WO2020051728A1 - 一种陶瓷球自动分拣系统及方法 - Google Patents

一种陶瓷球自动分拣系统及方法 Download PDF

Info

Publication number
WO2020051728A1
WO2020051728A1 PCT/CN2018/104787 CN2018104787W WO2020051728A1 WO 2020051728 A1 WO2020051728 A1 WO 2020051728A1 CN 2018104787 W CN2018104787 W CN 2018104787W WO 2020051728 A1 WO2020051728 A1 WO 2020051728A1
Authority
WO
WIPO (PCT)
Prior art keywords
ceramic ball
ceramic
image
ball
robot arm
Prior art date
Application number
PCT/CN2018/104787
Other languages
English (en)
French (fr)
Inventor
宋健
孙峰
张伟儒
董廷霞
徐学敏
张宝存
司尚进
Original Assignee
中材高新氮化物陶瓷有限公司
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 中材高新氮化物陶瓷有限公司 filed Critical 中材高新氮化物陶瓷有限公司
Priority to JP2019535303A priority Critical patent/JP2021502229A/ja
Priority to EP18884856.8A priority patent/EP3650129A4/en
Priority to US16/479,019 priority patent/US11327028B2/en
Priority to PCT/CN2018/104787 priority patent/WO2020051728A1/zh
Publication of WO2020051728A1 publication Critical patent/WO2020051728A1/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
    • G01N21/951—Balls
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
    • B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
    • B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
    • B07C5/34—Sorting according to other particular properties
    • B07C5/342—Sorting according to other particular properties according to optical properties, e.g. colour
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J9/00—Programme-controlled manipulators
    • B25J9/16—Programme controls
    • B25J9/1694—Programme controls characterised by use of sensors other than normal servo-feedback from position, speed or acceleration sensors, perception control, multi-sensor controlled systems, sensor fusion
    • B25J9/1697—Vision controlled systems
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B19/00—Programme-control systems
    • G05B19/02—Programme-control systems electric
    • G05B19/18—Numerical control [NC], i.e. automatically operating machines, in particular machine tools, e.g. in a manufacturing environment, so as to execute positioning, movement or co-ordinated operations by means of programme data in numerical form
    • G05B19/4155—Numerical control [NC], i.e. automatically operating machines, in particular machine tools, e.g. in a manufacturing environment, so as to execute positioning, movement or co-ordinated operations by means of programme data in numerical form characterised by programme execution, i.e. part programme or machine function execution, e.g. selection of a programme
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00—Program-control systems
    • G05B2219/30—Nc systems
    • G05B2219/40—Robotics, robotics mapping to robotics vision
    • G05B2219/40269—Naturally compliant robot arm

