WO2019014937A1 - 用于汽车生产线上激光焊保护镜片的缺陷检测方法和装置 - Google Patents

用于汽车生产线上激光焊保护镜片的缺陷检测方法和装置 Download PDF

Info

Publication number
WO2019014937A1
WO2019014937A1 PCT/CN2017/093915 CN2017093915W WO2019014937A1 WO 2019014937 A1 WO2019014937 A1 WO 2019014937A1 CN 2017093915 W CN2017093915 W CN 2017093915W WO 2019014937 A1 WO2019014937 A1 WO 2019014937A1
Authority
WO
WIPO (PCT)
Prior art keywords
mirror
image
lens
dead
laser welding
Prior art date
Application number
PCT/CN2017/093915
Other languages
English (en)
French (fr)
Inventor
吕猛
郭孟宇
郭磊
吴晓鹏
陈仙勇
Original Assignee
易思维 (天津) 科技有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 易思维 (天津) 科技有限公司 filed Critical 易思维 (天津) 科技有限公司
Priority to PCT/CN2017/093915 priority Critical patent/WO2019014937A1/zh
Priority to US16/073,241 priority patent/US10794838B2/en
Priority to CN201710867489.6A priority patent/CN107643296B/zh
Publication of WO2019014937A1 publication Critical patent/WO2019014937A1/zh

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
    • G01N21/958—Inspecting transparent materials or objects, e.g. windscreens
    • 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
    • G01N21/94—Investigating contamination, e.g. dust
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
    • B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
    • B23K26/00—Working by laser beam, e.g. welding, cutting or boring
    • B23K26/70—Auxiliary operations or equipment
    • B23K26/702—Auxiliary equipment
    • B23K26/705—Beam measuring device
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
    • G01N21/47—Scattering, i.e. diffuse reflection
    • 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
    • G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
    • G01N21/55—Specular reflectivity
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/01—Arrangements or apparatus for facilitating the optical investigation
    • G01N21/15—Preventing contamination of the components of the optical system or obstruction of the light path
    • G01N2021/151—Gas blown
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G01N21/88—Investigating the presence of flaws or contamination
    • G01N21/95—Investigating the presence of flaws or contamination characterised by the material or shape of the object to be examined
    • G01N21/958—Inspecting transparent materials or objects, e.g. windscreens
    • G01N2021/9583—Lenses

