S5, and dividing the detection area;
S8, performing first annular light image defect detection, judging whether a defect result exists or not, and if yes, ending;
s9, if not, turning to a reverse camera for detection;
S10, only turning on a back white point light source;
S11, collecting a third image: the back camera, the back 2 times of the back camera with the coaxial telecentric lens and the back white point light source acquire an ruby bearing image which is a second coaxial light image;
S12, the back white point light source is turned off, the ruby is motionless, and the back 0-degree blue ring side light and the back 90-degree blue ring backlight ring combined light are simultaneously lightened;
S13, triggering and collecting a fourth image: the back camera, the back 2 times of the back coaxial telecentric lens, the back 0-degree blue ring side light and the back 90-degree blue ring type backlight ring combined light acquire a ruby bearing image which is a second ring light image;
s14, loading a second coaxial light image for image preprocessing: loading the upward facing marker bit of the S5, obtaining a ruby region, judging the position, calculating sample information, and storing the sample information including the circle center and the radius of each ring into a database; dividing a detection area;
S15, performing second coaxial light image defect detection, judging whether a defect result exists, if so, ending;
S16, otherwise, loading a second annular light image for image preprocessing; loading the center coordinates in the S14 to obtain fitted rings, and dividing the detection area;
S17, performing defect detection on the second annular light image, judging whether a defect result exists, if so, ending,
The image preprocessing process in S5 is as follows:
S501, obtaining a ruby region;
S502, judging the view field position of the ruby in the image: multiple pieces, offset, empty field;
s503, calculating position information: calculating sample information by fitting circle centers and circles, wherein the sample information comprises circle centers and circles
The ring radius is saved to a database;
S504, judging which surface of the ruby faces upwards according to the radius of the circular ring, and marking the position 1 facing upwards;
S505, performing detection region segmentation: dividing to obtain all ring areas, and aiming at defect detection: external collapse, loading collapse, scratch, pit, upper collapse, kong Beng; corresponding to the outer collapse area, the mounting collapse area, the plane area, the upper collapse area and the Kong Beng area;
In S503, fitting calculation is performed on the circle center and each circular ring, and the specific steps are as follows;
S5031, before detection, collecting a first coaxial light image with the upper surface of a ruby bearing standard piece upwards, obtaining physical size values of an outer diameter, a groove diameter and a hole diameter and theoretical values of pixel points on the image, and obtaining the proportion of the physical size values and the theoretical values of the pixel points on the image, wherein the proportion is the image magnification Mag;
s5032, automatically acquiring a plurality of black-and-white rings in the current region of the ruby to be detected;
S5033, tracking edge points of the circular ring areas respectively, bringing the edge points into a circular polar coordinate series formula, respectively obtaining the center of gravity of the areas, and carrying out weighted average on the center of gravity of the areas to obtain center coordinates;
s5034, calculating the distance from the contour point of each ring area to the center of the circle by taking the center coordinates as the center, and calculating the radius of each ring;
S5035, obtaining the number of theoretical values of the pixel points of the outer diameter, the groove diameter and the aperture radius image according to the S5031, respectively comparing the theoretical values with the radius of each ring in the S5034, and sequentially determining the ring radius to which the outer diameter, the groove diameter and the aperture belong;
The specific steps for determining which surface of ruby faces upwards in S504 are as follows:
s5041, calculating to obtain a second ring radius R2 and a third ring radius R3 from the circle center outwards;
S5042, if R2 is more than 0.4 x R3, the upper surface of the ruby is upward, and the upper surface upward flag bit is 1;
s5043, otherwise, the ruby is with the lower surface facing upwards, and the upper surface facing upwards is with the flag bit of 0.
2. The method for detecting defects of a complex surface ruby bearing according to claim 1, wherein the specific steps of region segmentation in S505 are as follows:
S5051, dividing a detection area according to the obtained circle center coordinates and the radius of each circular ring;
S5052, exotic region: the outer diameter edge R4 deviates from a circle of 5 outwards pixel points from the circle center, and an outer contour annular area with the width of 10 pixel points is formed by the circle which is close to the inwards 5 pixel points, so as to detect outwards collapse;
S5053, collapse area: the edge R3 deviates from a circle of 10 pixel points outwards from the circle center, and a circle close to 10 pixel points inwards encloses an edge area with 10 pixel points, which is used for detecting the collapse;
s5054, planar area: the method is used for detecting scratches, dirt, cracks and pock defects;
S5055, under the condition that the upper surface of the ruby is upward, the mounting edge R3 is close to a circle of 10 inward pixel points, and the plane area surrounded by the circle of 10 outward pixel points is used for detecting scratches, dirt, cracks and pock defects;
S5056, under the condition that the lower surface of the ruby is upward, the mounting edge R3 is close to a circle of 10 inward pixel points, and the plane area surrounded by the circle of 10 outward pixel points is used for detecting scratches, dirt, cracks and pock defects;
S5057, upper collapse region: only under the condition that the lower surface of the ruby is upward, the area from the circle of 5 pixel points with the radius of the upper edge R2 inward to the area surrounded by 5 pixel points with the radius of the upper edge R2 outward is an upper edge area, and the upper edge area is used for detecting upper collapse;
S5058, kong Beng region: the aperture R1 is outward 5 pixel points, and the area surrounded by the R1 inward 5 pixel points is used for detecting hole collapse.
3. The method for detecting defects of a complex surface ruby bearing according to claim 1, wherein the first coaxial optical image defect detection process in S6 is as follows:
S601, if the upper surface of the ruby is upward, sequentially detecting bad groove size and eccentric groove;
s602, performing upward collapse defect detection, and if yes, judging that the product is unqualified, and exiting;
s603, if not, further, detecting the outward collapse defect, if yes, judging that the product is unqualified, and then exiting;
s604, otherwise, further, carrying out the detection of the loading collapse defect, and if yes, judging that the loading collapse defect is unqualified, and then exiting;
S605, if the lower surface of the ruby is upward, hole collapse detection defects, outward collapse, assembly collapse, cracks and pits are detected.