Definitions

  • the invention relates to the technical field of surface quality detection of bearing rolling bodies, and in particular to an automatic sorting system and method for ceramic balls.
  • Ceramic balls are the most important parts of slewing bodies in various bearings. According to statistics, 90% of bearing damage is slewing parts, especially caused by appearance quality problems, such as ceramic balls with surface defects such as pits and pores. Under high temperature and high pressure, they are first locally heated and damaged. In production, surface defects must be detected on finished balls, and balls with quality problems in appearance must be rejected 100%.
  • the traditional detection method is to place the ceramic ball under a microscope and magnify it 20 times, and then compare it with the standard visual inspection. This manual detection method has low degree of automation, high rate of false detection and missed detection, and it is difficult to improve production efficiency.
  • the technical problem to be solved by the present invention is to provide a ceramic ball automatic sorting system and method, thereby realizing the automation of ceramic ball defect recognition and sorting, and improving the accuracy of ceramic ball defect recognition and sorting efficiency.
  • the present invention provides a ceramic ball automatic sorting system and method.
  • the ceramic ball automatic sorting system includes a computer and a ceramic ball feeding track, a robot arm, an image acquisition device, a ceramic ball clamping and turning device, and a ball storage device respectively connected to the computer;
  • the computer controls the robot arm to be positioned at the position where the ceramic ball is sucked, suck the ceramic ball on the ceramic ball feed track, and transfer the ceramic ball to the acquisition area of the image acquisition device for image acquisition;
  • the image acquisition device performs image acquisition on the ceramic ball, and sends the acquired image information to the computer in real time;
  • the computer controls the robot arm to transfer the ceramic ball to the ceramic ball gripping and turning device to flip 90 degrees in a warp direction, and then perform image acquisition on the ceramic ball after the flip, and acquire the acquired image information. Send to the computer in real time;
  • the computer identifies the ball storage device into which the ceramic ball is placed according to the image information collected by the image acquisition device, and sends a control instruction to control the robot arm to place the ceramic ball into the ball storage device to realize the ceramic ball. Automatic sorting.
  • the ceramic ball feeding track includes a ceramic ball automatic positioning device, and the ceramic ball feeding track is connected to the computer through the ceramic ball automatic positioning device;
  • the ceramic ball feeding track is inclined, and the ceramic ball rolls from a high point to a low point of the inclined ceramic ball feeding track.
  • a vacuum gas source is connected to the robot arm, and the robot arm sucks the ceramic ball through the connected vacuum gas source;
  • a rotary motor is connected to the robot arm, and the rotary motor drives the robot arm Rotating in the axial direction, the angular range of the axial rotation is 120 degrees.
  • the image acquisition device includes a pneumatic slider, a camera fixed on the pneumatic slider, a camera position control system and a position sensor connected to the pneumatic slider; the image acquisition device passes the camera A position control system is connected to the computer.
  • the ball storage device includes a defective ceramic ball storage device and a qualified ceramic ball storage device; the position of the defective ceramic ball storage device and the qualified ceramic ball storage device is fixed, and the computer according to the The position of the defective ceramic ball storage device and the qualified ceramic ball storage device controls the robot arm to place the ceramic ball into a corresponding ball storage device.
  • the ceramic ball automatic sorting method includes:
  • the calculating the parallel distance L between the highest point of the ceramic ball and the end point of the ceramic ball feed track according to the diameter D of the ceramic ball specifically includes:
  • calculating the horizontal distance S from the front end of the image acquisition device to the center of the ceramic ball according to the diameter D of the ceramic ball specifically includes:
  • the stitching of the primary image of the ceramic ball and the secondary image of the ceramic ball to obtain a full coverage image of the ceramic ball specifically includes:
  • the primary image includes three images of the ceramic ball continuously flipped three times in the weft direction at 120 °;
  • the secondary image of the ceramic ball is stitched to obtain a secondary image after the stitching;
  • the secondary image includes three images taken after the ceramic ball is turned 90 degrees in a warp direction and then flipped three times in a parallel direction at 120 ° ;
  • the stitched primary image and the stitched secondary image are stitched to obtain a full coverage image of the ceramic ball.
  • the using a threshold segmentation algorithm to identify defects in the full coverage image of the ceramic ball according to the set image feature value threshold and defect threshold specifically includes:
  • a threshold segmentation algorithm is used to identify a full-cover image of the ceramic ball in the segmented background image to obtain a surface image of the ceramic ball after recognition;
  • the recognition result is that the ceramic ball is a qualified ceramic ball.
  • the present invention has the beneficial effects that the ceramic ball automatic sorting system and method disclosed in the present invention include a computer and a ceramic ball feed track, a robot arm, and image acquisition connected to the computer, respectively.
  • Device, ceramic ball clamping and turning device, and ball storage device the computer controls the robot arm to suck the ceramic ball on the ceramic ball feed track, and transfer the ceramic ball to the acquisition area of the image acquisition device Perform image acquisition, and send the acquired image information to the computer in real time.
  • the computer identifies the ball storage device into which the ceramic ball is placed according to the image information collected by the image acquisition device, and sends control instructions to control the robot
  • the arm puts the ceramic ball into the ball storage device to realize automatic sorting of the ceramic ball.
  • the system automatically picks up the ceramic balls on the ceramic ball feeding track for image acquisition, and identifies whether the ceramic balls are defective balls according to the collected image information, and determines the ball storage device into which the ceramic balls are put.
  • the whole process is realized without manual participation.
  • the automation of ceramic ball defect recognition and sorting improves the accuracy of ceramic ball defect recognition and the efficiency of sorting.
  • the ceramic ball automatic sorting method disclosed by the present invention uses automatic image stitching technology to automatically stitch the surface images of the balls at various angles to achieve full coverage. According to the set image feature value threshold, the ceramic ball is completely covered by a threshold segmentation algorithm. The image is identified and segmented from the background image.
  • the defect is identified using the threshold segmentation algorithm based on the set defect threshold to determine whether the ceramic ball is a defective ball and the storage should be placed in it.
  • Ball device thereby realizing the automation of ceramic ball defect recognition and sorting, and improving the accuracy and sorting efficiency of ceramic ball defect recognition.
  • FIG. 1 is a structural diagram of an embodiment of an automatic sorting system of ceramic balls according to the present invention.
  • FIG. 2 is a flowchart of an embodiment of a method for automatically sorting ceramic balls according to the present invention
  • FIG. 3 is a calculation principle diagram of the L value in the embodiment of the method for automatically sorting ceramic balls according to the present invention.
  • FIG. 4 is a calculation principle diagram of the S value in the embodiment of the ceramic ball automatic sorting method according to the present invention.
  • FIG. 1 is a structural diagram of an embodiment of a ceramic ball automatic sorting system according to the present invention.
  • the automatic ceramic ball sorting system includes a computer 5 and a ceramic ball feeding track 1 connected to the computer 5, a robot arm 3, an image acquisition device 4, a ceramic ball clamping and turning device 6, and a storage device. ⁇ ⁇ 7 ⁇ Ball device 7.
  • the ceramic ball automatic sorting system is fixed on the equipment platform and is in a closed positive pressure environment (a positive pressure environment created by compressed gas inside the equipment to prevent foreign impurities from entering), avoiding the interference of external dust impurities during the detection process. .
  • the ceramic ball feeding track 1 includes a ceramic ball automatic positioning device 2.
  • the ceramic ball feeding track 1 is connected to the computer 5 through the ceramic ball automatic positioning device 2; the ceramic ball feeding track 1 is inclined The ceramic ball rolls from the high point to the low point of the inclined ceramic ball feed track 1, and when the previous ceramic ball is removed, the next ceramic ball will roll to the end of the track by gravity, The end position of the track is the position of the ceramic ball automatic positioning device 2.
  • the robot arm 3 is composed of a servo motor, a grating linear displacement sensor, a single-chip microcomputer control system, and a feedback signal processing circuit, etc., to realize signal feedback, signal processing, and operation control.
  • the servo motor controls the robot arm 3 to achieve a linear displacement action
  • the grating linear displacement sensor performs positioning.
  • the specific model of the single-chip microcomputer control system is STC89C54.
  • the robot arm 3 can move left and right, up and down, and rotate 120 degrees in the axial direction (rotate 120 degrees in the parallel direction). The left and right movement is realized by a servo motor, and the up and down movement is realized by the motor shaft and the silk hole.
  • the robot arm 3 is connected There is a rotation motor, which drives the robot arm 3 to rotate in the axial direction, and the angle range of the axial rotation is 120 degrees.
  • a vacuum gas source is connected to the robot arm 3, the robot arm 3 sucks the ceramic ball through the connected vacuum gas source, uses a negative pressure air pressure to realize the suction of the ceramic ball, and pauses the compressed gas to realize the ceramic ball and the robot arm The separation avoids the damage caused by the mechanical structure to the surface of the ceramic ball. After the robot arm 3 finishes sorting a ceramic ball, it will directly return to a fixed starting position to sort the next ceramic ball again.
  • the image acquisition device 4 includes a pneumatic slider, a camera fixed on the pneumatic slider, a camera position control system and a position sensor connected to the pneumatic slider; the image acquisition device 4 is controlled by the camera position
  • the system is connected to the computer 5.
  • the camera position control system automatically selects the corresponding sensor position to achieve automatic positioning, automatic focus adjustment and automatic adjustment of the camera position.
  • the camera is connected to the motor shaft through a threaded hole, and the camera is moved back and forth by rotating the motor shaft. Different ceramic ball sizes correspond to different shooting positions.
  • the ball storage device 7 includes a defective ceramic ball storage device and a qualified ceramic ball storage device; the position of the defective ceramic ball storage device and the qualified ceramic ball storage device is fixed, and the computer 5 according to the defect The position of the ceramic ball storage device and the qualified ceramic ball storage device controls the robot arm 3 to place the ceramic ball in the corresponding ball storage device 7.
  • the computer 5 controls the robot arm 3 to be positioned at a position where the ceramic ball is sucked, sucks the ceramic ball on the ceramic ball feed track 1, and transfers the ceramic ball to a collection area of the image acquisition device 4 for Image Acquisition.
