CN113686892B - Novel bearing surface defect intelligent detection system - Google Patents
Novel bearing surface defect intelligent detection system Download PDFInfo
- Publication number
- CN113686892B CN113686892B CN202110960356.XA CN202110960356A CN113686892B CN 113686892 B CN113686892 B CN 113686892B CN 202110960356 A CN202110960356 A CN 202110960356A CN 113686892 B CN113686892 B CN 113686892B
- Authority
- CN
- China
- Prior art keywords
- bearing
- motor
- stepping motor
- light source
- box body
- 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 47
- 238000001514 detection method Methods 0.000 title claims abstract description 44
- 230000000007 visual effect Effects 0.000 claims description 17
- 210000003813 thumb Anatomy 0.000 claims description 15
- 230000001360 synchronised effect Effects 0.000 claims description 9
- 238000003384 imaging method Methods 0.000 claims description 4
- VYPSYNLAJGMNEJ-UHFFFAOYSA-N Silicium dioxide Chemical compound O=[Si]=O VYPSYNLAJGMNEJ-UHFFFAOYSA-N 0.000 claims description 3
- 230000002950 deficient Effects 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 abstract description 2
- 238000012634 optical imaging Methods 0.000 abstract description 2
- 101100190618 Arabidopsis thaliana PLC3 gene Proteins 0.000 description 12
- 238000000034 method Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 230000005484 gravity Effects 0.000 description 6
- 230000004913 activation Effects 0.000 description 4
- 238000005286 illumination Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 239000000428 dust Substances 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- JEIPFZHSYJVQDO-UHFFFAOYSA-N iron(III) oxide Inorganic materials O=[Fe]O[Fe]=O JEIPFZHSYJVQDO-UHFFFAOYSA-N 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 230000009191 jumping Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000001960 triggered effect Effects 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
-
- 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
- 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
- G01N21/8806—Specially adapted optical and illumination features
-
- 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
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)
Abstract
The invention belongs to the technical field of bearing surface defect detection, and particularly discloses a novel intelligent bearing surface defect detection system, which comprises a bottom plate and a lower box body; the base plate is fixedly provided with a power supply module, an industrial personal computer, a PLC, a camera power supply controller, a light source power supply controller, a first stepping motor controller and a second stepping motor controller; the lower box body is connected with a vertical base, a vertical rod connected with a strip light source and a line scanning vision sensor, a first motor support plate connected with a first stepping motor, a second motor support plate connected with a second stepping motor and a motor bearing seat support plate; the base is provided with a bearing guide groove, a bearing sliding groove, a first sliding bearing connected with a cam shaft and a second sliding bearing connected with a friction wheel shaft, and prism supports used for fixing prisms are arranged on two sides of the base. According to the invention, an optical imaging technology is fused with machine vision, and the integral surface information acquisition and defect detection of the bearing are realized through one station, so that the problem of low bearing detection efficiency is solved.
Description
Technical Field
The invention belongs to the technical field of bearing surface defect detection, and particularly discloses a novel intelligent bearing surface defect detection system.
Background
In rotating machinery, the quality of the bearings plays a critical role in the stability of the operation of the machinery. In particular in automotive axles, both loads from both axial and radial directions and precise guidance for rotation are required. If the bearing has a problem, the performance of the automobile is affected by light weight, and serious accidents are caused by heavy weight. Bearing manufacturers should ensure that bearings with quality problems are detected before shipping as much as possible, and if they are found after an accident occurs during use by users, huge economic and reputation losses will be incurred. Therefore, the quality detection of the bearings becomes an important link of strict control in the bearing processing process of each bearing production enterprise.
At present, the domestic bearing production enterprises mostly adopt manual mode for detecting the defects on the surfaces of the bearings. Typically, workers look for defects by visually inspecting the bearing surface on a high intensity light table. The efficiency and accuracy of the detection method are affected by individual factors of workers, strict unification of detection standards is difficult to ensure, and detection omission and false detection are easy to occur. Meanwhile, under the working environment of strong light, the eyes are greatly injured, and the health of workers is seriously affected.
Disclosure of Invention
The invention aims to provide a novel intelligent detection system for bearing surface defects, which aims to solve the problem of low bearing detection efficiency.
