CN108918543B - Dynamic detection device and method for mirror surface scratch - Google Patents
Dynamic detection device and method for mirror surface scratch Download PDFInfo
- Publication number
- CN108918543B CN108918543B CN201810455914.5A CN201810455914A CN108918543B CN 108918543 B CN108918543 B CN 108918543B CN 201810455914 A CN201810455914 A CN 201810455914A CN 108918543 B CN108918543 B CN 108918543B
- Authority
- CN
- China
- Prior art keywords
- lens
- ccd camera
- detected
- inspected
- lifting rod
- 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
- G01N21/89—Investigating the presence of flaws or contamination in moving material, e.g. running paper or textiles
- G01N21/892—Investigating the presence of flaws or contamination in moving material, e.g. running paper or textiles characterised by the flaw, defect or object feature 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
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/89—Investigating the presence of flaws or contamination in moving material, e.g. running paper or textiles
- G01N21/892—Investigating the presence of flaws or contamination in moving material, e.g. running paper or textiles characterised by the flaw, defect or object feature examined
- G01N21/896—Optical defects in or on transparent materials, e.g. distortion, surface flaws in conveyed flat sheet or rod
Landscapes
- Engineering & Computer Science (AREA)
- Textile Engineering (AREA)
- 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)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The invention belongs to the technical field of machine vision defect detection, and particularly relates to a dynamic detection device and a method for scratches on the surface of a mirror, wherein the device is an intelligent and automatic detection platform which is provided with a camera bellows, a movable scanning light source, a high-speed array CCD camera which are symmetrically arranged, and can realize multi-station transformation and other structures of a lens to be detected, and can quickly, accurately and repeatedly acquire gray level images of the lens to be detected at different positions, different reflection angles and different high-brightness areas; meanwhile, a Gauss low-pass filtering algorithm, a Laplace operator, binarization and a Hough transform image visual algorithm which are sequentially carried out are adopted, the scratch of the lens to be detected is quickly, accurately and stably extracted, and the method is particularly suitable for extracting the infinitesimal scratch on the surface of the mirror, so that the detection efficiency and the detection quality are greatly improved, and the cost is saved; the dynamic detection device and the method are easy to implement, meet the production mode of online detection, are stable and reliable, and are convenient to popularize and apply.
Description
Technical Field
The invention belongs to the technical field of machine vision defect detection, and particularly relates to a dynamic detection device and method for scratches on the surface of a mirror.
Background
With the continuous development of the glass industry and the continuous increase of the demand, the quality requirement of glass products is higher and higher. Because of the influence of manufacturing process, human factors and the like, the glass original plate has the possibility of generating defects at any stage of the production process, and the glass quality detection plays a very important role at the moment. According to the regulations of the current standards of glass, the common defects of glass mainly comprise: air bubbles, tin sticking, scratches, inclusions and the like.
The traditional mirror quality detection methods are manual detection methods, but along with the requirements of high quality of users and the aggravation of market competition, the requirements on the quality of mirrors are higher and higher, for example, the automobile rearview mirror requires that the mirror surface has no defects of deformation, blurring, stripes, bubbles, scars, cracks, inclusions and the like, and strict requirements on the flatness of the mirror surface and even some curvature radiuses of the edge of the mirror are also provided, at the moment, the traditional manual detection methods cannot meet the production requirements of high quality and high efficiency.
Nowadays, in the mirror industrial production process, the problems of surface defects of products such as scratches are often encountered, and the problems are very challenging for both manual detection and machine vision detection. The detection difficulty of the scratch on the surface of the mirror is as follows: the scratch shape is irregular, the depth contrast is low, and the reflection of the light on the surface of the mirror interferes the detection process and result. Therefore, in the process of detecting the scratch on the surface of the mirror, very high requirements are placed on correct lighting, camera resolution, relative positions of the detected part and an industrial camera, complex machine vision algorithms and the like.
Disclosure of Invention
The invention aims to provide a dynamic detection device for scratches on the surface of a mirror, which provides a dark field environment for image acquisition, can repeatedly acquire images of a to-be-detected lens under the conditions of different positions, different reflection angles and different highlight areas, reduces the probability of missed detection and false detection, and improves the precision and accuracy of scratch detection; the device can realize intelligent automatic control, when improving detection quality and detection efficiency, practices thrift the cost.
The invention also provides a dynamic detection method for the scratch on the surface of the mirror, which collects the gray level image in the dark field environment, adopts the unique noise reduction and image enhancement algorithm of the gray level image, improves the extraction accuracy, accuracy and stability of the machine vision algorithm, simplifies the machine vision algorithm, realizes the intelligent automatic control of the scratch detection of the lens, and meets the requirements of high quality and high efficiency of the scratch detection of the lens.
The technical scheme of the invention is as follows: a dynamic detection device for scratches on the surface of a mirror comprises a camera bellows, a first conveyor belt and a second conveyor belt, wherein the first conveyor belt and the second conveyor belt are used for conveying a lens to be detected; the scanning lamp tube positioned right above the lifting rod is arranged in the scanning lamp tube dynamic slide rail and can move back and forth along the scanning lamp tube dynamic slide rail; a feeding port and a discharging port are respectively arranged on two side surfaces of the black box, and a first conveyor belt and a second conveyor belt are arranged between the feeding port and the discharging port;
the PLC system controls the lifting rod, the first CCD camera, the second CCD camera and the scanning lamp tube; the first CCD camera and the second CCD camera are connected with an image acquisition card, and the image acquisition card transmits the acquired image of the lens to be detected to image analysis software on the computer;
in the material conveying direction from a feeding port to a discharging port, the same lens to be detected is simultaneously placed on a first conveying belt and a second conveying belt, and the first conveying belt and the second conveying belt which synchronously and circularly move are driven by a stepping motor to jointly convey;
the lifting rod is a hollow tube, and a universal vacuum chuck is fixed at the top of the lifting rod; the lifting rod horizontally fixes the lens to be detected on the universal vacuum chuck through negative pressure adsorption, vertically lifts the lens to be detected upwards, and can drive the lens to be detected to make an inclined motion; when the lens to be detected is not adsorbed and fixed, the sucking disc surface of the universal vacuum sucking disc is kept horizontal and is flush with the conveying surfaces of the first conveying belt and the second conveying belt, which are positioned at the upper ends;
the first CCD camera and the second CCD camera are respectively arranged at two sides of the reciprocating movement direction of the scanning lamp tube and are positioned on the central line of the lens to be detected; after the lifting rod vertically and upwards lifts the horizontally fixed lens to be inspected, the lens of the first CCD camera and the lens of the second CCD camera respectively face the lens to be inspected and respectively form included angles with the lens to be inspected; when the universal vacuum chuck at the top of the lifting rod drives the lens to be inspected to do inclined motion, any one of the lens of the first CCD camera and the lens of the second CCD camera can be perpendicular to the lens to be inspected.