  • the image acquisition device 4 performs image acquisition on the ceramic ball, and sends the acquired image information to the computer 5 in real time.
  • the computer 5 controls the robot arm 3 to transfer the ceramic ball to the ceramic ball gripping and turning device 6 to turn 90 degrees in the warp direction, and then perform image acquisition on the inverted ceramic ball, and The image information is sent to the computer 5 in real time.
  • the computer 5 recognizes the ball storage device 7 into which the ceramic ball is placed according to the image information collected by the image acquisition device 4, and sends a control instruction to control the robot arm 3 to place the ceramic ball into the ball storage device. 7 to achieve automatic sorting of ceramic balls.
  • the robot arm is controlled by the computer to suck the ceramic balls on the ceramic ball feed track, and transfer the ceramic balls to the acquisition area of the image acquisition device to perform Image acquisition, sending the acquired image information to the computer in real time, the computer identifying the ball storage device into which the ceramic ball is placed according to the image information collected by the image acquisition device, and sending control instructions to control the robot arm
  • the ceramic balls are put into the ball storage device to realize automatic sorting of the ceramic balls.
  • the ceramic ball automatic sorting system disclosed by the invention can automatically pick up the ceramic balls on the ceramic ball feeding track for image collection, and identify whether the ceramic balls are defective balls according to the collected image information, and determine the storage balls placed in the ceramic balls.
  • the device eliminates the need for manual participation during the whole process, realizes the automation of ceramic ball defect identification and sorting, and improves the accuracy and sorting efficiency of ceramic ball defect identification.
  • FIG. 2 is a flowchart of an embodiment of a method for automatically sorting ceramic balls according to the present invention.
  • the ceramic ball automatic sorting method includes:
  • Step 201 Obtain the diameter D of the ceramic ball, and calculate a parallel distance L between the highest point of the ceramic ball and the end point of the feeding path of the ceramic ball according to the diameter D of the ceramic ball.
  • the step 201 specifically includes:
  • FIG. 3 is a calculation principle diagram of the L value in the embodiment of the method for automatically sorting ceramic balls according to the present invention.
  • L is the horizontal distance from the center of the ceramic ball to the end of the track (ceramic ball automatic positioning device).
  • the center of the ceramic ball and the highest point and the lowest point of the ceramic ball are on a line. Therefore, the highest point of the ceramic ball is away from the ceramic ball
  • the parallel distance of the end point of the feed track is also L, and the value of L is 1/2 of the diameter D of the ceramic ball. Different ceramic ball sizes correspond to different robot arm suction positions.
  • the size of the ceramic ball to be detected is input into the computer, and the computer automatically calculates and positions the suction position.
  • the robot arm After the L value is input to the microcontroller control system in the robot arm, the robot arm passes the grating line.
  • the displacement sensor performs positioning and realizes a linear displacement action under the action of a servo motor. The positioning is started at the position L from the end of the track and the suction action is started.
  • Step 202 The robot arm is controlled to suck the ceramic ball at a position having a parallel distance L from the end point of the ceramic ball feeding track.
  • Step 203 Control the robot arm to move the ceramic ball to the acquisition area of the image acquisition device.
  • Step 204 Calculate the horizontal distance S of the collection lens of the image acquisition device from the center of the ceramic ball according to the diameter D of the ceramic ball.
  • the step 204 specifically includes:
  • FIG. 4 is a calculation principle diagram of the S value in the embodiment of the ceramic ball automatic sorting method according to the present invention.
  • S is the horizontal distance from the center of the ceramic ball to the acquisition lens (camera) of the image acquisition device.
  • the camera position control system automatically adjusts the distance between the camera and the ceramic ball. The distance between the cameras can be adjusted automatically to ensure that at least 15% of the images are overlapped each time.
  • Step 205 Control the image acquisition device to acquire a primary image of the ceramic ball at a position where the horizontal distance of the acquisition lens of the image acquisition device from the center of the ceramic ball is S.
  • Step 206 Control the robot arm to move the ceramic ball to a ceramic ball gripping and flipping device to flip the ceramic ball 90 degrees in the warp direction, and move the flipped ceramic ball to the image acquisition device. Collection area.
  • Step 207 Control the image acquisition device to acquire a secondary image of the ceramic ball at a position where the acquisition lens of the image acquisition device is at a horizontal distance S from the center of the ceramic ball.
  • Step 208 stitch the primary image of the ceramic ball and the secondary image of the ceramic ball to obtain a full coverage image of the ceramic ball.
  • the step 208 specifically includes:
  • the primary image includes three images of the ceramic ball successively flipped three times in a weft direction at 120 °;
  • the secondary image of the ceramic ball is stitched to obtain a secondary image after the stitching;
  • the secondary image includes three images taken after the ceramic ball is turned 90 degrees in a warp direction and then flipped three times in a parallel direction at 120 ° ;
  • the stitched primary image and the stitched secondary image are stitched to obtain a full coverage image of the ceramic ball.
  • the stitching uses an automatic image stitching technology.
  • By automatically stitching the 6 captured images a 360-degree full coverage of the surface of the ceramic ball can be achieved, and a full coverage image of the ceramic ball can be obtained.
  • the pneumatic slider fixed by the camera and the position sensor cooperate to control the camera to automatically adjust the distance from the detected ceramic ball to ensure that each image is 15 % Overlap.
  • Step 209 Use a threshold segmentation algorithm to identify defects in the full coverage image of the ceramic ball according to the set image feature value threshold and defect threshold to obtain a recognition result.
  • the step 209 specifically includes:
  • a threshold segmentation algorithm is used to identify a full-cover image of the ceramic ball in the segmented background image to obtain a surface image of the ceramic ball after recognition;
  • the recognition result is that the ceramic ball is a qualified ceramic ball.
  • the image feature value threshold is a threshold segmentation algorithm based on the nature of the area, which segment the ball surface image from the background image recognition.
  • the defect threshold is also a threshold segmentation algorithm based on the nature of the area, which removes the ball surface defects from the identified ball surface image. Then perform segmentation recognition.
  • the computer distinguishes the defect from the surface background according to a set threshold, so as to identify the defect.
  • a silicon nitride ceramic ball is used as a test object, and the set image characteristic value threshold is generally selected to be about 15, that is, a region with a gray value of 15 or more is black, and a region less than 15 is white;
  • the set defect threshold is generally taken from 20 to 30.
  • the area smaller than the set defect threshold is the defect area, which is displayed as white, which is greater than or equal to the defect threshold.
  • the area is a matrix and the display remains black.
  • common defects on the surface of ceramic balls include pits, cracks, scratches, snowflakes, and impurities.
  • the thresholds set for different defects are different, but there are large gray differences from the substrate, so manual debugging is required (input different thresholds to determine whether the identified defects meet the requirements; if not, continue to adjust the thresholds, thresholds (It will change according to the change of the scene light intensity and other factors), and finally find the most suitable threshold that can identify most of the defects.
  • the threshold can be manually adjusted according to the accuracy requirements, so as to adjust the acceptability of different defects.
  • the image feature value threshold and the defect threshold can be set to a threshold or a threshold range. Specifically, whether to set the threshold or the threshold range needs to be determined according to factors such as the scene lighting environment.
  • the ceramic ball automatic sorting method disclosed by the present invention uses automatic image stitching technology to automatically stitch the surface images of the balls at various angles to achieve full coverage.
  • a threshold segmentation algorithm is used to cover the full coverage of the ceramic balls. The image is identified and segmented from the background image. In the full coverage image of the identified ceramic ball, the defect is identified using the threshold segmentation algorithm based on the set defect threshold to determine whether the ceramic ball is a defective ball and the storage should be placed in it. Ball device, thereby realizing the automation of ceramic ball defect recognition and sorting, and improving the accuracy and sorting efficiency of ceramic ball defect recognition.
  • the ceramic ball automatic sorting method disclosed by the present invention applies image stitching technology and thresholding technology to the field of ceramic ball detection, realizes the automation of ceramic ball defect recognition and sorting, and improves the accuracy and sorting of ceramic ball defect recognition. s efficiency.
  • computer-based automatic image stitching technology and threshold segmentation algorithm automatic collection and recognition of ceramic ball surface images are achieved. According to the image characteristics corresponding to different defects on the surface of the ceramic ball, that is, different defects are reflected in the image The light and dark levels are different, that is, the gray levels are different.
  • Using the collected images first manually set a threshold value, and after the degree of defect recognition is reached, determine the threshold value, and use the set threshold value for the collected ball surface picture.
  • defects are identified from the background of the surface of the ceramic ball.
  • the identified defective balls and qualified balls are automatically sorted by the computer and placed in different positions. Automatic detection of surface defects of ceramic balls and automatic sorting of ceramic balls are realized.
  • This fully automatic inspection method without human participation significantly improves the efficiency of ceramic ball surface quality inspection, and can accurately and efficiently inspect the surface defects of ceramic balls.
  • the automatic ceramic ball sorting system used by them has moderate price, sensitivity and accuracy. Both are high, and can be used in large quantities for the appearance inspection of ceramic balls
  • the ceramic ball automatic sorting system can realize automatic identification and sorting of defects of ⁇ 3 to ⁇ 10mm ceramic balls.
  • the ceramic ball automatic sorting system was used to test silicon nitride ceramic balls with a diameter of 3.175 mm to 9.525 mm, and the following test results were obtained:
  • the ceramic ball automatic sorting system and method disclosed in the present invention can accurately detect ceramic balls with various defects, and the detection rate reaches 100%.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Pathology (AREA)
  • Immunology (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Chemical & Material Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Biochemistry (AREA)
  • Automation & Control Theory (AREA)
  • Manufacturing & Machinery (AREA)
  • Human Computer Interaction (AREA)
  • Robotics (AREA)
  • Mechanical Engineering (AREA)
  • Length Measuring Devices By Optical Means (AREA)
  • Manipulator (AREA)
  • Sorting Of Articles (AREA)
  • Image Analysis (AREA)