In order to achieve the above purpose, the technical scheme of the invention is as follows: the novel intelligent detection system for the surface defects of the bearings comprises a bottom plate and a lower box body, wherein the lower box body is fixed above the bottom plate; a power module, an industrial personal computer, a PLC, a camera power controller, a light source power controller, a first stepping motor controller and a second stepping motor controller are fixed on one side, close to the lower box body, of the bottom plate;
a power switch is arranged on one side of the outer part of the lower box body, which is vertical to the bottom plate; one end, far away from the bottom plate, of the outer part of the lower box body is provided with a vertical upright rod, one end, far away from the lower box body, of the upright rod is vertically connected with a camera support frame through a camera support adapter block, the camera support frame is connected with a linear array camera support plate, and a linear scanning vision sensor is connected between the linear array camera support plates; the middle part of the upright rod is connected with a strip-shaped light source fixing frame parallel to the camera support frame through a light source support adapter, and a strip-shaped light source is fixed on the strip-shaped light source fixing frame;
the lower box body is provided with a lower bottom plate, a lower bottom plate is arranged on the lower bottom plate, a bearing guide groove is arranged at the upper end of the lower bottom plate, a bearing guide groove is arranged at the middle part of the lower bottom plate, a bearing guide groove is arranged below the bearing guide groove, a first sliding bearing and a second sliding bearing are arranged on the lower bottom plate, the first sliding bearing is connected with a cam shaft in a sliding manner, the second sliding bearing is connected with a friction wheel shaft in a sliding manner, a synchronous pulley is fixed on the cam shaft and the friction wheel shaft, a synchronous belt is connected on the synchronous pulley in a sliding manner, and the normal line of the central lines of the cam shaft and the friction wheel shaft coincides with the central line of the line scanning vision sensor; the two sides of the base are provided with prism supports for fixing the prisms, and rectangular planes corresponding to the right-angle sides of the prisms are parallel to the bearing side planes; the inside of the base is also provided with a first proximity switch and a second proximity switch;
the lower box body is provided with a first motor support plate, a second motor support plate and a motor bearing seat support plate which are vertical at one end far away from the bottom plate, wherein the first motor support plate and the second motor support plate are positioned at one side of the base close to the vertical rod; a first stepping motor is fixed on one side, close to the upright post, of the top end of the first motor support plate, a second motor bearing seat is arranged on one side, far away from the upright post, of the top end of the first motor support plate, and a motor coupler is arranged between the first stepping motor and the second motor bearing seat; the top end of the motor bearing seat supporting plate is provided with a first motor bearing seat, and a poking wheel matched with the bearing guide groove is connected between the first motor bearing seat and the second motor bearing seat in a sliding manner.
The working principle of the technical scheme is as follows:
a novel intelligent detection system for bearing surface defects is characterized in that a line scanning visual sensor of a bearing image acquisition device is used for acquiring bearing images, and bearing surface defects are detected by utilizing specific algorithm processing. In the bearing image acquisition device, the central line of the line scanning visual sensor coincides with the normal line of the central lines of the cam shaft and the friction wheel shaft, and the target surface line of the line scanning visual sensor is parallel to the central line of the bearing, so that the line scanning visual sensor is ensured to acquire complete bearing surface images in rows; the rectangular plane corresponding to the right-angle side of the prism is parallel to the side plane of the bearing, and when the bearing rotates, the image of the upper half part of each side can be reflected to the upper surface of the prism through the rectangular plane corresponding to the inclined side of the prism, and the upper half part of each side is on the same plane with the uppermost end of the cylindrical surface, and the line scanning visual sensor is arranged at a station to collect visual information of the whole outer surface of the bearing once.
The beneficial effects of this technical scheme lie in:
(1) A novel intelligent detection system for defects on the surface of a bearing fuses an optical imaging technology with machine vision, a rectangular plane corresponding to a right-angle side of the prism is parallel to a side plane of the bearing, when the bearing rotates, an image of the upper half part of each side can be reflected to the vicinity of the upper surface of the prism through the rectangular plane corresponding to the inclined side of the prism, and one station is used for realizing information acquisition and defect detection on the whole surface of the bearing, so that the detection efficiency of the bearing is greatly improved, and the working environment of workers is improved.
(2) The defect detection is realized through the acquisition of the line scanning visual sensor and the processing of a specific algorithm, the center line of the line scanning visual sensor coincides with the normal line of the center lines of the cam shaft and the friction wheel shaft, the target line of the line scanning visual sensor is parallel to the center line of the bearing, the line scanning visual sensor is ensured to acquire complete bearing surface images according to rows, the signal to noise ratio of the images is improved, the cylindrical surface area of the bearing and the side surface areas of the two bearings are rectangular in the acquired images, which is equivalent to the expansion of the bearing around the radius of the bearing, and the extraction of the bearing surface area in the algorithm is facilitated.
Furthermore, an imaging method of visual information of the outer surface of the bearing is used, and the imaging method is realized through a prism with a rectangular plane corresponding to the right-angle side and parallel to the side plane of the bearing; the prism is an equilateral right-angle triangular prism made of quartz glass, and the surface of the triangular surface is frosted; the two rectangular surfaces corresponding to the triangular right-angle sides are transparent surfaces, the rectangular surface corresponding to the triangular hypotenuse is a mirror surface, and the upper half part of images on the two side surfaces of the bearing are respectively reflected to the same plane with the uppermost end of the cylindrical surface through the two prism mirror surfaces. The prisms on two sides of the bearing can reflect image information on the side surfaces of the bearing to two planes parallel to the upper cylindrical surface, and the line scanning visual sensor can collect information on three surfaces simultaneously.