Definitions

  • the present application relates to the field of lasers, and in particular to a method and apparatus for detecting defects in laser welding protection lenses for use in automobile production lines.
  • the body quality of welding involves the appearance, quality and ease of assembly of the vehicle body. Therefore, the welding production process is in four major processes. From the role of the link up and down. Compared with traditional welding technology, laser welding has unparalleled advantages in welding precision, efficiency, reliability and automation. In recent years, with the continuous development of high-power, high-performance laser processing equipment, laser technology has been rapidly developed in the automotive, energy, electronics and other industrial fields in Japan, the United States, Germany and other developed countries. Laser welding is considered to be the most in the 21st century. One of the promising manufacturing technologies. As the demand for modern automobile manufacturing increases, the dependence on laser welding is also increasing.
  • Fig. 1 is an internal optical path diagram of a laser welding head according to the related art.
  • the high temperature during laser welding can cause "spark" splash, and the spatter will randomly adhere to the protective lens in the laser welding head, causing the optical path of the laser beam to be blocked, resulting in a decrease in the quality of the welding.
  • workers in the downstream of the production line perform manual inspections. By manually inspecting the laser welding head to protect the lenses, the welding accidents caused by the protection of the lens defects are prevented.
  • the related technology also provides a solution, that is, a little automobile manufacturer does not hesitate to spend the cost in order to reduce the repair rate of the body-in-white, and the protective lens is periodically replaced regardless of whether the laser welding head protection mirror is defective, but this still brings many problems. : 1) The cost of periodically replacing the laser welding head to protect the lens is higher than the manual inspection. 2) The internal optical path of the laser welding head is in a vacuum state, and frequent replacement of the unboxing and replacement of the protective lens may cause impurities such as dust to enter, thereby affecting the conductivity of the laser in the internal optical path.
  • the main purpose of the present application is to provide a defect detecting method and device for laser welding protective lens on an automobile production line, so as to solve the problem that the protective lens is not contaminated in time during laser welding.
  • a defect detecting apparatus for a laser welding protective lens for an automobile production line comprising: a coaxial light source configured to emit collimated detecting light, the detecting light a direction perpendicular to the preset horizontal direction; a half mirror disposed above the coaxial light source, configured to reflect the probe light to a preset position; and a mirror to receive the half mirror reflection And reflecting the light reflected by the half mirror to the protective lens; the industrial camera, and the half mirror and the mirror are horizontally disposed in the predetermined horizontal direction, and are arranged to receive the incident Light, obtaining a detected image, wherein the incident light is incident light formed after the protective lens reflects the probe light and passes through the mirror and the half mirror; the processor is set according to the The detected image is calculated to determine whether there is a dead spot on the protective lens.
  • the apparatus further includes: an air blowing device that blows air to the lens of the industrial camera during laser welding, and is configured to remove dust on the industrial camera lens.
  • the apparatus further includes: a black light absorbing cotton located above the semi-transparent mirror and the industrial camera, configured to absorb incident light of the coaxial light source transmitted by the half mirror.
  • a defect detecting method for a laser welding protective lens for an automobile production line comprising: acquiring a detection image, wherein the detection image is claim 1 a detection image obtained by an industrial camera of a defect detecting device for laser welding protection lens on an automobile production line; determining a circumscribed rectangle of a mirror region in the detected image by feature matching; performing a threshold setting in the circumscribed rectangle The point contour is retrieved to obtain the dead point area and the number of dead pixels.
  • the method further includes: determining whether an image corresponding to the mirror region exists in the detected image; wherein, if it is determined that the existence exists The image corresponding to the mirror area determines the circumscribed rectangle of the mirror area in the detected image by feature matching; if it is determined that the image corresponding to the mirror area does not exist, the detected image is reacquired.
  • performing dead point contour retrieval by threshold setting in the circumscribed rectangle includes: calculating a dead point area and converting the bad point area into a number of pixels; obtaining the dead point area and the dead pixel After the number, the method further includes: determining whether the number of dead pixels converted into the bad point area is greater than or equal to the number of user-defined dead pixels, if the number of dead pixels converted into the bad point area is greater than or equal to If the user customizes the number of dead pixels, the bad point contour is marked, and it is determined whether the number of the bad point contours is greater than or equal to the user-defined bad points.
  • the protection lens is abnormal; if the bad point area is converted into a number of bad pixels is smaller than the user-defined dead point The number of pixels indicates that the protective lens is normal; if the number of the bad point contours is less than the user-defined dead point value, the protective lens is prompted to be normal.
  • the method further includes: receiving an instruction sent by the robot; determining an identifier value of the instruction, wherein the identifier value is two; wherein, the identifier value is The detection image is acquired at one time, and when the identification value is the second type, the lens of the industrial camera is blown by the air blowing device.
  • a defect detecting device provided as a laser welding protective lens on an automobile production line, the device comprising: an acquiring unit configured to acquire a detection image, wherein the detecting The image is the detection image obtained by the industrial camera of the defect detecting device of the laser welding protection lens on the automobile production line of the present application; the determining unit is configured to determine the circumscribed rectangle of the mirror area in the detection image by feature matching; the retrieval unit, It is set to perform a dead point contour search by threshold setting in the circumscribed rectangle, and obtain a bad point area and a bad point pixel number.
  • the device further includes: a determining unit, configured to determine whether an image corresponding to the mirror area exists in the detected image before determining a circumscribed rectangle of the mirror area in the detected image by feature matching; If it is determined that the image corresponding to the mirror region exists, the circumscribed rectangle of the mirror region in the detected image is determined by feature matching; if it is determined that the image corresponding to the mirror region does not exist, the detected image is reacquired.
  • a determining unit configured to determine whether an image corresponding to the mirror area exists in the detected image before determining a circumscribed rectangle of the mirror area in the detected image by feature matching; If it is determined that the image corresponding to the mirror region exists, the circumscribed rectangle of the mirror region in the detected image is determined by feature matching; if it is determined that the image corresponding to the mirror region does not exist, the detected image is reacquired.
  • the application uses a coaxial light source for emitting collimated detection light, and the direction of the detection light is perpendicular to a preset horizontal direction; a semi-transparent mirror is disposed above the coaxial light source for reflecting the detection light to a preset position a mirror that receives the light reflected by the half mirror and reflects the light reflected by the half mirror to the protective lens; the industrial camera, and the half mirror and the mirror are horizontally oriented in a predetermined horizontal direction, Receiving the incident light to obtain a detection image, wherein the incident light is incident light formed by the protective lens after being reflected by the detecting light and passing through the mirror and the half mirror; the processor is configured to perform calculation according to the detected image to determine the protective lens Whether there is a dead point on the surface solves the problem that the protective lens is not contaminated in time when laser welding is obtained, thereby achieving the effect of timely knowing the bad point on the protective lens in order to replace the protective lens.
  • 1 is an internal optical path diagram of a laser welding head according to the related art
  • FIG. 2 is a schematic view of a defect detecting device for a laser welding protective lens for an automobile production line according to a first embodiment of the present application;
  • FIG. 3 is a schematic diagram showing the principle of total reflection and diffuse reflection of a protective mirror according to an embodiment of the present application
  • FIG. 4 is a schematic diagram of an optical path principle according to an embodiment of the present application.
  • FIG. 5 is a schematic cross-sectional view of a defect detecting device for a laser welding protective lens for an automobile production line according to an embodiment of the present application;
  • FIG. 6 is a schematic view showing a position where a light absorbing cotton is disposed according to an embodiment of the present application
  • FIG. 7 is a flow chart of a defect detecting method for a laser welding protective lens for an automobile production line according to a first embodiment of the present application
  • FIG. 8 is a flow chart of a defect detecting method for a laser welding protective lens for an automobile production line according to a second embodiment of the present application.
  • FIG. 9 is a schematic diagram showing the hardware configuration of a defect detecting device for a laser welding protective lens on an automobile production line according to an embodiment of the present application.
  • FIG. 10 is a flow diagram of hardware communication in accordance with an embodiment of the present application.
  • FIG. 11 is a flow chart of a blowing interaction process between a robot and a sensor according to an embodiment of the present application
  • FIG. 12 is a flowchart of a robot control sensor detecting an image according to an embodiment of the present application
  • FIG. 13 is a schematic diagram of an interface of embedded human-computer interaction software according to an embodiment of the present application.
  • FIG. 14 is a schematic diagram of a defect detecting device for a laser welding protective lens for an automobile production line according to an embodiment of the present application.
  • the embodiment of the present application provides a defect detecting device for laser welding protection lens on an automobile production line.
  • FIG. 2 is a schematic diagram of a defect detecting device for a laser welding protective lens for an automobile production line according to a first embodiment of the present application. As shown in FIG. 2, the device includes:
  • a coaxial light source 10 for emitting collimated detection light, the direction of the detection light being perpendicular to a preset horizontal direction;
  • a half mirror 20 disposed above the coaxial light source 10 for reflecting the probe light to a preset position
  • the mirror 30 receives the light reflected by the half mirror 20 and reflects the light reflected by the half mirror 20 to the protective lens;
  • An industrial camera 40 horizontally facing the half mirror 20 and the mirror in the predetermined horizontal direction, for receiving incident light to obtain a detected image, wherein the incident light is the protective lens pair
  • the incident light formed after the probe light is reflected and passed through the mirror and the half mirror;
  • the processor 50 is configured to perform calculation according to the detected image to determine whether there is a dead pixel on the protective lens.
  • FIG. 2 is a schematic diagram showing the principle of total reflection and diffuse reflection of a protective mirror according to an embodiment of the present application. As shown in FIG.
  • the protective lens is free from defects, that is, a smooth flat surface, total reflection will occur on the protective mirror surface. If the protective mirror is defective, diffuse reflection occurs on the protective mirror surface. If the protective mirror is a smooth mirror, the coaxial light will totally reflect its surface, and the light will return according to the original light path; if the protective mirror has a dead pixel, it will appear as a concave and convex mirror surface, and the coaxial light will be diffusely reflected here, and the light cannot be as original. The light path is all returned. When imaging at the camera, only a portion of the light entering the protective lens defect is dark.
  • This embodiment adopts a coaxial light source 10 for emitting collimated detecting light, the direction of the detecting light is perpendicular to a preset horizontal direction; a half mirror 20 is disposed above the coaxial light source 10, Reflecting the probe light to a preset position; the mirror 30 receives the light reflected by the half mirror 20 and reflects the light reflected by the half mirror 20 to the protective lens; the industrial camera 40 And the semi-transparent mirror 20 and the mirror are horizontal in the predetermined horizontal direction for receiving incident light to obtain a detection image, wherein the incident light is the protective lens pair The incident light formed after the light is reflected by the mirror and the half mirror; the processor 50 is configured to perform calculation according to the detected image, determine whether there is a dead point on the protective lens, and solve the laser
  • the problem that the protective lens is contaminated during welding cannot be known in time, thereby achieving the effect of timely knowing the dead spots on the protective lens in order to replace the protective lens.
  • the defect detecting device for laser welding protection lens for automobile production line by the embodiment of the present application can be used for visual mirror surface defect detection of laser welding protection lens on automobile production line, and can solve the defect of automatically detecting laser welding head protection lens in laser welding operation