Abstract

本发明公开一种陶瓷球自动分拣系统及方法，该系统自动吸取陶瓷球进给轨道上的陶瓷球进行图像采集，并根据采集的图像信息对陶瓷球是否为缺陷球进行识别，确定陶瓷球放入的储球装置，全程无需人工参与，实现了陶瓷球缺陷识别和分拣的自动化，提高了陶瓷球缺陷识别的准确性和分拣的效率。该方法利用图像自动拼接技术将拍摄的各角度球表面图像进行自动拼接实现全覆盖，根据设定的阈值，利用阈值分割算法对缺陷进行识别，确定陶瓷球是否为缺陷球以及应放入的储球装置，从而实现陶瓷球缺陷识别和分拣的自动化，提高陶瓷球缺陷识别的准确性和分拣的效率。

Description

一种陶瓷球自动分拣系统及方法 技术领域
本发明涉及轴承滚动体表面质量检测技术领域，特别是涉及一种陶瓷球自动分拣系统及方法。
背景技术
随着工业的高速发展，对陶瓷球轴承的要求越来越高。陶瓷球是各种轴承中最为重要的回转体零件，据统计轴承的损坏，90％是回转体零件的损坏，特别是外观质量问题引起的，如凹坑、气孔等表面缺陷的陶瓷球，在高温高压下，它们最先局部发热而损坏。生产中必须对成品球进行表面缺陷检测，对外观有质量问题的球要100％剔除。传统的检测方法是将陶瓷球置于显微镜下放大20倍，再与标准目视对照，这种人工检测方法自动化程度低，误检和漏检率高，生产效率难以提高。因此，对轴承失效影响最大的陶瓷球缺陷控制与自动化识别挑选，己成为十分关注的技术关键问题，本领域迫切需要一种陶瓷球自动检测装置，以正确高效地进行陶瓷球表面缺陷检查，这就需要大批价格适中、灵敏度和准确度都较高的陶瓷球外观检测设备，而此类设备目前还没有应用实例。由于之前一直只是靠人工分拣，自动识别分拣装置还处在研究阶段。因此，本领域技术人员急需一种陶瓷球自动分拣系统及方法以实现陶瓷球缺陷识别和分拣的自动化。
发明内容
本发明要解决的技术问题是提供一种陶瓷球自动分拣系统及方法，从而实现陶瓷球缺陷识别和分拣的自动化，提高陶瓷球缺陷识别的准确性和分拣的效率。
为解决上述技术问题，本发明提供了一种陶瓷球自动分拣系统及方法。
该陶瓷球自动分拣系统，包括：计算机以及与所述计算机分别连接的陶瓷球进给轨道、机器人手臂、图像采集装置、陶瓷球夹取翻转装置、储球装置；
所述计算机控制所述机器人手臂定位在吸取陶瓷球的位置，吸取所述陶瓷球进给轨道上的陶瓷球，并将所述陶瓷球转移至所述图像采集装置的采集区进行图像采集；
所述图像采集装置对所述陶瓷球进行图像采集，并将采集后的图像信息实时发送给所述计算机；
所述计算机控制所述机器人手臂将所述陶瓷球转移至所述陶瓷球夹取翻转装置进行经线方向90度翻转，再对翻转后的所述陶瓷球进行图像采集，并将采集后的图像信息实时发送给所述计算机；
所述计算机根据所述图像采集装置采集的图像信息识别所述陶瓷球放入的储球装置，并发出控制指令控制所述机器人手臂将所述陶瓷球放入所述储球装置实现陶瓷球的自动分拣。
可选的，所述陶瓷球进给轨道包括陶瓷球自动定位装置，所述陶瓷球进给轨道通过所述陶瓷球自动定位装置与所述计算机连接；
所述陶瓷球进给轨道倾斜设置，所述陶瓷球由倾斜的所述陶瓷球进给轨道的高点至低点滚动。
可选的，所述机器人手臂连接有真空气源，所述机器人手臂通过连接的所述真空气源吸取所述陶瓷球；所述机器人手臂连接有转动电机，所述转动电机驱动所述机器人手臂沿轴向转动，所述轴向转动的角度范围为120度。
可选的，所述图像采集装置包括气动滑块、固定在所述气动滑块上的相机、与所述气动滑块连接的相机位置控制系统和位置传感器；所述图像采集装置通过所述相机位置控制系统与所述计算机连接。
可选的，所述储球装置包括缺陷陶瓷球储球装置和合格陶瓷球储球装置；所述缺陷陶瓷球储球装置与所述合格陶瓷球储球装置的位置固定，所述计算机根据所述缺陷陶瓷球储球装置与所述合格陶瓷球储球装置的位置控制所述机器人手臂将所述陶瓷球放入对应的储球装置中。
该陶瓷球自动分拣方法，包括：
获取陶瓷球的直径D，并根据所述陶瓷球的直径D计算陶瓷球的最高点距离陶瓷球进给轨道终点的平行距离L；
控制机器人手臂在距离陶瓷球进给轨道终点的平行距离为L的位置 吸取陶瓷球；
控制所述机器人手臂将所述陶瓷球移动至图像采集装置的采集区；
根据所述陶瓷球的直径D计算图像采集装置的采集镜头距离陶瓷球中心的水平距离S；
控制所述图像采集装置在所述图像采集装置的采集镜头距离所述陶瓷球中心的水平距离为S的位置采集所述陶瓷球的一次图像；
控制所述机器人手臂将所述陶瓷球移动至陶瓷球夹取翻转装置对所述陶瓷球进行经线方向90度翻转，并将翻转后的所述陶瓷球移动至所述图像采集装置的采集区；
控制所述图像采集装置在所述图像采集装置的采集镜头距离所述陶瓷球中心的水平距离为S的位置采集所述陶瓷球的二次图像；
对所述陶瓷球的一次图像以及所述陶瓷球的二次图像进行拼接，得到所述陶瓷球的全覆盖图像；
根据设定的图像特征值阈值和缺陷阈值利用阈值分割算法对所述陶瓷球的全覆盖图像中的缺陷进行识别，得到识别结果；
当识别结果表示所述陶瓷球为缺陷陶瓷球时，控制所述机器人手臂将所述陶瓷球放入缺陷陶瓷球储球装置；
当识别结果表示所述陶瓷球为合格陶瓷球时，控制所述机器人手臂将所述陶瓷球放入合格陶瓷球储球装置。
可选的，所述根据所述陶瓷球的直径D计算陶瓷球的最高点距离陶瓷球进给轨道终点的平行距离L，具体包括：
根据公式L＝D/2计算陶瓷球的最高点距离陶瓷球进给轨道终点的平行距离L。
可选的，所述根据所述陶瓷球的直径D计算图像采集装置前端距离陶瓷球中心的水平距离S，具体包括：
根据公式S＝D/2sin12°计算图像采集装置的采集镜头距离陶瓷球中心的水平距离S。
可选的，所述对所述陶瓷球的一次图像以及所述陶瓷球的二次图像进行拼接，得到所述陶瓷球的全覆盖图像，具体包括：
拼接所述陶瓷球的一次图像，得到拼接后的一次图像；所述一次图像 包括所述陶瓷球在纬线方向连续翻转三次120°拍摄的三幅图像；
拼接所述陶瓷球的二次图像，得到拼接后的二次图像；所述二次图像包括所述陶瓷球在经线方向翻转90度后，再在纬线方向连续翻转三次120°拍摄的三幅图像；
拼接所述拼接后的一次图像和所述拼接后的二次图像，得到所述陶瓷球的全覆盖图像。
可选的，所述根据设定的图像特征值阈值和缺陷阈值利用阈值分割算法对所述陶瓷球的全覆盖图像中的缺陷进行识别，具体包括：
根据所述设定的图像特征值阈值，利用阈值分割算法识别分割背景图中的所述陶瓷球的全覆盖图像，得到识别后的所述陶瓷球的表面图像；
根据所述设定的缺陷阈值，利用阈值分割算法识别分割所述识别后的所述陶瓷球的表面图像的缺陷，得到识别结果；
当所述识别后的所述陶瓷球的表面图像存在缺陷时，确定所述识别结果为所述陶瓷球为缺陷陶瓷球；
当所述识别后的所述陶瓷球的表面图像不存在缺陷时，确定所述识别结果为所述陶瓷球为合格陶瓷球。