After the lifting rod vertically and upwards lifts the horizontally fixed lens to be inspected, the lens of the first CCD camera and the lens of the second CCD camera are intersected at the center of the lens to be inspected and respectively form included angles with the lens to be inspected; when the universal vacuum chuck at the top of the lifting rod drives the lens to be inspected to do inclined motion, any one of the lens of the first CCD camera and the lens of the second CCD camera can be perpendicular to the center of the lens to be inspected.
After the horizontally fixed lens to be inspected is vertically and upwards supported by the supporting rod, the lens of the first CCD camera and the lens of the second CCD camera form an included angle of 45 degrees with the lens to be inspected respectively.
The reciprocating moving direction of the scanning lamp tube is parallel to the material conveying direction, and the first CCD camera and the second CCD camera are symmetrically distributed on the central line of the lens to be detected. Furthermore, the first CCD camera and the second CCD camera are symmetrically distributed on a central line of the lens to be inspected, wherein the central line is parallel to the material conveying direction.
The scanning lamp tube dynamic slide rail is arranged right above the lifting rod, and the lifting rod corresponds to the center of the scanning lamp tube dynamic slide rail; the scanning lamp tube dynamic slide rail comprises a first elliptical ring and a second elliptical ring, the first elliptical ring and the second elliptical ring are respectively and vertically fixed on the inner wall of the top surface of the dark box along the material conveying direction, and the first elliptical ring and the second elliptical ring are horizontally spaced and have flush end surfaces; the two ends of the scanning lamp tube are respectively fixed in the first elliptical ring and the second elliptical ring and slide along the first elliptical ring and the second elliptical ring in a reciprocating manner.
The lifting rod is positioned at the interval between the first conveyor belt and the second conveyor belt and is arranged in the middle in the material conveying direction; the universal vacuum chuck fixed on the top of the lifting rod corresponds to the center of the lens to be detected fixed on the universal vacuum chuck.
The scanning light source adopts an LED strip light source, and the back of the scanning light source is provided with a shading plate.
The method for dynamically detecting the scratch on the surface of the mirror by using the dynamic detection device for the scratch on the surface of the mirror comprises the following steps:
I. two conveyor belts which are arranged at intervals in the material conveying direction synchronously and circularly move under the driving of a stepping motor, and when the lens to be detected which is placed on the two conveyor belts moves to the upper part of the lifting rod along with the two conveyor belts, a material inlet and a material outlet of the camera bellows are closed, and the PLC system controls the lifting rod, the scanning light source, the first CCD camera and the second CCD camera to simultaneously start working;
the method comprises the following steps that II, under the control of a PLC system, a lifting rod vertically and upwards supports a lens to be inspected, the top end of the lens to be inspected is horizontally fixed, a scanning light source starts to emit light and performs reciprocating scanning along the whole course of a dynamic sliding rail of a scanning lamp tube, when the scanning light source which starts to emit light is respectively positioned at two end parts of the dynamic sliding rail of the scanning lamp tube, the lens to be inspected respectively performs one-time tilting motion until the lens to be inspected is perpendicular to a lens of a first CCD camera or a lens of a second CCD camera, the end parts of the lens to be inspected and the scanning light source are positioned at the same side, the first CCD camera or the second CCD camera which is perpendicular to the lens to be inspected collects images of the lens to be inspected in different positions, different reflection angles and different;
when the scanning light source does reciprocating movement scanning along the whole course of the scanning lamp tube dynamic slide rail, the lens to be detected is kept in a horizontal placing state; the first CCD camera and the second CCD camera are respectively arranged at two sides of the reciprocating movement direction of the scanning light source, are positioned on the central line of the lens to be detected and form an included angle with the lens to be detected, respectively collect images of the lens to be detected at different positions, different reflection angles and different highlight areas in real time, and transmit the images to image analysis software on a computer through an image acquisition card;
image analysis software sequentially adopts a visual algorithm of a Gauss low-pass filtering algorithm, a Laplace operator, binarization and Hough transform to perform noise reduction, filtering, binarization and Hough transform processing, and rapidly, accurately and stably extracts scratches of the lens to be detected;
when the lens to be detected is scratched, the lens to be detected is delivered to one of the first conveyor belt and the second conveyor belt and is delivered out of the discharge port; when the lens to be detected is not scratched, the lens to be detected is delivered to the other one of the first conveyor belt and the second conveyor belt and is delivered out of the discharge port.
In the step II, when the scanning light sources for starting to emit light are respectively positioned at the two end parts of the dynamic slide rail of the scanning lamp tube, the lens of the first CCD camera or the lens of the second CCD camera which is positioned at the same side with the end part where the scanning light sources are positioned is vertical to the center of the inclined lens to be detected;
when the scanning light source does reciprocating movement scanning along the whole course of the dynamic slide rail of the scanning lamp tube, the lens of the first CCD camera and the lens of the second CCD camera are intersected at the center of the lens to be detected and form included angles with the lens to be detected respectively.
In the step II, when the scanning light source does reciprocating movement scanning along the whole course of the dynamic slide rail of the scanning lamp tube, the first CCD camera and the second CCD camera respectively form an included angle of 45 degrees with the lens to be detected.