  • the problem can be applied to all laser welding heads with protective lenses on the automobile production line, which ensures timely warning of defects in the laser welding head protection lens during welding, and greatly reduces the return rate of the body-in-white return line. .
  • the defect detecting device for laser welding protection lens on the automobile production line further comprises an air blowing device for blowing air to the lens of the industrial camera during the laser welding process, and the dust on the industrial camera lens can be clearly understood .
  • an air blowing device for blowing air to the camera lens, the error caused by the sputtering of the magazine particles into the camera lens during the welding process can be reduced, and the detection result can be more accurate.
  • 5 is a schematic cross-sectional view of a defect detecting device for a laser welding protective lens on an automobile production line according to an embodiment of the present application. As shown in FIG. 5, when the laser welding head is operated online, the device will turn on the blowing mode, and utilize The gas source of the factory is blown and dusted at the exit of the optical path of the wedge-shaped box. After the laser welding head completes the operation, the defect detection is directly performed, and the time cost can be saved by setting the air blowing device.
  • the defect detecting device for the laser welding protection lens of the automobile production line of the embodiment of the present application further includes black light absorbing cotton
  • FIG. 6 is a schematic diagram of the position of the light absorbing cotton according to the embodiment of the present application, as shown in FIG.
  • the cotton is located above the semi-transparent mirror and the industrial camera for absorbing the incident light of the coaxial light source transmitted by the transflectoscope; the black light absorbing cotton can prevent the original light of the coaxial light source from passing through the transflective for the first time.
  • the transmitted light produced by the mirror has an effect on the imaging of industrial cameras.
  • Special black light-absorbing cotton can be installed above the half mirror to reduce the transmission interference of the light source and improve the accuracy of the test results.
  • FIG. 7 is a flowchart of a defect detecting method for a laser welding protective lens for an automobile production line according to the first embodiment of the present application. As shown in FIG. 7, the method includes the following steps:
  • Step S102 acquiring a detection image, wherein the detection image is used for an inspection image obtained by an industrial camera of a defect detection device for laser welding protection lens on an automobile production line;
  • Step S104 determining, by feature matching, a circumscribed rectangle of the mirror area in the detected image
  • step S106 the bad point contour search is performed by the threshold setting in the circumscribed rectangle, and the dead point area and the number of dead pixels are obtained.
  • the detected image is the detected image in the industrial camera in the defect detecting device for the laser welding protective lens on the automobile production line in the above embodiment, and is detected by the feature matching method after the detected image is acquired.
  • the circumscribed rectangle of the mirrored circular area in the image if it is detected that the mirrored circular area of the detected image is incomplete, re-detected, if it can locate the circular mirror area in the detected image, determine the circumscribed rectangle of the circular mirror area Then, in the circumscribed rectangle, the dead point contour retrieval is performed on the region by the threshold setting, and the dead point area and the number of dead pixels in the circumscribed rectangle are obtained.
  • FIG. 8 is a flow chart of a defect detecting method for a laser welding protective lens for an automobile production line according to a second embodiment of the present application, which embodiment can be used as a preferred embodiment of the first embodiment, as shown in FIG.
  • the method includes the following steps:
  • the lens inside the image should be bright white. If found, the feature matching method is used to detect the mirror area, and it is judged whether or not the positioning area is out of bounds. If the out of bounds occurs, the lens is not present, and if the boundary is not crossed, the circumscribed rectangle of the positioning area is taken. Threshold setting is performed in the circumscribed rectangle to retrieve the dead point contour.
  • the final bad point judgment result is obtained, if the dead point area is larger than the user-defined dead pixel value (that is, the preset Value), mark the bad point contour and determine whether the number of bad points is greater than or equal to the user-defined pixel value. If the judgment result is yes, it indicates that the protection lens is abnormal (GlassIsBad). If the judgment result is no, the protection lens is normal. (GlassIsGood).
  • the detection image is reacquired.
  • performing a defect point contour search by using a threshold setting in the circumscribed rectangle includes: calculating a bad point area and converting the bad point area into a number of pixels; and determining the bad point area after obtaining the bad point area and the number of bad points Is it greater than It is equal to the number of user-defined dead pixels. If the bad point area is greater than or equal to the user-defined dead pixel number, mark the bad point contour and judge whether the number of bad points is greater than or equal to the user-defined bad point value. If the bad point is judged. If the number of pixels is greater than or equal to the user-defined bad point value, the protection lens is abnormal.
  • the protective lens In determining whether the area of the dead point is greater than or equal to the number of pixels of the user-defined dead pixel, if the area of the dead point is smaller than the number of pixels of the user-defined dead pixel, the protective lens is normal, and it is determined whether the number of dead pixels is greater than or equal to the user-defined dead point. After the value, if the number of dead pixels is less than the user-defined bad point value, the protective lens is normal.
  • receiving an instruction sent by the robot before receiving the detection image, receiving an instruction sent by the robot; determining an identification value of the instruction, where the identification value includes two types; wherein, when the flag value is the first type, acquiring the detection image, where When the value of the flag is the second type, the lens of the industrial camera is blown by the air blowing device to achieve the effect of dust removal.
  • the embodiment of the present application further provides a hardware structure of a defect detecting device for laser welding protection lens on an automobile production line
  • FIG. 9 is a defect detecting device for laser welding protection lens on an automobile production line according to an embodiment of the present application.
  • a schematic diagram of a hardware structure is shown in FIG. 9.
  • the hardware platform of the embodiment of the present application uses an ARM chip as a main core, and other types of processors are also within the scope of protection of the present application.
  • the sensor of the embodiment of the present application can directly communicate with a robot or a PLC, is compatible with the current mainstream industrial communication protocol, and is connected with a series of peripherals such as a touch screen and CAMERA to implement core functions.
  • the external storage medium of the application can be an SD card.
  • FIG. 10 is a flow diagram of hardware communication in accordance with an embodiment of the present application. As shown in Figure 10, the entire process includes two branch processes:
  • the selective opening of the inspection process and the blowing process is completely done by the robot sending instructions.
  • the detection process branch is entered; when the flag Tmp is 2, the blowing process branch is entered.
  • the dust of the sensor lens will affect the test results, and it is necessary to perform dust removal work at each measurement cycle.
  • the dust removal method uses a gas blow method, and the sensor needs to be introduced into the gas source of the factory.
  • the sensor control solenoid valve performs air blown dust removal before each measurement. When Tmp is not 2, the error is indicated (Default).
  • the device comes with a dust removal function.
  • the blowing process is carried out simultaneously with the laser welding, in order to prevent the welding slag from splashing onto the camera of the detecting device during the laser welding process, causing a bad point misjudgment.
  • the blowing process stops.
  • FIG. 11 is a flow chart of a blowing interaction process between a robot and a sensor according to an embodiment of the present application.
  • the robot sends an instruction to the sensor, and the blowing process is performed synchronously with the laser welding, and the sensor will be after the acquisition is completed.
  • the result is sent to the robot, and the robot receives the result of the blow completion at the same time, and the robot sends the sensor to the sensor.
  • the clear instruction “0” is sent, and after the sensor receives the clear instruction, the stored result is cleared.
  • FIG. 12 is a flow chart of a robot control sensor detecting an image according to an embodiment of the present application.
  • a coaxial light source is used to illuminate a lens
  • a camera collects a measurement image
  • an embedded image processing algorithm is combined with a customized image analysis algorithm. Detect whether there is a dead pixel on the lens, display the measurement image and the detection result (with or without dead pixels, dead point position and size) on the LED display of the sensor, and feed back the detection result to the robot.
  • the inspection process begins each time the laser welding head is required to reach the detection position.
  • FIG. 13 is a schematic diagram of an interface of embedded human-computer interaction software according to an embodiment of the present application.
  • the human-computer interaction function of the device is completed by using a touch screen and embedded software.
  • the image processing algorithm is accurately simulated to find the bad points on the lens in real time.
  • the user can observe the dead point detection situation in real time through the user interface, define the area and number of dead pixels, browse the bad point image, and so on.
  • the defect detecting method for the laser welding protection lens of the automobile production line in the embodiment of the present application can perform the mirror surface defect detection of the laser welding head protection lens for all the automobile production lines adopting the laser welding.
  • the static defect detection is performed on the protective lens of the laser welding head in the operation, and the detection result is given in the first time and the feedback is given to the upper robot.
  • the method has the advantages of simple structure, low algorithm complexity and small workload of the robot track debugging.
  • the calibration process is simple and convenient, and takes less time.
  • the embodiment of the present application provides a defect detecting device for a laser welding protective lens on an automobile production line, and the device can be used for performing a defect detecting method for a laser welding protective lens on an automobile production line according to an embodiment of the present application.
  • FIG. 14 is a schematic diagram of a defect detecting device for a laser welding protective lens for an automobile production line according to an embodiment of the present application. As shown in FIG. 14, the device includes:
  • the acquiring unit 110 is configured to acquire a detection image, wherein the detection image is a detection image obtained by an industrial camera of a defect detecting device for laser welding protection lens on an automobile production line of the embodiment of the present application;
  • a determining unit 120 configured to determine, by feature matching, a circumscribed rectangle of the mirror area in the detected image
  • the searching unit 130 is configured to perform a dead-point contour search by threshold setting in the circumscribed rectangle to obtain a bad point area and a bad point pixel number.
  • the device further includes: a determining unit, configured to determine, before the circumscribed rectangle of the mirror region in the detected image by the feature matching, whether there is an image corresponding to the mirror region in the detected image; wherein, if the mirror region is determined to exist Corresponding image, determining the external moment of the mirror area in the detected image by feature matching If it is determined that there is no image corresponding to the mirror area, the detected image is reacquired.
  • a determining unit configured to determine, before the circumscribed rectangle of the mirror region in the detected image by the feature matching, whether there is an image corresponding to the mirror region in the detected image; wherein, if the mirror region is determined to exist Corresponding image, determining the external moment of the mirror area in the detected image by feature matching If it is determined that there is no image corresponding to the mirror area, the detected image is reacquired.
  • the defect detecting device for the laser welding protective lens for the automobile production line of the embodiment of the present application solves the problem that the protective lens is not contaminated in time when the laser welding is obtained, thereby achieving the timely recognition of the dead point on the protective lens to replace the protective lens. Effect.
  • embodiments of the present application can be provided as a method, system, or computer program product.
  • the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment in combination of software and hardware.
  • embodiments of the present application can be provided as a method, system, or computer program product.
  • the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment in combination of software and hardware.
  • the application emits collimated detection light through a coaxial light source, and the direction of the detection light is perpendicular to a preset horizontal direction; the semi-transparent mirror is disposed above the coaxial light source for reflecting the detection light to a preset position; the mirror Receiving light reflected by the half mirror and reflecting the light reflected by the half mirror to the protective lens; the industrial camera, and the half mirror and the mirror are horizontally arranged in a preset horizontal direction for receiving the incident Light, obtaining a detection image, wherein the incident light is incident light formed by the protective lens after reflecting the detection light and passing through the mirror and the half mirror;
  • the device is used for calculating the image according to the detected image to determine whether there is a dead point on the protective lens, and solves the problem that the protective lens is not contaminated in time when the laser welding is performed, thereby achieving the timely recognition of the dead point on the protective lens in order to replace the protective lens. effect.

