CN110082365B - Robot for detecting quality of edge of steel coil based on machine vision - Google Patents
Robot for detecting quality of edge of steel coil based on machine vision Download PDFInfo
- Publication number
- CN110082365B CN110082365B CN201910275490.9A CN201910275490A CN110082365B CN 110082365 B CN110082365 B CN 110082365B CN 201910275490 A CN201910275490 A CN 201910275490A CN 110082365 B CN110082365 B CN 110082365B
- Authority
- CN
- China
- Prior art keywords
- unit
- image
- robot
- processing unit
- sequence
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- 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
Abstract
The invention discloses a robot for detecting the edge quality of a steel coil based on machine vision, and mainly relates to an image acquisition unit, an image processing unit, a PC unit and an alarm processing unit. The image acquisition unit consists of an LED light source, a CCD camera and a mechanical arm, and can quickly adjust different shooting angles according to different working conditions. The direction adjustment is realized by connecting a motor main shaft with a driving shaft in a key mode, and the driving shaft is connected with a mechanical supporting arm. The camera is connected with the image processing unit and used for acquiring image information of the surface of the product in real time and sending the acquired image information to the image processing unit, the image processing unit receives the image information sent by the camera and performs compression processing on the image information, the processed image information is displayed on the display, and the alarm can give an alarm for unqualified products. Finally, all the detected data are stored in a storage unit, so that data management is facilitated; the PC unit includes a storage unit, a control unit, and a display unit.
Description
Technical Field
The invention relates to a robot for detecting the quality of the edge of a steel coil, which can detect the flaw of the edge of the steel coil.
Background
With the increase of the production of steel coils, the speed of manual detection can not meet the production requirement, and the quality level of manual detection is uneven, which becomes an important factor for restricting the improvement of steel quality. Secondly, because the relative surface illumination of the edge of the steel coil is insufficient, more severe requirements are put forward for inspectors, and the detection in a dark light environment for a long time is harmful to the health of production workers. Therefore, automatic detection techniques have been developed in recent years.
For example, CN208188046U discloses a robot and system for product quality detection, which includes a six-axis robot, an ultrasonic detection device, a terahertz detection device and a control device, wherein the detection system detects the product quality based on ultrasonic waves and reflection.
In recent years, the product inspection technology based on machine vision has been rapidly developed with the advantages of high efficiency, high reliability and low cost. The machine vision detection system is applied to a plurality of industrial production lines, and the prior patent technology such as CN208341168U introduces a product surface detection system based on machine vision, which elaborates product surface detection in detail and mainly aims at product surface quality detection, while the invention mainly aims at steel coil edge quality detection and designs a detection robot which is directly driven by a motor to turn and is convenient to move and install.
The invention content is as follows:
the invention aims to solve the problem of edge quality detection on an automatic steel plate production line and efficiently and accurately identify edge defects.
The invention relates to a robot for detecting the edge quality of a steel coil based on machine vision, which comprises an image acquisition unit, an image processing unit, a PC unit and an alarm unit, wherein the image acquisition unit comprises an LED light source, a CCD camera and a mechanical arm, the LED light source and the CCD camera are fixed on the mechanical arm in a screw connection mode, the mechanical arm comprises a large supporting arm and a small supporting arm, a shock absorption beam is arranged between the large supporting arm and the small supporting arm, and the mechanical arm is directly driven by a motor spindle to shorten the response time and simplify the system structure; the image acquisition unit is connected with the image processing unit through an I/O port to transmit image data, the image processing unit removes residual noise and optical noise in an image by applying median filtering, and based on image segmentation processing of a threshold value, edge defects are identified by applying wavelet energy, specifically:
during median filtering, a template containing odd number of points is used, the gray value of the central pixel point of the template is replaced by the median of the gray values of all the points in the template, and a group of one-dimensional sequences f is set1，f2，f3，…，fnTaking the length of the template as n, carrying out median filtering, and extracting n numbers f from the one-dimensional sequencei-v，…，fi-1，fi，fi+1，…，fi+vWherein f isiV is (n-1)/2, v represents the one-way length of the sequence, the n numbers are rearranged into a new number sequence according to the size sequence, and then the value y at the center of the sequence number of the new number sequence is takeniAs a filtered output; expressed by the following expression:
yi＝Med{fi-v，…，fi-1，fi，fi+1，…，fi+v}i∈Z，v＝(n-1)/2
meanwhile, the numerical value y of the central point of the serial number of the new sequence during two-dimensional median filteringijComprises the following steps:
yij＝Med{fij}，
wherein f is1，f2，f3，…，fnRepresents a one-dimensional sequence: med represents the median of the sequence number center points of the new sequence,
when the threshold-based image segmentation processing is performed, a preset gray value is used as a threshold t, and a segmented image g (x, y) is represented as:
the image processing unit adopts wavelet energy method to detect the energy of the divided image, when the wavelet basis functions are mutually orthogonal, the wavelet energy E under single scalejComprises the following steps:
in the expression, j and k are both positive integers, j represents a demarcation scale, k is the number of sampling points, Cj(k) K-layer decomposition as a vector;
judging whether the defect exists according to the wavelet energy, wherein the defect point has high-energy information, and the background point has low-energy information; the PC unit comprises a control unit, a storage unit and a display unit, the alarm unit comprises an LED strobe light and an alarm bell, and when the PC unit detects that the wavelet energy of the edge of the steel coil reaches a set value, the alarm unit gives an alarm.
Preferably, the threshold is set to a range t1，t2]In the range [ t1，t2]The gradation value within the range [ t ] becomes 1 or 01，t2]The outer one becomes 0 or 1.
Preferably, the LED light source adopts a strip light source, the strip light source is connected with the PC unit through a power line, and the switch and the brightness of the strip light source are controlled by the PC unit; the CCD camera is connected with the image processing unit through an I/O port.
Preferably, all be equipped with the sensor on the motor, with the data access of gathering PC unit to real-time supervision robot operating condition and operational aspect.
Preferably, a base of the mechanical arm is provided with a threaded hole to fix the image acquisition unit.
Preferably, the buffering beam is made of a polyurethane buffering material, so that the noise influence of production vibration on shooting is weakened.
The invention has the beneficial effects that:
(1) by using a machine vision technology, an industrial camera is used for capturing the edge defects of the steel coil instead of manpower, the quality of the edge of the steel coil can be detected more quickly and efficiently, defective product images can be displayed, and the positions of the defects can be identified;
(2) the robot mechanical arm can be fixed on a production line and can be fixed independently of the production line, and a shock absorption beam is arranged in the middle of the mechanical arm, so that on one hand, the noise caused by the vibration of the production line can be weakened, and on the other hand, the mass of the mechanical arm is lightened;
(3) a sensor is arranged on the motor and connected with an industrial PC, so that the working state of the robot can be monitored in real time and whether the robot needs to be repaired or not can be monitored;
(4) the light source adopts an LED matrix light source, so that the illumination quality of a shooting area is improved, and the influence of optical noise in the shooting process is reduced; the camera adopts a CCD camera, is stable and efficient, and can adjust the shooting speed according to the production speed.
Drawings
FIG. 1 is a schematic diagram of a robot system;
FIG. 2 is a schematic diagram of an image acquisition architecture;
fig. 3 is a schematic structural diagram of a second motor driving unit.
Reference numerals:
a base 1; a first electric machine 2; a first motor rotating shaft 3; a drive shaft clamp plate 4; a bearing seat 5; a second motor 6; a large support arm 7; a large cushioning beam 8; a third motor 9; a fourth motor 10; a small support arm 11; a small cushioning beam 12; a fifth motor 13; an LED light source 14; a CCD camera 15; a drive shaft 16; a deep groove ball bearing 17; a motor spindle 18; a key groove 19.
Detailed Description
The following detailed description of the preferred embodiments of the present invention, taken in conjunction with the accompanying drawings, will make the advantages and features of the invention easier to understand by those skilled in the art, and thus will clearly and clearly define the scope of the invention.
Fig. 1 is a schematic structural diagram of a robot system for detecting edge quality of a steel plate based on machine vision. The invention comprises an image acquisition unit, an image processing unit, a high-performance PC unit and an alarm unit. The image acquisition unit is installed on the mechanical arm by an LED light source 14 and a CCD camera 15.