4. The method for detecting defects of a complex surface ruby bearing according to claim 1, wherein the process of detecting the chipping-off defects in S603 is as follows:
s60301, binarizing the image, and extracting a region with a gray threshold value between 8 and 255;
S60302, calculating all connected areas of the extracted area;
s60303, calculating the minimum circumscribing circle and the minimum circumscribing circle radius OR of all the communication areas;
s60304, screening out an area with the OR size within +/-15 pixel points of the radius of the standard ruby according to the OR size;
S60305, performing open operation on the regions left after screening to remove edge burrs;
s60306, calculating all connected areas of the residual area;
S60307, calculating equivalent ellipses of all the communication areas, and major axis ra, minor axis rb and equidistance anisametry of the equivalent ellipses;
s60308, screening out the areas with ra and anisametry sizes within a threshold range in all the connected areas;
s60309, judging that the area of the residual area is larger than 0, namely, judging that the package is broken, namely, the package is unqualified and exits;
s60310, carrying out Huber robustness fitting circle on the ruby region;
s60311, calculating the difference between the fitted circle area and the ruby original area;
s60312 obtaining all regions of all region extraction differences;
S60313, region conversion; the width of the area, the size of the circular ring, according to tearing with the center of the area, the circular ring is unfolded into a rectangle;
s60314, area drying treatment: performing open operation on all the extracted areas to remove edge burrs, and performing closed operation to remove edge gaps;
s60315, calculating equivalent ellipses of all the connected areas, and calculating the major axis ra of the equivalent ellipses;
S60316, screening out a communication region with ra size between a threshold value and a maximum value;
S60317, calculating the area of the residual area;
S60318, the area of the remaining area >0 is the outward collapse.
The invention has the following beneficial effects: the invention can realize the defect detection of the ruby bearing with the complex surface, and can detect and classify different defects. Compared with the current manual detection, the invention can improve the accuracy and the stability and reduce the labor cost on the premise of not influencing the operation; under the condition of complex surfaces, the polishing scheme of the ruby bearing defects is solved, better prominent defect characteristics are obtained, and the ruby bearing image recognition and detection are carried out according to the defect characteristics.
Drawings
FIG. 1 is a schematic axial view of a detection device according to an embodiment of the present invention;
FIG. 2 is a front view of a sample ruby bearing structure according to an embodiment of the present invention;
FIG. 3 is a graph showing a defect distribution of a sample ruby bearing according to an embodiment of the present invention;
FIG. 4 is a flow chart of a method for detecting defects of a complex surface ruby bearing according to an embodiment of the present invention;
FIG. 5 is a flowchart of an algorithm for detecting defects in a complex surface ruby bearing according to an embodiment of the present invention;
FIG. 6 is an upward view of the upper surface of the ruby under first coaxial light in this reverse embodiment;
FIG. 7 is a view showing the first coaxial light lower ruby of the present reverse embodiment with the lower surface facing upward;
FIG. 8 is a flowchart of an algorithm for detecting an outward collapse defect according to an embodiment of the present invention;
FIG. 9 is a flowchart of an algorithm for detecting a breakage defect according to an embodiment of the present invention;
FIG. 10 is a flowchart of a method algorithm for Kong Beng defect detection in accordance with an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
In specific application implementation, referring to fig. 1, a front camera 1, a 2-fold coaxial telecentric lens 2 and a white point light source 3 are used for collecting a first coaxial light image, calculating an image coordinate of a ruby bearing 6 and the radius of each ring, and detecting defects of a plane and an outline; the front camera 1, the 2 times with the coaxial telecentric lens 2, the 0-degree blue-color ring-shaped side light 5 and the 90-degree blue-color ring-shaped backlight 8 are used for collecting a first ring-shaped light image and detecting defects on the outer diameter, the groove diameter, the aperture and the curved surface; the red filter 4 is used for filtering ambient stray light; the transparent detection disc 7 is used for bearing the ruby bearing 6, and sequentially rotates the ruby bearing 6 to the positions of the front camera 1 and the back camera 14 for detection. The back camera 14, the back 2-fold coaxial telecentric lens 13 and the back white point light source are used for collecting a second coaxial light image and detecting defects of the working surface; the back side camera 14, the 2-fold coaxial telecentric lens 13, the back side 0-degree blue ring-shaped side light and the back side 90-degree blue ring-shaped backlight are combined together to collect a second ring-shaped light image, and working face outer diameter edge chamfering and working hole site defects are carried out.
In the implementation of the specific application, referring to fig. 2, a ruby bearing with a groove having an outer diameter of 1.1-1.6mm is detected, and the ruby bearing mainly comprises an outer diameter chamfer 15, a mounting edge 16, an upper surface 17, a groove edge 18, a groove curved surface 19, a hole site 20 and a lower surface 21. In particular implementations, referring to fig. 3, the detected ruby defect distribution includes an outward break 22, a scratch 23, an inward break 24, an upward break 25, kong Beng, a pit 27, a raised foot 28, and a crack 29.