Landscapes

  • Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Chemical & Material Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Biochemistry (AREA)
  • General Health & Medical Sciences (AREA)
  • General Physics & Mathematics (AREA)
  • Immunology (AREA)
  • Pathology (AREA)
  • Engineering & Computer Science (AREA)
  • Optics & Photonics (AREA)
  • Plasma & Fusion (AREA)
  • Mechanical Engineering (AREA)
  • Laser Beam Processing (AREA)

Abstract

一种用于汽车生产线上激光焊保护镜片的缺陷检测装置。该装置包括：同轴光源（10），用于发出准直探测光，探测光的方向与预设的水平方向垂直；半透半反射镜（20），设置在同轴光源（10）上方，用于将探测光反射到预设位置；反射镜（30），接收半透半反射镜（20）反射的光，并将半透半反射镜（20）反射的光反射到保护镜片；工业相机（40），与半透半反射镜（20）和反射镜（30）在预设的水平方向水平，用于接收入射光，得到检测图像，其中，入射光为保护镜片对探测光反射后经反射镜（30）和半透半反射镜（20）之后形成的入射光；处理器（50），用于根据检测图像进行计算，确定保护镜片上是否存在坏点。以及一种采用上述装置对激光焊保护镜片的缺陷进行检测的方法。

Description

用于汽车生产线上激光焊保护镜片的缺陷检测方法和装置 技术领域
本申请涉及激光领域，具体而言，涉及一种用于汽车生产线上激光焊保护镜片的缺陷检测方法和装置。
背景技术
在汽车生产的冲压、焊装、涂装、总装的四大生产工艺中，焊装的车身质量牵涉车身的外观、质量及总装装配的难易程度，因此，焊装生产工艺在四大工艺中起承上启下的作用。激光焊接与传统焊接技术相比，在焊接精度、效率、可靠性、自动化各方面都具有无可比拟的优越性。近年来，随着大功率、高性能激光加工设备的不断研制成功，激光技术在日本、美国、德国等发达国家的汽车、能源、电子等工业领域得到快速发展，激光焊接被认为是21世纪最具有发展前景的制造技术之一。随着现代汽车制造需求量的提升，对激光焊接的依赖程度也越来越高。
然而，图1是根据相关技术的一种激光焊接头内部光路图。激光焊作业时的高温会产生“火花”飞溅，飞溅物会随机性地粘附在激光焊头内的保护镜片上，导致激光束的光路受阻，造成焊接质量的下降。针对此问题，现阶段没有良好的解决案例，通常都是生产线下游的工人来进行人工检测，通过人工检查激光焊接头保护镜片的方法来预防因保护镜片缺陷而差生的焊接事故。但是，这种人工操作存在两个问题：1)由于下游的工人发现焊接问题具有一定的滞后性，生产线上工人往往在白车身经过激光焊接后对其进行整体检查，当发现由于激光焊接未达到标准条件从而导致车身焊接产生坏焊的现象时，流水线上已经有多台白车身的焊接也出现了问题。这将导致大量坏焊白车身重新进行补焊，降低汽车生产效率。这些问题白车身要么重新焊接，要么直接报废，无形中降低了生产效率，提高了制造成本。2)工人一般是无法进入正在作业的汽车焊接现场的，所以进行人工检查往往需要暂停焊接现场作业。这将导致整个汽车生产流水线产生停滞，进而使整个生产系统瘫痪。
相关技术还提供了一种解决方案，就是少许汽车厂商为了减小白车身的返修率，不惜花费成本，无论激光焊接头保护镜是否存在缺陷都定期更换其保护镜片，但是这依然带来很多问题：1)周期性更换激光焊接头保护镜片的成本要比人工检查高。2)激光焊接头内部光路为真空状态，频繁地更换开箱更换保护镜片会使灰尘等杂质进入，进而影响激光在内部光路的传导性。
针对相关技术中激光焊接时保护镜片被污染无法及时获知的问题，目前尚未提出 有效的解决方案。
发明内容
本申请的主要目的在于提供一种用于汽车生产线上激光焊保护镜片的缺陷检测方法和装置，以解决激光焊接时保护镜片被污染无法及时获知的问题。
为了实现上述目的，根据本申请的一个方面，提供了一种用于汽车生产线上激光焊保护镜片的缺陷检测装置，该装置包括：同轴光源，设置为发出准直探测光，所述探测光的方向与预设的水平方向垂直；半透半反射镜，设置在所述同轴光源上方，设置为将所述探测光反射到预设位置；反射镜，接收所述半透半反射镜反射的光，并将所述半透半反射镜反射的光反射到保护镜片；工业相机，与所述半透半反射镜和所述反射镜在所述预设的水平方向水平，设置为接收入射光，得到检测图像，其中，所述入射光为所述保护镜片对所述探测光反射后经所述反射镜和所述半透半反射镜之后形成的入射光；处理器，设置为根据所述检测图像进行计算，确定所述保护镜片上是否存在坏点。
进一步地，所述装置还包括：吹气装置，在激光焊接过程中向所述工业相机的镜头处吹气，设置为清除所述工业相机镜头上的灰尘。
进一步地，所述装置还包括：黑色吸光棉，位于所述半透半反镜与所述工业相机上方，设置为吸收所述半透半反镜透射的所述同轴光源的入射光。
为了实现上述目的，根据本申请的另一方面，还提供了一种用于汽车生产线上激光焊保护镜片的缺陷检测方法，该方法包括：获取检测图像，其中，所述检测图像为权利要求1中用于汽车生产线上激光焊保护镜片的缺陷检测装置的工业相机得到的检测图像；通过特征匹配确定所述检测图像中的镜面区域的外接矩形；在所述外接矩形内通过阈值设定进行坏点轮廓检索，得到坏点面积和坏点像素数。
进一步地，在通过特征匹配确定所述检测图像中的镜面区域的外接矩形之前，所述方法还包括：判断所述检测图像中是否存在所述镜面区域对应的图像；其中，如果判断出存在所述镜面区域对应的图像，则通过特征匹配确定所述检测图像中的镜面区域的外接矩形；如果判断出不存在所述镜面区域对应的图像，则重新获取检测图像。
进一步地，在所述外接矩形内通过阈值设定进行坏点轮廓检索包括：计算坏点面积并将所述坏点面积转化成像素个数；在得到所述坏点面积和所述坏点像素数之后，所述方法还包括：判断所述坏点面积转化成的坏点像素数是否大于等于用户自定义坏点像素数量，如果所述坏点面积转化成的坏点像素数大于等于所述用户自定义坏点像素数量，则标记坏点轮廓，判断所述坏点轮廓的个数是否大于等于用户自定义坏点个 数值，如果判断出所述坏点轮廓的个数大于等于所述用户自定义坏点个数值，则提示保护镜片异常；如果所述坏点面积转化成的坏点像素数小于用户自定义坏点像素数量，则提示所述保护镜片正常；如果所述坏点轮廓的个数小于所述用户自定义坏点个数值，则提示所述保护镜片正常。
进一步地，在获取所述检测图像之前，所述方法还包括：接收机器人发送的指令；判断所述指令的标识数值，其中，所述标识数值有两种；其中，在所述标识数值为第一种时获取所述检测图像，在所述标识数值为第二种时通过吹气装置对所述工业相机的镜头处吹气。
为了实现上述目的，根据本申请的另一方面，还提供了一种设置为汽车生产线上激光焊保护镜片的缺陷检测装置，该装置包括：获取单元，设置为获取检测图像，其中，所述检测图像为本申请的设置为汽车生产线上激光焊保护镜片的缺陷检测装置的工业相机得到的检测图像；确定单元，设置为通过特征匹配确定所述检测图像中的镜面区域的外接矩形；检索单元，设置为在所述外接矩形内通过阈值设定进行坏点轮廓检索，得到坏点面积和坏点像素数。
进一步地，所述装置还包括：判断单元，设置为在通过特征匹配确定所述检测图像中的镜面区域的外接矩形之前，判断所述检测图像中是否存在所述镜面区域对应的图像；其中，如果判断出存在所述镜面区域对应的图像，则通过特征匹配确定所述检测图像中的镜面区域的外接矩形；如果判断出不存在所述镜面区域对应的图像，则重新获取检测图像。
本申请通过同轴光源，用于发出准直探测光，探测光的方向与预设的水平方向垂直；半透半反射镜，设置在同轴光源上方，用于将探测光反射到预设位置；反射镜，接收半透半反射镜反射的光，并将半透半反射镜反射的光反射到保护镜片；工业相机，与半透半反射镜和反射镜在预设的水平方向水平，用于接收入射光，得到检测图像，其中，入射光为保护镜片对探测光反射后经反射镜和半透半反射镜之后形成的入射光；处理器，用于根据检测图像进行计算，确定保护镜片上是否存在坏点，解决了激光焊接时保护镜片被污染无法及时获知的问题，进而达到了及时获知保护镜片上的坏点以便更换保护镜片的效果。
