WO2018214661A1 - 一种织物疵点自动检测方法、系统和计算机可读存储介质 - Google Patents

一种织物疵点自动检测方法、系统和计算机可读存储介质 Download PDF

Info

Publication number
WO2018214661A1
WO2018214661A1 PCT/CN2018/082661 CN2018082661W WO2018214661A1 WO 2018214661 A1 WO2018214661 A1 WO 2018214661A1 CN 2018082661 W CN2018082661 W CN 2018082661W WO 2018214661 A1 WO2018214661 A1 WO 2018214661A1
Authority
WO
WIPO (PCT)
Prior art keywords
defect
fabric
marking
detection
defects
Prior art date
Application number
PCT/CN2018/082661
Other languages
English (en)
French (fr)
Inventor
黄伟强
Original Assignee
香港纺织及成衣研发中心
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 香港纺织及成衣研发中心 filed Critical 香港纺织及成衣研发中心
Priority to US16/611,183 priority Critical patent/US10942133B2/en
Publication of WO2018214661A1 publication Critical patent/WO2018214661A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84Systems specially adapted for particular applications
    • G01N21/88Investigating the presence of flaws or contamination
    • G01N21/89Investigating the presence of flaws or contamination in moving material, e.g. running paper or textiles
    • G01N21/892Investigating the presence of flaws or contamination in moving material, e.g. running paper or textiles characterised by the flaw, defect or object feature examined
    • G01N21/898Irregularities in textured or patterned surfaces, e.g. textiles, wood
    • G01N21/8983Irregularities in textured or patterned surfaces, e.g. textiles, wood for testing textile webs, i.e. woven material
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84Systems specially adapted for particular applications
    • G01N21/88Investigating the presence of flaws or contamination
    • G01N21/89Investigating the presence of flaws or contamination in moving material, e.g. running paper or textiles
    • G01N21/892Investigating the presence of flaws or contamination in moving material, e.g. running paper or textiles characterised by the flaw, defect or object feature examined
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84Systems specially adapted for particular applications
    • G01N21/88Investigating the presence of flaws or contamination
    • G01N21/8851Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N13/20Image signal generators
    • H04N13/204Image signal generators using stereoscopic image cameras
    • H04N13/239Image signal generators using stereoscopic image cameras using two 2D image sensors having a relative position equal to or related to the interocular distance
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N13/20Image signal generators
    • H04N13/204Image signal generators using stereoscopic image cameras
    • H04N13/254Image signal generators using stereoscopic image cameras in combination with electromagnetic radiation sources for illuminating objects
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/60Control of cameras or camera modules
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84Systems specially adapted for particular applications
    • G01N21/88Investigating the presence of flaws or contamination
    • G01N21/8851Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
    • G01N2021/8854Grading and classifying of flaws
    • G01N2021/8861Determining coordinates of flaws
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84Systems specially adapted for particular applications
    • G01N21/88Investigating the presence of flaws or contamination
    • G01N21/8851Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
    • G01N2021/8854Grading and classifying of flaws
    • G01N2021/888Marking defects
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84Systems specially adapted for particular applications
    • G01N21/88Investigating the presence of flaws or contamination
    • G01N21/8851Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
    • G01N2021/8887Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques

Definitions

  • the present invention relates to the textile industry detection technology, and in particular to a fabric defect automatic detection method, system and computer readable storage medium.
  • the quality of incoming fabrics is directly related to the subsequent cutting process and garment quality.
  • garment factories currently rely on the naked eye of the staff to inspect the incoming fabric. Due to the large amount of incoming materials, garment factories generally assemble at least one fabric inspection machine, and one fabric inspection machine requires at least one operator. During the inspection process, once the cloth defects are found, the operator will stop the machine for confirmation and make a paper record and a defect mark. Due to the limitations of the human eye, the speed of manual inspection is generally slow. If the speed of the inspection machine is too fast, more missed inspections will occur, which will affect the accuracy of the objective scoring and rating of the cloth. On the other hand, due to the repeatability and strength of the cloth inspection operation, the long-term inspection process will cause the operator's visual fatigue, resulting in more missed detection, and the operator's working time is longer, and the missing inspection occurs. The probability will be greater.
  • the present invention provides a fabric defect automatic detecting method, system and computer readable storage medium.
  • a method for automatically detecting a fabric defect comprising: preprocessing a fabric area to be detected during a fabric transmission process by a noise removing device, eliminating noise interference during the detecting process;
  • the rotation information of the roller motor triggers the image acquisition of the fabric to be detected by the camera; and automatically identifies the defect on the fabric to be detected based on the acquired image.
  • the rotational speed of the roller motor of the fabric inspection machine can be adjusted to reduce the fabric transport speed, and the detected defect position on the fabric is marked by the defect marking device.
  • the rotation speed of the roller motor of the fabric inspection machine is increased to restore the conveying speed of the fabric.
  • Automatically identifying the defects on the fabric to be inspected may also include performing a point image stitching on the defects that span two or more adjacent images.
  • the method further comprises: obtaining a global coordinate position of the defect in the overall fabric, comprising: obtaining a local coordinate of the defect on the sampled picture, a size of the defect area, a length of the defect, and a direction of the defect, according to the roller motor of the inspection machine
  • the rotation information obtains the global ordinate of the defect in the overall fabric, and obtains the global abscissa of the defect in the overall fabric according to the number of the camera corresponding to the sampled image, thereby accurately marking the position of the defect according to the global coordinate position of the defect in the overall fabric.
  • Marking the defect position may further include: controlling the lateral translation of the defect marking device to the top of the defect to perform the defect marking according to the global coordinate position of the defect in the overall fabric, and the longitudinal movement speed of the defect marking device matches the fabric transmission speed.
  • marking the defect position further comprises: controlling the defect marking device to locate the edge of the fabric corresponding to the identified defect according to the global coordinate position of the defect in the overall fabric, and marking the edge of the fabric without Translating the defect marking device.
  • the noise removing device used in the method of the present invention may include a blowing device and a cleaning device.
  • the blowing speed and strength of the blowing device can be adjusted according to the characteristics of the fabric to be tested, including but not limited to fabric color, fabric pattern, fabric fiber degree and fabric structure; and the cleaning device can be controlled by adjusting the cleaning device.
  • the degree of tightness or gap between the cleaning surface of the cleaning device and the surface of the fabric is included, and the cleaning surface includes, but is not limited to, a fibrous structure, a brush, a sponge, and the like.
  • the method of the present invention may further include classifying and scoring the defects according to the characteristics of the defects, including but not limited to the length of the defect, the size of the defect, the direction of the defect, the density of the unit code length of the defect, and the like.