The embodiment of the invention also provides a method for detecting the defects of the ruby bearing with the complex surface, which is applied to the device, and in a specific application example, referring to fig. 4, the method comprises the following steps:
S1, automatic detection and initialization: finishing feeding, and rotating the disc to a specified front detection station;
S2, acquiring a first image: the method comprises the steps that a front camera, a 2-fold coaxial telecentric lens and a white point light source acquire an ruby bearing image which is a first coaxial light image;
S3, the white point light source 1 is turned off, the ruby is motionless, and the 0-degree blue ring-shaped side light and the back ring-shaped 90-degree blue backlight (ring-shaped combined light) are simultaneously lightened;
S4, acquiring a second image: the method comprises the steps of obtaining an ruby bearing image by the combined action of a front camera, a 2-fold coaxial telecentric lens, 0-degree blue ring-shaped side light and a back ring-shaped 90-degree blue backlight (ring-shaped combined light), wherein the ruby bearing image is a first ring-shaped light image;
S5, loading a first coaxial light image for image preprocessing: obtaining a ruby area, judging the position of the ruby in the image, calculating position information, calculating sample information, and storing the sample information including the circle center and the radius of each ring into a database; obtaining which surface of the ruby faces upwards, and storing a surface-up information zone bit; dividing the region;
S6, performing first coaxial light image defect detection, judging whether a defect result exists or not, and if yes, ending;
S7, otherwise, loading a first annular light image, and carrying out image preprocessing: loading the center coordinates and the radius of each circular ring in the S5, and dividing the detection area;
S8, performing first annular light image defect detection, judging whether a defect result exists or not, and if yes, ending;
s9, if not, turning to a reverse camera for detection;
S10, only turning on a back white point light source;
S11, collecting a third image: the back camera, the back 2 times of the back camera with the coaxial telecentric lens and the back white point light source acquire an ruby bearing image which is a second coaxial light image;
S12, the back white point light source is turned off, the ruby is motionless, and the back 0-degree blue ring side light and the back 90-degree blue ring backlight (ring combined light) are simultaneously turned on;
S13, triggering and collecting a fourth image: a back camera, a back 2 times of coaxial telecentric lens, back 0-degree blue ring side light and back 90-degree blue ring backlight (ring combined light) acquire a ruby bearing image which is a second ring light image,
S14, loading a second coaxial light image for image preprocessing: loading the upward facing marker bit of the S5, obtaining a ruby region, judging the position, calculating sample information, and storing the sample information including the circle center and the radius of each ring into a database; dividing a detection area;
S15, performing second coaxial light image defect detection, judging whether a defect result exists, if so, ending;
S16, otherwise, loading a second annular light image for image preprocessing; loading the center coordinates in the S14 to obtain fitted rings, and dividing the detection area;
and S17, performing defect detection on the second annular light image, judging whether a defect result exists, and if so, ending.
Further, the image preprocessing process is as follows:
S501, obtaining a ruby region;
S502, judging the view field position of the ruby in the image: multiple pieces, offset, empty field;
s503, calculating position information: calculating sample information by fitting circle centers and circles, wherein the sample information comprises circle centers and circles
The ring radius is saved to a database;
S504, judging which surface of the ruby faces upwards according to the radius of the circular ring, and marking the position 1 facing upwards;
s505, performing detection region segmentation: dividing to obtain all ring areas, and aiming at defect missing detection: external collapse, loading collapse, scratch, pit, upper collapse, kong Beng; corresponding to the outer collapse area, the mounting collapse area, the plane area, the upper collapse area and the Kong Beng area.
In specific implementation application, the fitting calculation of the circle center and each ring in S503 comprises the following specific steps;
S5031, before detection, collecting a first coaxial light image with the upper surface of a ruby bearing standard piece upwards, obtaining physical size values of an outer diameter, a groove diameter and a hole diameter and theoretical values of pixel points on the image, and obtaining the proportion of the physical size values and the theoretical values of the pixel points on the image, wherein the proportion is the image magnification Mag;
s5032, automatically acquiring a plurality of black-and-white rings for the current region of ruby to be detected
S5033, tracking edge points of the circular ring areas respectively, bringing the edge points into a circular polar coordinate series formula, acquiring the center of gravity of the areas respectively, and carrying out weighted average on the center of gravity of the areas to obtain center coordinates.
S5034, calculating the distance from the contour point of each circular ring area to the center by taking the center coordinates as the center,
And counting to obtain the radius of each ring.
S5035, obtaining the number of theoretical values of the pixel points of the outer diameter, the groove diameter and the aperture radius according to the S5031, respectively comparing the theoretical values with the radii of the circular rings in the S5034, and sequentially determining the circular ring radii of the outer diameter, the groove diameter and the aperture. In specific implementation application: the specific steps for determining which surface of ruby faces upwards in S504 are as follows:
s5041, calculating to obtain a second ring radius R2 and a third ring radius R3 from the circle center outwards;
s5042, if R2 is greater than 0.4R 3, the upper surface of the ruby is upward, as shown in FIG. 6, the upper surface upward flag is 1;
S5043, otherwise, the ruby is with its lower surface facing upward, as shown in fig. 7, and the upper surface facing upward has a flag bit of 0.
In specific implementation application: the specific steps of the region segmentation in S505 are as follows:
S5051, obtaining a circle center coordinate and the radius of each circular ring according to 503, and dividing a detection area;
S5052, exotic region: the outer diameter edge R4 deviates from a circle of 5 outwards pixel points from the circle center, and an outer contour annular area with the width of 10 pixel points is formed by the circle which is close to the inwards 5 pixel points, so as to detect outwards collapse;
S5053, collapse area: the edge R3 deviates from a circle of 10 pixel points outwards from the circle center, and a circle close to 10 pixel points inwards encloses an edge area with 10 pixel points, which is used for detecting the collapse;
s5054, planar area: the method is used for detecting defects such as scratches, dirt, cracks, pits and the like;
s5055, under the condition that the upper surface of the ruby is upward, the mounting edge R3 is close to a circle of 10 pixel points inward, and the plane area surrounded by the circle of 10 pixel points outward is used for detecting defects such as scratches, dirt, cracks, pits and the like;
s5056, under the condition that the lower surface of the ruby is upward, the mounting edge R3 is close to a circle of 10 pixels inwards, and the plane area surrounded by the circle of 10 pixels outwards is defined by R2, so that defect detection such as scratch, dirt, crack, pit and the like is detected;
S5057, upper collapse region: only under the condition that the lower surface of the ruby is upward, the area from the circle of 5 pixel points with the radius of the upper edge R2 inward to the area surrounded by 5 pixel points with the radius of the upper edge R2 outward is an upper edge area, and the upper edge area is used for detecting upper collapse;
S5058, kong Beng region: the aperture R1 is outward 5 pixel points, and the area surrounded by the R1 inward 5 pixel points is used for detecting hole collapse.
In a specific application example, referring to fig. 5, the first coaxial light image defect detection processing procedure is as follows:
S601, if the upper surface of ruby is upward, sequentially performing groove size failure, groove eccentricity,
S602, performing upward collapse defect detection, and if yes, judging that the product is unqualified, and exiting;
s603, if not, further, detecting the outward collapse defect, if yes, judging that the product is unqualified, and then exiting;
s604, otherwise, further, carrying out the detection of the loading collapse defect, and if yes, judging that the loading collapse defect is unqualified, and then exiting;
S605, if the lower surface of the ruby is upward, hole collapse detection defects, outward collapse, assembly collapse, cracks and pits are detected.
Further, in a specific application example, referring to fig. 8, the process of performing the outward collapse defect detection in S603 is as follows:
s60301, binarizing the image, and extracting a region with a gray threshold value between 8 and 255;
S60302, calculating all connected areas of the extracted area;
s60303, calculating the minimum circumscribing circle and the minimum circumscribing circle radius OR of all the communication areas;
s60304, screening out the area with the OR size within +/-15 pixel points of the radius of the standard ruby according to the OR size
S60305, performing open operation on the regions remained after screening to remove edge burrs
S60306, calculating all connected areas of the residual area;
s60307 calculating equivalent ellipses of all connected regions, and major axis ra, minor axis rb and equidistant anisametry (major axis/minor axis) of equivalent ellipses
S60308, screening out the areas with ra and anisametry in the threshold range in all the connected areas
S60309 judging that the residual area is greater than 0 as the loading collapse, namely the unqualified exit
S60310, huber robustness fitting circle is carried out on ruby region
S60311, calculating the difference between the fitted circle area and the ruby original area
S60312 obtaining all regions of all region extraction differences
S60313, region conversion; the width of the area, the size of the circular ring, according to tearing with the center of the area, the circular ring is unfolded into a rectangle;
s60314, area drying treatment: performing open operation on all the extracted areas to remove edge burrs, and performing close operation to remove edge gaps
S60315, calculating equivalent ellipses of all the connected areas, and calculating the major axis ra of the equivalent ellipses
S60316, screening out a communication region with ra size between a threshold value and a maximum value
S60317 calculating the remaining area
S60318, if the area of the residual area is more than 0, the area is qualified and is an outward collapse
In a specific application embodiment, as shown in fig. 9, the specific process of the disintegrating defect in S604 is as follows:
s60401 the ruby region shears off the outer ring leaving the in-ring region
S60402, carrying out Huber robustness fitting circle on the ring-in region
S60403, calculating the difference between the fitted circle region and the in-loop original region
S60404, calculating all connected areas of the difference area
S60405, calculating equivalent ellipses of all connected areas, and major axis ra, minor axis rb and length ratio anisametry (major axis/minor axis) of the equivalent ellipses
S60406, screening out the areas with ra, rb and anisametry in the threshold range in all the connected areas
S60407 filtering all regions with intersections exceeding one or less than one
S60408, calculating the length, width and included angle between rectangle and X axis of the circumscribed rectangle of the residual region
S60409, if the area length is greater than the set threshold value, the area is broken;
S60410, if 80 ° < angle <100 ° or the area width is not equal to 0, the package collapse;
S60411, in the case of a ring light image: extracting an annular light image ruby region;
s60412, extracting the ruby outer ring to create a standard circle by taking the ruby center as a center point and taking the standard radius as a radius
S60413, calculating a difference zone ZB01 of the standard circle and the ruby zone
S60414 Small impurity point of filtration zone ZB01
S60415, calculating convex hulls of ruby areas
S60416 calculating the difference zone ZB02 of convex hull and ruby
S60417, smaller miscellaneous points of the filtration zone ZB02
S60418, calculating a union region of the region ZB01 and the region ZB02
S60419, calculating all connected areas of the union area
S60420, calculating the most circumscribed rectangle of all the connected areas, and calculating the intersection of the circumscribed rectangle and the ruby area
S60421 filtering all regions with intersections of more than one or less than one
S60422, calculating the length, width and included angle between rectangle and X axis of the circumscribed rectangle of the residual region
S60423, if the area length is greater than the set threshold value, the area is broken;
s60424, if 80 ° < angle <100 ° or the area width is not equal to 0, the package is collapsed.
In a specific application example, as shown in fig. 10, kong Beng defect detection in S605 is as follows:
s60501, carrying out 2-group mean filtering on the hole site region, and taking a local bright boundary for binarization according to the 2-group result after filtering;
S60502, calculating the area with the largest area in the binarized area to obtain an area KLR;
S60503, filling holes with the area size of 1 to 100 pixels in the KLR of the area to obtain an area KLRFUS, and calculating the number of the remaining holes in the area;
s60504, filling all holes in the area KLR to obtain an area KLRFU;
S60505, calculating the difference between the region KLR and the region KLRFU to obtain a difference region, and obtaining a minimum circumcircle of the difference region to obtain a minimum circumcircle region RT;
S60506, the number of remaining holes in the region KLRFUS is more than 0;
S60507, calculating a fitting circle of the region KLRFU, and calculating a radius of the fitting circle;
S60508, KLRFU fitting a circle radius >0;
S60509, taking the center of the area KLRFU as the center to form a standard hole circular area;
s60510, calculating the difference between the standard hole area and the area KLRFU, and filtering small-area miscellaneous points to obtain an area A;
s60511, performing open operation on the region KLRFU to remove edge burrs, so as to obtain a region RO;
S60512, calculating the difference between the region RO and the region KLRFU, and filtering small-area miscellaneous points to obtain a region B;
s60513, respectively carrying out 2 groups of mean filtering on the original image of the region KLRFU, and taking a local bright boundary according to the obtained 2 groups of results after filtering to carry out binarization to obtain a region RDT;
S60514, calculating the difference between the region RDT and the region RT, and filtering small-area miscellaneous points to obtain a region C;
S60515, fusing the regions A, B, C to obtain a region SKBR, and calculating the equivalent ellipses and the equivalent ellipse length and the equivalent ellipse minor axis of all the subregions in the SKBR;
S60516, screening according to the equivalent elliptical long axis length of all the subareas in the area SKBR within a threshold range;
s60517, the area of the residual area after screening is >0, if yes, kong Beng is obtained, or the area is qualified.
It should be understood that the exemplary embodiments described herein are illustrative and not limiting. Although one or more embodiments of the present invention have been described with reference to the accompanying drawings, it will be understood by those of ordinary skill in the art that various changes in form and details may be made therein without departing from the spirit and scope of the present invention as defined by the following claims.
Claims (4)
1. The front detection station of the device for detecting the defects of the complex surface ruby bearing comprises a front detection camera, a 2-time coaxial telecentric lens, a white point light source, a red filter, 0-degree blue ring-shaped side light and 90-degree blue ring-shaped backlight; a transparent detection disc; the back detection station comprises a back 90-degree blue ring-shaped backlight, a back red filter, a back 0-degree blue ring-shaped side light, a back 2-time zone coaxial telecentric lens and a back detection camera, wherein:
The transparent detection disc is used for bearing the ruby bearing, and sequentially rotating the ruby bearing to the positions of the front camera and the back camera for detection; the front camera, the 2 times of the coaxial telecentric lens and the white point light source are used for collecting a first coaxial light image of the ruby bearing, calculating the image coordinates and the radius of each ring of the ruby bearing, and detecting defects of the plane and the outline; the front camera, the 2-time coaxial telecentric lens, the 0-degree blue ring-shaped side light and the 90-degree blue ring-shaped backlight are used for collecting a first annular light image of the ruby bearing and detecting defects on the outer diameter, the groove diameter, the aperture and the curved surface; the red filter is used for filtering ambient stray light; the back camera, the back 2 times of the back camera with the coaxial telecentric lens and the back white point light source are used for collecting a second coaxial light image and detecting defects of the working surface; the back camera, the back 2 times of the coaxial telecentric lens, the back 0-degree blue annular side light and the back 90-degree blue annular backlight are combined together to collect a second annular light image for working face outer diameter chamfering and working hole position defect detection, and the method is characterized by comprising the following steps:
S1, automatic detection and initialization: finishing feeding, and rotating the disc to a specified front detection station;
S2, acquiring a first image: the method comprises the steps that a front camera, a 2-fold coaxial telecentric lens and a white point light source acquire an ruby bearing image which is a first coaxial light image;
s3, the white point light source is turned off, the ruby is motionless, and the annular combined light of the 0-degree blue annular side light and the back annular 90-degree blue backlight is lightened simultaneously;
s4, acquiring a second image: the method comprises the steps that a front camera, a 2-fold coaxial telecentric lens, 0-degree blue ring-shaped side light And and a 90-degree blue backlight ring-shaped combined light on the back face jointly act to obtain a ruby bearing image which is a first ring-shaped light image;
S5, loading a first coaxial light image for image preprocessing: obtaining a ruby area, judging the position of the ruby in the image, calculating position information, calculating sample information, and storing the sample information including the circle center and the radius of each ring into a database; obtaining which surface of the ruby faces upwards, and storing a surface-up information zone bit; dividing the region;
S6, performing first coaxial light image defect detection, judging whether a defect result exists or not, and if yes, ending;
S7, otherwise, loading a first annular light image, and carrying out image preprocessing: loading the center coordinates and the radius of each circular ring in the S5, and dividing the detection area;
S8, performing first annular light image defect detection, judging whether a defect result exists or not, and if yes, ending;
s9, if not, turning to a reverse camera for detection;
S10, only turning on a back white point light source;
S11, collecting a third image: the back camera, the back 2 times of the back camera with the coaxial telecentric lens and the back white point light source acquire an ruby bearing image which is a second coaxial light image;
S12, the back white point light source is turned off, the ruby is motionless, and the back 0-degree blue ring side light and the back 90-degree blue ring backlight ring combined light are simultaneously lightened;
S13, triggering and collecting a fourth image: the back camera, the back 2 times of the back coaxial telecentric lens, the back 0-degree blue ring side light and the back 90-degree blue ring type backlight ring combined light acquire a ruby bearing image which is a second ring light image;
s14, loading a second coaxial light image for image preprocessing: loading the upward facing marker bit of the S5, obtaining a ruby region, judging the position, calculating sample information, and storing the sample information including the circle center and the radius of each ring into a database; dividing a detection area;
S15, performing second coaxial light image defect detection, judging whether a defect result exists, if so, ending;
S16, otherwise, loading a second annular light image for image preprocessing; loading the center coordinates in the S14 to obtain fitted rings, and dividing the detection area;
S17, performing defect detection on the second annular light image, judging whether a defect result exists, if so, ending,
The image preprocessing process in S5 is as follows:
S501, obtaining a ruby region;
S502, judging the view field position of the ruby in the image: multiple pieces, offset, empty field;
s503, calculating position information: calculating sample information by fitting the circle center and each circle, wherein the sample information comprises the circle center and the radius of each circle;
S504, judging which surface of the ruby faces upwards according to the radius of the circular ring, and marking the position 1 facing upwards;
S505, performing detection region segmentation: dividing to obtain all ring areas, and aiming at defect detection: external collapse, loading collapse, scratch, pit, upper collapse, kong Beng; corresponding to the outer collapse area, the mounting collapse area, the plane area, the upper collapse area and the Kong Beng area;
In S503, fitting calculation is performed on the circle center and each circular ring, and the specific steps are as follows;
S5031, before detection, collecting a first coaxial light image with the upper surface of a ruby bearing standard piece upwards, obtaining physical size values of an outer diameter, a groove diameter and a hole diameter and theoretical values of pixel points on the image, and obtaining the proportion of the physical size values and the theoretical values of the pixel points on the image, wherein the proportion is the image magnification Mag;
s5032, automatically acquiring a plurality of black-and-white rings in the current region of the ruby to be detected;
S5033, tracking edge points of the circular ring areas respectively, bringing the edge points into a circular polar coordinate series formula, respectively obtaining the center of gravity of the areas, and carrying out weighted average on the center of gravity of the areas to obtain center coordinates;
s5034, calculating the distance from the contour point of each ring area to the center of the circle by taking the center coordinates as the center, and calculating the radius of each ring;
S5035, obtaining the number of theoretical values of the pixel points of the outer diameter, the groove diameter and the aperture radius image according to the S5031, respectively comparing the theoretical values with the radius of each ring in the S5034, and sequentially determining the ring radius to which the outer diameter, the groove diameter and the aperture belong;
The specific steps for determining which surface of ruby faces upwards in S504 are as follows:
s5041, calculating to obtain a second ring radius R2 and a third ring radius R3 from the circle center outwards;
S5042, if R2 is more than 0.4 x R3, the upper surface of the ruby is upward, and the upper surface upward flag bit is 1;
s5043, otherwise, the ruby is with the lower surface facing upwards, and the upper surface facing upwards is with the flag bit of 0.