与现有技术相比，本发明的有益效果在于：本发明公开的陶瓷球自动分拣系统及方法，系统包括：计算机以及与所述计算机分别连接的陶瓷球进给轨道、机器人手臂、图像采集装置、陶瓷球夹取翻转装置、储球装置，所述计算机控制所述机器人手臂吸取所述陶瓷球进给轨道上的陶瓷球，并将所述陶瓷球转移至所述图像采集装置的采集区进行图像采集，将采集后的图像信息实时发送给所述计算机，所述计算机根据所述图像采集装置采集的图像信息识别所述陶瓷球放入的储球装置，并发出控制指令控制所述机器人手臂将所述陶瓷球放入所述储球装置实现陶瓷球的自动分拣。该系统自动吸取陶瓷球进给轨道上的陶瓷球进行图像采集，并根据采集的图像信息对陶瓷球是否为缺陷球进行识别，确定陶瓷球放入的储球装置，全程无需人工参与，实现了陶瓷球缺陷识别和分拣的自动化，提高了陶瓷球缺陷识别的准确性和分拣的效率。本发明公开的陶瓷球自动分拣方法，利用图像自动拼接技术将拍摄的各角度球表面图像进行自动拼接实现全覆盖，根据设定的图像特征值阈值，利用阈值分割算法将陶瓷球的全覆盖图像从 背景图中识别分割出来，在识别出的陶瓷球的全覆盖图像中，根据设定的缺陷阈值，利用阈值分割算法对缺陷进行识别，确定陶瓷球是否为缺陷球以及应放入的储球装置，从而实现陶瓷球缺陷识别和分拣的自动化，提高陶瓷球缺陷识别的准确性和分拣的效率。
说明书附图
下面结合附图对本发明作进一步说明：
图1为本发明陶瓷球自动分拣系统实施例的结构图；
图2为本发明陶瓷球自动分拣方法实施例的流程图；
图3为本发明陶瓷球自动分拣方法实施例中L值的计算原理图；
图4为本发明陶瓷球自动分拣方法实施例中S值的计算原理图。
具体实施方式
图1为本发明陶瓷球自动分拣系统实施例的结构图。
参见图1，该陶瓷球自动分拣系统，包括：计算机5以及与所述计算机5分别连接的陶瓷球进给轨道1、机器人手臂3、图像采集装置4、陶瓷球夹取翻转装置6、储球装置7。该陶瓷球自动分拣系统固定于设备平台上，整体处于封闭式正压环境(设备内部是有压缩气营造的正压环境，防止外界杂质进入)中，避免了检测过程中外界灰尘杂质的干扰。
所述陶瓷球进给轨道1包括陶瓷球自动定位装置2，所述陶瓷球进给轨道1通过所述陶瓷球自动定位装置2与所述计算机5连接；所述陶瓷球进给轨道1倾斜设置，所述陶瓷球由倾斜的所述陶瓷球进给轨道1的高点至低点滚动，当上一个陶瓷球被取走后，下一个陶瓷球会在重力作用下自己滚动到轨道终点位置，其中，轨道终点位置即为所述陶瓷球自动定位装置2的位置。
所述机器人手臂3由伺服电机、光栅线位移传感器、单片机控制系统、反馈信号处理电路等组成，实现信号反馈、信号处理、运算控制。其中，伺服电机控制所述机器人手臂3实现直线位移动作，光栅线位移传感器进行定位，单片机控制系统的具体型号为STC89C54。所述机器人手臂3可以左右移动、上下移动和轴向120度旋转(沿纬线方向120度旋转)，左右移动通过伺服电机实现，上下移动通过电机轴与丝孔配合实现，所述机器人手臂3连接有转动电机，所述转动电机驱动所述机器人手臂3沿轴向 转动，所述轴向转动的角度范围为120度。所述机器人手臂3连接有真空气源，所述机器人手臂3通过连接的所述真空气源吸取所述陶瓷球，利用负压气压实现陶瓷球的吸取，通过暂停压缩气实现陶瓷球与机器人手臂的分离，避免了机械结构对陶瓷球表面带来的损伤。所述机器人手臂3分拣完成一个陶瓷球后会直接回到一个固定的起始位置再次对下一个陶瓷球进行分拣。
所述图像采集装置4包括气动滑块、固定在所述气动滑块上的相机、与所述气动滑块连接的相机位置控制系统和位置传感器；所述图像采集装置4通过所述相机位置控制系统与所述计算机5连接。所述相机位置控制系统自动选定对应传感器位置，实现相机位置的自动定位、自动调焦和自动调节。所述相机通过螺纹孔与电机轴连接，通过电机轴旋转实现相机的前后移动。不同陶瓷球尺寸对应不同的拍摄位置。
所述储球装置7包括缺陷陶瓷球储球装置和合格陶瓷球储球装置；所述缺陷陶瓷球储球装置与所述合格陶瓷球储球装置的位置固定，所述计算机5根据所述缺陷陶瓷球储球装置与所述合格陶瓷球储球装置的位置控制所述机器人手臂3将所述陶瓷球放入对应的储球装置7中。
所述计算机5控制所述机器人手臂3定位在吸取陶瓷球的位置，吸取所述陶瓷球进给轨道1上的陶瓷球，并将所述陶瓷球转移至所述图像采集装置4的采集区进行图像采集。
所述图像采集装置4对所述陶瓷球进行图像采集，并将采集后的图像信息实时发送给所述计算机5。
所述计算机5控制所述机器人手臂3将所述陶瓷球转移至所述陶瓷球夹取翻转装置6进行经线方向90度翻转，再对翻转后的所述陶瓷球进行图像采集，并将采集后的图像信息实时发送给所述计算机5。
所述计算机5根据所述图像采集装置4采集的图像信息识别所述陶瓷球放入的储球装置7，并发出控制指令控制所述机器人手臂3将所述陶瓷球放入所述储球装置7实现陶瓷球的自动分拣。
本发明公开的陶瓷球自动分拣系统，通过所述计算机控制所述机器人手臂吸取所述陶瓷球进给轨道上的陶瓷球，并将所述陶瓷球转移至所述图像采集装置的采集区进行图像采集，将采集后的图像信息实时发送给所述 计算机，所述计算机根据所述图像采集装置采集的图像信息识别所述陶瓷球放入的储球装置，并发出控制指令控制所述机器人手臂将所述陶瓷球放入所述储球装置实现陶瓷球的自动分拣。本发明公开的陶瓷球自动分拣系统能够自动吸取陶瓷球进给轨道上的陶瓷球进行图像采集，并根据采集的图像信息对陶瓷球是否为缺陷球进行识别，确定陶瓷球放入的储球装置，全程无需人工参与，实现了陶瓷球缺陷识别和分拣的自动化，提高了陶瓷球缺陷识别的准确性和分拣的效率。
图2为本发明陶瓷球自动分拣方法实施例的流程图。
参见图2，该陶瓷球自动分拣方法，包括：
步骤201：获取陶瓷球的直径D，并根据所述陶瓷球的直径D计算陶瓷球的最高点距离陶瓷球进给轨道终点的平行距离L。
所述步骤201具体包括：
根据公式L＝D/2计算陶瓷球的最高点距离陶瓷球进给轨道终点的平行距离L。图3为本发明陶瓷球自动分拣方法实施例中L值的计算原理图。参见图3，L值为陶瓷球中心距离轨道终点(陶瓷球自动定位装置)的水平距离，陶瓷球中心和陶瓷球最高点以及最低点位于一条线上，因此，陶瓷球的最高点距离陶瓷球进给轨道终点的平行距离也为L，而L值为陶瓷球直径D的1/2。不同陶瓷球尺寸对应了不同的机器人手臂吸取位置，将需检测陶瓷球尺寸输入计算机，由计算机自动计算并进行吸取位置定位；L值输入给机器人手臂中的单片机控制系统后，机器人手臂通过光栅线位移传感器进行定位，在伺服电机作用下实现直线位移动作，定位在距离轨道终点L的位置开始吸取动作。
步骤202：控制机器人手臂在距离陶瓷球进给轨道终点的平行距离为L的位置吸取陶瓷球。
步骤203：控制所述机器人手臂将所述陶瓷球移动至图像采集装置的采集区。
步骤204：根据所述陶瓷球的直径D计算图像采集装置的采集镜头距离陶瓷球中心的水平距离S。
所述步骤204具体包括：