  • the identified defect information of the fabric may be stored, and the stored defect information includes but is not limited to the defect level, the defect point, the local coordinates of the defect, the global coordinate of the defect, the storage path of the original image of the defect, the storage path of the defect detection image, and the overall detection of the fabric. Score, fabric detection code length, and unit code length average score.
  • test report of the fabric to be inspected can be automatically generated and output.
  • a fabric defect automatic detecting system comprising: a noise removing device configured to pre-process a fabric area to be detected during a fabric transfer process, to eliminate noise interference during the detecting process; and the camera is configured to The image of the fabric to be inspected is triggered according to the rotation information of the roller motor of the fabric inspection machine; and the defect recognition device is configured to automatically identify the defect on the fabric to be detected based on the acquired image.
  • the system of the present invention may further comprise a defect marking device, wherein when the defect recognition device recognizes the defect, the rotational speed of the roller motor of the fabric inspection machine is lowered to reduce the fabric transmission speed, and the fabric is marked by the defect marking device The detected defect position is marked. When the defect position mark is completed, the rotational speed of the roller motor of the cloth inspection machine is increased to restore the high speed transmission of the fabric.
  • the defect recognition device can also be configured to perform a point image stitching on a defect that spans two or more adjacent images.
  • the defect recognition device is further configured to obtain the global coordinate position of the defect in the overall fabric, by obtaining the local coordinates of the defect on the sampled image, the size of the defect region, the length of the defect, and the direction of the defect, according to the inspection machine.
  • the rotation information of the roller motor obtains the global ordinate of the defect in the overall fabric, and obtains the global abscissa of the defect in the overall fabric according to the number of the camera corresponding to the sample picture, and the defect marking device according to the global coordinate position of the defect in the overall fabric Accurately mark the location of the defect.
  • the defect marking device can also perform lateral marking according to the lateral position of the defect in the global coordinate position in the overall fabric to the top of the defect, and the longitudinal moving speed of the defect marking device matches the fabric conveying speed.
  • the defect marking device can also position the fabric edge corresponding to the identified defect according to the global coordinate position of the defect in the overall fabric, and mark the edge of the fabric while the defect marking device does not shift.
  • the noise removing device in the system of the present invention may include a blowing device and a cleaning device.
  • the air blowing speed and strength of the blowing device can be adjusted according to the characteristics of the fabric to be detected, and the fabric characteristics include, but are not limited to, fabric color, fabric pattern, fabric fiber degree and fabric structure; and the cleaning surface of the cleaning device is The degree of tightness of the surface contact of the fabric or the height of the gap can be controlled by adjusting the cleaning device, including but not limited to fibrous tissue, a brush, a sponge, and the like.
  • the system of the present invention may further comprise a defect grading device configured to classify and score the defects according to the defect characteristics, including but not limited to the defect length, the defect region size, the defect direction, the unit code length density at which the defect occurs, and the like.
  • a defect grading device configured to classify and score the defects according to the defect characteristics, including but not limited to the defect length, the defect region size, the defect direction, the unit code length density at which the defect occurs, and the like.
  • the system of the present invention may further comprise a database configured to store information on the identified defects of the fabric, the stored defect information including but not limited to a defect level, a defect point, a local point of the defect, a global coordinate of the defect, and a storage path of the original image of the defect.
  • the system can also include a detection reporting device configured to automatically generate and output a test report of the fabric to be inspected.
  • a fabric defect automatic detecting apparatus comprising a memory and a controller, wherein the control is configured to perform the method as described above.
  • a computer readable storage medium having stored thereon a computer program, which when executed, performs the method as described above.
  • the system of the invention can effectively solve the defects of low accuracy, high detection rate, duplication of work and high strength of the artificial fabric defect detection.
  • the method of the invention can effectively improve the accuracy of the defect detection, ensure that the detection operation can be continued for a long time, and the automatically generated defect detection report can be used as the basis for subsequent fabric cutting.
  • FIG. 1 is a flow chart of a method for automatically detecting a fabric defect according to an embodiment of the present invention
  • FIG. 2 is a flow chart of a method for automatically detecting a fabric defect according to another embodiment of the present invention.
  • FIG. 3 is a schematic view showing the overall arrangement of a fabric defect automatic detecting system according to an embodiment of the present invention.
  • FIG. 4 is a schematic illustration of the distribution of cloth regions processed by various subsystems in a fabric defect automatic detection system in accordance with an embodiment of the present invention.
  • the method mainly comprises: S101, preprocessing the fabric area to be detected during the fabric transmission process by the noise removing device, eliminating noise interference during the detecting process; next, in S102, according to the roller motor of the fabric inspection machine
  • the rotation information triggers the image acquisition of the fabric to be detected by the camera; then, at S103, based on the acquired image, the defect on the fabric to be detected is automatically identified. Based on the above method, the automatic detection and recognition of fabric defects can be completed.
  • FIG. 2 is a flow chart showing a method of automatically detecting a fabric defect according to another embodiment of the present invention.
  • the method of this embodiment includes more improved features, and provides a more complete automatic detection method for defects.
  • the method shown in FIG. 2 includes: S101, preprocessing the fabric area to be detected during the fabric transmission process by the noise removing device, eliminating noise interference during the detecting process; next, in S102, according to the roller motor of the fabric inspection machine Rotating the information triggers the camera to perform image acquisition on the fabric to be inspected; then, in S103, based on the acquired image, the defect on the fabric to be detected is automatically identified.
  • the defect When the defect is recognized, at S204, it is determined whether the defect crosses two or more adjacent images, and if the determination result is YES, the image is spliced to the defect of the two or more adjacent images in S206. If the result of the determination is no, proceed directly to S205, detect the size of the defect, classify, score, etc. the defect. It is worth mentioning that the classification and scoring of the defects can be performed according to the characteristics of the defects, including but not limited to the length of the defect, the size of the defect, the direction of the defect, the density of the unit code length of the defect, etc., measuring the size of the defect and evaluating each The detected defect ranking may also include determining the type of defect.
  • it may further include, at S207, obtaining a global coordinate position of the defect in the entire fabric.
  • obtaining a global coordinate position of the defect in the entire fabric For example, by obtaining the local coordinates of the defect on the sampled image, the size of the defect area, the length of the defect, and the direction of the defect, the global vertical coordinate of the defect in the overall fabric is obtained according to the rotation information of the roller motor of the inspection machine, according to The number of the camera corresponding to the sampled image obtains the global abscissa of the defect in the overall fabric, thereby obtaining the global coordinate position of the defect, thereby accurately marking the position of the defect according to the global coordinate position of the defect in the overall fabric.