The robot arm is directly driven by the motor spindle 18 to shorten response time and increase system flexibility. All motors are provided with sensors, monitoring data are transmitted into a high-performance PC, and the running conditions of the motors and the robot can be monitored and analyzed in real time.
The middle of the large support arm of the mechanical arm is provided with a large shock absorption beam 8, the middle of the small support arm of the mechanical arm is provided with a small shock absorption beam 12, and the shock absorption beams are fastened on the large support arm 6 and the small support arm 11 through bolts.
The large cushioning beam 8 and the small cushioning beam 12 are made of polyurethane cushioning rubber, the polyurethane cushioning rubber ensures that the influence of noise on image quality is reduced in the image acquisition process, and the weight can be reduced on the premise of keeping the rigidity of the mechanical arm.
The large shock absorption beam 8 is of a cuboid structure with the same width as the two large supporting arms, and a threaded hole is formed in the large shock absorption beam 8, corresponds to the threaded hole of the large supporting arm, and is fastened on the large supporting arm through a bolt structure.
The structure of the small cushioning beam 12 is similar to that of a large cushioning beam, but in order to increase the rigidity of the small arm, the top of the small cushioning beam 12 has a radian and is slightly higher than the small supporting arm part, and the length is slightly shorter than the length of the small supporting arm, so that a rotating space is reserved for the fifth motor 13. The small cushioning beam 12 and the small support arm are also connected by bolts.
The LED light source 14 adopts a strip light source with uniform luminous intensity, which can reduce the distortion and distortion of the image in the image acquisition process and ensure the accurate image acquisition.
Furthermore, the alarm unit consists of an LED strobe light and an alarm bell, the LED strobe light flashes when the defect is found and the alarm bell rings, and only when the defect is found by manual confirmation, the operator can cancel the LED alarm light and the alarm bell through the PC terminal key.
The high-performance PC can be divided into a display unit, a storage unit and a control unit, wherein the display unit mainly realizes the function that on one hand, an image processed by the image processing card is displayed on a display screen; on the other hand, the running state of the whole robot is displayed through sensors on parts such as a motor and the like; the storage unit stores the detected defect information; the control unit controls the operation of the whole robot system, including the adjustment of the angle of the mechanical arm, the photographing frequency of the CCD camera 15, and the like.
Furthermore, the display unit is a high-precision liquid crystal display, the storage unit is a high-performance PC built-in memory, and the control unit is a Keynes KV-7000 series PLC control system.
The image processing unit mainly comprises an image processing card, and the positioning, denoising and entropy energy curve of the image are completed by the image processing card. The image acquisition unit is connected with the image processing unit through the I/O to transmit data.
And the processed image is transmitted to a high-performance PC display through an I/O port, and the high-performance PC display is used for displaying the defects of the edge quality of the detected steel plate.
The specific working process of the invention is as follows:
the mechanical arm base 1 is installed on a production line or independently installed. The high performance PC adjusts the camera angle and aims at the production line. When the mechanical arm needs to adjust the shooting angle, the mechanical arm only needs to adjust and control the motor through the high-performance PC to rotate. As shown in the motor unit of FIG. 3, a stepping motor main shaft 18 is connected with a driving shaft 16 through a key slot 19, a high-performance PC controls the motor to rotate, the motor main shaft 18 rotates to drive the driving shaft 16 to rotate, the driving shaft 16 is connected with a large supporting arm side plate 7, the large supporting arm 7 is installed on a deep groove ball bearing 17, the support arm 7 is driven to rotate by the rotation of the driving shaft 16, and the deep groove ball bearing 17 reduces the rotation resistance to drive the large supporting plate to rotate.
Furthermore, the rotation modes of the first motor 2, the third motor 9, the fourth motor 10 and the fifth motor 13 are similar to those of the second motor, the first motor 2 rotates to drive the first single-machine rotating shaft 3 to rotate, the rotation of the mechanical arm on the horizontal plane is realized, the driving shaft clamping plate 4 is connected with the rotating shaft 3 and the large supporting arm 7, and the bearing seat 5 is provided with a rotating bearing.