In the step II, the reciprocating movement scanning direction of the scanning light source is parallel to the material conveying direction, and the first CCD camera and the second CCD camera are symmetrically distributed on the central line of the lens to be detected. Furthermore, the first CCD camera and the second CCD camera are symmetrically distributed on a central line of the lens to be inspected, wherein the central line is parallel to the material conveying direction.
In the step II, the lifting rod horizontally fixes the lens to be detected on the universal vacuum chuck at the top end of the lifting rod through negative pressure adsorption under the control of the PLC system. Under the control of the PLC system, when the lens to be detected is moved to the lifting rod and corresponds to the center of the lens to be detected, the lens to be detected is horizontally fixed on the universal vacuum chuck at the top end of the lifting rod through negative pressure adsorption.
And step II, the scanning light source is initially positioned at the end part of the scanning lamp tube dynamic slide rail, and the scanning light source starts to emit light and performs reciprocating scanning along the initial end to the terminal end of the scanning lamp tube dynamic slide rail under the control of the PLC system. When the scanning light source is located at the starting end and the terminating end of the dynamic sliding rail along the scanning lamp tube, the lens to be detected respectively makes one-time tilting motion.
Compared with the prior art, the invention has the beneficial effects that:
the invention provides a dynamic detection device and a detection method for scratches on the surface of a mirror, and provides an intelligent and automatic detection platform which is provided with a camera bellows, a movable scanning light source, symmetrically arranged high-speed display CCD cameras and can realize multi-station transformation and other structures of a lens to be detected, so that gray images of the lens to be detected under the conditions of different positions, different reflection angles and different highlight areas can be rapidly, accurately and repeatedly acquired, and the high-efficiency acquisition of the gray images on the surface of the lens to be detected is realized; meanwhile, a Gaussian low-pass filtering algorithm, a Laplace operator, binarization and Hough transform image visual algorithm are sequentially adopted, the scratch of the lens to be detected is extracted quickly, accurately and stably while the method is reasonable and simple, the method is particularly suitable for extracting the very weak scratch detection on the surface of the mirror, the detection efficiency and the detection quality are greatly improved, and the cost is saved.
Compared with the existing manual detection and common scratch detection platform, the dynamic detection platform and the method for scratches on the surface of the mirror are intelligent, efficient, fast, high in accuracy, easy to implement, stable, reliable and convenient to popularize and apply, and meet the production mode of online detection.
Drawings
Fig. 1 is a perspective schematic view of a mirror surface scratch dynamic detection device of the present invention.
FIG. 2 is a comparison image of the images of the lens to be inspected before and after the dynamic detection method for the scratch on the surface of the mirror. In the figure, A is the image of the lens to be detected before the dynamic detection method for the scratch on the surface of the mirror is used for processing, and B is the image of the lens to be detected after the dynamic detection method for the scratch on the surface of the mirror is used for processing.
The system comprises a camera obscura, a camera lens, a camera fixing rod.
Detailed Description
The following detailed description of embodiments of the invention refers to the accompanying drawings.
As shown in fig. 1 and 2, the dynamic mirror surface scratch detection platform includes a camera bellows 1, a first conveyor belt 2 and a second conveyor belt 3 for conveying a lens 9 to be detected, a lifting rod 4 is disposed between the first conveyor belt 2 and the second conveyor belt 3, the camera bellows 1 is further provided therein with a first camera fixing rod 5, a second camera fixing rod 6, a scanning lamp dynamic slide rail 7 and a scanning lamp 8, the first camera fixing rod 5 and the second camera fixing rod 6 are disposed above the lifting rod 4, the first camera fixing rod 5 is fixed with a first CCD camera 51, the second camera fixing rod 6 is fixed with a second CCD camera 61, and the first CCD camera 51 and the second CCD camera 61 are also located above the lifting rod 4; the scanning lamp tube 8 positioned right above the lifting rod 4 is arranged in the scanning lamp tube dynamic slide rail 7 and can reciprocate along the scanning lamp tube dynamic slide rail 7; a feeding port 11 and a discharging port 12 are respectively arranged on two side surfaces of the camera bellows 1, and the first conveyor belt 2 and the second conveyor belt 3 are arranged between the feeding port 11 and the discharging port 12.
The first conveyor belt 2 and the second conveyor belt 3 are positioned in the camera bellows 1, and two ends of the first conveyor belt 2 and the second conveyor belt 3 are respectively and closely adjacent to the feeding port 11 and the discharging port 12; alternatively, the first conveyor belt 2 and the second conveyor belt 3 may also pass through the camera bellows 1 from the material inlet 11 and the material outlet 12, respectively.
The camera bellows 1 isolates outside natural light, and after the lens 9 to be detected enters the camera bellows 1, the feeding port 11 and the discharging port 12 can be closed, and a dark field environment is formed inside the camera bellows 1. In the material conveying direction from a feeding port 11 to a discharging port 12, the same lens to be detected is simultaneously placed on a first conveying belt 2 and a second conveying belt 3, the first conveying belt 2 and the second conveying belt 3 which are driven by a stepping motor to move synchronously and circularly are conveyed together, and the first conveying belt 2 and the second conveying belt 3 are arranged at intervals.
The lifting rod 4 can lift the lens 9 to be inspected, which is conveyed by the first conveyor belt 2 and the second conveyor belt 3 together, vertically upwards for photographing and imaging under a subsequent scanning light source. The lifting rod 4, the first CCD camera 51, the second CCD camera 61 and the scanning lamp tube 8 are matched with each other for imaging detection of the lens to be detected.
The lifting rod 4 is located at the interval between the first conveyor belt 2 and the second conveyor belt 3 and is arranged in the middle in the material conveying direction from the material inlet 11 to the material outlet 12. The lifting rod 4 is a hollow tube, and a universal vacuum chuck is fixed at the top of the lifting rod 4. Hold in the palm 4 vacuum pumping down in negative pressure states, adsorb firmly through the negative pressure and adsorb the universal vacuum chuck that will examine lens 9 level and set up in the top of holding in the palm 4, universal vacuum chuck drives under micromotor control and examines lens 9 and make the tilt motion.