  • the rotation speed of the roller motor of the fabric inspection machine is adjusted to reduce the fabric conveying speed, the detected defect position on the fabric is marked by the defect marking device, and when the defect position marking is completed, the inspection machine is improved.
  • the speed of the roller motor to restore the speed of the fabric.
  • Marking the defect position can control the lateral marking of the defect marking device to the top of the defect point according to the global coordinate position of the defect point in the overall fabric, and the longitudinal movement speed of the defect marking device matches the fabric transmission speed.
  • marking the defect position may also control the defect marking device to locate the edge of the fabric corresponding to the identified defect according to the global coordinate position of the defect in the overall fabric, and mark the edge of the fabric in the alternative. In the middle, the defect marking device is not translated.
  • the noise removing device used in the method of the present invention may include a blowing device and a cleaning device.
  • the air blowing speed and strength of the blowing device can be adjusted according to the characteristics of the fabric to be tested, including but not limited to fabric color, fabric pattern, fabric fiber degree and fabric structure; and the cleaning device can be controlled by adjusting the cleaning device.
  • the degree of tightness or clearance of the cleaning surface in contact with the surface of the fabric includes, but is not limited to, fibrous structure, brush, sponge, and the like.
  • the identified defect information of the fabric may be stored, and the stored defect information includes but is not limited to the defect level, the defect point, the defect local coordinate, the defect global coordinate, the storage path of the defect original image, the storage path of the defect detection image, The overall detection score of the fabric, the length of the fabric detection code, and the average score of the unit code length, etc.
  • the detection report of the fabric to be inspected can be automatically generated and output.
  • FIGS. 1 and 2 are merely an exemplary embodiment, and the above steps are not necessarily performed in order, and steps may be added according to actual needs or unnecessary steps may be reduced from the above steps.
  • FIG. 3 is a schematic view showing the overall arrangement of a fabric defect automatic detecting system according to an embodiment of the present invention, the system comprising: a noise removing device 9 configured to pretreat a fabric area to be detected during fabric transfer, and eliminate the detecting process Noise interference in the camera; the camera 6 may include N cameras (C1 ... CN), wherein N is greater than or equal to 1, configured to be triggered according to the rotation information of the motor 5 of the roller 8 of the fabric inspection machine to perform an image on the fabric to be inspected And the defect recognition device 1 is configured to automatically identify a defect on the fabric to be detected based on the acquired image.
  • a noise removing device 9 configured to pretreat a fabric area to be detected during fabric transfer, and eliminate the detecting process Noise interference in the camera
  • the camera 6 may include N cameras (C1 ... CN), wherein N is greater than or equal to 1, configured to be triggered according to the rotation information of the motor 5 of the roller 8 of the fabric inspection machine to perform an image on the fabric to be inspected
  • the system of the present invention may further comprise a defect marking device 7 which, when the defect recognition device 1 recognizes the defect, reduces the rotational speed of the motor 5 of the roller 8 of the fabric inspection machine to reduce the speed of the fabric transfer, by means of the defect marking device 7 on the fabric
  • the detected defect position is marked.
  • the rotation speed of the motor 5 of the roller 8 of the cloth inspection machine is increased to restore the conveying speed of the fabric.
  • the system can further perform splicing of the image of the smear point of the two or more adjacent images, and obtain the global coordinate position of the ⁇ point in the overall fabric, by obtaining the local coordinates of the ⁇ point on the sampled image, the size of the defect area, and the defect point.
  • the length and the direction of the defect according to the rotation information of the roller motor 5 of the fabric inspection machine, obtain the global ordinate of the defect in the overall fabric, and obtain the global abscissa of the defect in the overall fabric according to the number of the camera corresponding to the sampled image,
  • the defect marking device accurately marks the position of the defect according to the global coordinate position of the defect in the overall fabric.
  • the defect marking device 7 can be laterally translated to the point above the defect according to the global coordinate position of the defect in the overall fabric, and the longitudinal movement speed of the defect marking device 7 matches the fabric transmission speed.
  • the lateral direction means a direction perpendicular to the moving direction of the fabric on the plane of the fabric
  • the longitudinal direction means the moving direction of the fabric.
  • the defect marking device 7 can also position the edge of the fabric corresponding to the identified defect according to the global coordinate position of the defect in the overall fabric, and mark the edge of the fabric while the defect marking device 7 does not translate.
  • the noise removing device 9 in the system of the present invention may include a blowing device and a cleaning device (not shown).
  • the air blowing speed and strength of the blowing device can be adjusted according to the characteristics of the fabric to be tested, including but not limited to fabric color, fabric pattern, fabric fiber degree and fabric structure; and the cleaning surface of the cleaning device is in contact with the fabric surface.
  • the degree of tightness or the height of the gap can be controlled by adjusting the cleaning device, including but not limited to fibrous tissue, brush, sponge, and the like.
  • the system may further comprise a defect grading device configured to classify and score the defects according to the defect characteristics, including but not limited to the length of the defect, the size of the defect, the direction of the defect, the density of the unit code length of the defect, and the like.
  • a defect grading device configured to classify and score the defects according to the defect characteristics, including but not limited to the length of the defect, the size of the defect, the direction of the defect, the density of the unit code length of the defect, and the like.
  • the system of the present invention may further comprise a database configured to store information on the identified defects of the fabric, the stored defect information including but not limited to a defect level, a defect point, a local point of the defect, a global coordinate of the defect, and a storage path of the original image of the defect.
  • the system can also include a detection reporting device configured to automatically generate and output a test report of the fabric to be inspected.
  • a detection reporting device configured to automatically generate and output a test report of the fabric to be inspected.
  • the defect recognition device, the defect classification device, the database, the detection report device, and the like can be realized by a computer, a central controller, a processor, a microprocessor, or the like.
  • the fabric defect automatic detection system is divided into three parts according to the function: fabric noise removal subsystem, fabric defect detection subsystem and defect automatic marking subsystem.
  • the fabric noise removal subsystem includes a blowing device and a cleaning device.
  • the fabric noise removal subsystem can be controlled by the central control computer according to the detection process, and can also be controlled to start and stop. Through the fabric noise removal subsystem, secondary pollution such as dust, yarn breakage, fibers, and other external interference factors attached to the fabric surface can be eliminated, thereby improving the accuracy of detection.
  • the defect detection subsystem includes one or more cameras 6, an encoder 4, a flip flop 3, and a central control computer 1.
  • the defect detection subsystem is an important part of the system of the invention, and is divided into two parts: defect detection and auxiliary device control.
  • the defect detection subsystem, the fabric noise removal subsystem and the defect automatic marking subsystem work together and are controlled by the central control computer.
  • the defect automatic marking subsystem includes a frequency converter 2, a defect marking device 7, and a translation motor.