After the optimal angle is adjusted, the CCD camera continuously shoots production line products, shot pictures are directly transmitted into the image processing unit through the I/O port, the pictures directly shot through the industrial camera generally cannot have ideal effects, the quality of the pictures is low due to noise, illumination and the like, and therefore the directly shot pictures are generally subjected to preprocessing such as filtering or denoising. The preprocessing can reduce noise interference of an original image, interference caused by light source problems and the like, so that useless information of the image is removed, useful information is reserved, and a good basis is provided for subsequent image segmentation and feature recognition.
The image preprocessing is to eliminate noise interference and maintain various non-negligible details in the image, and the image processing unit adopts median filtering. Median filtering can not only make the image smoother, but also make the image detail clearer under certain conditions.
The main working steps of median filtering are: in the process that the template moves in the graph, the center of the template corresponds to a single pixel point; reading all pixel gray values in the template; sorting the read gray values according to the size sequence; finding out the median value; and assigning the median value to a pixel point at the center of the template as a processed gray value. And because median filtering is not simply an average, it produces less blurring.
Further, it can be seen from the principle of median filtering that only one template with odd number of points is used, and the gray value of the central pixel point of the template is replaced by the median of the gray values of the points in the template, assuming that the template is 3 × 3, the median is 23 if the values in the template are 21, 24, 34, 15, 35, 23, 30, 43, and 17, respectively.
Further, there is a set of one-dimensional sequences f1，f2，f3，…，fnTaking the length of the template as n, carrying out median filtering, and extracting n numbers f from the one-dimensional sequencei-v，…，fi-1，fi，fi+1，…，fi+vWherein f isiThe number v is (n-1)/2, the n numbers are rearranged according to the size sequence, and then the numerical value of the center point of the serial number is taken as the filtering output; expressed by the following mathematical expression:
yi＝Med{fi-v，…，fi-1，fi，fi+1，…，fi+v}i∈Z，v＝(n-1)/2
meanwhile, the two-dimensional median filtering can be expressed as:
yij＝Med{fij}
the image processing unit of the invention performs image segmentation on the preprocessed image. Image segmentation is the division of an image into regions that do not overlap and have their characteristic meaning. The purpose of image segmentation is to segment the defective part from the whole image and calculate it separately. The image processing unit adopts image segmentation based on a threshold, the threshold segmentation utilizes the gray characteristic difference of a target point and a background point in an image, the image is regarded as the combination of two types of areas with different gray levels, and each pixel point in the image is divided into the target point or the background point by utilizing one threshold so as to generate a corresponding binary image;
further, assuming an image, a gray value is found as a threshold t by a certain method, and then the segmented image can be represented as:
further, the threshold may also be set to a range t1,t2]As long as the gradation value in this range becomes 1 or 0, the other becomes 0 or 1 on the contrary.
Further, the image processing unit detects the energy of the segmented image by using a wavelet energy method. When the wavelet basis functions are orthogonal to each other, the wavelet transform energy is conserved, and then the wavelet energy under a single scale can be defined
Further, j and k in the energy expression are both positive integers, j represents a boundary scale, and k is the number of sampling points.
Further, the total energy of the signal is
Further, the defect points are high-energy information, and the background points belong to low-energy information. The high-energy information and the low-energy information are relative concepts, and the information to which the background point belongs is the low-energy information. In other words, it is determined whether there is a defect according to the level of wavelet energy, where the defect point has high energy information and the background point has low energy information.
Furthermore, the alarm unit gives an alarm, when the stop button is manually pressed down by the control unit, the alarm device stops giving an alarm, and the computer stores the defect data of the product, so that the data management is facilitated.
Finally, it should be noted that: the above-mentioned embodiments are only used for illustrating the technical solution of the present invention, and not for limiting the same; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.