When the top end of the lifting rod is not adsorbed and fixed with the lens to be detected, the horizontal sucking disc surface of the universal vacuum sucking disc at the top of the lifting rod 4 is flush with the conveying surfaces of the first conveying belt 2 and the second conveying belt 3, which are positioned at the upper ends; when the top of holding in the palm the lifting rod adsorbs fixedly waits to examine the lens, the universal vacuum chuck adsorption at 4 tops of lifting rod fixes the central point that waits to examine lens 9 and puts, and lifting rod 4 can be with fixed waiting to examine on it that lens 9 is vertical upwards to hold up 100 mm.
A scanning lamp tube dynamic sliding rail 7 is arranged right above the lifting rod 4, and the lifting rod 4 corresponds to the center of the scanning lamp tube dynamic sliding rail 7. The scanning lamp tube dynamic slide rail 7 comprises a first elliptical ring 71 and a second elliptical ring 72; the first elliptical ring 71 and the second elliptical ring 72 are vertically fixed on the inner wall of the top surface of the dark box 1 along the material conveying direction from the feeding port 11 to the discharging port 12 respectively, and the first elliptical ring 71 and the second elliptical ring 72 are horizontally spaced and flush in end surface.
Both ends of the scanning lamp tube 8 are fixed in the first elliptical ring 71 and the second elliptical ring 72, respectively, and reciprocate along the first elliptical ring 71 and the second elliptical ring 72. The reciprocating direction of the scanning lamp tube 8 along the dynamic slide rail 7 of the scanning lamp tube is parallel to the material conveying direction.
The back of the scanning lamp tube 8 is provided with a light shielding plate, such as a C-shaped light shielding plate, which seals the back of the scanning lamp tube 8 without light leakage; the scanning lamp tube 8 adopts an LED strip light source.
The first camera fixing rod 5 and the second camera fixing rod 6 are respectively vertically fixed on the inner wall of the top surface of the camera bellows 1 and symmetrically distributed at two ends of the scanning lamp tube dynamic slide rail 7 along the material conveying direction; when the lens to be inspected is fixed at the top end of the lifting rod in an adsorption manner, a first CCD camera 51 fixed on the first camera fixing rod 5 and a second CCD camera 61 fixed on the second camera fixing rod 6 are symmetrically distributed on a central line parallel to the material conveying direction of the lens to be inspected 9.
When the lens to be inspected is adsorbed and fixed at the top end of the lifting rod but is not vertically lifted upwards, the vertical heights of the plane where the lens to be inspected 9 is horizontally fixed at the top end of the lifting rod 4 from the center of the lens of the first CCD camera 51 and the center of the lens of the second CCD camera 61 are both 350mm, and the horizontal distances from the center of the lens of the first CCD camera 51 and the center of the lens of the second CCD camera 61 to the vertical central line of the lifting rod 4 are both 250 mm.
When the lifting rod drives the lens to be inspected 9 with the top end horizontally fixed to the lifting rod to vertically lift upwards by 100mm, as shown in fig. 1, the lens of the first CCD camera 51 and the lens of the second CCD camera 61 intersect at the center of the lens to be inspected 9, and an included angle β between the lens of the first CCD camera 51 and the horizontally fixed lens to be inspected 9 is 45 degrees; the angle α between the lens of the second CCD camera 61 and the horizontally fixed lens to be inspected 9 is 45 degrees.
The first CCD camera 51 and the second CCD camera 61 are connected with an image acquisition card, the image acquisition card transmits the images of the lens to be detected, which are acquired by the first CCD camera 51 and the second CCD camera 61, to image analysis software on a computer, and the visual algorithms of noise reduction, filtering, binaryzation and Hough transformation of the images of the lens to be detected are carried out, so that scratches on the lens to be detected are accurately extracted.
The specific process of mirror surface detection by using the dynamic mirror surface scratch detection platform comprises the following steps:
(1) the first conveyor belt 2 and the second conveyor belt 3 synchronously and circularly move under the drive of the stepping motor, the lens 9 to be detected is placed on the first conveyor belt 2 and the second conveyor belt 3 from the feeding port 11 at the same time, and the lens 9 to be detected is preferably placed in the middle on the first conveyor belt 2 and the second conveyor belt 3; when the lens 9 to be detected moves to the upper part of the lifting rod 4 along with the first conveyor belt 2 and the second conveyor belt 3 synchronously, the material inlet 11 and the material outlet 12 of the camera bellows 1 are closed, and natural light outside the camera bellows 1 is isolated; the PLC system controls the lifting rod 4, the scanning light source 8, the first CCD camera 51 and the second CCD camera 61 to start working simultaneously, and intelligent automatic acquisition of images of the lens to be detected is achieved.
(2) Under the control of a PLC system, a lifting rod 4 drives a lens to be inspected 9 fixed on the lifting rod to move to a station required for detection, a scanning light source 8 located at one end of a dynamic sliding rail 7 of a scanning lamp tube is started to emit light and scan movably, a first CCD camera 51 and a second CCD camera 61 simultaneously capture images of the lens to be inspected in different positions, different reflection angles and different high-brightness areas in real time, the images are transmitted to image analysis software on a computer through an image acquisition card, and the image analysis software sequentially passes through a Gauss low-pass filtering algorithm, a Laplacian operator, binarization and Hough transformation, so that scratches of the lens to be inspected are extracted quickly, accurately and stably. The method comprises the following specific steps:
s1: starting a scanning light source 8 positioned at one end of a dynamic sliding rail 7 of a scanning lamp tube to emit light; the lifting rod 4 horizontally fixes the lens 9 to be detected on a universal vacuum chuck of the lifting rod 4 through negative pressure adsorption, the universal vacuum chuck corresponds to the center of the lens 9 to be detected, and the lens 9 to be detected is vertically lifted upwards by 100 mm; then, under the control of a micro motor, the universal vacuum chuck drives the lens 9 to be detected to do 45-degree tilting motion, and the lens of the first CCD camera 51 which is positioned at the same side with the starting end of the scanning light source 8 is tilted to be vertical to the center of the lens 9 to be detected; at this time, the first CCD camera 51 captures the image of the lens 9 to be inspected, and transmits the image to the image analysis software on the computer through the image acquisition card;
s2: after the first CCD camera 51 captures the image of the lens 9 to be detected, the lens 9 to be detected returns to the horizontal placement state from the inclined state, the scanning light source 8 moves from the starting end to the ending end of the dynamic slide rail 7 of the scanning lamp tube along the direction parallel to the material conveying, and in the process, the first CCD camera 51 and the second CCD camera 61 simultaneously capture the image of the lens 9 to be detected at different positions, different reflection angles and different highlight areas in real time and transmit the image to image analysis software on a computer through an image acquisition card;
s3: when the scanning light source 8 moves to the termination end of the dynamic sliding rail 7 of the scanning lamp tube, the scanning light source stops moving, under the control of a micro-motor, the universal vacuum chuck drives the lens 9 to be detected to do 45-degree inclined motion, and the lens of the second CCD camera 61 which is inclined to the same side as the termination end of the scanning light source 8 is vertical to the center of the lens 9 to be detected; at this time, the second CCD camera 61 captures the image of the lens 9 to be inspected, and transmits the image to the image analysis software on the computer through the image acquisition card;
s4: after the second CCD camera 61 finishes capturing the image, the lens 9 to be inspected returns to the horizontal placement state from the inclined state, the scanning light source 8 moves from the termination end of the scanning lamp tube dynamic slide rail 7 to the starting end along the direction parallel to the material conveying direction to finish resetting, and in the process, the first CCD camera 51 and the second CCD camera 61 simultaneously capture the image of the lens 9 to be inspected at different positions, different reflection angles and different highlight areas in real time and transmit the image to image analysis software on a computer through an image acquisition card;
in the process from S1 to S4, the positions and angles of the first CCD camera 51 and the second CCD camera 61 are fixed, and only the lens 9 to be inspected makes a tilting motion; first CCD camera 51 and second CCD camera 61 symmetric distribution are on waiting to examine the central line parallel with material direction of transfer of lens 9 to distribute in scanning lamp tube dynamic slide rail 7 along the both sides of material direction of transfer, wait to examine the vertical back that upwards holds up of lens 9, first CCD camera 51 and second CCD camera 61 intersect in waiting to examine the center of lens 9, and respectively with waiting to examine lens 9 and form 45 degrees contained angles.
In the reciprocating movement scanning process of the scanning light source 8, the first CCD camera 51 and the second CCD camera 61 which are positioned at two sides of the reciprocating movement direction of the scanning light source 8 can repeatedly acquire images of the lens to be detected under the conditions of different positions, different reflection angles and different highlight areas, the repeated acquisition method reduces the probability of missed detection and false detection, and improves the detection precision and accuracy of scratches.
In the reciprocating scanning process of the scanning light source 8, different highlight areas are generated on the surface of the lens to be detected, brightness difference (namely larger gray value gradient) exists between the highlight areas and the rest areas on the surface of the lens to be detected, and at the moment, if a scratch or a scratch is located at the light-dark junction of the highlight areas and the rest areas, the scratch is very obvious in characteristic and is easier to detect.
(3) And the image analysis software sequentially passes through a Gauss low-pass filtering algorithm, a Laplace operator, binarization and Hough transformation to realize the rapid, accurate and stable extraction of the scratch of the lens to be detected.
If the lens 9 to be detected has a scratch, the lens 9 to be detected is delivered to one of the first conveyor belt 2 and the second conveyor belt 3 and is delivered out of the discharge port 12; if the lens 9 to be inspected has no scratch, the lens 9 to be inspected is delivered to the other of the first conveyor 2 and the second conveyor 3 and is delivered from the discharge port 12.
The image analysis software accurately extracts the scratch of the lens to be detected through a Gauss low-pass filtering algorithm, a Laplace operator, binarization and Hough transform in sequence as follows:
and step I, denoising the collected mirror surface gray level image by adopting a Gaussian low-pass filtering algorithm. Because the image is collected in a dark room, the light quantity is insufficient, and a lot of noise exists on the image. The noise points have certain interference on the detection of the scratch, so that a Gaussian Low Pass Filter algorithm (Gaussian Low Pass Filter) is adopted, a Gaussian kernel is utilized to perform sliding convolution on the whole gray level image, Low-frequency energy (such as noise) in the gray level image is filtered out, the image is smoother, and a clear image with Low noise is obtained.
The method is characterized in that a Laplace Operator (Laplace Operator) is adopted to filter the noise-reduced image, the Laplace Operator is the simplest differential Operator, programming difficulty is low, calculated amount is small, and due to the characteristics of rotation invariance, sensitivity to isolated points or end points and the like, the method is particularly suitable for highlighting isolated points or isolated lines (such as scratches) in the image, so that the boundary information of the image is more obvious, and the scratched boundary information is more prominent in the image. The image after filtering only has two components of a background and a scratch, and the difference of the gray values of the two components is large, so that the accuracy of subsequent Hough transform on scratch detection can be improved.
Since the laplacian operator can enhance the noise in the image, the image is subjected to noise reduction smoothing by using the gaussian low-pass filtering algorithm before the image enhancement by using laplacian, so that the scratch in the image is highlighted, meanwhile, the interference factors such as noise are reduced, and the accuracy of image extraction is improved.
And II, carrying out binarization on the image subjected to noise reduction and filtering processing. Binarization simplifies the image, reduces the data size, and highlights the contour of the target (scratch). Most scratches on the surface of the mirror are linear or arc scratches left by collision or scratch, so that in order to detect scratches more accurately, local binarization processing is performed on the image after preprocessing in the step I, scratches in different areas on the image are unified, different threshold values are set, and a binarized image with a pure black background and high scratches is obtained after binarization processing. The method has the advantages that the scratch in the image is highlighted through Laplace operation, binarization is conducted after Laplace operation, threshold values of different scratch areas can be conveniently set, and a more accurate binarization image is obtained.
And III, extracting scratches of the image subjected to binarization processing in the step II by adopting Hough Transform. The Hough transform directly extracts images which are not subjected to binarization processing, and if the size of the Hough transform is increased, short and light scratches are lost while most of interference is filtered; if the Hough transform size is adjusted to be small, some noisy backgrounds can be mistaken for scratches while scratches are fine; the scratches in the image cannot be accurately extracted. The image after binarization only has two characteristics of pure black and pure white, the background is pure, and interference hardly exists, the optimal scratch detection result can be obtained by subsequent Hough change, and the extraction of Hough transform is more efficient and accurate by binarization processing.
The specific operation of hough transform is: first, a buffer is initialized, and all data of the buffer is set to 0 corresponding to the parameter plane. For each foreground point on the binary image, obtaining source pixel point data on an image space, starting Hough transform algorithm, converting each pixel coordinate point P (x, y) to the curve point of (r, theta), accumulating the curve points to the corresponding grid data points, calculating the corresponding straight line in the parameter plane, counting the occurrence times of all the points on the straight line, and finding the point position with the maximum occurrence times on the parameter plane, wherein the position is the parameter of the straight line on the original image. And finally, searching a maximum Hough value, setting a threshold value, and inversely transforming the straight line corresponding to the scratch on the parameter space into an image space.
The comparison result of the images of the lens to be detected before and after being processed by the dynamic detection method for the scratch on the surface of the mirror is shown in FIG. 2, wherein A is the image of the lens to be detected before being processed by the dynamic detection method for the scratch on the surface of the mirror, and B is the image of the lens to be detected after being processed by the dynamic detection method for the scratch on the surface of the mirror; in A and B, white oblique line scratches are arranged in the white coil winding part. The comparison between the A and the B shows that the scratch on the surface of the lens processed by the dynamic detection method for the scratch on the surface of the mirror is more clearly and prominently strengthened, which shows that the dynamic detection device and the method for the scratch on the surface of the mirror extract the scratch on the surface of the lens to be detected more accurately and stably, and are particularly suitable for detecting the infinitesimal scratch on the surface of the mirror.
Although the present invention has been described in detail with reference to the specific embodiments, it will be appreciated by those skilled in the art that the above embodiments are merely illustrative of the present invention, and that various changes or modifications may be made thereto without departing from the spirit and scope of the invention.
Claims (6)
1. A dynamic detection device for scratches on the surface of a mirror is characterized by comprising a camera bellows (1), a first conveyor belt (2) and a second conveyor belt (3) which are used for conveying a lens (9) to be detected, wherein the first conveyor belt (2) and the second conveyor belt (3) are arranged in parallel at intervals, a lifting rod (4) is arranged between the first conveyor belt (2) and the second conveyor belt (3), a first camera fixing rod (5), a second camera fixing rod (6), a scanning lamp tube dynamic sliding rail (7) and a scanning lamp tube (8) are further arranged in the camera bellows (1), the first camera fixing rod (5) and the second camera fixing rod (6) are arranged above the lifting rod (4), a first CCD camera (51) is fixed on the first camera fixing rod (5), and a second CCD camera (61) is fixed on the second camera fixing rod (6); the scanning lamp tube (8) positioned right above the lifting rod (4) is arranged in the scanning lamp tube dynamic slide rail (7) and can reciprocate along the scanning lamp tube dynamic slide rail (7); a feeding port (11) and a discharging port (12) are respectively arranged on two side surfaces of the camera bellows (1), and the first conveyor belt (2) and the second conveyor belt (3) are arranged between the feeding port (11) and the discharging port (12);
the PLC system controls the lifting rod (4), the first CCD camera (5), the second CCD camera (6) and the scanning lamp tube (8); the first CCD camera (51) and the second CCD camera (61) are connected with an image acquisition card, and the image acquisition card transmits the acquired image of the lens to be detected to image analysis software on a computer;
in the material conveying direction from a feeding port (11) to a discharging port (12), the same lens to be detected is simultaneously placed on a first conveying belt (2) and a second conveying belt (3), and the first conveying belt (2) and the second conveying belt (3) which are driven by a stepping motor to synchronously and circularly move are conveyed together;
the lifting rod (4) is a hollow tube, and a universal vacuum chuck is fixed at the top of the lifting rod (4); the lifting rod (4) horizontally fixes the lens (9) to be detected on the universal vacuum chuck through negative pressure adsorption, vertically lifts the lens (9) to be detected upwards, and can drive the lens (9) to be detected to make an inclined motion; when the lens to be detected is not adsorbed and fixed, the sucking disc surface of the universal vacuum sucking disc is kept horizontal and is flush with the conveying surfaces of the first conveying belt (2) and the second conveying belt (3) which are positioned at the upper ends;
the first CCD camera (51) and the second CCD camera (61) are respectively arranged at two sides of the reciprocating direction of the scanning lamp tube (8) and are positioned on the central line of the lens (9) to be detected; after the horizontally fixed lens (9) to be inspected is vertically and upwards supported by the supporting and lifting rod (4), the lens of the first CCD camera (51) and the lens of the second CCD camera (61) respectively face the lens (9) to be inspected and respectively form included angles with the lens (9) to be inspected; when the universal vacuum chuck at the top of the lifting rod (4) drives the lens (9) to be inspected to do tilting motion, any one of the lens of the first CCD camera (51) and the lens of the second CCD camera (61) can be perpendicular to the lens (9) to be inspected.
2. The dynamic detection device according to claim 1, wherein after the lifting rod (4) lifts the horizontally fixed lens (9) to be inspected vertically upwards, the lens of the first CCD camera (51) and the lens of the second CCD camera (61) intersect at the center of the lens (9) to be inspected and respectively form an included angle with the lens (9) to be inspected; when the universal vacuum chuck at the top of the lifting rod (4) drives the lens (9) to be inspected to do tilting motion, any one of the lens of the first CCD camera (51) and the lens of the second CCD camera (61) can be perpendicular to the center of the lens (9) to be inspected.
3. The dynamic detection device according to claim 2, wherein after the lifting rod (4) lifts the horizontally fixed lens (9) to be inspected vertically upwards, the lens of the first CCD camera (51) and the lens of the second CCD camera (61) form an angle of 45 degrees with the lens (9) to be inspected, respectively.
4. The dynamic detection device according to claim 1, wherein the reciprocating direction of the scanning lamp (8) is parallel to the material conveying direction, and the first CCD camera (51) and the second CCD camera (61) are symmetrically distributed on the central line of the lens (9) to be detected.
5. The dynamic detection device according to claim 1, wherein the scanning lamp dynamic sliding rail (7) is arranged right above the lifting rod (4), the scanning lamp dynamic sliding rail (7) comprises a first elliptical ring (71) and a second elliptical ring (72), the first elliptical ring (71) and the second elliptical ring (72) are respectively vertically fixed on the inner wall of the top surface of the black box (1) along the material conveying direction, and the first elliptical ring (71) and the second elliptical ring (72) are horizontally spaced and flush in end surface; two ends of the scanning lamp tube (8) are respectively fixed in the first elliptical ring (71) and the second elliptical ring (72) and slide along the first elliptical ring (71) and the second elliptical ring (72) in a reciprocating manner.
6. A dynamic sensing method for use in the dynamic sensing device of any one of claims 1-5, the method comprising the steps of:
i, two conveyor belts which are arranged at intervals in the material conveying direction synchronously and circularly move under the driving of a stepping motor, when a lens (9) to be detected which is placed on the two conveyor belts moves to the position above a lifting rod (4) along with the two conveyor belts, a feeding port (11) and a discharging port (12) of a camera bellows (1) are closed, and a PLC system controls the lifting rod (4), a scanning light source (8), a first CCD camera (51) and a second CCD camera (61) to simultaneously start working;
II, under the control of a PLC system, a lifting rod (4) vertically and upwards supports a lens (9) to be detected, the top end of which is horizontally fixed, a scanning light source (8) starts to emit light and performs reciprocating scanning along the whole course of a dynamic slide rail (7) of a scanning lamp tube, when the scanning light sources (8) for starting to emit light are respectively positioned at the two end parts of the scanning lamp tube dynamic slide rail (7), the lens (9) to be inspected makes a tilting motion once respectively until the lens (9) to be inspected is vertical to the lens of the first CCD camera (51) or the lens of the second CCD camera (61) which is positioned at the same side with the end part where the scanning light source (8) is positioned, a first CCD camera (51) or a second CCD camera (61) which is vertical to the lens (9) to be detected collects images of the lens to be detected at different positions, different reflection angles and different highlight areas, and the images are transmitted to image analysis software on a computer through an image acquisition card;
when the scanning light source (8) does reciprocating movement scanning along the whole course of the scanning lamp tube dynamic slide rail (7), the lens (9) to be detected is kept in a horizontal placement state; the first CCD camera (51) and the second CCD camera (61) are respectively arranged at two sides of the reciprocating movement direction of the scanning light source (8), are positioned at the central line of the lens (9) to be detected and form an included angle with the lens (9) to be detected, respectively collect images of the lens to be detected at different positions, different reflection angles and different highlight areas in real time, and transmit the images to image analysis software on a computer through an image acquisition card;
image analysis software sequentially adopts a visual algorithm of a Gauss low-pass filtering algorithm, a Laplace operator, binarization and Hough transform to perform noise reduction, filtering, binarization and Hough transform processing, and rapidly, accurately and stably extracts scratches of the lens (9) to be detected;
when the lens to be detected is scratched, the lens to be detected (9) is delivered to one of the first conveyor belt (2) and the second conveyor belt (3) and is delivered out of the discharge port (12); when the lens to be inspected is not scratched, the lens to be inspected (9) is delivered to the other of the first conveyor belt (2) and the second conveyor belt (3) and is delivered from the discharge port (12).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810455914.5A CN108918543B (en) | 2018-05-14 | 2018-05-14 | Dynamic detection device and method for mirror surface scratch |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810455914.5A CN108918543B (en) | 2018-05-14 | 2018-05-14 | Dynamic detection device and method for mirror surface scratch |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108918543A CN108918543A (en) | 2018-11-30 |
| CN108918543B true CN108918543B (en) | 2021-03-23 |
Family
ID=64402448
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810455914.5A Active CN108918543B (en) | 2018-05-14 | 2018-05-14 | Dynamic detection device and method for mirror surface scratch |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108918543B (en) |
Families Citing this family (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110044919A (en) * | 2019-04-23 | 2019-07-23 | 上海工程技术大学 | A kind of detection device and its detection method for mirror surfaces scratch |
| CN110177268A (en) * | 2019-05-31 | 2019-08-27 | 苏州五方光电材料有限公司 | A kind of camera eyeglass characterization processes |
| CN110238080A (en) * | 2019-06-19 | 2019-09-17 | 罗博特科智能科技股份有限公司 | Mobile phone screen defect detecting system based on ultrasonic atomization |
| CN110487821A (en) * | 2019-08-30 | 2019-11-22 | 太原科技大学 | A kind of clear sheet defect detecting device and its detection method based on machine vision |