  • the defect automatic marking subsystem accurately marks the position of the defect on the fabric according to the type and coordinates of the defect detected by the defect detection subsystem.
  • the fabric defect automatic detecting system of the present invention When the fabric defect automatic detecting system of the present invention is in operation, as shown in FIG. 3, the fabric is conveyed by the fabric inspection drive wheel 8, and the central control computer 1 activates the noise removal subsystem and the camera system 6.
  • the noise removal subsystem 9 can initiate and stop separately during the fabric drive process, simply by ensuring that the detected fabric area passes through the noise removal subsystem.
  • the noise removal subsystem 9 removes interference factors attached to the surface of the fabric by means of a blowing and cleaning device, including but not limited to dust, yarn breakage, fibers, etc., to ensure that the surface of the fabric is clean and authentic.
  • the blowing device can adjust the output wind speed in combination with the characteristics of the detecting fabric.
  • the cleaning device can be adjusted to control the degree of tightness (or gap height) in contact with the surface of the fabric.
  • the encoder 4 samples the distance of the rotation of the inspection machine roller motor 5 in real time, and the distance recorded by the encoder 4 includes the total accumulated distance and the single cumulative distance. When the encoder 4's single cumulative distance reaches the set trigger distance value, the encoder 4 sends a signal to the trigger 3, which triggers the camera 6 to capture the fabric picture. After the encoder 4 outputs the signal, the single cumulative distance is reset to zero, and the total accumulated distance remains unchanged.
  • the number N of camera 6 activations can be set according to the width of the fabric, and it is necessary to ensure that the combination of pictures taken by the activated camera covers the overall width of the fabric.
  • the image of the fabric captured by the camera 6 is transmitted to the defect detection subsystem of the central control computer 1.
  • the defect detection subsystem first selects the corresponding defect detection algorithm according to the set fabric type.
  • the defect detection algorithm is pre-integrated in the defect detection subsystem.
  • the fabric types include but are not limited to fabric color, fabric pattern, fabric fiber degree, fabric structure, etc. information.
  • the sampled picture is analyzed for defect detection according to the selected defect detection algorithm. If there is a defect in the sampled picture, the coordinates of the defect in the sampled picture, that is, the local coordinate (x, y), are acquired, and the sampled original picture and the detected defect picture are stored.
  • the points are graded and scored, and the subsequent process is automatically detected; if there is no defect, the sampled picture Do not perform any subsequent operations.
  • the defect type can be evaluated and scored using a four-point or ten-point standard, and the local coordinates (x, y) of the defect are converted to global coordinates (X, Y) on the overall fabric. Based on the scores of individual defects, statistical information such as the total score of the fabric and the average score of the code length is generated.
  • the database is used to store information for each detection point of the fabric, including but not limited to the defect level, the defect point, the local coordinates of the defect, the global coordinates of the defect, the storage path of the defect, the storage path of the defect detection map, the overall detection score of the fabric, and the fabric. Check the code length and the average code length per unit code length. A fabric inspection report is generated based on the stored fabric inspection information, and a paper report can be printed by the printer.
  • the central control computer Based on the detection result of the automatic detection subsystem of the defect, if the fabric defect is detected during the detection process, the central control computer sets the low frequency value to the frequency converter 2, and reduces the rotation speed of the roller motor of the inspection machine, thereby reducing the fabric transmission speed and ensuring defects.
  • the marker device translates above the defect and then performs a marker marking operation.
  • the defect marking device can also perform the marking without marking at the edge of the fabric.
  • the central control computer sets the high frequency value to the inverter, and increases the rotation speed of the roller motor of the inspection machine to ensure the fabric detection speed.
  • the cloth inspection machine roller motor 5 When the defect detection system of this example starts to operate, first, the cloth inspection machine roller motor 5 is activated, the cloth inspection machine roller 8 is rotated, and the fabric is conveyed by the roller 8.
  • the distance at which the cloth inspection machine roller motor 5 rotates is sampled in real time by the encoder 4.
  • the encoder 4 When the distance accumulated by the encoder 4 reaches the set trigger distance h, the encoder 4 outputs a pulse signal to the trigger 3, and the single cumulative distance of the encoder 4 is reset to zero, and the overall cumulative distance remains unchanged.
  • the trigger 3 After receiving the pulse signal, the trigger 3 triggers the camera 6 to capture the fabric image by outputting a high level.
  • the noise removing device 9 After the cloth inspection machine roller motor 5 is started, the noise removing device 9 will be activated.
  • the output wind speed of the blowing device 9_1 can be adjusted according to the characteristics of the fabric.
  • the blowing device 9_1 can remove image interference factors such as dust, fibers, and yarn breakage attached to the surface of the fabric.
  • the cleaning device 9_2 can further remove the interference factor of strong adhesion.
  • the operation of the blowing device 9_1 and the cleaning device 9_2 can provide a basis for the camera to capture high-quality real images, thereby improving the accuracy of subsequent defect recognition.
  • the camera 6 can be composed of one or more cameras, ensuring that the camera sample picture combination can cover the overall width of the fabric.
  • the number of startups of the camera 6 can be preset according to the actual fabric width, or can be automatically selected according to the image information captured by the camera, that is, if there is fabric information in the image captured by the camera, it is activated, otherwise the camera is paused for image capture.
  • the image information captured by the camera 6 is transmitted to the central control computer 1 through the camera-specific transmission protocol.
  • the central control computer 1 selects a detection algorithm according to the characteristics of the fabric, and the fabric features include, but are not limited to, fabric color, fabric pattern, fabric material and the like. All detection algorithms are pre-integrated into the inspection software system.
  • the sampled picture is analyzed for defect detection according to the selected defect detection algorithm. If there is a defect, the local coordinates (x, y) of the defect on the sampled picture, the size of the statistical defect area, the length of the determination point, and the linear defect point are further determined.
  • the global coordinate and the local coordinate of the defect point satisfy the following relationship:
  • the type of defect is evaluated and scored using a four-point or ten-point standard. Based on the scores of individual defects, statistical information such as the total score of the fabric, the average score of the code length, and the like is generated. Taking the four-point system as an example, the defect type and corresponding score are as follows:
  • the average score for detecting the code length is:
  • the database is used to store information for each detection point of the fabric, including but not limited to the defect level, the defect point, the local coordinates of the defect, the global coordinates of the defect, the storage path of the defect, the storage path of the defect detection map, the overall detection score of the fabric, and the fabric detection. Code length and unit code length average score, etc.
  • a test report of the detected fabric can be generated, and a paper test report can be printed by the printer.
  • the central control computer 1 reduces the rotational speed of the roller motor 5 of the fabric inspection machine by setting the lower frequency value of the frequency converter 2, thereby reducing the transmission speed of the fabric.
  • the central control computer 1 controls the defect marking device 7 to laterally translate to the defect global coordinate X, and controls the defect marking device 7 to position the detected defect (X, Y) for the defect marking.
  • the defect marking device 7 can also be non-translated, and only needs to position the defect marking device 7 to the global ordinate Y of the detected defect to mark the edge of the fabric. Marking methods include, but are not limited to, fluorescent labels, label paper labels, and the like.
  • the system can operate in fully automated and interactive mode.
  • the noise removal subsystem, the defect detection subsystem, and the defect automatic marking subsystem in the defect detection system of the present invention are distributed in the processing area of the fabric transfer process as shown in FIG.
  • noise factors such as fibers, small yarns, dust, etc. can be eliminated. Therefore, a clean image can be obtained for subsequent processing. If there is no noise filtering device, the captured image may be unreliable due to contamination, resulting in false alarms.
  • the noise removal subsystem ensures high-quality fidelity pictures captured by the camera, providing a basis for the defect detection subsystem and reducing the impact of noise factors on the detection results.
  • the detection result of the defect detection subsystem provides an accurate coordinate reference for the defect automatic marking subsystem, so that the defect can be accurately marked, which provides important guidance and reference significance for subsequent cutting.
  • the automatic detection method based on computer vision image can effectively eliminate the defects of low detection efficiency, high missed detection rate and unsuitable for long-term continuous operation, and can greatly improve the detection speed under the premise of ensuring the detection precision, thereby improving the overall fabric. Inspection efficiency.
  • noise elimination, defect detection, scoring and storage, and defect marking are automatically completed, so that an operator can manage multiple automatic inspection systems, greatly reducing the labor cost of the factory. Invest.
  • a fabric defect automatic detecting apparatus may include a memory and a controller, wherein the control is configured to perform the method as described above.
  • a computer readable storage medium having stored thereon a computer program, which when executed, performs the method as described above.
  • the modules or units involved in the embodiments of the present application may be implemented by software or by hardware.
  • Those skilled in the art will appreciate that all or a portion of the steps of the above-described embodiments may be implemented as a computer program or instructions executed by a CPU.
  • the computer program is executed by the CPU, the above-described functions defined by the above-described methods provided by the present invention are performed.
  • the program may be stored in a computer readable storage medium, which may be a read only memory, a magnetic disk or an optical disk, or the like.

Landscapes

  • Engineering & Computer Science (AREA)
  • Textile Engineering (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Physics & Mathematics (AREA)
  • Analytical Chemistry (AREA)
  • Pathology (AREA)
  • Chemical & Material Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Biochemistry (AREA)
  • General Health & Medical Sciences (AREA)
  • General Physics & Mathematics (AREA)
  • Immunology (AREA)
  • Signal Processing (AREA)
  • Multimedia (AREA)
  • Wood Science & Technology (AREA)
  • Electromagnetism (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
  • Treatment Of Fiber Materials (AREA)

Abstract

本发明公开了一种织物疵点自动检测方法和系统,该方法包括:通过噪声清除设备在织物传输过程中对待检测织物区域做预处理,消除检测过程中的噪声干扰;根据验布机的滚轮电机的转动信息,触发摄像头对待检测织物进行图像采集;以及基于采集的图像,自动识别待检测织物上的疵点。本发明的方法和系统可有效解决人工织物疵点检测面临的准确率低、漏检率高和强度大等问题,有效提高疵点检测的准确率。

Description

一种织物疵点自动检测方法、系统和计算机可读存储介质 技术领域
本发明涉及纺织业检测技术,具体而言,涉及一种织物疵点自动检测方法、系统和计算机可读存储介质。
背景技术
长期以来,在纺织行业中,来料布匹(在本文中也称为织物)的质量直接关系到后续裁剪流程和成衣质量。目前几乎所有服装工厂都是依赖于工作人员裸眼来检验来料布匹。由于来料数量较大,服装工厂一般会装配至少一台验布机,且一台验布机需要至少一名操作员。在验布过程中,一旦发现布匹疵点,操作员会停机进行疵点确认,同时做好纸质记录和疵点标记。由于人眼的局限性,使得人工验布速度一般较慢,若验布机速度过快,则必然会出现较多的漏检,从而影响对布匹客观的打分和等级评价的准确性。另一方面,由于验布操作的重复度和强度较大,长时间的验布过程会导致操作员的视力疲劳,从而导致较多漏检的产生,并且操作员工作时间越长,出现漏检的概率会越大。
近年来,业界中提出了一些自动检测技术。例如,通过利用摄像机从各个角度拍摄来进行疵点检测。然而,由于拍摄的布匹上有噪点(诸如,纤维,断纱,灰尘等)存在,而现有系统中没有提供消除噪声干扰的设备,噪点会使得检测准确度大大降低。另外一些现有技术提出了使用正交小波变换检测算法来检测疵点的技术,然而,由于这种算法的运算成本非常高,无法实现实时检测。因此,需要一种实时并且准确地对织物疵点进行自动检测和标记的设备。
发明内容
为了解决现有技术中存在的一个或多个问题,本发明提供一种织物疵点自动检测方法、系统及计算机可读存储介质。
根据本发明的第一方面,提供了一种织物疵点自动检测方法,包括:通过噪声清除设备在织物传输过程中对待检测织物区域做预处理,消除检测过程中的噪声干扰;根据验布机的滚轮电机的转动信息,触发摄像头对待检测织物进行图像采集;以及基于采集的图像,自动识别待检测织物上的疵点。当识别出疵点时,可以调节验布机的滚轮电机的转速以降低织物传输速度,通过疵点标记装置对所述织物上的检测出的疵点位置进行标记。当疵点位置标记完成时,提高验布机的滚轮电机的转速,以恢复织物的传输速度。
自动识别待检测织物上的疵点还可以包括,对跨接两个或多个相邻图像的疵点进行疵点图像拼接。
在该方法中,还包括获取疵点在整体织物中的全局坐标位置,包括:获取疵点在采样图片上的局部坐标、疵点区域大小、疵点的长度和疵点的方向,根据验布机的滚轮电机的转动信息获得疵点在整体织物中的全局纵坐标,根据采样图片所对应的摄像头的编号获得疵点在 整体织物中的全局横坐标,从而根据疵点在整体织物中的全局坐标位置对疵点位置进行准确标记。
对疵点位置进行标记还可以包括:根据所述疵点在整体织物中的全局坐标位置,控制疵点标记装置横向平移到疵点上方进行疵点标记,并且疵点标记装置的纵向移动速度与织物传输速度相匹配。可替代地,对疵点位置进行标记还包括:根据所述疵点在整体织物中的全局坐标位置,控制疵点标记装置定位到所识别出的疵点对应的织物边缘,并在织物边缘进行标记,而不平移所述疵点标记装置。
此外,本发明的方法中使用的噪声清除设备可包括吹风装置和清扫装置。可以根据待检测织物特性调节所述吹风装置的出风速度和强弱,织物特性包括但不限于织物颜色、织物花型、织物纤维程度和织物结构;并且可以通过调节所述清扫装置来控制所述清扫装置的清扫面与织物表面接触的松紧程度或者间隙高低,清扫面包括但不限于纤维组织、毛刷、海绵等。
本发明的方法还可以包括,根据疵点特性对疵点进行分级和打分,疵点特性包括但不限于疵点长度、疵点区域大小、疵点方向、疵点出现的单位码长密度等。
识别出的织物的疵点信息可以被存储,存储的疵点信息包括但不限于疵点等级、疵点分数、疵点局部坐标、疵点全局坐标、疵点原始图像的存储路径、疵点检测图像的存储路径、织物整体检测分数、织物检测码长以及单位码长平均分数等。
最后,可以自动生成并输出待检测织物的检测报告。
根据本发明的第二方面,提供一种织物疵点自动检测系统,包括:噪声清除设备,配置成在织物传输过程中对待检测织物区域做预处理,消除检测过程中的噪声干扰;摄像头,配置成根据验布机的滚轮电机的转动信息被触发以对待检测织物进行图像采集;以及疵点识别装置,配置成基于采集的图像自动识别待检测织物上的疵点。本发明的系统还可以包括疵点标记装置,当所述疵点识别装置识别出疵点时,验布机的滚轮电机的转速被降低,以降低织物传输速度,通过所述疵点标记装置对所述织物上的检测出的疵点位置进行标记。当疵点位置标记完成时,验布机的滚轮电机的转速被提高,以恢复织物的高速传输。
疵点识别装置还可以配置成对跨接两个或多个相邻图像的疵点进行疵点图像拼接。
在该系统中,疵点识别装置还配置成获取疵点在整体织物中的全局坐标位置,通过获取疵点在采样图片上的局部坐标、疵点区域大小、疵点的长度和疵点的方向,根据验布机的滚轮电机的转动信息获得疵点在整体织物中的全局纵坐标,并根据采样图片所对应的摄像头的编号获得疵点在整体织物中的全局横坐标,疵点标记装置根据疵点在整体织物中的全局坐标位置对疵点位置进行准确标记。
所述疵点标记装置还可以根据所述疵点在整体织物中的全局坐标位置横向平移到疵点上方进行疵点标记,并且疵点标记装置的纵向移动速度与织物传输速度相匹配。可替代地,疵点标记装置还可以根据所述疵点在整体织物中的全局坐标位置定位到所识别出的疵点对应的织物边缘,并在织物边缘进行标记,同时所述疵点标记装置不发生平移。
此外,本发明的系统中的噪声清除装置可包括吹风装置和清扫装置。所述吹风装置的出风速度和强弱可以根据待检测织物特性来调节,织物特性包括但不限于织物颜色、织物花型、 织物纤维程度和织物结构;并且,所述清扫装置的清扫面与织物表面接触的松紧程度或者间隙高低可以通过调节所述清扫装置来控制,清扫面包括但不限于纤维组织、毛刷、海绵等。
本发明的系统还可以包括疵点分级装置,配置成根据疵点特性对疵点进行分级和打分,疵点特性包括但不限于疵点长度、疵点区域大小、疵点方向、疵点出现的单位码长密度等。
本发明的系统还可以包括数据库,配置成对识别出的织物的疵点进行信息存储,存储的疵点信息包括但不限于疵点等级、疵点分数、疵点局部坐标、疵点全局坐标、疵点原始图像的存储路径、疵点检测图像的存储路径、织物整体检测分数、织物检测码长以及单位码长平均分数等。
最后,该系统还可以包括检测报告装置,配置成自动生成并输出待检测织物的检测报告。
根据本发明的第三方面,提供一种织物疵点自动检测设备,包括存储器和控制器,其中,所述控制配置成执行如上所述的方法。
根据本发明的第四方面,提供一种计算机可读存储介质,该存储介质上存储有计算机程序,该计算机程序被执行时,执行如上文所述的方法。
本发明的系统可有效解决人工织物疵点检测面临的准确率低、漏检率高、工作重复和强度大等缺陷。利用本发明的方法可有效提高疵点检测的准确率,确保检测操作可长时间持续进行,并且自动生成的疵点检测报表,可作为后续织物裁剪的依据。
应当理解的是,以上的一般性描述和后文的详细描述仅是示例性的,并不能限制本发明。
附图说明
下面将参照附图详细描述本发明的示例实施例,本发明的上述和其它目标、特征和优点将变得更加显而易见。
图1是根据本发明的一个实施例的织物疵点自动检测方法的流程图;
图2是根据本发明的另一实施例的织物疵点自动检测方法的流程图;
图3是根据本发明的一个实施例的织物疵点自动检测系统的整体布置示意图;以及
图4是根据本发明的一个实施例的织物疵点自动检测系统中各个子系统所处理的布匹区域分布的示意图。
具体实施方式
现将参考附图更全面地描述本发明的示例性实施例。应理解,本文中的示例性实施例仅是提供用来帮助理解本发明,而不应以任何形式限制本发明。提供这些实施例是为了使本发明的描述更加全面和完整,并将示例性实施例的构思全面地传达给本领域的技术人员。附图仅为本发明的示意性图解,并非一定是按比例绘制。图中相同的附图标记表示相同或类似的部分,因而将省略对它们的重复描述。
此外,本文描述的特征、结构或优点可以以任何合适的方式结合在一个或更多实施例中。在下面的描述中,提供许多具体细节从而给出对本发明的实施方式的充分理解。然而,本领域技术人员将意识到,可以实践本发明的技术方案而省略特定细节中的一个或多个,或者可 以采用其它等效的方法、方式、装置、步骤等来代替。为了简明起见,对于本领域中公知的结构、方法、装置、实现或者操作,将不再赘述。
图1示出了根据本发明的一个实施例的织物疵点自动检测方法的流程图。如图所示,该方法主要包括:S101,通过噪声清除设备在织物传输过程中对待检测织物区域做预处理,消除检测过程中的噪声干扰;接下来,在S102,根据验布机的滚轮电机的转动信息,触发摄像头对待检测织物进行图像采集;然后,在S103,基于采集的图像,自动识别待检测织物上的疵点。基于以上方法,可以完成织物疵点的自动检测识别。
图2示出了根据本发明的另一实施例的织物疵点自动检测方法的流程图。该实施例的方法除了图1所示的步骤外,还包括更多的改进特征,提供了一种更加完备的疵点自动检测方法。如图2所示的方法包括:S101,通过噪声清除设备在织物传输过程中对待检测织物区域做预处理,消除检测过程中的噪声干扰;接下来,在S102,根据验布机的滚轮电机的转动信息,触发摄像头对待检测织物进行图像采集;然后,在S103,基于采集的图像,自动识别待检测织物上的疵点。当识别出疵点时,在S204,判断疵点是否跨接两个或更多个相邻图像,如果判断结果为是,则在S206对跨接两个或多个相邻图像的疵点进行疵点图像拼接;如果判断结果为否,则直接前进到S205,检测疵点尺寸,对疵点进行分级、打分等。值得一提的是,对疵点分级、打分可以根据疵点特性来进行,疵点特性包括但不限于疵点长度、疵点区域大小、疵点方向、疵点出现的单位码长密度等,测量疵点尺寸并评估每个检测到的疵点分级还可以包括判断疵点的类型。
在该方法中,优选地,还可以包括,在S207,获取疵点在整体织物中的全局坐标位置。举例来说,可以通过获取疵点在采样图片上的局部坐标、疵点区域大小、疵点的长度和疵点的方向,根据验布机的滚轮电机的转动信息获得疵点在整体织物中的全局纵坐标,根据采样图片所对应的摄像头的编号获得疵点在整体织物中的全局横坐标,从而获得疵点的全局坐标位置,从而根据疵点在整体织物中的全局坐标位置对疵点位置进行准确标记。
接下来,在S208,调节验布机的滚轮电机的转速以降低织物传输速度,通过疵点标记装置对织物上的检测出的疵点位置进行标记,并且当疵点位置标记完成时,提高验布机的滚轮电机的转速,以恢复织物的传输速度。对疵点位置进行标记可以根据疵点在整体织物中的全局坐标位置,控制疵点标记装置横向平移到疵点上方进行疵点标记,并且疵点标记装置的纵向移动速度与织物传输速度相匹配。可替代地,对疵点位置进行标记还可以根据疵点在整体织物中的全局坐标位置,控制疵点标记装置定位到所识别出的疵点对应的织物边缘,并在织物边缘进行标记,在该可替代方案中,不平移疵点标记装置。
此外,本发明的方法中使用的噪声清除设备可包括吹风装置和清扫装置。可以根据待检测织物特性调节所述吹风装置的出风速度和强弱,织物特性包括但不限于织物颜色、织物花型、织物纤维程度和织物结构;并且可以通过调节清扫装置来控制清扫装置的清扫面与织物表面接触的松紧程度或者间隙高低,清扫面包括但不限于纤维组织、毛刷、海绵等。
在S209,识别出的织物的疵点信息可以被存储,存储的疵点信息包括但不限于疵点等级、疵点分数、疵点局部坐标、疵点全局坐标、疵点原始图像的存储路径、疵点检测图像的 存储路径、织物整体检测分数、织物检测码长以及单位码长平均分数等,最后,可以自动生成并输出待检测织物的检测报告。
以上根据图1和图2示出的仅是示例性的实施例,以上步骤不一定按照顺序执行,可以根据实际需要增添步骤或从以上步骤中减少不需要的步骤。
下面结合图3,通过举例方式来介绍本发明的织物疵点自动检测系统。
图3示出了根据本发明的一个实施例的织物疵点自动检测系统的整体布置示意图,该系统包括:噪声清除设备9,配置成在织物传输过程中对待检测织物区域做预处理,消除检测过程中的噪声干扰;摄像头6,可包括N个摄像头(C1……CN),其中N大于或等于1,配置成根据验布机的滚轮8的电机5的转动信息被触发以对待检测织物进行图像采集;以及疵点识别装置1,配置成基于采集的图像自动识别待检测织物上的疵点。本发明的系统还可以包括疵点标记装置7,当疵点识别装置1识别出疵点时,验布机的滚轮8的电机5的转速被降低,以降低织物传输速度,通过疵点标记装置7对织物上的检测出的疵点位置进行标记。当疵点位置标记完成时,验布机的滚轮8的电机5的转速被提高,以恢复织物的传输速度。
该系统可以进一步对跨接两个或多个相邻图像的疵点进行疵点图像拼接,并且获取疵点在整体织物中的全局坐标位置,通过获取疵点在采样图片上的局部坐标、疵点区域大小、疵点的长度和疵点的方向,根据验布机的滚轮电机5的转动信息获得疵点在整体织物中的全局纵坐标,并根据采样图片所对应的摄像头的编号获得疵点在整体织物中的全局横坐标,疵点标记装置根据疵点在整体织物中的全局坐标位置对疵点位置进行准确标记。