根据公式S＝D/2sin12°计算图像采集装置的采集镜头距离陶瓷球中 心的水平距离S。图4为本发明陶瓷球自动分拣方法实施例中S值的计算原理图。参见图4，S为陶瓷球中心距离图像采集装置的采集镜头(相机)的水平距离，S值输入给图像采集装置的相机位置控制系统后，所述相机位置控制系统自动调整相机与陶瓷球之间的距离以实现相机位置的自动调节，保证每次拍摄图像至少15％重合。该实施例中设计的每次图像重合度为30％，S计算公式为：S＝D/2sin12°，其中D为陶瓷球直径。该计算公式所得S值是为了保证每次拍摄图像至少15％重合，以保证图像拼接，得到陶瓷球的全覆盖图像。至少15％重合就要求相机拍摄覆盖角度至少138°，为保证更高的重合度30％，相机拍摄覆盖角度取156°，那么其对角即为24°，S值即为L＝D/2sin12°。
步骤205：控制所述图像采集装置在所述图像采集装置的采集镜头距离所述陶瓷球中心的水平距离为S的位置采集所述陶瓷球的一次图像。
步骤206：控制所述机器人手臂将所述陶瓷球移动至陶瓷球夹取翻转装置对所述陶瓷球进行经线方向90度翻转，并将翻转后的所述陶瓷球移动至所述图像采集装置的采集区。
步骤207：控制所述图像采集装置在所述图像采集装置的采集镜头距离所述陶瓷球中心的水平距离为S的位置采集所述陶瓷球的二次图像。
步骤208：对所述陶瓷球的一次图像以及所述陶瓷球的二次图像进行拼接，得到所述陶瓷球的全覆盖图像。
所述步骤208具体包括：
拼接所述陶瓷球的一次图像，得到拼接后的一次图像；所述一次图像包括所述陶瓷球在纬线方向连续翻转三次120°拍摄的三幅图像；
拼接所述陶瓷球的二次图像，得到拼接后的二次图像；所述二次图像包括所述陶瓷球在经线方向翻转90度后，再在纬线方向连续翻转三次120°拍摄的三幅图像；
拼接所述拼接后的一次图像和所述拼接后的二次图像，得到所述陶瓷球的全覆盖图像。
所述拼接利用的是图像自动拼接技术，通过对拍摄到的6幅图像进行自动拼接，即可实现陶瓷球表面360度全覆盖，得到所述陶瓷球的全覆盖图像。此外，为了保证6次图像拼接的准确性，根据陶瓷球尺寸大小，由 相机所固定在的气动滑块以及位置传感器配合控制相机自动调整与被检测陶瓷球间的距离，以保证每张图像15％以上的重合度。
步骤209：根据设定的图像特征值阈值和缺陷阈值利用阈值分割算法对所述陶瓷球的全覆盖图像中的缺陷进行识别，得到识别结果；
当识别结果表示所述陶瓷球为缺陷陶瓷球时，控制所述机器人手臂将所述陶瓷球放入缺陷陶瓷球储球装置；
当识别结果表示所述陶瓷球为合格陶瓷球时，控制所述机器人手臂将所述陶瓷球放入合格陶瓷球储球装置。
所述步骤209具体包括：
根据所述设定的图像特征值阈值，利用阈值分割算法识别分割背景图中的所述陶瓷球的全覆盖图像，得到识别后的所述陶瓷球的表面图像；
根据所述设定的缺陷阈值，利用阈值分割算法识别分割所述识别后的所述陶瓷球的表面图像的缺陷，得到识别结果；
当所述识别后的所述陶瓷球的表面图像存在缺陷时，确定所述识别结果为所述陶瓷球为缺陷陶瓷球；
当所述识别后的所述陶瓷球的表面图像不存在缺陷时，确定所述识别结果为所述陶瓷球为合格陶瓷球。
图像特征值阈值是基于区域性质的阈值分割算法，将球表面图像从背景图中识别分割出来，缺陷阈值同样是基于区域性质的阈值分割算法，将球表面缺陷从已识别出的球表面图像中再进行分割识别。当球表面存在缺陷时，计算机根据设定的阈值将缺陷与表面背景区分，从而实现对该缺陷进行识别。该实施例中以氮化硅陶瓷球为检验对象，所述设定的图像特征值阈值一般选定为15左右，即灰度值大于等于15的区域为黑色，小于15的区域为白色；所述设定的缺陷阈值一般取20～30，由于氮化硅陶瓷球基体为深色，一般缺陷颜色为浅色，故小于设定的缺陷阈值区域为缺陷区域，显示为白色，大于等于缺陷阈值区域为基体，显示仍为黑色。实际分拣过程中，陶瓷球表面常见缺陷有凹坑、裂纹、划痕、雪花和杂质等，陶瓷球表面有缺陷时，计算机会根据设定好的阈值，自动将缺陷从表面背景中识别并分离出来，不同缺陷所设置的阈值不同，但都与基体存在较大灰度差异，所以需要人工调试(输入不同的阈值，判断所识别的缺陷是否 满足要求；如不满足，继续调整阈值，阈值会根据现场光线强度等因素的变化而变化)，最终找到能够识别大部分缺陷的最适合阈值，可根据精度要求人工调整阈值，从而调整不同缺陷的可接受度。此外，图像特征值阈值和缺陷阈值既可以设置阈值也可以设置阈值范围，具体是要设置阈值还是阈值范围需要根据现场照明环境等因素来确定。
本发明公开的陶瓷球自动分拣方法，利用图像自动拼接技术将拍摄的各角度球表面图像进行自动拼接实现全覆盖，根据设定的图像特征值阈值，利用阈值分割算法将陶瓷球的全覆盖图像从背景图中识别分割出来，在识别出的陶瓷球的全覆盖图像中，根据设定的缺陷阈值，利用阈值分割算法对缺陷进行识别，确定陶瓷球是否为缺陷球以及应放入的储球装置，从而实现陶瓷球缺陷识别和分拣的自动化，提高陶瓷球缺陷识别的准确性和分拣的效率。
本发明公开的陶瓷球自动分拣方法，将图像拼接技术和阈值化技术应用于陶瓷球检测领域，实现了陶瓷球缺陷识别和分拣的自动化，提高了陶瓷球缺陷识别的准确性和分拣的效率。通过陶瓷球自动分拣系统，基于计算机的图像自动拼接技术和阈值分割算法，实现了陶瓷球表面图像自动采集和识别，根据陶瓷球表面不同缺陷所对应的图像特征，即不同缺陷在图像中反映的明暗程度不一样，也就是灰度不一样，利用采集到的图像，首先人为设定阈值，达到缺陷被识别的程度后，确定阈值，将采集到的球表面图片，利用设定好的阈值通过计算机处理，将缺陷从陶瓷球表面背景中识别出来，识别出来的缺陷球和合格球由计算机自动分选并放置于不同位置，实现了陶瓷球表面缺陷的自动检测以及陶瓷球的自动分拣，这种无需人工参与、全自动化的检测方法显著提高了陶瓷球表面质量检测的效率，能够正确高效地进行陶瓷球表面缺陷检查，其采用的陶瓷球自动分拣系统价格适中、灵敏度和准确度都较高，可大批量用于陶瓷球的外观检测。
本实施例中，陶瓷球自动分拣系统可实现φ3～φ10mm陶瓷球的缺陷自动识别和分选。采用该陶瓷球自动分拣系统对Φ3.175mm～Φ9.525mm的氮化硅陶瓷球进行检测，得到如下的检测结果：
Figure PCTCN2018104787-appb-000001
Figure PCTCN2018104787-appb-000002
由上述检测结果可以明显看出，本发明公开的陶瓷球自动分拣系统及方法，能够准确检测出各种缺陷的陶瓷球，检出率达到了100％。
上述实施方式旨在举例说明本发明能够被本领域专业技术人员实现或使用，对上述实施方式进行常规修改对本领域技术人员来说将是显而易见的，故本发明包括但不限于上述实施方式，任何符合本申请文件的描述，符合与本文所公开的原理相同或相似的方法、工艺、产品，均落入本发明的保护范围之内。