Further, the upper end of the lower box body is also connected with an upper box body, and a display is arranged at one side, close to the power switch, of the outer part of the upper box body.
Further, the power supply module is connected with a power supply switch, a display, a camera power supply controller, a line scanning vision sensor, an industrial personal computer, a light source power supply controller, a first stepping motor and a second stepping motor; the industrial personal computer is connected with the power supply module, the line scanning visual sensor, the PLC and the display; the PLC is connected with the power supply module, the industrial personal computer, the line scanning visual sensor, the strip light source, the first stepping motor controller, the first proximity switch and the second proximity switch; the camera power supply controller is connected with the power supply module and the line scanning vision sensor; the light source power supply controller is connected with the power supply module and the strip-shaped light source; the first stepping motor controller is connected with the power supply module, the PLC and the first stepping motor; the second stepping motor controller is connected with the power supply module, the PLC and the second stepping motor; the power switch is connected with the power module; the display is connected with the power module and the industrial personal computer.
Further, the bearing guide groove is in a U-shaped long strip shape, and the width of the concave area of the bearing guide groove is equal to the width of the bearing; the thumb wheel is ratchet-shaped, with a total of 6 notches, each of which allows only one bearing to pass through. The bearing can roll in only along the strip direction, so that the accuracy of image acquisition is ensured, and the running stability of the system is improved.
Further, the intelligent detection of the bearing surface defects is realized through an algorithm, and the method comprises the following steps:
(1) Collecting a large number of defect-free bearing images by using a bearing image acquisition device, constructing a bearing surface defect model based on a reconstruction network in an industrial personal computer, and performing network training;
(2) Deploying a model with the capability of reconstructing a defect-free bearing image into an industrial personal computer;
(3) And an algorithm module in the industrial personal computer receives the bearing image, respectively inputs different sum algorithms I and II, and considers the bearing as a defective bearing if any one of the algorithm I and the algorithm II judges that the image has defects.
Drawings
FIG. 1 is a schematic diagram of an embodiment of a novel bearing surface defect intelligent detection system of the present invention;
FIG. 2 is a schematic diagram of an embodiment of the present invention;
FIG. 3 is a schematic external construction of an embodiment of the present invention;
FIG. 4 is a hardware block diagram of an embodiment;
FIG. 5 is a schematic diagram of the principal principles of embodiment defect detection;
FIG. 6 is a schematic illustration of an embodiment thumbwheel and bearing guide slot configuration;
FIG. 7 is a schematic view of an embodiment prism installation;
FIG. 8 is a schematic view of an embodiment prism;
FIG. 9 is an embodiment acquisition schematic;
FIG. 10 is an information image acquired by an embodiment;
FIG. 11 is a schematic view of the structure of the detection area of the embodiment;
FIG. 12 is a flowchart of an embodiment algorithm;
FIG. 13 is a schematic diagram of an embodiment reconstruction algorithm;
FIG. 14 is a diagram of an embodiment reconfiguration network architecture;
FIG. 15 is a flowchart of an embodiment algorithm two.
Detailed Description
The following is a further detailed description of the embodiments:
reference numerals in the drawings of the specification include: the power module 1, the industrial personal computer 2, the PLC3, the camera power controller 4, the line scanning vision sensor 5, the light source power controller 6, the bar-shaped light source 7, the first stepping motor controller 8, the first stepping motor 9, the second stepping motor controller 10, the second stepping motor 11, the first proximity switch 12, the second proximity switch 13, the display 14, the power switch 15, the thumb wheel 16, the motor coupling 17, the line camera support plate 18, the first motor bearing housing 19, the second motor bearing housing 20, the bearing guide groove 21, the first second motor support plate 2, the bearing slide groove 23, the motor bearing housing support plate 24, the bar-shaped light source support adapter block 25, the friction wheel shaft 26, the cam shaft 27, the first and second slide bearings 8, the timing pulley 29, the timing belt 30, the prism support 31, the base 32, the prism 33, the bar-shaped light source fixing frame 34, the camera support 35, the upper case 36, the lower case 37, the bottom plate 38, the second motor support plate 39, the second slide bearings 40, the bar-shaped light source support frames 41, the upright bars 42, and the camera support adapter block 43.