附图说明
构成本申请的一部分的附图用来提供对本申请的进一步理解，本申请的示意性实施例及其说明用于解释本申请，并不构成对本申请的不当限定。在附图中：
图1是根据相关技术的一种激光焊接头内部光路图；
图2是根据本申请第一实施例的用于汽车生产线上激光焊保护镜片的缺陷检测装置的示意图；
图3是根据本申请实施例的保护镜面的全反射和漫反射的原理示意图；
图4是根据本申请实施例的光路原理的示意图；
图5是根据本申请实施例的一种用于汽车生产线上激光焊保护镜片的缺陷检测装置的截面示意图；
图6是根据本申请实施例的吸光棉设置位置的示意图；
图7是根据本申请第一实施例的用于汽车生产线上激光焊保护镜片的缺陷检测方法的流程图；
图8是根据本申请第二实施例的用于汽车生产线上激光焊保护镜片的缺陷检测方法的流程图；
图9是根据本申请实施例的一种用于汽车生产线上激光焊保护镜片的缺陷检测装置的硬件结构的示意图；
图10是根据本申请实施例的硬件通信的流程图；
图11是根据本申请实施例的一种机器人与传感器的吹气交互过程的流程图；
图12是根据本申请实施例的一种机器人控制传感器检测图像的流程图；
图13是根据本申请实施例的嵌入式人机交互软件的界面示意图；
图14是根据本申请实施例的用于汽车生产线上激光焊保护镜片的缺陷检测装置的示意图。
具体实施方式
需要说明的是，在不冲突的情况下，本申请中的实施例及实施例中的特征可以相互组合。下面将参考附图并结合实施例来详细说明本申请。
为了使本技术领域的人员更好地理解本申请方案，下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本申请一部分的实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都应当属于本申请保护的范围。
需要说明的是，本申请的说明书和权利要求书及上述附图中的术语“第一”、“第二”等是用于区别类似的对象，而不必用于描述特定的顺序或先后次序。应该理解这样使用的数据在适当情况下可以互换，以便这里描述的本申请的实施例。此外，术语“包括”和“具有”以及他们的任何变形，意图在于覆盖不排他的包含，例如，包含了一系列步骤或单元的过程、方法、系统、产品或设备不必限于清楚地列出的那些步骤或单元，而是可包括没有清楚地列出的或对于这些过程、方法、产品或设备固有的其它步骤或单元。
本申请实施例提供了一种用于汽车生产线上激光焊保护镜片的缺陷检测装置。
图2是根据本申请第一实施例的用于汽车生产线上激光焊保护镜片的缺陷检测装置的示意图，如图2所示，该装置包括：
同轴光源10，用于发出准直探测光，所述探测光的方向与预设的水平方向垂直；
半透半反射镜20，设置在所述同轴光源10上方，用于将所述探测光反射到预设位置；
反射镜30，接收所述半透半反射镜20反射的光，并将所述半透半反射镜20反射的光反射到保护镜片；
工业相机40，与所述半透半反射镜20和所述反射镜在所述预设的水平方向水平，用于接收入射光，得到检测图像，其中，所述入射光为所述保护镜片对所述探测光反射后经所述反射镜和所述半透半反射镜之后形成的入射光；
处理器50，用于根据所述检测图像进行计算，确定所述保护镜片上是否存在坏点。
如图2所示，同轴光源发出的探测光通过半透半反射镜的与反射镜的反射，将光源发出的光垂直射到焊接头的保护镜片上，同轴光源在测量光路中能够提供比传统光源更准直的探测光，提高了视觉测量的准确性和灵敏度，由于光源为同轴光源，所以在内部光路中将经由半透半反射镜与反射镜垂直入射至保护镜片上。图3是根据本申请实施例的保护镜面的全反射和漫反射的原理示意图，如图3所示，如果保护镜片无缺陷，即为平整光滑平面，将在保护镜面上发生全反射。如果保护镜面有缺陷，则在保护镜面上回发生漫反射。如果保护镜面为光滑镜面，同轴光将其表面发生全反射，光线按原光路返回；如果保护镜面有坏点，呈现为凹凸镜面，同轴光将在此处发生漫反射，光线无法按原光路全部返回。这样在相机处成像时，保护镜片缺陷处只有部分光线进入，呈暗色。在全反射情况下，由于光路的可逆性，将有部分光先后经由发射镜和半透半反射镜垂直入射至工业相机内，光路原理图如图4所示，保护镜片反射的光将经过反射镜的反射，透过半透半反射镜到达工业相机，在相机中成像。通过光路可以看出，如果保护镜片上附着有杂质，则凹凸缺陷处无法按照原光路反射入社光， 导致该处只有少量部分光线最终进入相机。所以凹凸处在相机中成像的对应位置应该为暗点，也称坏点，因此就可以根据相机成像判断出保护镜片上是否存在坏点。
该实施例采用通过同轴光源10，用于发出准直探测光，所述探测光的方向与预设的水平方向垂直；半透半反射镜20，设置在所述同轴光源10上方，用于将所述探测光反射到预设位置；反射镜30，接收所述半透半反射镜20反射的光，并将所述半透半反射镜20反射的光反射到保护镜片；工业相机40，与所述半透半反射镜20和所述反射镜在所述预设的水平方向水平，用于接收入射光，得到检测图像，其中，所述入射光为所述保护镜片对所述探测光反射后经所述反射镜和所述半透半反射镜之后形成的入射光；处理器50，用于根据所述检测图像进行计算，确定所述保护镜片上是否存在坏点，解决了激光焊接时保护镜片被污染无法及时获知的问题，进而达到了及时获知保护镜片上的坏点以便更换保护镜片的效果。
通过本申请实施例的用于汽车生产线上激光焊保护镜片的缺陷检测装置可以用于汽车生产线上激光焊接保护镜片的视觉镜面缺陷检测，可以解决在激光焊接作业中自动检测激光焊接头保护镜片缺陷的问题，可适用于汽车生产线上所有的带有保护镜片的激光焊接头，保证了对焊接作业中激光焊接头保护镜片产生缺陷的及时预警，也大大减小了白车身重回生产线的返修率。
可选地，该用于汽车生产线上激光焊保护镜片的缺陷检测装置还包括吹气装置，用于在激光焊接过程中向所述工业相机的镜头处吹气，可以清楚工业相机镜头上的灰尘。通过向相机镜头处吹气可以减少焊接过程中的杂志颗粒溅射到相机镜头处带来的误差，使检测结果更加准确。图5是根据本申请实施例的一种用于汽车生产线上激光焊保护镜片的缺陷检测装置的截面示意图，如图5所示，当激光焊头在线作业时，设备将开启吹气模式，利用工厂的气源，在楔形盒体的光路出口进行气吹除尘，待激光焊头完成作业后，直接进行缺陷检测，通过设置吹气装置可以节约时间成本。
可选地，本申请实施例的汽车生产线上激光焊保护镜片的缺陷检测装置还包括黑色吸光棉，图6是根据本申请实施例的吸光棉设置位置的示意图，如图6所示，黑色吸光棉位于半透半反镜与工业相机上方，用于吸收半透半反镜透射的同轴光源的入射光；通过设置黑色吸光棉可以防止同轴光源的原始光第一次经过半透半反射镜产生的透射光对工业相机成像产生影响，特殊的黑色吸光棉可以安装在半透半反射镜的上方以减少光源透射干扰，提高测试结果准确性。
本申请实施例还提供了一种用于汽车生产线上激光焊保护镜片的缺陷检测方法，图7是根据本申请第一实施例的用于汽车生产线上激光焊保护镜片的缺陷检测方法的流程图，如图7所示，该方法包括以下步骤：
步骤S102，获取检测图像，其中，检测图像用于汽车生产线上激光焊保护镜片的缺陷检测装置的工业相机得到的检测图像；
步骤S104，通过特征匹配确定检测图像中的镜面区域的外接矩形；
步骤S106，在外接矩形内通过阈值设定进行坏点轮廓检索，得到坏点面积和坏点像素数。
在本申请实施例中，检测图像为上述实施例中用于汽车生产线上激光焊保护镜片的缺陷检测装置中的工业相机中的检测图像，在获取到检测图像之后通过特征匹配的方式定位到检测图像中的镜面圆形区域的外接矩形，如果检测到检测图像的镜面圆形区域不完整，则重新检测，如果能够定位到检测图像中的圆形镜面区域，则确定圆形镜面区域的外接矩形，然后在外接矩形内通过阈值设定对区域内进行坏点轮廓检索，得到外接矩形内的坏点面积和坏点像素数。通过这样的方法能够及时获取到保护镜片上的杂质产生的坏点，及时确定是否需要更换保护镜片。
在定位图像中的镜面区域之后，可以取其外接矩形作为感兴趣区域，并将该感兴趣区域作为图像识别的工作区域。