Claims (10)

  1. 一种陶瓷球自动分拣系统，其特征在于，包括：计算机以及与所述计算机分别连接的陶瓷球进给轨道、机器人手臂、图像采集装置、陶瓷球夹取翻转装置、储球装置；
    所述计算机控制所述机器人手臂定位在吸取陶瓷球的位置，吸取所述陶瓷球进给轨道上的陶瓷球，并将所述陶瓷球转移至所述图像采集装置的采集区进行图像采集；
    所述图像采集装置对所述陶瓷球进行图像采集，并将采集后的图像信息实时发送给所述计算机；
    所述计算机控制所述机器人手臂将所述陶瓷球转移至所述陶瓷球夹取翻转装置进行经线方向90度翻转，再对翻转后的所述陶瓷球进行图像采集，并将采集后的图像信息实时发送给所述计算机；
    所述计算机根据所述图像采集装置采集的图像信息识别所述陶瓷球放入的储球装置，并发出控制指令控制所述机器人手臂将所述陶瓷球放入所述储球装置实现陶瓷球的自动分拣。
  2. 根据权利要求1所述的陶瓷球自动分拣系统，其特征在于，所述陶瓷球进给轨道包括陶瓷球自动定位装置，所述陶瓷球进给轨道通过所述陶瓷球自动定位装置与所述计算机连接；
    所述陶瓷球进给轨道倾斜设置，所述陶瓷球由倾斜的所述陶瓷球进给轨道的高点至低点滚动。
  3. 根据权利要求1所述的陶瓷球自动分拣系统，其特征在于，所述机器人手臂连接有真空气源，所述机器人手臂通过连接的所述真空气源吸取所述陶瓷球；所述机器人手臂连接有转动电机，所述转动电机驱动所述机器人手臂沿轴向转动，所述轴向转动的角度范围为120度。
  4. 根据权利要求1所述的陶瓷球自动分拣系统，其特征在于，所述图像采集装置包括气动滑块、固定在所述气动滑块上的相机、与所述气动滑块连接的相机位置控制系统和位置传感器；所述图像采集装置通过所述相机位置控制系统与所述计算机连接。
  5. 根据权利要求1所述的陶瓷球自动分拣系统，其特征在于，所述 储球装置包括缺陷陶瓷球储球装置和合格陶瓷球储球装置；所述缺陷陶瓷球储球装置与所述合格陶瓷球储球装置的位置固定，所述计算机根据所述缺陷陶瓷球储球装置与所述合格陶瓷球储球装置的位置控制所述机器人手臂将所述陶瓷球放入对应的储球装置中。
  6. 一种应用于权利要求1所述系统的陶瓷球自动分拣方法，其特征在于，包括：
    获取陶瓷球的直径D，并根据所述陶瓷球的直径D计算陶瓷球的最高点距离陶瓷球进给轨道终点的平行距离L；
    控制机器人手臂在距离陶瓷球进给轨道终点的平行距离为L的位置吸取陶瓷球；
    控制所述机器人手臂将所述陶瓷球移动至图像采集装置的采集区；
    根据所述陶瓷球的直径D计算图像采集装置的采集镜头距离陶瓷球中心的水平距离S；
    控制所述图像采集装置在所述图像采集装置的采集镜头距离所述陶瓷球中心的水平距离为S的位置采集所述陶瓷球的一次图像；
    控制所述机器人手臂将所述陶瓷球移动至陶瓷球夹取翻转装置对所述陶瓷球进行经线方向90度翻转，并将翻转后的所述陶瓷球移动至所述图像采集装置的采集区；
    控制所述图像采集装置在所述图像采集装置的采集镜头距离所述陶瓷球中心的水平距离为S的位置采集所述陶瓷球的二次图像；
    对所述陶瓷球的一次图像以及所述陶瓷球的二次图像进行拼接，得到所述陶瓷球的全覆盖图像；
    根据设定的图像特征值阈值和缺陷阈值利用阈值分割算法对所述陶瓷球的全覆盖图像中的缺陷进行识别，得到识别结果；
    当识别结果表示所述陶瓷球为缺陷陶瓷球时，控制所述机器人手臂将所述陶瓷球放入缺陷陶瓷球储球装置；
    当识别结果表示所述陶瓷球为合格陶瓷球时，控制所述机器人手臂将所述陶瓷球放入合格陶瓷球储球装置。
  7. 根据权利要求6所述的陶瓷球自动分拣方法，其特征在于，所述根据所述陶瓷球的直径D计算陶瓷球的最高点距离陶瓷球进给轨道终点 的平行距离L，具体包括：
    根据公式L＝D/2计算陶瓷球的最高点距离陶瓷球进给轨道终点的平行距离L。
  8. 根据权利要求6所述的陶瓷球自动分拣方法，其特征在于，根据所述陶瓷球的直径D计算图像采集装置的采集镜头距离陶瓷球中心的水平距离S，具体包括：
    根据公式S＝D/2sin12°计算图像采集装置的采集镜头距离陶瓷球中心的水平距离S。
  9. 根据权利要求6所述的陶瓷球自动分拣方法，其特征在于，所述对所述陶瓷球的一次图像以及所述陶瓷球的二次图像进行拼接，得到所述陶瓷球的全覆盖图像，具体包括：
    拼接所述陶瓷球的一次图像，得到拼接后的一次图像；所述一次图像包括所述陶瓷球在纬线方向连续翻转三次120°拍摄的三幅图像；
    拼接所述陶瓷球的二次图像，得到拼接后的二次图像；所述二次图像包括所述陶瓷球在经线方向翻转90度后，再在纬线方向连续翻转三次120°拍摄的三幅图像；
    拼接所述拼接后的一次图像和所述拼接后的二次图像，得到所述陶瓷球的全覆盖图像。
  10. 根据权利要求6所述的陶瓷球自动分拣方法，其特征在于，所述根据设定的图像特征值阈值和缺陷阈值利用阈值分割算法对所述陶瓷球的全覆盖图像中的缺陷进行识别，得到识别结果，具体包括：
    根据所述设定的图像特征值阈值，利用阈值分割算法识别分割背景图中的所述陶瓷球的全覆盖图像，得到识别后的所述陶瓷球的表面图像；
    根据所述设定的缺陷阈值，利用阈值分割算法识别分割所述识别后的所述陶瓷球的表面图像的缺陷，得到识别结果；
    当所述识别后的所述陶瓷球的表面图像存在缺陷时，确定所述识别结果为所述陶瓷球为缺陷陶瓷球；
    当所述识别后的所述陶瓷球的表面图像不存在缺陷时，确定所述识别结果为所述陶瓷球为合格陶瓷球。
PCT/CN2018/104787 2018-09-10 2018-09-10 一种陶瓷球自动分拣系统及方法 WO2020051728A1 (zh)