Claims (6)
1. The utility model provides a coil of strip limit portion quality testing robot based on machine vision, includes image acquisition unit, image processing unit, PC unit and alarm unit, its characterized in that:
the image acquisition unit comprises an LED light source, a CCD camera and a mechanical arm, the LED light source and the CCD camera are fixed on the mechanical arm in a threaded connection mode, the mechanical arm comprises a large supporting arm and a small supporting arm, a cushioning beam is arranged between the large supporting arm and the small supporting arm, and the mechanical arm is directly driven by a main shaft of a motor to shorten response time;
the image acquisition unit is connected with the image processing unit through an I/O port to transmit image data, the image processing unit removes residual noise and optical noise in an image by applying median filtering, and based on image segmentation processing of a threshold value, edge defects are identified by applying wavelet energy, specifically:
during median filtering, a template containing odd number of points is used, the gray value of the central pixel point of the template is replaced by the median of the gray values of all the points in the template, and a group of one-dimensional sequences f is set1，f2，f3，…，fnTaking the length of the template as n, carrying out median filtering, and extracting n numbers f from the one-dimensional sequencei-v，…，fi-1，fi，fi+1，…，fi+vWherein f isiV is (n-1)/2, v represents the one-way length of the sequence, the n numbers are rearranged into a new number sequence according to the size sequence, and then the value y at the center of the sequence number of the new number sequence is takeniAs a filtered output; expressed by the following expression:
yi＝Med{fi-v，...，fi-1，fi，fi+1，...，fi+v}i∈Z，v＝(n-1)/2
meanwhile, the numerical value y of the central point of the serial number of the new sequence during two-dimensional median filteringijComprises the following steps:
yij＝Med{fij}，
wherein f is1，f2，f3，…，fnRepresents a one-dimensional sequence: med represents the median of the sequence number center points of the new sequence,
when performing the threshold-based image segmentation process, a preset gray value is taken as a threshold t, and a segmented image g (x, y) is expressed as:
the image processing unit adopts wavelet energy method to detect the energy of the divided image, when the wavelet basis functions are mutually orthogonal, the wavelet energy E under single scalejComprises the following steps:
in the expression, j and k are both positive integers, j represents a demarcation scale, k is the number of sampling points, Cj(k) K-layer decomposition as a vector;
judging whether the defect exists according to the wavelet energy, wherein the defect point has high-energy information, and the background point has low-energy information;
the PC unit comprises a control unit, a storage unit and a display unit, the alarm unit comprises an LED strobe light and an alarm bell, and when the PC unit detects that the wavelet energy of the edge of the steel coil reaches a set value, the alarm unit gives an alarm.
2. The robot for detecting the quality of the edge of the steel coil based on the machine vision as claimed in claim 1, wherein: setting the threshold to a range t1，t2]In the range [ t1，t2]The gradation value within becomes 1 within the range [ t ]1，t2]Outside becomes 0, or in the range [ t ]1，t2]The gradation value within becomes 0 in the range [ t ]1，t2]The outer one becomes 1.
3. The robot for detecting the quality of the edge of the steel coil based on the machine vision as claimed in claim 2, wherein: the LED light source adopts a strip light source, the strip light source is connected with the PC unit through a power line, and the switch and the brightness of the strip light source are controlled by the PC unit; the CCD camera is connected with the image processing unit through an I/O port.
4. The robot for detecting the quality of the edge of the steel coil based on the machine vision as claimed in claim 3, wherein: all be equipped with the sensor on the motor, with the data access of gathering PC unit to real-time supervision robot operating condition and operational aspect.
5. The robot for detecting the quality of the edge of the steel coil based on the machine vision as claimed in claim 4, wherein: the base of arm is equipped with the screw hole to realize image acquisition unit's is fixed.