图8是根据本申请第二实施例的用于汽车生产线上激光焊保护镜片的缺陷检测方法的流程图，该实施例可以作为上述第一实施例的优选实施方式，如图8所示，该方法包括以下步骤：
当进入图像处理算法流程后，首先在图像内寻找保护镜片。因为同轴光会经过透射反射进入工业相机，所以图像内的镜片应该为亮白色。如果找到则利用特征匹配方法检测镜面区域，并判断是否发生定位区域越界。如果发生越界则提示镜片不存在，未发生越界则取定位区域的外接矩形。在外接矩形内进行阈值设定进而检索坏点轮廓。最后经过计算检索到的坏点像素面积以及比较其与预先人工设定的坏点面积的大小，得到最终的坏点判断结果，如果坏点面积大于用户自定义坏点像素值(也即预设值)，则标记坏点轮廓并判断坏点个数是否大于等于用户自定义像素个数值，如果判断结果为是，则提示保护镜片异常(GlassIsBad)，如果判断结果为否，则提示保护镜片正常(GlassIsGood)。
可选地，在通过特征匹配定位检测图像中的镜面区域之前，判断在检测图像中是否存在镜面区域对应的图像；其中，如果判断出存在镜面区域对应的图像，则通过特征匹配定位检测图像中的镜面区域；如果判断出不存在镜面区域对应的图像，则重新获取检测图像。
可选地，在外接矩形内通过阈值设定进行坏点轮廓检索包括：计算坏点面积并将坏点面积转化成像素个数；在得到坏点面积和坏点个数之后，判断坏点面积是否大于 等于用户自定义坏点像素数，如果坏点面积大于等于用户自定义坏点像素数，则标记坏点轮廓，判断坏点个数是否大于等于用户自定义坏点个数值，如果判断出坏点像素数大于等于用户自定义坏点个数值，则提示保护镜片异常。
在判断坏点面积是否大于等于用户自定义坏点像素数，如果坏点面积小于用户自定义坏点像素数，则提示保护镜片正常，在判断坏点像素数是否大于等于用户自定义坏点个数值之后，如果坏点像素数小于用户自定义坏点个数值，则提示保护镜片正常。
可选地，在获取检测图像之前，接收机器人发送的指令；判断指令的标识数值，其中，标识数值包括两种；其中，在所述标志数值为第一种时获取所述检测图像，在所述标志数值为第二种时通过吹气装置对所述工业相机的镜头处吹气，达到除尘的效果。
本申请实施例还提供了一种用于汽车生产线上激光焊保护镜片的缺陷检测装置的硬件结构，图9是根据本申请实施例的一种用于汽车生产线上激光焊保护镜片的缺陷检测装置的硬件结构的示意图，如图9所示，本申请实施例的硬件平台利用ARM芯片作为主核，其他类型的处理器也在本申请的保护范围内。本申请实施例的传感器可以直接与机器人或者PLC通信，兼容目前主流的工业通讯协议，同时连接触摸屏、CAMERA等一系列外设来实现核心功能。应用的外部存储介质可以为SD卡。
图10是根据本申请实施例的硬件通信的流程图。如图10所示，整个流程包括两个分支流程：
1)检测流程；2)吹气流程。
检测流程和吹气流程的选择性开启完全是由机器人发送指令来完成。当标志Tmp为1时，进入检测流程分支；当标志Tmp为2时，进入吹气流程分支。
吹气流程：
传感器镜片的尘土会影响测试效果，需要在每个测量节拍进行除尘工作。除尘方法采用气吹的方式，传感器需要引入工厂的气源。传感器控制电磁阀在每次测量之前，进行气吹除尘工作。当Tmp不为2时，提示错误(Default)。
设备自带除尘功能。吹气流程与激光焊接同时进行，目的是为了防止在激光焊接的过程中有焊渣飞溅到检测设备的相机上，造成坏点误判。当焊接停止时，吹气流程随之停止。
图11是根据本申请实施例的一种机器人与传感器的吹气交互过程的流程图，如图11所示，机器人向传感器发送指令，吹气过程与激光焊接同步进行，传感器在采集完成之后将结果发送到机器人，机器人同时接收到吹气完成的结果，机器人向传感器发 送清零指令“0”，传感器接收到清零指令之后，清除存储的结果。
图12是根据本申请实施例的一种机器人控制传感器检测图像的流程图，如图12所示，利用同轴光源对镜片照明，相机采集测量图像，通过嵌入式处理系统结合定制的图像分析算法，检测镜片上是否存在坏点，将测量图像和检测结果(有无坏点、坏点位置及尺寸)显示在传感器的LED显示屏上，同时将检测结果反馈给机器人。检测流程每次需激光焊头到达检测位置后开始。
图13是根据本申请实施例的嵌入式人机交互软件的界面示意图，如图13所示，该设备的人机交互功能是通过触摸屏以及嵌入式软件来完成。嵌入式软件中，通过精确仿真的图像处理算法，准确实时的找到镜片上的坏点。并且用户可以通过用户界面实时观察坏点检测情况，自行定义坏点的面积和个数，浏览坏点图像等。
本申请实施例的用于汽车生产线上激光焊保护镜片的缺陷检测方法可以对一切采用激光焊接的汽车生产线进行激光焊头保护镜片的镜面缺陷检测。对在作业中的激光焊头的保护镜片进行静态的缺陷检测并且第一时间给出检测结果并对上位机器人给出反馈，本方法结构简单，算法复杂度低，机器人轨迹调试工作量较小，校准过程方便简洁，占用时间较短。
需要说明的是，在附图的流程图示出的步骤可以在诸如一组计算机可执行指令的计算机系统中执行，并且，虽然在流程图中示出了逻辑顺序，但是在某些情况下，可以以不同于此处的顺序执行所示出或描述的步骤。
本申请实施例提供了一种用于汽车生产线上激光焊保护镜片的缺陷检测装置，该装置可以用于执行本申请实施例的用于汽车生产线上激光焊保护镜片的缺陷检测方法。
图14是根据本申请实施例的用于汽车生产线上激光焊保护镜片的缺陷检测装置的示意图，如图14所示，该装置包括：
获取单元110，用于获取检测图像，其中，检测图像为本申请实施例的用于汽车生产线上激光焊保护镜片的缺陷检测装置的工业相机得到的检测图像；
确定单元120，用于通过特征匹配确定检测图像中的镜面区域的外接矩形；
检索单元130，用于在外接矩形内通过阈值设定进行坏点轮廓检索，得到坏点面积和坏点像素数。
可选地，该装置还包括：判断单元，用于在通过特征匹配确定检测图像中的镜面区域的外接矩形之前，判断检测图像中是否存在镜面区域对应的图像；其中，如果判断出存在镜面区域对应的图像，则通过特征匹配确定检测图像中的镜面区域的外接矩 形；如果判断出不存在镜面区域对应的图像，则重新获取检测图像。
通过本申请实施例的用于汽车生产线上激光焊保护镜片的缺陷检测装置，解决了激光焊接时保护镜片被污染无法及时获知的问题，进而达到了及时获知保护镜片上的坏点以便更换保护镜片的效果。
本领域内的技术人员应明白，本申请的实施例可提供为方法、系统、或计算机程序产品。因此，本申请可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。
本申请是参照根据本申请实施例的方法、设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器，使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。
还需要说明的是，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、商品或者设备不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、商品或者设备所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括要素的过程、方法、商品或者设备中还存在另外的相同要素。
本领域技术人员应明白，本申请的实施例可提供为方法、系统或计算机程序产品。因此，本申请可采用完全硬件实施例、完全软件实施例或结合软件和硬件方面的实施例的形式。
以上仅为本申请的实施例而已，并不用于限制本申请。对于本领域技术人员来说，本申请可以有各种更改和变化。凡在本申请的精神和原理之内所作的任何修改、等同替换、改进等，均应包含在本申请的权利要求范围之内。
工业实用性
本申请通过同轴光源发出准直探测光，探测光的方向与预设的水平方向垂直；半透半反射镜，设置在同轴光源上方，用于将探测光反射到预设位置；反射镜，接收半透半反射镜反射的光，并将半透半反射镜反射的光反射到保护镜片；工业相机，与半透半反射镜和反射镜在预设的水平方向水平，用于接收入射光，得到检测图像，其中，入射光为保护镜片对探测光反射后经反射镜和半透半反射镜之后形成的入射光；处理 器，用于根据检测图像进行计算，确定保护镜片上是否存在坏点，解决了激光焊接时保护镜片被污染无法及时获知的问题，进而达到了及时获知保护镜片上的坏点以便更换保护镜片的效果。

Claims (9)

  1. 一种用于汽车生产线上激光焊保护镜片的缺陷检测装置，包括：
    同轴光源，设置为发出准直探测光，所述探测光的方向与预设的水平方向垂直；
    半透半反射镜，设置在所述同轴光源上方，设置为将所述探测光反射到预设位置；
    反射镜，接收所述半透半反射镜反射的光，并将所述半透半反射镜反射的光反射到保护镜片；
    工业相机，与所述半透半反射镜和所述反射镜在所述预设的水平方向水平，设置为接收入射光，得到检测图像，其中，所述入射光为所述保护镜片对所述探测光反射后经所述反射镜和所述半透半反射镜之后形成的入射光；
    处理器，设置为根据所述检测图像进行计算，确定所述保护镜片上是否存在坏点。
  2. 根据权利要求1所述的装置，其中，所述装置还包括：
    吹气装置，在激光焊接过程中向所述工业相机的镜头处吹气，设置为清除所述工业相机镜头上的灰尘。
  3. 根据权利要求1所述的装置，其特征在于，所述装置还包括：
    黑色吸光棉，位于所述半透半反镜与所述工业相机上方，设置为吸收所述半透半反镜透射的所述同轴光源的入射光。
  4. 一种用于汽车生产线上激光焊保护镜片的缺陷检测方法，包括：
    获取检测图像，其中，所述检测图像为权利要求1中用于汽车生产线上激光焊保护镜片的缺陷检测装置的工业相机得到的检测图像；
    通过特征匹配确定所述检测图像中的镜面区域的外接矩形；
    在所述外接矩形内通过阈值设定进行坏点轮廓检索，得到坏点面积和坏点像素数。
  5. 根据权利要求4所述的方法，其中，在通过特征匹配确定所述检测图像中的镜面区域的外接矩形之前，所述方法还包括：
    判断所述检测图像中是否存在所述镜面区域对应的图像；
    其中，如果判断出存在所述镜面区域对应的图像，则通过特征匹配确定所述 检测图像中的镜面区域的外接矩形；如果判断出不存在所述镜面区域对应的图像，则重新获取检测图像。
  6. 根据权利要求4所述的方法，其中，
    在所述外接矩形内通过阈值设定进行坏点轮廓检索包括：计算坏点面积并将所述坏点面积转化成像素个数；
    在得到所述坏点面积和所述坏点像素数之后，所述方法还包括：判断所述坏点面积转化成的坏点像素数是否大于等于用户自定义坏点像素数量，如果所述坏点面积转化成的坏点像素数大于等于所述用户自定义坏点像素数量，则标记坏点轮廓，
    判断所述坏点轮廓的个数是否大于等于用户自定义坏点个数值，如果判断出所述坏点轮廓的个数大于等于所述用户自定义坏点个数值，则提示保护镜片异常；
    如果所述坏点面积转化成的坏点像素数小于用户自定义坏点像素数量，则提示所述保护镜片正常；如果所述坏点轮廓的个数小于所述用户自定义坏点个数值，则提示所述保护镜片正常。
  7. 根据权利要求4所述的方法，其中，在获取所述检测图像之前，所述方法还包括：
    接收机器人发送的指令；
    判断所述指令的标识数值，其中，所述标识数值有两种；
    其中，在所述标识数值为第一种时获取所述检测图像，在所述标识数值为第二种时通过吹气装置对所述工业相机的镜头处吹气。
  8. 一种用于汽车生产线上激光焊保护镜片的缺陷检测装置，包括：
    获取单元，设置为获取检测图像，其中，所述检测图像为权利要求1中用于汽车生产线上激光焊保护镜片的缺陷检测装置的工业相机得到的检测图像；
    确定单元，设置为通过特征匹配确定所述检测图像中的镜面区域的外接矩形；
    检索单元，设置为在所述外接矩形内通过阈值设定进行坏点轮廓检索，得到坏点面积和坏点像素数。
  9. 根据权利要求8所述的装置，其中，所述装置还包括：
    判断单元，设置为在通过特征匹配确定所述检测图像中的镜面区域的外接矩形之前，判断所述检测图像中是否存在所述镜面区域对应的图像；
    其中，如果判断出存在所述镜面区域对应的图像，则通过特征匹配确定所述检测图像中的镜面区域的外接矩形；如果判断出不存在所述镜面区域对应的图像，则重新获取检测图像。
PCT/CN2017/093915 2017-07-21 2017-07-21 用于汽车生产线上激光焊保护镜片的缺陷检测方法和装置 WO2019014937A1 (zh)

Priority Applications (3)

Application Number Priority Date Filing Date Title
PCT/CN2017/093915 WO2019014937A1 (zh) 2017-07-21 2017-07-21 用于汽车生产线上激光焊保护镜片的缺陷检测方法和装置
US16/073,241 US10794838B2 (en) 2017-07-21 2017-07-21 Method and device for detecting defect of cover lens in laser welding system on automobile production line
CN201710867489.6A CN107643296B (zh) 2017-07-21 2017-09-22 用于汽车生产线上激光焊保护镜片的缺陷检测方法和装置

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2017/093915 WO2019014937A1 (zh) 2017-07-21 2017-07-21 用于汽车生产线上激光焊保护镜片的缺陷检测方法和装置

Publications (1)

Publication Number Publication Date
WO2019014937A1 true WO2019014937A1 (zh) 2019-01-24

Family

ID=61111344

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/093915 WO2019014937A1 (zh) 2017-07-21 2017-07-21 用于汽车生产线上激光焊保护镜片的缺陷检测方法和装置

Country Status (3)

Country Link
US (1) US10794838B2 (zh)
CN (1) CN107643296B (zh)
WO (1) WO2019014937A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112548321A (zh) * 2020-12-04 2021-03-26 哈尔滨工业大学 一种基于同轴监测的真空激光焊接焊缝缺陷识别方法
CN115979597A (zh) * 2023-02-13 2023-04-18 杭州简并激光科技有限公司 一种激光水平仪锥镜的缺陷检测方法