2. The method for detecting defects of a complex surface ruby bearing according to claim 1, wherein the specific steps of region segmentation in S505 are as follows:
S5051, dividing a detection area according to the obtained circle center coordinates and the radius of each circular ring;
S5052, exotic region: the outer diameter edge R4 deviates from a circle of 5 outwards pixel points from the circle center, and an outer contour annular area with the width of 10 pixel points is formed by the circle which is close to the inwards 5 pixel points, so as to detect outwards collapse;
S5053, collapse area: the edge R3 deviates from a circle of 10 pixel points outwards from the circle center, and a circle close to 10 pixel points inwards encloses an edge area with 10 pixel points, which is used for detecting the collapse;
s5054, planar area: the method is used for detecting scratches, dirt, cracks and pock defects;
S5055, under the condition that the upper surface of the ruby is upward, the mounting edge R3 is close to a circle of 10 inward pixel points, and the plane area surrounded by the circle of 10 outward pixel points is used for detecting scratches, dirt, cracks and pock defects;
S5056, under the condition that the lower surface of the ruby is upward, the mounting edge R3 is close to a circle of 10 inward pixel points, and the plane area surrounded by the circle of 10 outward pixel points is used for detecting scratches, dirt, cracks and pock defects;
S5057, upper collapse region: only under the condition that the lower surface of the ruby is upward, the area from the circle of 5 pixel points with the radius of the upper edge R2 inward to the area surrounded by 5 pixel points with the radius of the upper edge R2 outward is an upper edge area, and the upper edge area is used for detecting upper collapse;
S5058, kong Beng region: the aperture R1 is outward 5 pixel points, and the area surrounded by the R1 inward 5 pixel points is used for detecting hole collapse.
3. The method for detecting defects of a complex surface ruby bearing according to claim 1, wherein the first coaxial optical image defect detection process in S6 is as follows:
S601, if the upper surface of the ruby is upward, sequentially detecting bad groove size and eccentric groove;
s602, performing upward collapse defect detection, and if yes, judging that the product is unqualified, and exiting;
s603, if not, further, detecting the outward collapse defect, if yes, judging that the product is unqualified, and then exiting;
s604, otherwise, further, carrying out the detection of the loading collapse defect, and if yes, judging that the loading collapse defect is unqualified, and then exiting;
S605, if the lower surface of the ruby is upward, hole collapse detection defects, outward collapse, assembly collapse, cracks and pits are detected.
4. The method for detecting defects of a complex surface ruby bearing according to claim 1, wherein the process of detecting the chipping-off defects in S603 is as follows:
s60301, binarizing the image, and extracting a region with a gray threshold value between 8 and 255;
S60302, calculating all connected areas of the extracted area;
s60303, calculating the minimum circumscribing circle and the minimum circumscribing circle radius OR of all the communication areas;
s60304, screening out an area with the OR size within +/-15 pixel points of the radius of the standard ruby according to the OR size;
S60305, performing open operation on the regions left after screening to remove edge burrs;
s60306, calculating all connected areas of the residual area;
S60307, calculating equivalent ellipses of all the communication areas, and major axis ra, minor axis rb and equidistance anisametry of the equivalent ellipses;
s60308, screening out the areas with ra and anisametry sizes within a threshold range in all the connected areas;
s60309, judging that the area of the residual area is larger than 0, namely, judging that the package is broken, namely, the package is unqualified and exits;
s60310, carrying out Huber robustness fitting circle on the ruby region;
s60311, calculating the difference between the fitted circle area and the ruby original area;
s60312 obtaining all regions of all region extraction differences;
S60313, region conversion; the width of the area, the size of the circular ring, according to tearing with the center of the area, the circular ring is unfolded into a rectangle;
s60314, area drying treatment: performing open operation on all the extracted areas to remove edge burrs, and performing closed operation to remove edge gaps;
s60315, calculating equivalent ellipses of all the connected areas, and calculating the major axis ra of the equivalent ellipses;
S60316, screening out a communication region with ra size between a threshold value and a maximum value;
S60317, calculating the area of the residual area;
S60318, the area of the remaining area >0 is the outward collapse.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110393320.8A CN113218960B (en) | 2018-12-27 | 2018-12-27 | Method for detecting defects of ruby bearing with complex surface |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110393320.8A CN113218960B (en) | 2018-12-27 | 2018-12-27 | Method for detecting defects of ruby bearing with complex surface |
| CN201811614523.XA CN110006920B (en) | 2018-12-27 | 2018-12-27 | Device and method for detecting defects of ruby bearing with complex surface |