The embodiment is basically as shown in fig. 1, and the novel intelligent detection system for the surface defects of the bearing comprises a bottom plate 38 and a lower box 37, wherein the upper end of the bottom plate 38 is fixedly provided with a power module 1, an industrial personal computer 2, a PLC3, a camera power controller 4, a light source power controller 6, a first stepping motor controller 8 and a second stepping motor controller 10 through bolts; the lower box 37 is fixed above the bottom plate 38 through bolts, and the power module 1, the industrial personal computer 2, the PLC3, the camera power controller 4, the light source power controller 6, the first stepping motor controller 8 and the second stepping motor controller 10 are positioned in the lower box 37;
referring to fig. 3, a power switch 15 is connected to the outer front end of the lower case 37 through bolts; the upper end of the lower box 37 is connected with a vertical upright rod 42 through threads, the upper end of the upright rod 42 is vertically connected with a camera support frame 35 through a camera support frame adapter block 43, the camera support frame 35 is connected with a linear array camera support plate 18, and the linear array camera support plates 18 are connected with a linear scanning vision sensor 5 through bolts; the middle part of the upright rod 42 is connected with a strip light source fixing frame 34 parallel to the camera support frame 35 through a light source support adapter, the strip light source fixing frame 34 is connected with a strip light source 7 through a bolt, and the length direction of the light emitting surface of the strip light source 7 is parallel to the center line of the bearing; the upper end of the lower box 37 is provided with a vertical hollow base 32, the top end of the base 32 is provided with a bearing guide groove 21, the middle part of the base 32 is positioned below the bearing guide groove 21 and provided with a bearing chute 23, and the bearing guide groove 21 and the bearing chute 23 are fixed on two walls of the base 32 through bolts; the base 32 is provided with a first sliding bearing 28 and a second sliding bearing 40, the first sliding bearing 28 is connected with a cam shaft 27 in a sliding way, the first sliding bearing 28 is connected with a friction wheel shaft 26 in a sliding way, a synchronous pulley 29 is fixed on the cam shaft 27 and the friction wheel shaft 26, and a synchronous belt 30 is connected on the synchronous pulley 29 in a sliding way; the prism supports 31 for fixing the prism 33 are respectively fixed on the front side and the rear side of the base 32 through bolts; the prism 33 is placed in the prism holder 31, the prism 33 is fixed by bolts on both sides, and the prism holder 31 is fixed on both sides of the base 32. The rectangular plane corresponding to the right-angle side of the prism 33 is parallel to the bearing side plane; the first proximity switch 12 and the second proximity switch 13 are fixed inside the base 32 through bolts; the upper end of the lower box 37 is welded with a first vertical motor support plate 22, a second vertical motor support plate 39 and a motor bearing support plate 24, the first motor support plate 22 and the second motor support plate 39 are positioned at the front end of the base 32, the motor bearing support plate 24 is positioned at the rear end of the base 32, and the first motor support plate 22 and the motor bearing support plate 24 are positioned on the same straight line; a first stepping motor 9 is fixed at one side, close to the upright post, of the upper end of the first motor support plate 22 through a bolt, a second motor bearing seat 20 is arranged at one side, far away from the upright post, of the upper end of the first motor support plate 22 through a bolt, and a motor coupler 17 is connected between the first stepping motor 9 and the second motor bearing seat 20 in a sliding manner; the top end of the motor bearing seat supporting plate 24 is fixed with a first motor bearing seat 19 through bolts, and a thumb wheel 16 matched with the bearing guide groove 21 is connected between the first motor bearing seat 19 and the second motor bearing seat 20 in a sliding manner.
The novel intelligent bearing surface defect detection system is characterized in that a hardware connection relation is shown in fig. 4, a power switch 15 is connected with a power module 1, a display 14 is connected with the power module 1, a camera power controller 4 is connected with the power module 1, a line scanning vision sensor 5 is connected with the camera power controller 4, an industrial personal computer 2 is connected with the line scanning vision sensor 5, the industrial personal computer 2 is connected with the display 14, the power module 1 is connected with the industrial personal computer 2, a light source power controller 6 is connected with the power module 1, a strip light source 7 is connected with the light source power controller 6, a PLC3 is connected with the power module 1, the industrial personal computer 2 is connected with the PLC3, the PLC3 is connected with the line scanning vision sensor 5, the PLC3 is connected with the strip light source 7, the PLC3 is connected with a first stepping motor controller 8, the first stepping motor controller 8 is connected with a first stepping motor 9, a second stepping motor controller 10 is connected with a second stepping motor 11, a power supply is connected with the first stepping motor controller 8, the power module 1 is connected with the second stepping motor controller 10, the first stepping motor controller 12 is connected with the second stepping motor 3, and the PLC3 is connected with the second stepping motor controller 11, and the PLC3 is connected with the power module 1.