6. The robot for detecting the quality of the edge of the steel coil based on the machine vision as claimed in claim 5, wherein: the buffering beam is made of a polyurethane buffering material so as to weaken noise influence on shooting caused by production vibration.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910275490.9A CN110082365B (en) | 2019-04-08 | 2019-04-08 | Robot for detecting quality of edge of steel coil based on machine vision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910275490.9A CN110082365B (en) | 2019-04-08 | 2019-04-08 | Robot for detecting quality of edge of steel coil based on machine vision |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110082365A CN110082365A (en) | 2019-08-02 |
| CN110082365B true CN110082365B (en) | 2020-07-03 |
Family
ID=67414414
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910275490.9A Active CN110082365B (en) | 2019-04-08 | 2019-04-08 | Robot for detecting quality of edge of steel coil based on machine vision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110082365B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113465511B (en) * | 2021-06-19 | 2022-12-13 | 精锐视觉智能科技(上海)有限公司 | Steel coil size online measurement and omnibearing end surface defect online detection method |
| JP7327675B2 (en) * | 2021-07-01 | 2023-08-16 | 東芝三菱電機産業システム株式会社 | data collection device |
Family Cites Families (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150185147A1 (en) * | 2012-08-30 | 2015-07-02 | Jason Miller | Apparatus for determining gauge profile for flat rolled material with laser-based lap counter |
| CN103207185A (en) * | 2012-01-11 | 2013-07-17 | 宝山钢铁股份有限公司 | Steel coil end portion quality detection system and method thereof |
| CN104535586B (en) * | 2014-11-18 | 2017-05-17 | 华东理工大学 | Strip steel edge defect detection identification method |
| CN104777224A (en) * | 2015-04-28 | 2015-07-15 | 华南理工大学 | Defect detecting method for junction surface of metal alloy |
| CN108698774A (en) * | 2016-02-25 | 2018-10-23 | 康宁股份有限公司 | The method and apparatus that edge surface inspection is carried out to movable glass coiled material |
| FR3053792B1 (en) * | 2016-07-06 | 2023-07-14 | Tiama | METHOD, DEVICE AND INSPECTION LINE FOR DETERMINING A BURR AT THE LOCATION OF AN INTERNAL EDGE OF A RING SURFACE |
| CN207256220U (en) * | 2017-05-10 | 2018-04-20 | 岳佩瑜 | A kind of omnidirectional of substation crusing robot |
| CN109186912A (en) * | 2018-06-29 | 2019-01-11 | 浙江工业大学 | Mechanical arm damping test device based on camera |
| CN109540918B (en) * | 2018-11-28 | 2021-04-16 | 鞍钢集团自动化有限公司 | Hot-rolled coil edge defect detection device and method |
-
2019
- 2019-04-08 CN CN201910275490.9A patent/CN110082365B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN110082365A (en) | 2019-08-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US6791695B2 (en) | Shearographic imaging machine with archive memory for animation data and air handling system | |
| CN103913468B (en) | Many defects of vision checkout equipment and the method for large-scale LCD glass substrate on production line | |
| CN104007122A (en) | Cylindrical object lateral surface detection device and method based on turntable type assembly line | |
| CN203124215U (en) | Frame sealant coating machine | |
| CN110082365B (en) | Robot for detecting quality of edge of steel coil based on machine vision | |
| CN112945972B (en) | Steel wire rope state detection device and method based on machine vision | |
| JPS60119404A (en) | Inspecting device for distortion of plate glass | |
| CN106468668A (en) | Industrial camera cylinder detection method | |
| CN108225798B (en) | Tire slippage testing system and method based on edge enhancement identification | |
| CN112858321A (en) | Steel plate surface defect detection system and method based on linear array CCD | |
| CN111672773A (en) | Product surface defect detection system and method based on machine vision | |
| CN217505690U (en) | Film product detection device based on machine vision | |
| CN116718616A (en) | Machine vision detection system and detection method for flaw detection | |
| CN114170598B (en) | Colony height scanning imaging device, and automatic colony counting equipment and method capable of distinguishing atypical colonies | |
| CN113566733B (en) | Line laser vision three-dimensional scanning device and method | |
| CN114092471A (en) | Visual detection method and detection device for surface defects of rudder type structural member | |
| CN115355824B (en) | Light-transmitting tube image acquisition method, tube diameter measurement method and device | |
| CN110595959B (en) | Massecuite particle size analysis system and analysis method | |
| CN112289154A (en) | Machine vision teaching system and method for Newton's ring experiment | |
| CN114677597B (en) | Gear defect visual inspection method and system based on improved YOLOv5 network | |
| CN211014597U (en) | Glass surface line defect rapid detection device | |
| CN108225175B (en) | Device and method for acquiring tooth surface contact mark data of bevel gear of tractor transmission system | |
| CN112051272A (en) | High-pressure gas cylinder inner surface defect detection system based on machine vision | |
| CN111389743A (en) | Rubber sealing ring detection device and detection method | |
| CN114136987B (en) | Device and method for detecting deformation defect of lens |
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 |