Related Parent Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811614523.XA Division CN110006920B (en) | 2018-12-27 | 2018-12-27 | Device and method for detecting defects of ruby bearing with complex surface |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113218960A CN113218960A (en) | 2021-08-06 |
| CN113218960B true CN113218960B (en) | 2024-05-24 |
Family
ID=67165284
Family Applications (2)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110393320.8A Active CN113218960B (en) | 2018-12-27 | 2018-12-27 | Method for detecting defects of ruby bearing with complex surface |
| CN201811614523.XA Active CN110006920B (en) | 2018-12-27 | 2018-12-27 | Device and method for detecting defects of ruby bearing with complex surface |
Family Applications After (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811614523.XA Active CN110006920B (en) | 2018-12-27 | 2018-12-27 | Device and method for detecting defects of ruby bearing with complex surface |
Country Status (1)
| Country | Link |
|---|---|
| CN (2) | CN113218960B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113218960B (en) * | 2018-12-27 | 2024-05-24 | 浙江大学台州研究院 | Method for detecting defects of ruby bearing with complex surface |
| CN110672624B (en) * | 2019-10-22 | 2022-03-15 | 北京领邦智能装备股份公司 | Workpiece side imaging system, defect detection system and workpiece side imaging method |
| CN111294491B (en) * | 2020-02-13 | 2021-01-22 | 北京领邦智能装备股份公司 | Image forming apparatus and image forming method |
| CN111412949B (en) * | 2020-04-13 | 2022-04-01 | 杭州思元智能科技有限公司 | Electronic component detection method and detection device |
| CN111623961A (en) * | 2020-06-10 | 2020-09-04 | 太仓考斯茂石英有限公司 | Glass lens detector and control method |
| CN111812098A (en) * | 2020-06-22 | 2020-10-23 | 燕山大学 | Ceramic bowl defective product marking device and marking method based on machine vision |
| CN112330605B (en) * | 2020-10-20 | 2023-04-07 | 湖北爱默思智能检测装备有限公司 | Ruby color grading method |
| CN113205499A (en) * | 2021-04-30 | 2021-08-03 | 聚时科技(上海)有限公司 | Bearing defect modular detection device and method based on machine vision |
| CN114813744A (en) * | 2022-03-30 | 2022-07-29 | 征图新视(江苏)科技股份有限公司 | Multi-station image comprehensive re-judgment defect detection method |
| CN118297940B (en) * | 2024-05-30 | 2024-09-13 | 泰山石膏(宜宾)有限公司 | Quality control method, device, equipment and medium for gypsum board production line |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| NL7306561A (en) * | 1972-05-19 | 1973-11-21 | ||
| JP2002202264A (en) * | 2000-12-27 | 2002-07-19 | Toshiba Corp | Method and device for appearance inspection |
| JP2002318195A (en) * | 2001-04-19 | 2002-10-31 | Murata Mfg Co Ltd | External appearance inspection method and external appearance inspection device |
| WO2004001285A2 (en) * | 2002-06-21 | 2003-12-31 | Pressco Technology Inc. | Patterned illumination method and apparatus for machine vision systems |
| CN1774721A (en) * | 2003-04-16 | 2006-05-17 | 热生物之星公司 | Detection, resolution, and identification of arrayed elements |
| JP2012159321A (en) * | 2011-01-31 | 2012-08-23 | Hitachi Information & Control Solutions Ltd | Can management device and can management method |
| CN104081185A (en) * | 2012-02-03 | 2014-10-01 | 锡克拜控股有限公司 | Watch identification and authentication system and method |
| CN105987917A (en) * | 2015-02-09 | 2016-10-05 | 苏州兰叶光电科技有限公司 | Sapphire crystal growing defect and surface defect optical detecting method and detecting system thereof |
| CN107315012A (en) * | 2017-06-22 | 2017-11-03 | 福建省万龙新材料科技有限公司 | Composite polycrystal-diamond end face collapses the intelligent detecting method at angle |
| CN206725456U (en) * | 2017-04-20 | 2017-12-08 | 浙江工业大学 | A kind of retainer positive and negative pattern recognition device |
| CN107884413A (en) * | 2017-10-24 | 2018-04-06 | 华东交通大学 | A kind of device and detection method of automatic detection bearing roller missing |
| CN107945180A (en) * | 2017-12-26 | 2018-04-20 | 浙江大学台州研究院 | Come from the visible detection method of the shallow cut in quartz wafer surface of polishing |
| CN207992073U (en) * | 2018-01-25 | 2018-10-19 | 惠州高视科技有限公司 | Battery battery core open defect testing agency |
| CN110006920A (en) * | 2018-12-27 | 2019-07-12 | 浙江大学台州研究院 | A kind of device and method of complex surface ruby bearing defect detection |
Family Cites Families (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008082900A (en) * | 2006-09-27 | 2008-04-10 | Sumitomo Electric Ind Ltd | Appearance inspecting apparatus |
| JP5368247B2 (en) * | 2009-10-16 | 2013-12-18 | Ｎｔｎ株式会社 | Thrust bearing inspection method and inspection apparatus |
| JP2012161951A (en) * | 2011-02-04 | 2012-08-30 | Komori Corp | Offset printing machine for printing securities |
| CN102706882B (en) * | 2012-04-28 | 2014-04-02 | 东信和平科技股份有限公司 | Visual inspection method for SIM (Subscriber Identity Module) |
| CN204448629U (en) * | 2014-12-25 | 2015-07-08 | 衢州亿龙信息技术有限公司 | A kind of precision bearing ball automatic Surface Defect Detection System |
| CN205003098U (en) * | 2015-10-14 | 2016-01-27 | 中国计量学院 | Brake block appearance imperfections multistation on -line measuring device |
| CN105938110A (en) * | 2016-03-11 | 2016-09-14 | 杭州利珀科技有限公司 | Metal can image acquisition device and acquisition method |
| CN106442566A (en) * | 2016-11-03 | 2017-02-22 | 中航工业哈尔滨轴承有限公司 | Tool special for bearing appearance quality inspection and using method |
| CN107894424A (en) * | 2017-11-17 | 2018-04-10 | 重庆远创光电科技有限公司 | Efficient vision detection system for bearing open defect |
| CN108956618B (en) * | 2018-06-25 | 2020-08-11 | 衢州学院 | Tapered roller bearing surface detection method based on frequency domain vision |
-
2018
- 2018-12-27 CN CN202110393320.8A patent/CN113218960B/en active Active
- 2018-12-27 CN CN201811614523.XA patent/CN110006920B/en active Active
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| NL7306561A (en) * | 1972-05-19 | 1973-11-21 | ||
| JP2002202264A (en) * | 2000-12-27 | 2002-07-19 | Toshiba Corp | Method and device for appearance inspection |
| JP2002318195A (en) * | 2001-04-19 | 2002-10-31 | Murata Mfg Co Ltd | External appearance inspection method and external appearance inspection device |
| WO2004001285A2 (en) * | 2002-06-21 | 2003-12-31 | Pressco Technology Inc. | Patterned illumination method and apparatus for machine vision systems |
| CN1774721A (en) * | 2003-04-16 | 2006-05-17 | 热生物之星公司 | Detection, resolution, and identification of arrayed elements |
| JP2012159321A (en) * | 2011-01-31 | 2012-08-23 | Hitachi Information & Control Solutions Ltd | Can management device and can management method |
| CN104081185A (en) * | 2012-02-03 | 2014-10-01 | 锡克拜控股有限公司 | Watch identification and authentication system and method |
| CN105987917A (en) * | 2015-02-09 | 2016-10-05 | 苏州兰叶光电科技有限公司 | Sapphire crystal growing defect and surface defect optical detecting method and detecting system thereof |
| CN206725456U (en) * | 2017-04-20 | 2017-12-08 | 浙江工业大学 | A kind of retainer positive and negative pattern recognition device |
| CN107315012A (en) * | 2017-06-22 | 2017-11-03 | 福建省万龙新材料科技有限公司 | Composite polycrystal-diamond end face collapses the intelligent detecting method at angle |
| CN107884413A (en) * | 2017-10-24 | 2018-04-06 | 华东交通大学 | A kind of device and detection method of automatic detection bearing roller missing |
| CN107945180A (en) * | 2017-12-26 | 2018-04-20 | 浙江大学台州研究院 | Come from the visible detection method of the shallow cut in quartz wafer surface of polishing |
| CN207992073U (en) * | 2018-01-25 | 2018-10-19 | 惠州高视科技有限公司 | Battery battery core open defect testing agency |
| CN110006920A (en) * | 2018-12-27 | 2019-07-12 | 浙江大学台州研究院 | A kind of device and method of complex surface ruby bearing defect detection |
Non-Patent Citations (3)
| Title |
|---|
| Intelligent fault pattern recognition of aerial photovoltaic module images based on deep learning technique;Li X, Yang Q, Wang J, et al;《J. Syst. Cybern. Inf》;20181220(第16期);67-71 * |
| 一种基于图像展开与拼接的精密轴承表面缺陷光学检测方法;杨加东;谢明;;《机床与液压》;20170228;第45卷(第04期);2251-2253 * |
| 动态图像处理系统的设计与研究;陈浙泊, 林斌;《光学仪器》;20031230;第25卷(第06期);35-37 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110006920B (en) | 2021-06-15 |
| CN110006920A (en) | 2019-07-12 |
| CN113218960A (en) | 2021-08-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113218960B (en) | Method for detecting defects of ruby bearing with complex surface | |
| CN114723701B (en) | Gear defect detection method and system based on computer vision | |
| CN113362326B (en) | Method and device for detecting defects of welding spots of battery | |
| CN105784713B (en) | Sealing ring detection method of surface flaw based on machine vision | |
| CN107315012B (en) | Composite polycrystal-diamond end face collapses the intelligent detecting method at angle | |
| CN105548212A (en) | System and method for online automatically detecting defects on glass surface of touch screen | |
| CN105548208A (en) | Method for detecting surface defects of ceramic valve cores based on machine vision | |
| CN114913177B (en) | Automobile part defect detection method based on Hough circle | |
| CN115020267A (en) | Semiconductor surface defect detection method | |
| CN106780473A (en) | A kind of magnet ring defect multi-vision visual detection method and system | |
| CN111272768B (en) | Ceramic tube detection method | |
| CN114994060B (en) | Intelligent detection system and method for magnetic ring appearance defects under machine vision | |
| CN109934808A (en) | One kind being based on image Multiple Shape normal direction gradient difference value pcb board defect classification method | |
| CN115471466A (en) | Steel surface defect detection method and system based on artificial intelligence | |
| CN116012292A (en) | Wafer appearance defect detection method based on machine vision | |
| CN113554582A (en) | Defect detection method, device and system for functional hole in cover plate of electronic equipment | |
| CN110345874A (en) | A kind of new method based on mechanical vision inspection technology measurement pipe tobacco width | |
| CN111122590A (en) | Ceramic surface defect detection device and detection method | |
| CN111766245A (en) | Button battery negative electrode shell defect detection method based on machine vision | |
| CN114897881A (en) | Crystal grain defect detection method based on edge characteristics | |
| CN106529551A (en) | Intelligent recognition, counting, and detection algorithm for quasi-circle object facing packaging industry | |
| CN113781458A (en) | Artificial intelligence based identification method | |
| CN103177426B (en) | Morphology-based strong-interference laser edge image restoration method | |
| CN109816627A (en) | The small and weak defect object detection method in flat glass parts ink area | |
| CN115866502A (en) | Microphone part surface defect online detection process |
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 |