The specific implementation process is as follows:
in the novel intelligent detection system for the surface defects of the bearings, when a power switch 15 is turned on, a PLC3 controls a first stepping motor 9 to rotate for 30 degrees to drive the bearings to be detected to enter a detection area between a cam shaft 27 and a friction wheel shaft 26 through a thumb wheel 16, and at the moment, a first proximity switch 12 is triggered; after receiving the trigger signal of the first proximity switch 12, the PLC3 controls the second stepping motor 11 to rotate for one circle, meanwhile, the strip-shaped light source 7 is turned on, the line scanning vision sensor 5 starts to collect images of the surface of the bearing, and when the bearing rotates, the images of the upper half part of each side surface of the bearing can be reflected to the upper surface of the prism 33 through a rectangular surface (vertical surface) corresponding to the hypotenuse of the prism 33; after the image acquisition is completed, the image is automatically uploaded to the industrial personal computer 2 for image processing, meanwhile, the bearing falls down through the notch of the cam to trigger the second proximity switch 13, detection is completed, meanwhile, the second bearing enters the thumb wheel 16, the first bearing is subjected to the action of gravity and rolls into the notch of the thumb wheel 16 and is blocked by the right convex tip, when the thumb wheel 16 rotates anticlockwise, the right convex tip is lifted, the bearing is subjected to the thrust of gravity and the left convex tip, rolls into a detection area along the strip direction of the bearing guide groove 21, meanwhile, the next bearing enters the notch of the thumb wheel 16 due to the action of gravity, and the main principle of defect detection is shown in fig. 5.
The intelligent detection of the bearing surface defects is realized through an algorithm, a large number of defect-free bearing images are collected by the bearing image acquisition device, and meanwhile, a bearing surface defect model based on a reconstruction network is built in the industrial personal computer 2 for network training, so that the model has the capability of reconstructing the defect-free bearing images. The trained network model is further deployed into the industrial personal computer 2 to finish preparation work, when an algorithm module in the industrial personal computer 2 receives a bearing image, image information is respectively input into a first algorithm and a second algorithm which are different from each other, and if any algorithm in the first algorithm and the second algorithm judges that the image has defects, the bearing is considered to be a defective bearing.
In the embodiment, the strip light source 7 can rotate at 0 to 45 degrees, so that the optimal illumination angle can be conveniently adjusted. The prisms 33 on the two sides of the bearing can reflect the image information of the side surface of the bearing to two planes parallel to the upper cylindrical surface, and the line scanning vision sensor 5 can collect the information of three surfaces at the same time, so that the problem that the defect detection of the surface of the bearing can be realized only by three stations in the prior art is solved.
In the embodiment, as shown in fig. 6, the bearing guide groove 21 and the thumb wheel 16 are in a U-shaped long strip shape, and the width of the concave area is equal to the width of the bearing, so that the bearing can roll in only along the long strip direction; a thumb wheel 16 is arranged above the end of the bearing guide groove 21, the thumb wheel 16 is in a ratchet shape, 6 notches are formed, and each notch only allows one bearing to pass through. When detection begins, the first bearing rolls into the notch of the thumb wheel 16 under the action of gravity and is blocked by the right convex tip; when the thumb wheel 16 rotates anticlockwise, the right-side lobe is lifted, the bearing is forced by gravity and the thrust of the left-side lobe, rolls along the bearing guide 21 into the detection area, and the next bearing enters the notch of the thumb wheel 16 due to gravity. The mechanism is simple and small, and moves through the rotation of the bearing, so that the kinetic energy is not required to be additionally provided, and the energy consumption is saved.
In the embodiment, as shown in fig. 7, the prism 33 is an equilateral right triangle prism, the material is quartz glass, the surface of the triangle surface is frosted, two rectangular surfaces corresponding to the triangle right angle side are transparent surfaces, the rectangular surface corresponding to the triangle hypotenuse is a mirror surface, and the bearing side surface in the vertical direction is reflected to the direction of the bearing cylindrical surface through the prism 33; as shown in fig. 8, the prism 33 is placed in the prism holder 31, the prism 33 is fixed by bolts on both sides, and the prism holder 31 is fixed on both sides of the base 32. The rectangular plane corresponding to the right-angle side of the prism 33 is kept parallel to the bearing side plane. When the bearing rotates, the image of the upper half part of each side surface of the bearing can be reflected to the upper surface of the prism 33 through a rectangular surface (vertical surface) corresponding to the hypotenuse of the prism 33, and the upper half part and the uppermost end of the cylindrical surface are on the same plane.
In the embodiment, the strip light source 7 is a high-power blue strip light source, as shown in fig. 9, so that illumination can be more uniform, the influence caused by uneven illumination is prevented, the length direction of the light emitting surface of the strip light source 7 is parallel to the central line of the bearing, the light emitting surface is adjustable by 0-45 degrees with the surface of the bottom plate 38, the strip light source 7 is connected with the strip light source support 41 through bolts, the strip light source 7 can be guaranteed to rotate along the length direction of the strip light source 7, the angle between the light emitting surface and the surface of the bottom plate 38 is adjusted, and the optimal illumination collecting condition is obtained; the center line of the line scanning vision sensor 5 coincides with the normal line of the center lines of the cam shaft 27 and the friction wheel shaft 26, the imaging information (which is a pixel line) of the line scanning vision sensor 5 is parallel to the center line of the bearing, the line scanning vision sensor 5 can be ensured to acquire complete bearing surface images according to lines, and the signal to noise ratio of the images is improved. In the acquired image, as shown in fig. 10, the bearing cylindrical surface area and the two bearing side surface areas are rectangular, which is equivalent to the expansion of the bearing around the bearing radius, and is beneficial to the extraction of the bearing surface area in the algorithm.
In the embodiment, the control flow of the line scanning vision sensor 5 for collecting the bearing surface image in rows is as follows: the pulse frequency of a suitable second stepping motor 11 is set, the pulse frequency of the line scanning vision sensor 5 is initialized, and the acquisition frequency of the line scanning vision sensor 5 is selected by debugging. When the bearing passes through the first proximity switch 12, a trigger signal is sent out, and after a certain time delay, the line scanning vision sensor 5 is controlled to collect and the first stepping motor 9 is controlled to rotate at the same time.
In the embodiment, as shown in fig. 12, the first algorithm is to locate the bearing in the image and select the target area to be detected; because the bearing defects have larger difference, the defects are divided into five categories of deep and shallow pits, dust cover damage, inner ring damage, outer ring damage and rust, the five categories of defects are detected in parallel, and meanwhile, the defects of all parts are detected; firstly, adopting a blob analysis method, a difference method and a feature method to remove irrelevant interference, wherein pit defects are extracted from deep and shallow pits by adopting a mode of combining threshold segmentation, morphology and gray level co-occurrence matrix; the dust cover part detects defects in a mode of combining sub-pixels with features; detecting defects by the inner ring and the outer ring through a differential plus feature and gray level co-occurrence matrix method; after removing texture interference, the rust defect is mainly extracted by a method of threshold segmentation and feature and gray level co-occurrence matrix.
In an embodiment, as shown in fig. 13 and 14, the algorithm two comprises an input layer, an encoding layer, a semantic feature layer, a decoding layer and an output layer which are sequentially connected, wherein the output in the middle of the encoding layer is also connected with the decoding layer in a jumping manner; wherein the encoder comprises 4 convolution modules, each convolution module comprising 1 convolution layer, 1 BN layer and a nonlinear activation layer, the first three convolution modules further comprising a pooling layer capable of changing the scale of the image; the activation function in the activation layer is Relu; the decoder comprises 4 deconvolution modules, each deconvolution module comprises 1 deconvolution layer, one BN layer and a nonlinear activation layer, the first three deconvolution modules are further characterized by a splicing layer, and the corresponding layers of the encoder are obtained from the corresponding layers of the encoder to splice. And in combination with the illustration of fig. 15, inputting the images into a trained network model, outputting a reconstructed defect-free bearing image, then, differentiating the reconstructed image from the original image, positioning and detecting the defects of the bearing surface, and improving the accuracy of extracting the defects of the bearing surface through double algorithm judgment.
The foregoing is merely exemplary embodiments of the present invention, and specific structures and features that are well known in the art are not described in detail herein. It should be noted that modifications and improvements can be made by those skilled in the art without departing from the structure of the present invention, and these should also be considered as the scope of the present invention, which does not affect the effect of the implementation of the present invention and the utility of the patent.
Claims (5)
1. The intelligent detection system for the surface defects of the bearing is characterized by comprising a bottom plate and a lower box body, wherein the lower box body is fixed above the bottom plate; a power module, an industrial personal computer, a PLC, a camera power controller, a light source power controller, a first stepping motor controller and a second stepping motor controller are fixed on one side, close to the lower box body, of the bottom plate;
a power switch is arranged on one side of the outer part of the lower box body, which is vertical to the bottom plate; one end, far away from the bottom plate, of the outer part of the lower box body is provided with a vertical upright rod, one end, far away from the lower box body, of the upright rod is vertically connected with a camera support frame through a camera support adapter block, the camera support frame is connected with a linear array camera support plate, and a linear scanning vision sensor is connected between the linear array camera support plates; the middle part of the upright rod is connected with a strip-shaped light source fixing frame parallel to the camera support frame through a light source support adapter, and a strip-shaped light source is fixed on the strip-shaped light source fixing frame;
the lower box body is provided with a lower bottom plate, a lower bottom plate is arranged on the lower bottom plate, a bearing guide groove is arranged at the upper end of the lower bottom plate, a bearing guide groove is arranged at the middle part of the lower bottom plate, a bearing guide groove is arranged below the bearing guide groove, a first sliding bearing and a second sliding bearing are arranged on the lower bottom plate, the first sliding bearing is connected with a cam shaft in a sliding manner, the second sliding bearing is connected with a friction wheel shaft in a sliding manner, a synchronous pulley is fixed on the cam shaft and the friction wheel shaft, a synchronous belt is connected on the synchronous pulley in a sliding manner, and the normal line of the central lines of the cam shaft and the friction wheel shaft coincides with the central line of the line scanning vision sensor; the two sides of the base are provided with prism supports for fixing the prisms, and rectangular planes corresponding to the right-angle sides of the prisms are parallel to the bearing side planes; the inside of the base is also provided with a first proximity switch and a second proximity switch;
the lower box body is provided with a first motor support plate, a second motor support plate and a motor bearing seat support plate which are vertical at one end far away from the bottom plate, wherein the first motor support plate and the second motor support plate are positioned at one side of the base close to the vertical rod; a first stepping motor is fixed on one side, close to the upright post, of the top end of the first motor support plate, a second motor bearing seat is arranged on one side, far away from the upright post, of the top end of the first motor support plate, and a motor coupler is arranged between the first stepping motor and the second motor bearing seat; a first motor bearing seat is arranged at the top end of the motor bearing seat supporting plate, and a poking wheel matched with the bearing guide groove is connected between the first motor bearing seat and the second motor bearing seat in a sliding manner;
the detection system uses an imaging method of visual information of the outer surface of the bearing, and is realized through a prism with a rectangular plane corresponding to the right-angle side and parallel to the side plane of the bearing;
the prism is an equilateral right-angle triangular prism made of quartz glass, and the surface of the triangular surface is frosted; the two rectangular surfaces corresponding to the triangular right-angle sides are transparent surfaces, the rectangular surface corresponding to the triangular hypotenuse is a mirror surface, and the upper half part of images on the two side surfaces of the bearing are respectively reflected to the same plane with the uppermost end of the cylindrical surface through the two prism mirror surfaces.
2. The intelligent detection system for bearing surface defects according to claim 1, wherein the upper end of the lower box body is further connected with an upper box body, and a display is arranged on one side, close to the power switch, of the outer portion of the upper box body.
3. The intelligent bearing surface defect detection system of claim 1, wherein the power module is connected to a power switch, a display, a camera power controller, a line scan vision sensor, an industrial personal computer, a light source power controller, a first stepper motor and a second stepper motor; the industrial personal computer is connected with the power supply module, the line scanning visual sensor, the PLC and the display; the PLC is connected with the power supply module, the industrial personal computer, the line scanning visual sensor, the strip light source, the first stepping motor controller, the first proximity switch and the second proximity switch; the camera power supply controller is connected with the power supply module and the line scanning vision sensor; the light source power supply controller is connected with the power supply module and the strip-shaped light source; the first stepping motor controller is connected with the power supply module, the PLC and the first stepping motor; the second stepping motor controller is connected with the power supply module, the PLC and the second stepping motor; the power switch is connected with the power module; the display is connected with the power module and the industrial personal computer.
4. The intelligent detection system for bearing surface defects according to claim 1, wherein the bearing guide groove is in a U-shaped long strip shape, and the width of the concave area of the bearing guide groove is equal to the width of the bearing; the thumb wheel is ratchet-shaped, with a total of 6 notches, each of which allows only one bearing to pass through.
5. The intelligent detection system for bearing surface defects according to claim 1, wherein the intelligent detection of the bearing surface defects is realized by an algorithm, and the intelligent detection system comprises the following steps:
(1) Collecting a large number of defect-free bearing images by using a bearing image acquisition device, constructing a bearing surface defect model based on a reconstruction network in an industrial personal computer, and performing network training;
(2) Deploying a model with the capability of reconstructing a defect-free bearing image into an industrial personal computer;
(3) And an algorithm module in the industrial personal computer receives the bearing image, respectively inputs different sum algorithms I and II, and considers the bearing as a defective bearing if any one of the algorithm I and the algorithm II judges that the image has defects.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110960356.XA CN113686892B (en) | 2021-08-20 | 2021-08-20 | Novel bearing surface defect intelligent detection system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110960356.XA CN113686892B (en) | 2021-08-20 | 2021-08-20 | Novel bearing surface defect intelligent detection system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113686892A CN113686892A (en) | 2021-11-23 |
| CN113686892B true CN113686892B (en) | 2023-08-25 |
Family
ID=78581001
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110960356.XA Active CN113686892B (en) | 2021-08-20 | 2021-08-20 | Novel bearing surface defect intelligent detection system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113686892B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115753797B (en) * | 2022-11-23 | 2023-08-29 | 哈尔滨工程大学 | Three-dimensional rotating bearing appearance defect image acquisition device |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2016098872A1 (en) * | 2014-12-16 | 2016-06-23 | Canon Kabushiki Kaisha | Surface shape measuring apparatus and defect determining apparatus |
| CN107255911A (en) * | 2012-11-06 | 2017-10-17 | 株式会社尼康 | Exposure device |
| CN108548823A (en) * | 2018-05-15 | 2018-09-18 | 佛山市南海区广工大数控装备协同创新研究院 | A kind of large scale plate side surface of workpiece detection device and method |
| CN109211934A (en) * | 2018-08-29 | 2019-01-15 | 南京理工大学 | Based on interference micro- microballoon planar defect detection device and its detection method |
| CN111537488A (en) * | 2020-06-08 | 2020-08-14 | 山西中谱能源科技有限公司 | Device and method for eliminating mercury element measurement interference by utilizing Zeeman fluorescence |
| CN111965192A (en) * | 2020-09-18 | 2020-11-20 | 桂林电子科技大学 | Multi-surface imaging visual detection system and detection method |
| CN215678251U (en) * | 2021-05-18 | 2022-01-28 | 广东机电职业技术学院 | Acid-base neutralization automatic titration system based on color identification |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2005083399A1 (en) * | 2004-02-27 | 2005-09-09 | Technical Co., Ltd | Multidirectional simultaneous observation optical system, image reader, image reading method, multidirectional simultaneous observation optical system composite body |
| FR2925706B1 (en) * | 2007-12-19 | 2010-01-15 | Soc Tech Michelin | DEVICE FOR EVALUATING THE SURFACE OF A TIRE. |
| JP6671938B2 (en) * | 2014-12-16 | 2020-03-25 | キヤノン株式会社 | Surface shape measuring device, defect determination device, and surface shape measuring method |
| CN109073197A (en) * | 2016-03-16 | 2018-12-21 | 新加坡科技研究局 | Lighting device and inspection equipment |
-
2021
- 2021-08-20 CN CN202110960356.XA patent/CN113686892B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107255911A (en) * | 2012-11-06 | 2017-10-17 | 株式会社尼康 | Exposure device |
| WO2016098872A1 (en) * | 2014-12-16 | 2016-06-23 | Canon Kabushiki Kaisha | Surface shape measuring apparatus and defect determining apparatus |
| CN108548823A (en) * | 2018-05-15 | 2018-09-18 | 佛山市南海区广工大数控装备协同创新研究院 | A kind of large scale plate side surface of workpiece detection device and method |
| CN109211934A (en) * | 2018-08-29 | 2019-01-15 | 南京理工大学 | Based on interference micro- microballoon planar defect detection device and its detection method |
| CN111537488A (en) * | 2020-06-08 | 2020-08-14 | 山西中谱能源科技有限公司 | Device and method for eliminating mercury element measurement interference by utilizing Zeeman fluorescence |
| CN111965192A (en) * | 2020-09-18 | 2020-11-20 | 桂林电子科技大学 | Multi-surface imaging visual detection system and detection method |
| CN215678251U (en) * | 2021-05-18 | 2022-01-28 | 广东机电职业技术学院 | Acid-base neutralization automatic titration system based on color identification |
Non-Patent Citations (1)
| Title |
|---|
| Smooth Surface Visual Imaging Method for Eliminating High Reflection Disturbance;Wei Shao et al;《Sensor》;第1-11页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113686892A (en) | 2021-11-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN204064972U (en) | optical lens surface defect detecting system | |
| CN103163141A (en) | Strip steel surface online detection system and method based on embedded image processing system | |
| CN101825582B (en) | Method and device for detecting wall of cylindrical transparent bottle | |
| CN116256373B (en) | Perovskite battery film surface defect detection method | |
| CN113686892B (en) | Novel bearing surface defect intelligent detection system | |
| CN102284428A (en) | Corn real-time detection and grading machine vision system | |
| CN102890093B (en) | Machine vision detecting device of surface quality of master cylinder piston of automobile brake | |
| CN105119571A (en) | Silicon solar cell defect on-line detection device | |
| CN115753797B (en) | Three-dimensional rotating bearing appearance defect image acquisition device | |
| CN205941399U (en) | A imaging system for lithium - ion battery pole pieces surface defect detection device | |
| CN110702699B (en) | Rolled glass defect detection device and method | |
| CN111729860B (en) | Online automatic detection system and method for appearance defects of washing machine | |
| CN206146851U (en) | Intellectual detection system industrial part surface quality's device | |
| CN202166618U (en) | Detection device for solar cell piece of satellite solar wing | |
| CN115289990B (en) | Deformation detector for appearance cover plate of display module caused by deformation of mobile phone display screen | |
| CN106841211A (en) | Platform and method of a kind of utilization machine vision to cell piece surface defects detection | |
| CN218995107U (en) | New energy automobile battery defect visual inspection device | |
| CN111122596A (en) | Bearing roller surface defect detection system | |
| CN117373938A (en) | Device and method for detecting PL of silicon wafer | |
| CN113828541A (en) | Visual high sorting rate intelligence sorting machine of magnetic shoe outward appearance | |
| CN112255236A (en) | Full-intelligent strip-shaped light source appearance detection equipment | |
| CN203037588U (en) | Line scanning visual detection system for detecting large-breadth high-speed surface | |
| CN221707321U (en) | Machine vision defect detector for high-speed production line | |
| CN219997457U (en) | Universal light source structure for water meter detection shooting | |
| CN221351239U (en) | Industrial product surface defect detection equipment |
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 |