| CN110779474B (en) * | 2019-10-16 | 2021-06-01 | 徐州市铜山区嘉量计量检测创新中心 | Workpiece abrasion detection multipoint control positioning closed type light beam detection equipment |
| CN110779920A (en) * | 2019-10-30 | 2020-02-11 | 中国计量大学 | Pearl detection device |
| KR102147470B1 (en) * | 2020-02-20 | 2020-08-24 | 정규성 | Vision inspection device |
| CN111156931A (en) * | 2020-02-25 | 2020-05-15 | 成都威博恩科技有限公司 | High-precision flatness detection device and method |
| CN112093477A (en) * | 2020-06-16 | 2020-12-18 | 丹阳市精通眼镜技术创新服务中心有限公司 | Automatic feeding device for optical lenses |
| CN111650212B (en) * | 2020-07-03 | 2021-10-26 | 东北大学 | Metal surface normal direction three-dimensional information acquisition method based on linear array camera three-dimensional vision |
| CN112258448A (en) * | 2020-09-15 | 2021-01-22 | 郑州金惠计算机系统工程有限公司 | Fine scratch detection method, fine scratch detection device, electronic equipment and computer-readable storage medium |
| CN112557400B (en) * | 2020-11-30 | 2021-10-22 | 电子科技大学 | System and method for detecting surface defect contour of lens of satellite telescope |
| CN113125449B (en) * | 2021-04-20 | 2022-10-18 | 江苏善果缘智能科技有限公司 | Scanning device for detecting surface of integrated product and assembling method thereof |
| CN114113131B (en) * | 2021-12-03 | 2024-04-30 | 中科计算技术西部研究院 | Device and method for detecting scratch defects of reflecting lens |
| CN114055764B (en) * | 2022-01-17 | 2022-04-29 | 深圳市利和兴股份有限公司 | Pre-detection equipment for OLED flexible screen laminating |
| CN115616003B (en) * | 2022-10-17 | 2024-06-11 | 四川长虹电器股份有限公司 | Fresnel optical film inspection device and inspection method thereof |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004294088A (en) * | 2003-03-25 | 2004-10-21 | Dainippon Screen Mfg Co Ltd | Image acquiring device and image acquisition method |
| CN202735263U (en) * | 2012-08-23 | 2013-02-13 | 深圳市纳研科技有限公司 | Device for detecting minor defects of array graphs by machine vision |
| CN204924983U (en) * | 2015-07-20 | 2015-12-30 | 丹阳市精通眼镜技术创新服务中心有限公司 | Lens defect automatic checkout device of grating modulation |
| CN205844213U (en) * | 2016-06-28 | 2016-12-28 | 华南理工大学 | A kind of curved surface automatic defect detecting device |
| CN106353336A (en) * | 2016-10-27 | 2017-01-25 | 哈尔滨理工大学 | Lens coating automatic detection system |
| CN207280968U (en) * | 2017-08-02 | 2018-04-27 | 深圳市迪姆自动化有限公司 | Full-automatic glass surface blemish detection machine |
-
2018
- 2018-05-14 CN CN201810455914.5A patent/CN108918543B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004294088A (en) * | 2003-03-25 | 2004-10-21 | Dainippon Screen Mfg Co Ltd | Image acquiring device and image acquisition method |
| CN202735263U (en) * | 2012-08-23 | 2013-02-13 | 深圳市纳研科技有限公司 | Device for detecting minor defects of array graphs by machine vision |
| CN204924983U (en) * | 2015-07-20 | 2015-12-30 | 丹阳市精通眼镜技术创新服务中心有限公司 | Lens defect automatic checkout device of grating modulation |
| CN205844213U (en) * | 2016-06-28 | 2016-12-28 | 华南理工大学 | A kind of curved surface automatic defect detecting device |
| CN106353336A (en) * | 2016-10-27 | 2017-01-25 | 哈尔滨理工大学 | Lens coating automatic detection system |
| CN207280968U (en) * | 2017-08-02 | 2018-04-27 | 深圳市迪姆自动化有限公司 | Full-automatic glass surface blemish detection machine |
Non-Patent Citations (2)
| Title |
|---|
| 机器视觉在钢化玻璃缺陷检测中的应用研究;杨杰 等;《计算机技术与发展》;20130331;第23卷(第3期);第211-214页 * |
| 机器视觉技术在玻璃基板缺陷检测设备中的应用;李青 等;《玻璃与搪瓷》;20160630;第44卷(第3期);第26-31页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108918543A (en) | 2018-11-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108918543B (en) | Dynamic detection device and method for mirror surface scratch | |
| CN106670127B (en) | Full-automatic visual inspection system of screen defect | |
| CN107449782B (en) | Digital quantitative detection system for side surface defects of mobile phone shell | |
| CN108230324B (en) | Visual detection method for microdefect on surface of magnetic shoe | |
| CN107389701A (en) | A kind of PCB visual defects automatic checkout system and method based on image | |
| CN107290347B (en) | Automatic honeycomb carrier defect detection method | |
| CN208207914U (en) | PCB surface defect detecting device based on fast geometric alignment | |
| CN113504238B (en) | Glass surface defect acquisition device and detection method | |
| Liu et al. | Automatic detection technology of surface defects on plastic products based on machine vision | |
| CN110976338B (en) | Test paper sorting system and method based on machine vision | |
| CN111524101A (en) | Electronic screen defect detection method based on machine vision technology | |
| CN109406527B (en) | System and method for detecting fine appearance defects of micro camera module lens | |
| CN111474179A (en) | Lens surface cleanliness detection device and method | |
| CN102141379A (en) | Novel glass bottle and jar detection device | |
| CN104483320A (en) | Digitized defect detection device and detection method of industrial denitration catalyst | |
| CN112083002A (en) | Capacitance appearance detection device and method based on artificial intelligence technology | |
| CN111583193B (en) | Pistachio nut framework extraction device based on geometric contour template matching and algorithm thereof | |
| CN110648301A (en) | Device and method for eliminating imaging reflection | |
| CN110567968A (en) | part defect detection method and device | |
| CN107742119B (en) | Object contour extraction and matching device and method based on back-image imaging | |
| CN110044921A (en) | Lithium battery open defect detection system and method | |
| CN109387524A (en) | Thread defect detection method and device based on linearly polarized photon | |
| CN204422435U (en) | Camera module optical filter gluing pick-up unit | |
| CN115112682A (en) | Visual mobile phone curved glass detection device and method | |
| CN111330874A (en) | Detection device and detection method for pollution or impurity defects of bottom area of medicine bottle |
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 |