Priority Applications (4)

Application Number Priority Date Filing Date Title
JP2019535303A JP2021502229A (ja) 2018-09-10 2018-09-10 セラミックボールの自動選別システム及び方法
EP18884856.8A EP3650129A4 (en) 2018-09-10 2018-09-10 SYSTEM AND PROCEDURE FOR AUTOMATIC SORTING OF CERAMIC BALLS
US16/479,019 US11327028B2 (en) 2018-09-10 2018-09-10 Ceramic ball automatic sorting system and method
PCT/CN2018/104787 WO2020051728A1 (zh) 2018-09-10 2018-09-10 一种陶瓷球自动分拣系统及方法

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2018/104787 WO2020051728A1 (zh) 2018-09-10 2018-09-10 一种陶瓷球自动分拣系统及方法

Publications (1)

Publication Number Publication Date
WO2020051728A1 true WO2020051728A1 (zh) 2020-03-19

Family

ID=69776449

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/104787 WO2020051728A1 (zh) 2018-09-10 2018-09-10 一种陶瓷球自动分拣系统及方法

Country Status (4)

Country Link
US (1) US11327028B2 (zh)
EP (1) EP3650129A4 (zh)
JP (1) JP2021502229A (zh)
WO (1) WO2020051728A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114405872A (zh) * 2022-01-21 2022-04-29 长春理工大学 钢球直径高精度测量与分组系统
US11635346B1 (en) 2022-02-01 2023-04-25 General Electric Company Bearing element inspection system and method

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2005349239A (ja) * 2004-06-08 2005-12-22 Hitachi Metals Ltd 異常球検出装置
CN101131469A (zh) * 2007-09-19 2008-02-27 哈尔滨工业大学 陶瓷球表面图像自动采集装置
CN103990603A (zh) * 2014-06-05 2014-08-20 苏州大学 一种球体表面质量检测方法及装置
CN204523609U (zh) * 2015-03-09 2015-08-05 河南工业职业技术学院 一种plc控制机械臂分拣装置

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6630998B1 (en) * 1998-08-13 2003-10-07 Acushnet Company Apparatus and method for automated game ball inspection
JP5206936B2 (ja) * 2007-11-26 2013-06-12 日本電気硝子株式会社 球状体の直径不同の評価方法、選別方法および選別装置
US9719942B2 (en) * 2010-01-07 2017-08-01 Nikkato Corporation Sintered ceramic and ceramic sphere
CN102735693B (zh) * 2012-05-03 2014-04-30 天津大学 基于视觉的钢球表面缺陷检测装置及检测方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2005349239A (ja) * 2004-06-08 2005-12-22 Hitachi Metals Ltd 異常球検出装置
CN101131469A (zh) * 2007-09-19 2008-02-27 哈尔滨工业大学 陶瓷球表面图像自动采集装置
CN103990603A (zh) * 2014-06-05 2014-08-20 苏州大学 一种球体表面质量检测方法及装置
CN204523609U (zh) * 2015-03-09 2015-08-05 河南工业职业技术学院 一种plc控制机械臂分拣装置

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3650129A4 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114405872A (zh) * 2022-01-21 2022-04-29 长春理工大学 钢球直径高精度测量与分组系统
CN114405872B (zh) * 2022-01-21 2023-09-01 长春理工大学 钢球直径高精度测量与分组系统
US11635346B1 (en) 2022-02-01 2023-04-25 General Electric Company Bearing element inspection system and method

Also Published As

Publication number Publication date
EP3650129A1 (en) 2020-05-13
JP2021502229A (ja) 2021-01-28
US20210187556A1 (en) 2021-06-24
EP3650129A4 (en) 2020-11-25
US11327028B2 (en) 2022-05-10

Similar Documents

Publication Publication Date Title
JP2018017639A (ja) 表面欠陥検査方法及び表面欠陥検査装置
CN109433627B (zh) 基于机器视觉处理的机械手物流分拣系统及其工作方法
CN103090804A (zh) 成品磁环图像自动检测系统及检测方法
WO2020051728A1 (zh) 一种陶瓷球自动分拣系统及方法
KR100938318B1 (ko) 직물 검사방법과 그 장치
CN111307812A (zh) 基于机器视觉的焊点外观检测方法
CN114113129B (zh) 一种镜片微小缺陷识别抓取系统及方法
CN113916127A (zh) 一种气门导管成品外观视觉检测系统及检测方法
CN109013388B (zh) 一种陶瓷球自动分拣系统及方法
CN112705473A (zh) 一种基于机器视觉的多角度扫描检测装置
KR102646891B1 (ko) 렌즈유닛 검사장치
CN111198190A (zh) 光学检测系统
US20180176549A1 (en) Multi-view-angle image capturing device and multi-view-angle image inspection apparatus using the same
CN114674830A (zh) 一种高速生产线上瓶盖瑕疵检测模组
JP2800932B2 (ja) 円筒ころ軸受の検査装置
TWI705242B (zh) 光學檢測設備及方法
JP2002207011A (ja) ワークの外観検査装置
CN217237796U (zh) 一种透明产品检测设备
TW201814247A (zh) 一種光學檢測裝置
CN220022902U (zh) 一种芯片检测装置
CN110133000A (zh) 一种全自动显微镜视觉成像表面检测机
CN114136987B (zh) 一种镜片形变缺陷检测装置及方法
TWI833471B (zh) 瑕疵檢測方法及其裝置
JPS63241408A (ja) 陶磁器製品のクラツク検査装置
CN220508800U (zh) 一种芯片检测装置

Legal Events

Date Code Title Description
ENP Entry into the national phase

Ref document number: 2019535303

Country of ref document: JP

Kind code of ref document: A

ENP Entry into the national phase

Ref document number: 2018884856

Country of ref document: EP

Effective date: 20190614

ENP Entry into the national phase

Ref document number: 2018884856

Country of ref document: EP

Effective date: 20190614

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

Ref document number: 18884856

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE