CN115356261A - Defect detection system and method for automobile ball cage dust cover - Google Patents
Defect detection system and method for automobile ball cage dust cover Download PDFInfo
- Publication number
- CN115356261A CN115356261A CN202210906776.4A CN202210906776A CN115356261A CN 115356261 A CN115356261 A CN 115356261A CN 202210906776 A CN202210906776 A CN 202210906776A CN 115356261 A CN115356261 A CN 115356261A
- Authority
- CN
- China
- Prior art keywords
- industrial camera
- dust cover
- ball cage
- light source
- detection
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 87
- 239000000428 dust Substances 0.000 title claims abstract description 62
- 230000007547 defect Effects 0.000 title claims abstract description 48
- 238000000034 method Methods 0.000 title claims abstract description 13
- 239000000463 material Substances 0.000 claims abstract description 109
- 230000007246 mechanism Effects 0.000 claims abstract description 35
- 238000004891 communication Methods 0.000 claims abstract description 14
- 238000007689 inspection Methods 0.000 claims description 18
- 238000013135 deep learning Methods 0.000 claims description 6
- 230000003044 adaptive effect Effects 0.000 claims description 3
- 230000007797 corrosion Effects 0.000 claims description 3
- 238000005260 corrosion Methods 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 claims description 3
- 238000006073 displacement reaction Methods 0.000 claims description 2
- 238000004519 manufacturing process Methods 0.000 description 5
- 230000000007 visual effect Effects 0.000 description 5
- 230000002159 abnormal effect Effects 0.000 description 4
- 238000007664 blowing Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 2
- 239000004519 grease Substances 0.000 description 2
- 238000002347 injection Methods 0.000 description 2
- 239000007924 injection Substances 0.000 description 2
- 230000001050 lubricating effect Effects 0.000 description 2
- 238000005498 polishing Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 239000000243 solution Substances 0.000 description 2
- 238000005299 abrasion Methods 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000011796 hollow space material Substances 0.000 description 1
- 239000012535 impurity Substances 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 239000012779 reinforcing material Substances 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 229920006346 thermoplastic polyester elastomer Polymers 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Images
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
- G01N21/01—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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L69/00—Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass
- H04L69/16—Implementation or adaptation of Internet protocol [IP], of transmission control protocol [TCP] or of user datagram protocol [UDP]
- H04L69/163—In-band adaptation of TCP data exchange; In-band control procedures
-
- 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
- G01N2021/8841—Illumination and detection on two sides of object
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8887—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2201/00—Features of devices classified in G01N21/00
- G01N2201/04—Batch operation; multisample devices
- G01N2201/0484—Computer controlled
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2201/00—Features of devices classified in G01N21/00
- G01N2201/06—Illumination; Optics
- G01N2201/061—Sources
- G01N2201/06146—Multisources for homogeneisation, as well sequential as simultaneous operation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2201/00—Features of devices classified in G01N21/00
- G01N2201/10—Scanning
- G01N2201/102—Video camera
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2201/00—Features of devices classified in G01N21/00
- G01N2201/10—Scanning
- G01N2201/103—Scanning by mechanical motion of stage
- G01N2201/1035—3D motion
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Biochemistry (AREA)
- Immunology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Pathology (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Quality & Reliability (AREA)
- Theoretical Computer Science (AREA)
- Computer Security & Cryptography (AREA)
- Computer Networks & Wireless Communication (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The invention discloses a defect detection system and method for an automobile ball cage dust cover. And calling a corresponding detection algorithm to perform redundancy detection on the acquired image by the industrial control machine according to the machine position to which the image belongs. When all images collected by the first station are detected and no defect is found, the material can be transferred to the second station to continuously detect other parts. The industrial control machine controls the executing mechanism to throw the corresponding result materials of each station into the corresponding material frames in a TCP communication mode.
Description
Technical Field
The invention relates to the field of industrial defect detection, in particular to a system and a method for detecting defects of a dust cover of an automobile ball cage.
Background
Because the automobile ball cage bears the action of alternating load in the using process, the automobile ball cage is easy to damage if the protection is not in place, and even the automobile can not run. The most obvious characteristic of the common outer ball cage damage is that the automobile normally runs in a straight line, and the front wheel part makes 'rattling' sound when the automobile is steered, and even cannot steer when the automobile is severe; the most obvious characteristic of the damage of the inner ball cage is that the automobile can generate 'rattle' abnormal sound at the position of a gear box when accelerating or running on an uneven road surface, and simultaneously the phenomenon of oil leakage of the ball cage is accompanied, and the abnormal sound can also be generated when the automobile turns when the damage is serious.
The main function of the ball cage dust cover is to prevent dust, impurities enter the ball cage to cause mechanical abrasion inside the ball cage, abnormal sound is generated, even the ball cage is blocked to influence automobile transmission and steering, and damage is caused to driving safety. And the lubricating grease in the ball cage is prevented from leaking outwards, the abnormal running caused by the reduction of the lubricating grease is avoided, and the ball cage is a key component for protecting the ball cage and ensuring the normal running of the vehicle. Therefore, the quality detection of the cage dust cover is very important, and the defective dust cover is timely removed from the production end.
The time for manufacturing a dust cover by using a common TPEE ball cage dust cover injection-blowing machine is about 48s, and the time for manually inspecting the dust cover is about 40s due to the defects of the dust cover. Moreover, the injection blowing machine can work for 24 hours, the quality inspection time is long, and the quality inspection speed is not matched with the production speed, so that the injection blowing machine is an important factor causing capacity bottleneck of manufacturers. Moreover, the manual quality inspection is easily influenced by physiological and psychological factors of people, so that the quality inspection period is prolonged, and the quality inspection has the problems of misjudgment and the like.
The existing full-automatic industrial defect detection scheme generally adopts a multi-camera and deep learning scheme, and the scheme has the defects of high detection cost and long detection time.
Disclosure of Invention
The technical problem to be solved by the invention is to provide a system and a method for detecting the defects of the automobile ball cage dust cover, which realize that a machine replaces manual work to carry out full-automatic industrial detection on the defects of the automobile ball cage dust cover,
in order to solve the technical problems, the technical scheme adopted by the invention is as follows:
a defect detection system for an automobile ball cage dust cover comprises a first detection station, a second detection station, a light source bracket, a camera bracket, a background plate, an execution mechanism for performing operations such as rotation and displacement on materials and an industrial personal computer;
the first inspection station includes: a first ball cage dust cover material, a first industrial camera and a second industrial camera; the detected first ball cage dust cover material is clamped and stays in the air by the actuating mechanism, the first industrial camera is arranged and fixed in the air, the axis of the first industrial camera is flush with the bottom surface plane of the large opening of the first ball cage dust cover material, the second industrial camera is arranged on a station platform below the first ball cage dust cover material, and the axis of the second industrial camera deviates from the round center axis of the first material by a certain distance, so that the first ball cage dust cover material is positioned in the second industrial camera field far away from the center and close to the boundary;
the second inspection station comprises: the device comprises a second ball cage dust cover material, a third industrial camera, a fourth industrial camera and a bearing mechanism; the second ball cage dust cover material is arranged on the bearing mechanism, the axes of the second ball cage dust cover material and the bearing mechanism coincide, the bearing mechanism is a rotating mechanism capable of adjusting the rotating speed, the third industrial camera is arranged above the second ball cage dust cover material, the axis of the third industrial camera deviates from a round center shaft of the second ball cage dust cover material by a certain distance, so that the second ball cage dust cover material is positioned in the third industrial camera visual field far away from the position where the center is close to the boundary, the fourth industrial camera is arranged and fixed in the air, and the axis of the fourth industrial camera and the small-opening plane of the second ball cage dust cover material are on the same plane.
The technical scheme of the invention is further improved as follows: the first detection station is provided with a first strip-shaped light source and a first annular light source, the first strip-shaped light source is placed below the first industrial camera, the light direction faces the bottom of the first ball cage dust cover material in an inclined upward direction, the first annular light source is sleeved outside the second industrial camera, and the axes of the first strip-shaped light source and the second annular light source coincide; the second detection station is provided with a second strip-shaped light source and a second annular light source; the second strip-shaped light source is located above the third industrial camera by a certain distance, light rays of the second strip-shaped light source obliquely and downwards point to the small opening of the second ball cage dust cover material, the second annular light source is sleeved outside the third industrial camera, and the axes of the second strip-shaped light source and the second annular light source coincide.
The technical scheme of the invention is further improved as follows: the industrial computer is provided with at least 5 RJ45 interfaces, and the communication mode of the industrial computer, the camera and the executing mechanism is TCP communication.
The technical scheme of the invention is further improved as follows: the actuating mechanism is a manipulator.
An executing mechanism clamps a material to be detected to rotate, a bearing mechanism bears the material to be detected to rotate, an industrial personal computer controls four cameras to shoot the material to be detected to collect images, algorithm detection is carried out on the collected images after at least six images are collected by each camera, and the material is judged to be unqualified if one image is detected to be unqualified. And detecting the image acquired by the third industrial camera (6) by using an EDcircle-based defect detection algorithm, wherein the steps are as follows:
the method comprises the following steps: down-sampling the image acquired by the third industrial camera (6) by two layers of image pyramids, and reducing the size of the image under the condition of hardly losing defect features;
step two: a binaryzation self-adaptive threshold value is obtained through histogram statistics and a sliding window, and the step of manually adjusting parameters is omitted;
step three: performing threshold segmentation on the image by using an adaptive threshold;
step four: acquiring a segmented image ROI area, and further reducing the size of the image;
step five: then, carrying out corrosion operation on the ROI image to remove image holes which possibly influence the detection result;
step six: extracting all candidate circles and ellipses in the image by using an EDcircle algorithm;
step seven: taking the circle and the ellipse with the smallest radius and half axis in all candidate results;
step eight: taking the circle and the ellipse with the highest confidence level from the results as final results;
step nine: and comparing the finally obtained confidence coefficients of the circle and the ellipse with a confidence coefficient threshold value, wherein the circle and the ellipse are OK pieces if the confidence coefficient threshold value is larger than the confidence coefficient threshold value, and the NG pieces if the confidence coefficient threshold value is smaller than the confidence coefficient threshold value.
The technical scheme of the invention is further improved as follows: and an algorithm for detecting the graphs collected by the first industrial camera, the second industrial camera and the fourth industrial camera is a deep learning algorithm.
Due to the adoption of the technical scheme, the invention has the technical progress that:
the material is detected by two detection stations, namely four camera stations, so that the detection time of the material can be greatly shortened, the use number of industrial cameras is reduced, the image acquisition cost is reduced, and the detection visual field can cover all parts of the material. For a common image acquisition mode, if a detection function is to be realized, at least 10 industrial cameras are needed, namely at least 1 camera is arranged at the top and the bottom, at least 8 cameras are arranged at the cover body and the large opening and the small opening, images acquired by the 10 cameras can interfere with each other, and defect characteristics just positioned at the camera vision boundary can be missed.
Through making top and bottom industry camera sensor axle center all deviate from material circle central axis, combine the rotation of material, can utilize the burr defect characteristic of the inside minor-bore of increase visual angle reinforcing material ingeniously, improve the relevance ratio of this defect.
Aiming at the pure black characteristic of the material, the combined polishing mode of strip light and annular light and the polishing angle are adopted, so that the image acquisition quality of each camera position can be improved, and the defect characteristic is highlighted relative to the background.
The rotatable carrier and actuator are the necessary hardware structures to implement redundant sensing. Because the camera position is fixed, and the defect characteristic of material can appear on the optional position of material, so adopt the mode of redundant detection: the material is borne through the bearing mechanism or the executing mechanism and is driven to rotate 360 degrees, and in the rotating process, each camera position takes pictures of the material part which is responsible for the camera position respectively, and at least 6 images are taken. By means of a redundancy detection mode, defect characteristics of any position of the material can be captured, a detection algorithm is called for defect identification, and the material is judged to be unqualified as long as a defect is detected by one image. If the material is at the first detection station, the material can be transferred to the second detection station by the actuating mechanism to continuously detect other parts of the material, and if the material is at the second detection station, the material can be directly thrown to the good material frame by the actuating mechanism. The detection method can improve the detection speed, reduce the hardware cost and simultaneously reduce the probability of missed detection of defect detection to the maximum extent.
As the top machine position is only responsible for one defect of burrs inside the small opening, the dust cover top shot burr defect detection algorithm based on the EDcircle algorithm is used for the image acquired by the top camera, namely the third industrial camera. The algorithm belongs to a machine vision algorithm, and is shorter than the detection time of a deep learning algorithm. The experience of the total detection time can be reduced to the greatest extent by only calling one algorithm in combination with a camera at one part, and the total detection time can be reduced under the condition of ensuring the detection accuracy when the top machine position adopts the algorithm.
The industrial personal computer is used as a control center, and the TCP communication is used as a communication mode, so that the complexity of the system can be reduced to the maximum extent, the debugging and maintenance of later equipment are facilitated, the development of control software is facilitated, and the hardware control mode is simplified. And finally, the TCP control executing mechanism puts the materials into the corresponding material frames according to the corresponding judgment results, so that the full automation of the defect detection of the ball cage dust cover is realized.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the description in the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts;
FIG. 1 is a schematic structural view of two detection stations;
FIG. 2 is a schematic front view of a first inspection station structure;
FIG. 3 is a schematic isometric view of a first inspection station configuration from the southeast;
FIG. 4 is a schematic front view of a second inspection station structure;
FIG. 5 is a schematic isometric view of a second inspection station configuration from the south east;
FIG. 6 is a flow chart of a detection method;
FIG. 7 is a flow chart of an EDcircle-based dust cover top shot burr defect detection algorithm;
the device comprises a first ball cage dust cover material, a first industrial camera, a first strip light source, a second industrial camera, a first annular light source, a third industrial camera, a second annular light source, a second strip light source, a third industrial camera, a second ball cage dust cover material, a bearing mechanism, a first detection station, a second detection station and a third detection station, wherein the first ball cage dust cover material is 1, the first industrial camera is 2, the first industrial camera is 3, the first strip light source is 4, the second industrial camera is 5, the first annular light source is 6, the third industrial camera is 7, the second annular light source is 8, the second strip light source is 9, the third industrial camera is 10, the second ball cage dust cover material is 11, the bearing mechanism is 12, the first detection station is 13, and the second detection station is the second detection station.
Detailed Description
The present invention will be described in further detail with reference to the following examples:
example one
The automobile ball cage dust cover defect detection system comprises a first detection station 12, a second detection station 13, a light source support, a camera support, a background plate, a mechanical arm serving as an execution mechanism and an industrial personal computer. The first inspection station 12 includes: the device comprises a first ball cage dust cover material 1, a first industrial camera 2, a second industrial camera 4, a first linear light source 3 and a first annular light source 5. The second inspection station includes: the device comprises a second ball cage dust cover material 10, a third industrial camera 6, a fourth industrial camera 9, a second strip-shaped light source 8, a second annular light source 7 and a bearing mechanism 11.
As shown in fig. 1, which is a schematic diagram of two detection stations of an automobile ball cage dust cover defect detection system, a first ball cage dust cover material 1 (hereinafter referred to as a first material) clamped by a manipulator at a first detection station 12 to be detected is suspended in a half-hollow space, and a first industrial camera 2 is fixed in the air by a camera bracket, so that the axis of the first industrial camera 2 is flush with the bottom plane of a large opening of the first material. The first bar light source 3 is placed below the first industrial camera 1 with the light direction obliquely upwards towards the bottom of the first material. The second industrial camera 4 is arranged below the first material, the axis of the second industrial camera 4 deviates from the round center axis of the first material by a certain distance, so that the first material is positioned in the visual field of the second industrial camera 4 and far away from the center and close to the boundary, the first annular light source 5 is sleeved outside the second industrial camera 4, and the axes of the first annular light source and the second industrial camera coincide. The material detected on the second detection station 13 is a second ball cage dust cover material 10 (hereinafter referred to as a second material), the second material is placed on a bearing mechanism 11 of the second detection station 13, the axes of the second material and the second material are overlapped, the bearing mechanism 11 can adjust the rotating speed to rotate according to the instruction of an industrial personal computer, the third industrial camera 6 is arranged above the second material, and the axis of the third industrial camera 6 deviates from the round axis of the second material by a certain distance, so that the second material is positioned in the place, far away from the center and close to the boundary, in the visual field of the third industrial camera 6. The second annular light source 7 is sleeved outside the third industrial camera 6, and the axes of the second annular light source and the third industrial camera are superposed. The fourth industrial camera 9 is fixed in the midair, and the axis of the fourth industrial camera 9 is on the same plane with the small opening plane of the second material. The second strip light source 8 is fixedly arranged above the third industrial camera 9 at a certain distance, and light rays of the second strip light source obliquely and downwards point to the small opening part of the second material.
The detection system adopts an industrial personal computer as a control center, the industrial personal computer is provided with 5 RJ45 interfaces, and the communication mode between the industrial personal computer and the camera as well as the communication mode between the industrial personal computer and the manipulator is TCP communication. As shown in fig. 6, the industrial personal computer controls the manipulator to clamp the first material 1 to rotate at a preset speed in a soft triggering manner, the industrial personal computer controls the first industrial camera 2 and the second industrial camera 4 of the first detection station 12 to take a picture of the first material by rotating 360 degrees and acquire an image in a redundant detection manner, the defect characteristics of any position of the material can be captured, and after six pictures are acquired, the images acquired by the first industrial camera 2 and the second industrial camera 4 are all detected by using a deep learning algorithm. If only one image detects a defect, the material is judged to be unqualified and is a bad material, and when the detection result is the bad material, the industrial control machine controls the mechanical arm to throw the first material to the bad material frame in a TCP communication mode. When the detection result is that the material is good, the industrial control machine controls the manipulator to transfer the first material to the bearing mechanism 11 of the second detection station 13 in a TCP communication mode, at this time, the first material which is just transferred is referred to as the second material, and then the manipulator continues to take the material to be detected to the first detection station 12 for detection. After the second material reaches the detection position, the industrial personal computer triggers the third industrial camera 6 and the fourth industrial camera 9 to take pictures in a soft triggering mode, meanwhile, the bearing mechanism 11 can drive the second material to rotate, a redundant detection mode is still adopted to collect the material by rotating 360 degrees, the defect characteristics of any position of the material can be captured, and after enough six pictures are collected, the images collected by the fourth industrial camera 9 are detected by using a deep learning algorithm. The image acquired by the third industrial camera 6 is detected using an EDCircle-based defect detection algorithm. The defect detection algorithm based on the EDCircle performs detection as shown in fig. 7, and includes the following steps:
the method comprises the following steps: down-sampling the image acquired by the third industrial camera (6) through a two-layer image pyramid, and reducing the size of the image under the condition of hardly losing defect features;
step two: a binaryzation self-adaptive threshold value is obtained through histogram statistics and a sliding window, so that the step of manually adjusting parameters is omitted;
step three: performing threshold segmentation on the image by using an adaptive threshold;
step four: acquiring a segmented image ROI area, and further reducing the size of the image;
step five: then, carrying out corrosion operation on the ROI image to remove image holes which may influence the detection result;
step six: extracting all candidate circles and ellipses in the image by using an EDcircle algorithm;
step seven: taking the circle and the ellipse with the smallest radius and half axis in all candidate results;
step eight: taking the circle and the ellipse with the highest confidence level from the results as final results;
step nine: and comparing the confidence coefficients of the finally obtained circle and ellipse with a confidence coefficient threshold, wherein the circle and ellipse are OK pieces when the confidence coefficient threshold is larger than the confidence coefficient threshold, and the NG pieces when the confidence coefficient threshold is smaller than the confidence coefficient threshold.
And when the detection result is that the material is a bad material, the industrial control machine controls another mechanical arm to throw the second material to the bad material frame in a TCP communication mode. When the testing result is good material, industrial control machine will throw away the second material to good material frame through the mode control of TCP communication other manipulator. And finishing the defect detection of the spherical cage dust cover. It is worth noting that the calling of the algorithm can be realized by changing an industrial control software program while collecting and calling. The invention can realize full-automatic industrial detection of automobile dust cover defects by a machine instead of manpower, shortens the quality inspection period, has high quality inspection quality and is not easy to have error of false inspection. The machine detection is not limited by manpower, the production speed of the machine can be completely matched, the bottleneck of productivity is greatly broken through, and the production efficiency is improved.
The above-mentioned embodiments are merely illustrative of the preferred embodiments of the present invention, and do not limit the scope of the present invention, and various modifications and improvements of the technical solution of the present invention by those skilled in the art should fall within the protection scope defined by the claims of the present invention without departing from the spirit of the present invention.
Claims (6)
1. The utility model provides a defect detection system of car ball cage dust cover, includes first detection station (12), second detection station (13), light source support, camera support, background board, is used for carrying out actuating mechanism and the industrial computer of operations such as rotation, displacement to the material, its characterized in that:
the first inspection station (12) comprises: a first ball cage dust cover material (1), a first industrial camera (2) and a second industrial camera (4); the method comprises the following steps that a detected first ball cage dust cover material (1) is clamped and stays in the air by an executing mechanism, a first industrial camera (2) is arranged and fixed in the air, the axis of the first industrial camera (2) is flush with the large-opening bottom surface plane of the first ball cage dust cover material (1), a second industrial camera (4) is arranged on a station platform below the first ball cage dust cover material (1), and the axis of the second industrial camera (4) deviates from the circular central axis of the first material (1) by a certain distance;
the second inspection station (13) comprises: a second ball cage dust cover material (10), a third industrial camera (6), a fourth industrial camera (9) and a bearing mechanism (11); second ball cage dust cover material (10) set up on bearing mechanism (11), and both axle centers coincide, bearing mechanism (11) is for the rotary mechanism that can adjust the rotational speed, and third industrial camera (6) set up in second ball cage dust cover material (10) top, and the axle center of third industrial camera (6) is skew in the round axle certain distance of second ball cage dust cover material (10), and fourth industrial camera (9) set up to be fixed in the air, and the axle center of fourth industrial camera (9) and the osculum plane of second ball cage dust cover material (10) are on the coplanar.
2. The system for detecting the defects of the dust cover of the automobile ball cage according to claim 1, wherein: the first detection station (12) is provided with a first strip light source (3) and a first annular light source (5), the first strip light source (3) is placed below the first industrial camera (1), the light direction faces the bottom of the first ball cage dust cover material (1) in the oblique upward direction, the first annular light source (5) is sleeved outside the second industrial camera (4), and the axes of the first strip light source and the second annular light source coincide; the second detection station (13) is provided with a second strip light source (8) and a second annular light source (7); the second strip-shaped light source (8) is located above the third industrial camera (9) at a certain distance, light rays of the second strip-shaped light source obliquely and downwards point to a small opening of the second ball cage dust cover material (10), the second annular light source (7) is sleeved outside the third industrial camera (6), and the axes of the second strip-shaped light source and the third industrial camera coincide.
3. The system for detecting the defects of the dust cover of the automobile ball cage according to claim 2, is characterized in that: the industrial personal computer is provided with at least 5 RJ45 interfaces, and the communication mode of the industrial personal computer, the camera and the executing mechanism is TCP communication.
4. The system for detecting defects of a dust cover of an automobile ball cage according to claim 3, characterized in that: the actuating mechanism is a manipulator.
5. A defect detection method of an automobile ball cage dust cover is based on the defect detection system of the automobile ball cage dust cover as claimed in any one of claims 1-4, wherein an executing mechanism at a first detection station (12) clamps a material to be detected to rotate, a bearing mechanism (11) at a second detection station (13) bears the material to be detected to rotate, an industrial personal computer controls four cameras to photograph the material to be detected to collect images, after at least six images are collected by each camera, the collected images are subjected to algorithm detection, and if one image has a defect, the material is an unqualified material; the defect detection algorithm is characterized in that: and detecting the image acquired by the third industrial camera (6) by using an EDcircle-based defect detection algorithm, wherein the steps are as follows:
the method comprises the following steps: down-sampling the image acquired by the third industrial camera (6) through a two-layer image pyramid, and reducing the size of the image under the condition of hardly losing defect features;
step two: a binaryzation self-adaptive threshold value is obtained through histogram statistics and a sliding window, so that the step of manually adjusting parameters is omitted;
step three: performing threshold segmentation on the image by using an adaptive threshold;
step four: acquiring a ROI (region of interest) of the segmented image, and further reducing the size of the image;
step five: then, carrying out corrosion operation on the ROI image to remove image holes which may influence the detection result;
step six: extracting all candidate circles and ellipses in the image by using an EDcircle algorithm;
step seven: taking the circle and the ellipse with the smallest radius and half axis in all candidate results;
step eight: taking the circle and the ellipse with the highest confidence level from the results as final results;
step nine: and comparing the confidence coefficients of the finally obtained circle and ellipse with a confidence coefficient threshold, wherein the circle and ellipse are OK pieces when the confidence coefficient threshold is larger than the confidence coefficient threshold, and the NG pieces when the confidence coefficient threshold is smaller than the confidence coefficient threshold.
6. The method for detecting the defects of the dust cover of the automobile ball cage according to claim 5, wherein the method comprises the following steps: and an algorithm for detecting the images collected by the first industrial camera (2), the second industrial camera (4) and the fourth industrial camera (9) is a deep learning algorithm.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210906776.4A CN115356261A (en) | 2022-07-29 | 2022-07-29 | Defect detection system and method for automobile ball cage dust cover |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210906776.4A CN115356261A (en) | 2022-07-29 | 2022-07-29 | Defect detection system and method for automobile ball cage dust cover |
Publications (1)
Publication Number | Publication Date |
---|---|
CN115356261A true CN115356261A (en) | 2022-11-18 |
Family
ID=84031627
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210906776.4A Pending CN115356261A (en) | 2022-07-29 | 2022-07-29 | Defect detection system and method for automobile ball cage dust cover |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115356261A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117554374A (en) * | 2023-11-13 | 2024-02-13 | 杭州腾励传动科技股份有限公司 | Automatic defect detection classification system for ball cage and data tracking analysis method |
Citations (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2012177662A (en) * | 2011-02-28 | 2012-09-13 | Ntn Corp | Method and device for inspecting defect |
US20140267693A1 (en) * | 2013-03-15 | 2014-09-18 | Digital Wind Systems, Inc. | System and method for ground based inspection of wind turbine blades |
WO2016089431A1 (en) * | 2014-12-03 | 2016-06-09 | Faro Technologies, Inc. | Using depth-camera images to speed registration of three-dimensional scans |
CN106423916A (en) * | 2016-12-12 | 2017-02-22 | 燕山大学 | Machine-vision-based DELTA sorting manipulator |
US20170307544A1 (en) * | 2014-09-05 | 2017-10-26 | SCREEN Holdings Co., Ltd. | Inspection apparatus and inspection method |
CN109342445A (en) * | 2018-11-21 | 2019-02-15 | 中国质量认证中心 | A kind of ABS gear ring plate defects detection system and method |
US20190295290A1 (en) * | 2017-01-06 | 2019-09-26 | Position Imaging, Inc. | System and method of calibrating a directional light source relative to a camera's field of view |
WO2020024185A1 (en) * | 2018-08-01 | 2020-02-06 | SZ DJI Technology Co., Ltd. | Techniques for motion-based automatic image capture |
WO2020078440A1 (en) * | 2018-10-18 | 2020-04-23 | 北京中科虹霸科技有限公司 | Apparatus for collecting high-definition facial images and method for automatic pitch adjustment of camera gimbal |
WO2021028910A1 (en) * | 2019-08-12 | 2021-02-18 | Nanomotion Ltd | A gimbal apparatus system and method for automated vehicles |
WO2021068848A1 (en) * | 2019-10-09 | 2021-04-15 | 山东大学 | Tunnel structure disease multi-scale measurement and intelligent diagnosis system and method |
CN112858332A (en) * | 2021-02-22 | 2021-05-28 | 菲特(珠海横琴)智能科技有限公司 | Synchronizer gear hub surface defect detection method, system and terminal based on machine vision |
CN113000413A (en) * | 2021-05-06 | 2021-06-22 | 菲特(天津)检测技术有限公司 | System, method and terminal for detecting surface defects of synchronizer gear sleeve based on machine vision |
CN213633209U (en) * | 2020-10-21 | 2021-07-06 | 广州柔视智能科技有限公司 | Product defect detection device |
CN113643280A (en) * | 2021-08-30 | 2021-11-12 | 燕山大学 | Plate sorting system and method based on computer vision |
WO2022156381A1 (en) * | 2021-01-19 | 2022-07-28 | 陈贺龄 | Training robot, control method, and control system |
-
2022
- 2022-07-29 CN CN202210906776.4A patent/CN115356261A/en active Pending
Patent Citations (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2012177662A (en) * | 2011-02-28 | 2012-09-13 | Ntn Corp | Method and device for inspecting defect |
US20140267693A1 (en) * | 2013-03-15 | 2014-09-18 | Digital Wind Systems, Inc. | System and method for ground based inspection of wind turbine blades |
US20170307544A1 (en) * | 2014-09-05 | 2017-10-26 | SCREEN Holdings Co., Ltd. | Inspection apparatus and inspection method |
WO2016089431A1 (en) * | 2014-12-03 | 2016-06-09 | Faro Technologies, Inc. | Using depth-camera images to speed registration of three-dimensional scans |
CN106423916A (en) * | 2016-12-12 | 2017-02-22 | 燕山大学 | Machine-vision-based DELTA sorting manipulator |
US20190295290A1 (en) * | 2017-01-06 | 2019-09-26 | Position Imaging, Inc. | System and method of calibrating a directional light source relative to a camera's field of view |
WO2020024185A1 (en) * | 2018-08-01 | 2020-02-06 | SZ DJI Technology Co., Ltd. | Techniques for motion-based automatic image capture |
WO2020078440A1 (en) * | 2018-10-18 | 2020-04-23 | 北京中科虹霸科技有限公司 | Apparatus for collecting high-definition facial images and method for automatic pitch adjustment of camera gimbal |
CN109342445A (en) * | 2018-11-21 | 2019-02-15 | 中国质量认证中心 | A kind of ABS gear ring plate defects detection system and method |
WO2021028910A1 (en) * | 2019-08-12 | 2021-02-18 | Nanomotion Ltd | A gimbal apparatus system and method for automated vehicles |
WO2021068848A1 (en) * | 2019-10-09 | 2021-04-15 | 山东大学 | Tunnel structure disease multi-scale measurement and intelligent diagnosis system and method |
CN213633209U (en) * | 2020-10-21 | 2021-07-06 | 广州柔视智能科技有限公司 | Product defect detection device |
WO2022156381A1 (en) * | 2021-01-19 | 2022-07-28 | 陈贺龄 | Training robot, control method, and control system |
CN112858332A (en) * | 2021-02-22 | 2021-05-28 | 菲特(珠海横琴)智能科技有限公司 | Synchronizer gear hub surface defect detection method, system and terminal based on machine vision |
CN113000413A (en) * | 2021-05-06 | 2021-06-22 | 菲特(天津)检测技术有限公司 | System, method and terminal for detecting surface defects of synchronizer gear sleeve based on machine vision |
CN113643280A (en) * | 2021-08-30 | 2021-11-12 | 燕山大学 | Plate sorting system and method based on computer vision |
Non-Patent Citations (2)
Title |
---|
方琳等: "球笼内表面裂纹自动识别系统图像预处理研究", 《信息技术》, no. 4, 31 December 2015 (2015-12-31), pages 126 - 129 * |
李佳伟: "基于机器视觉的球笼防尘罩缺陷检测系统设计", 《硕士学位论文》, 31 May 2023 (2023-05-31), pages 2 - 4 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117554374A (en) * | 2023-11-13 | 2024-02-13 | 杭州腾励传动科技股份有限公司 | Automatic defect detection classification system for ball cage and data tracking analysis method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN115356261A (en) | Defect detection system and method for automobile ball cage dust cover | |
CN102636490A (en) | Method for detecting surface defects of dustproof cover of bearing based on machine vision | |
JP5302701B2 (en) | Tire shape inspection method, tire shape inspection device | |
US20100053790A1 (en) | Hard disk inspection apparatus and method, as well as program | |
CN112098419A (en) | System and method for detecting surface defects of automobile outer covering part | |
CN111308987B (en) | Automatic uncoiling control system of uncoiler based on image processing and detection method | |
CN114088738B (en) | High-precision defect detection method and device for annular welding seam | |
CN111127335A (en) | Bolt installation verticality head-up shooting visual detection method | |
CN112432949A (en) | Method for detecting surface defects of gravure roller by using 3D photographing technology | |
CN114581444B (en) | Glass cover polishing process monitoring method based on image processing | |
CN113670923A (en) | Omnidirectional high-precision detection device and method for notebook computer shell defects | |
CN114820475B (en) | Edge identification method and system, wafer processing device and method for determining concentric state of wafer and processing table | |
CN113744247A (en) | PCB welding spot defect identification method and system | |
CN114113129B (en) | Lens micro defect recognition and grabbing system and method | |
CN114612474A (en) | Method and device for detecting state of wafer cleaning and drying module and planarization equipment | |
CN113305037A (en) | Method for improving radial positioning precision of rim weld seam through turntable deceleration | |
CN112816495A (en) | Automobile tire defect detection system and method | |
US11327028B2 (en) | Ceramic ball automatic sorting system and method | |
KR102584592B1 (en) | Wheel cap automatic press-fit assembly system | |
CN206930449U (en) | A kind of robot detecting system | |
CN215339551U (en) | Detection equipment for external damage of ground trial operation aircraft engine | |
CN109013388B (en) | Automatic sorting system and method for ceramic balls | |
CN110948355A (en) | Polishing detection device for sealing surface of pressure container | |
CN117253026B (en) | Intelligent detection system and detection method based on image recognition | |
CN1301177C (en) | Detector for reduction of dust collection capacity in machining space of working machine |
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 |