Families Citing this family (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10643329B2 (en) * 2018-04-09 2020-05-05 The Boeing Company Automated paint quality control for aircraft
CN109211939B (zh) * 2018-11-26 2023-09-08 易思维（杭州）科技有限公司 激光焊保护镜片的缺陷检测装置、模块及检测方法
CN109636792B (zh) * 2018-12-14 2020-05-22 易思维（杭州）科技有限公司 一种基于深度学习的镜片缺陷检测方法
CN110044849B (zh) * 2019-05-24 2021-08-17 易思维（杭州）科技有限公司 半封闭腔体内部缺陷检测装置
CN110514410A (zh) * 2019-08-21 2019-11-29 奔腾楚天激光（武汉）有限公司 一种光纤激光切割机快速判断保护镜片质量的工艺方法
CN110530884B (zh) * 2019-10-09 2021-11-02 易思维（杭州）科技有限公司 激光焊保护镜片缺陷检测装置
CN110715935A (zh) * 2019-11-11 2020-01-21 佛山市新石器机器人有限公司 一种瓷砖缺陷检测设备及方法
CN110806412A (zh) * 2019-11-15 2020-02-18 中国工程物理研究院激光聚变研究中心 一种基于光学元件的缺陷尺寸检测方法及系统
CN111722408A (zh) * 2020-07-31 2020-09-29 中国人民解放军国防科技大学 大角度偏转收发一体光纤准直器
CN112666178B (zh) * 2020-12-14 2024-06-18 杭州当虹科技股份有限公司 一种户外led大屏坏点在线监控方法
CN113188474B (zh) * 2021-05-06 2022-09-23 山西大学 一种用于高反光材质复杂物体成像的图像序列采集系统及其三维形貌重建方法
CN115781004B (zh) * 2023-02-09 2023-05-16 广州三义激光科技有限公司 一种激光扩束镜保护镜防护系统
CN117900712B (zh) * 2024-03-18 2024-06-04 佛山市三乔焊接实业有限公司 一种焊接方法、装置、电子设备及存储介质

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6292260B1 (en) * 1997-07-03 2001-09-18 Isoa, Inc. System and method of optically inspecting surface structures on an object
CN101650307A (zh) * 2009-07-17 2010-02-17 山东富美科技有限公司 刮刀片表面检测系统
CN101776614A (zh) * 2009-01-09 2010-07-14 东捷科技股份有限公司 缺陷检查装置
CN201903499U (zh) * 2010-10-28 2011-07-20 松下电工电子材料(苏州)有限公司 一种能够自动吹扫的外观检查机
CN102353680A (zh) * 2011-07-08 2012-02-15 中国科学院宁波材料技术与工程研究所 小型工件表面缺陷的评估方法以及检测不合格工件的流程
CN102636490A (zh) * 2012-04-12 2012-08-15 江南大学 基于机器视觉的轴承防尘盖表面缺陷检测方法
CN103163149A (zh) * 2013-03-18 2013-06-19 上海理工大学 药瓶封口质量检测系统
CN104931505A (zh) * 2014-03-17 2015-09-23 荆弘明 机器视觉表面检测系统
CN106053479A (zh) * 2016-07-21 2016-10-26 湘潭大学 一种基于图像处理的工件外观缺陷的视觉检测系统

Family Cites Families (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TW325744U (en) * 1993-07-21 1998-01-21 Ciba Geigy Ag Two-sided contact lens mold
CA2279714A1 (en) * 1999-08-04 2001-02-04 Jds Fitel Inc. Method and system for measurement of a characteristic of lens
DE10158095B4 (de) * 2001-05-05 2012-03-22 Lpkf Laser & Electronics Ag Vorrichtung zur Kontrolle einer Schweißnaht in einem aus schweißfähigem Kunststoff bestehenden Werkstück
CN101089550B (zh) 2007-07-13 2010-05-19 上海大学 基于数字微镜和数据合成技术的透明物体或反射物体的形貌测量装置和方法
JP2009056507A (ja) 2007-09-03 2009-03-19 Olympus Corp レーザ加工装置
JP5090121B2 (ja) 2007-10-01 2012-12-05 オリンパス株式会社 調整装置、レーザ加工装置、調整方法、および調整プログラム
US8908167B2 (en) * 2010-05-05 2014-12-09 Afl Telecommunications Llc Fiber optic connector inspection microscope with integral optical power measurement
KR20130064479A (ko) 2011-12-08 2013-06-18 삼성테크윈 주식회사 피검체 검사용 렌즈 장치 및 이를 구비한 머신 비전 시스템
US9575008B2 (en) 2014-02-12 2017-02-21 ASA Corporation Apparatus and method for photographing glass in multiple layers
JP5929948B2 (ja) * 2014-02-28 2016-06-08 トヨタ自動車株式会社 溶接部の検査方法
CN104076039B (zh) 2014-03-28 2017-05-31 合波光电通信科技有限公司 滤光片外观缺陷自动检测方法

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6292260B1 (en) * 1997-07-03 2001-09-18 Isoa, Inc. System and method of optically inspecting surface structures on an object
CN101776614A (zh) * 2009-01-09 2010-07-14 东捷科技股份有限公司 缺陷检查装置
CN101650307A (zh) * 2009-07-17 2010-02-17 山东富美科技有限公司 刮刀片表面检测系统
CN201903499U (zh) * 2010-10-28 2011-07-20 松下电工电子材料(苏州)有限公司 一种能够自动吹扫的外观检查机
CN102353680A (zh) * 2011-07-08 2012-02-15 中国科学院宁波材料技术与工程研究所 小型工件表面缺陷的评估方法以及检测不合格工件的流程
CN102636490A (zh) * 2012-04-12 2012-08-15 江南大学 基于机器视觉的轴承防尘盖表面缺陷检测方法
CN103163149A (zh) * 2013-03-18 2013-06-19 上海理工大学 药瓶封口质量检测系统
CN104931505A (zh) * 2014-03-17 2015-09-23 荆弘明 机器视觉表面检测系统
CN106053479A (zh) * 2016-07-21 2016-10-26 湘潭大学 一种基于图像处理的工件外观缺陷的视觉检测系统

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112548321A (zh) * 2020-12-04 2021-03-26 哈尔滨工业大学 一种基于同轴监测的真空激光焊接焊缝缺陷识别方法
CN115979597A (zh) * 2023-02-13 2023-04-18 杭州简并激光科技有限公司 一种激光水平仪锥镜的缺陷检测方法
CN115979597B (zh) * 2023-02-13 2024-01-12 杭州简并激光科技有限公司 一种激光水平仪锥镜的缺陷检测方法

Also Published As

Publication number Publication date
US20200116652A1 (en) 2020-04-16
CN107643296B (zh) 2018-12-14
US10794838B2 (en) 2020-10-06
CN107643296A (zh) 2018-01-30

Similar Documents

Publication Publication Date Title
WO2019014937A1 (zh) 用于汽车生产线上激光焊保护镜片的缺陷检测方法和装置
JP3560694B2 (ja) レンズ検査のシステムと方法
Li et al. Measurement and defect detection of the weld bead based on online vision inspection
CN111830906B (zh) 一种旋转刀具失效状态在机监测系统及其检测方法
JP2007532910A (ja) 複合建造物の欠陥位置を示すために光を用いるシステムおよび方法
JPH07201946A (ja) 半導体装置等の製造方法及びその装置並びに検査方法及びその装置
JP2007501942A (ja) 好適に円形エッジを有する物体の品質を光学的に制御する光学的試験方法及び光学的試験装置
US11650126B2 (en) Systems and methods for automatic visual inspection of defects in ophthalmic lenses
CN111692968B (zh) 一种微细铣刀在机多视角视觉检测仪及其检测方法
TWI558999B (zh) 瑕疵檢測方法及其裝置
CN202433310U (zh) 用于检测阀座表面缺陷的系统
CN115035031A (zh) Pin针的缺陷检测方法、装置、电子设备及存储介质
IT201800003247A1 (it) Sistema di misura di gap e flush
CN202171575U (zh) 用于光学镜片外观的机械视觉检测装置
CN109931882B (zh) 换热翅片关键参数检测系统及测量方法
CN101726244A (zh) 柱体的视觉检验系统
US7899573B2 (en) Non-contact method and system for inspecting a multi-faceted machine surface
CN113686903A (zh) 一种光学元件缺陷检测系统及检测方法
KR101338443B1 (ko) 엔진 가공품의 무인 검사 방법
WO2020082714A1 (zh) 一种激光全反射式的3c透明构件缺陷检测装置及方法
KR20180103467A (ko) 3차원 카메라를 이용한 용접 로봇 끝단의 위치 추정 시스템 및 방법
TWI822209B (zh) 銲接品質檢測裝置
CN111183351A (zh) 图像传感器表面缺陷检测方法及检测系统
TWI837941B (zh) 引線框架的出貨方法
US11475553B1 (en) Production-speed component inspection system and method

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17918084

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17918084

Country of ref document: EP

Kind code of ref document: A1