疵点标记装置7可以根据疵点在整体织物中的全局坐标位置横向平移到疵点上方进行疵点标记,并且疵点标记装置7的纵向移动速度与织物传输速度相匹配。此处的横向方向是指在织物的平面上,与织物移动方向垂直的方向,纵向方向是指织物的移动方向。可替代地,疵点标记装置7还可以根据疵点在整体织物中的全局坐标位置定位到所识别出的疵点对应的织物边缘,并在织物边缘进行标记,同时疵点标记装置7不发生平移。
此外,本发明的系统中的噪声清除装置9可包括吹风装置和清扫装置(图中未示出)。吹风装置的出风速度和强弱可以根据待检测织物特性来调节,织物特性包括但不限于织物颜色、织物花型、织物纤维程度和织物结构;并且,清扫装置的清扫面与织物表面接触的松紧程度或者间隙高低可以通过调节清扫装置来控制,清扫面包括但不限于纤维组织、毛刷、海绵等。
该系统还可以包括疵点分级装置,配置成根据疵点特性对疵点进行分级和打分,疵点特性包括但不限于疵点长度、疵点区域大小、疵点方向、疵点出现的单位码长密度等。
本发明的系统还可以包括数据库,配置成对识别出的织物的疵点进行信息存储,存储的疵点信息包括但不限于疵点等级、疵点分数、疵点局部坐标、疵点全局坐标、疵点原始图像的存储路径、疵点检测图像的存储路径、织物整体检测分数、织物检测码长以及单位码长平均分数等。
最后,该系统还可以包括检测报告装置,配置成自动生成并输出待检测织物的检测报告。作为一个例子,疵点识别装置、疵点分级装置、数据库、检测报告装置等可以由计算机、中 央控制器、处理器、微处理器等来实现。
下面通过一个具体实例,更加详细地解释本发明的系统结构和工作原理。应理解,除非特别说明,实例中的任何组件或元件均不是必须的或者不可替代的,该实例仅是为了帮助理解本发明,而不应对本发明构成任何限制。
在该具体实例中,为了方便阐述,将织物疵点自动检测系统根据功能分为织物噪声清除子系统、织物疵点检测子系统和疵点自动标记子系统三个部分。
织物噪声清除子系统包括吹风装置和清扫装置两部分。织物噪声清除子系统可由中控电脑根据检测流程统一控制,亦可进行单独启停控制。通过织物噪声清除子系统,可消除布匹生产和运输中出现的二次污染,如灰尘、断纱、纤维以及其他附着在织物表面的外来干扰因素,从而提高检测的准确性。
仍参照图3,疵点检测子系统包括一个或多个摄像头6、编码器4、触发器3和中控电脑1。疵点检测子系统为本发明的系统的重要组成部分,分为疵点检测和辅助设备控制两部分。疵点检测子系统、织物噪声清除子系统和疵点自动标记子系统协同工作,统一由中控电脑进行控制。
疵点自动标记子系统包括变频器2、疵点标记装置7以及平移电机。疵点自动标记子系统根据疵点检测子系统检测到的疵点类型及坐标,在织物上对疵点位置进行精准标记。
本发明的织物疵点自动检测系统在工作时,如图3所示,织物通过验布机传动滚轮8进行输送,中控电脑1启动噪声清除子系统和摄像头系统6。噪声清除子系统9可以在织物传动过程中单独进行启停,只需确保检测的织物区域经过噪声清除子系统。噪声清除子系统9通过吹风及清扫装置,清除织物表面附着的干扰因素,包括但不限于灰尘、断纱、纤维等,确保织物表面干净真实。吹风装置可结合检测织物的特点调节输出风速强弱。清扫装置通过调节,可控制其与织物表面接触的松紧程度(或者间隙高低)。编码器4实时采样验布机滚轮电机5转动的距离,编码器4记录的输出的距离包括总累计距离和单次累计距离。当编码器4单次累计距离达到设定的触发距离值时,编码器4会输送信号给触发器3,触发器3触发摄像头6捕获织物图片。编码器4输出信号后,单次累计距离归零,总累计距离保持不变。摄像头6启动的数目N可以根据织物的宽度进行设定,需保证启动的摄像头拍摄的图片组合覆盖织物整体宽度。摄像头6捕获的织物图像被传送到中控电脑1的疵点检测子系统中。
疵点检测子系统首先根据设定的织物类型选择对应的疵点检测算法,疵点检测算法预先集成于疵点检测子系统中,织物类型包括但不限于织物颜色、织物花型、织物纤维程度、织物结构等信息。接着,根据选择的疵点检测算法对采样的图片进行疵点检测分析。若采样图片存在疵点,则获取疵点在采样图片中的坐标,即局部坐标(x,y),并存储采样的原始图片和检测出的疵点图片。根据疵点信息,包括但不限于疵点长度、疵点区域大小、疵点方向、疵点出现的单位码长密度等信息,对疵点进行分级打分,进入自动检测后续流程;若不存在疵点,则所采样的图片不进行任何后续操作。可以采用四分制或者十分制标准,对疵点类型进行评估和打分,同时将疵点局部坐标(x,y)转换为整体织物上的全局坐标(X,Y)。根据单个疵点的分数,生成织物的总分数以及关于码长的平均分数等统计信息。
采用数据库对织物的每个检出疵点进行信息存储,包括但不限于疵点等级、疵点分数、疵点局部坐标、疵点全局坐标、疵点原图存储路径、疵点检测图存储路径、织物整体检测分数、织物检测码长及单位码长平均分数等。根据存储的织物检测信息,生成织物检测报告,并可通过打印机打印出纸质报告。
基于疵点自动检测子系统的检测结果,检测过程中若检测出织物疵点,则中控电脑设定低频率值给变频器2,降低验布机滚轮电机的转速,从而降低织物传输速度,确保疵点标记装置平移到疵点上方,然后进行疵点标记操作。疵点标记装置也可以不进行平移,仅在织物的边缘进行疵点标记。疵点标记完成,中控电脑设定高频率值给变频器,提升验布机滚轮电机的转速,确保织物检测速度。
当该实例的疵点检测系统开始运行时,首先,验布机滚轮电机5启动,验布机滚轮8转动,织物通过滚轮8进行传输。验布机滚轮电机5转动的距离通过编码器4实时采样。当编码器4单次累计的距离达到设定的触发距离h时,编码器4输出脉冲信号给触发器3,编码器4的单次累计距离归零,其总体累计距离保持不变。触发器3收到脉冲信号后,通过输出高电平触发摄像头6进行织物图像捕获。
验布机滚轮电机5启动后,噪声清除装置9将启动。噪声清除装置9,如图2所示,包含吹风装置9_1(未示出)和清扫装置9_2(未示出)。吹风装置9_1的输出风速大小可根据织物的特点进行调整。吹风装置9_1可去除织物表面附着的灰尘、纤维、断纱等图像干扰因素。清扫装置9_2可进一步去除强力附着的干扰因素。通过吹风装置9_1和清扫装置9_2的工作,可为摄像头捕获高质量的真实图片提供了基础,从而提高后续疵点识别的准确性。
摄像头6可由一个或者多个摄像组成,确保摄像头采样图片组合可覆盖织物的整体宽度。摄像头6的启动数目可根据实际织物宽度预先设定,亦可根据摄像头捕获的图像信息进行自动选择,即若摄像头捕获的图片中有织物信息,则启动,否则暂停该摄像头进行图像捕获。
摄像头6捕获的图像信息通过摄像头专属传输协议传输给中控电脑1。中控电脑1收到摄像头6传输的图像信息后,首先,中控电脑1根据织物特点选择检测算法,织物特点包括但不限于织物颜色、织物花型、织物材质等。所有检测算法都预先集成于检测软件系统中。然后,根据选择的疵点检测算法对采样的图片进行疵点检测分析。若存在疵点,则判断疵点在采样图片上的局部坐标(x,y)、统计疵点区域大小、判断疵点的长度、线性疵点进一步判断疵点的方向。根据编码器4总体累计距离获得疵点在整体织物中的全局纵坐标Y,根据采样图片所属的摄像头编号获得疵点在整体织物中的全局横坐标X,从而得到疵点在整体织物中的全局坐标位置(X,Y)。对疵点所属的原始图及检测图,按预先设定的存储路径进行存储。
若采样图片大小为Wmm×Hmm,当前采样图片为第m个相机,触发器触发次数为n,则疵点全局坐标和局部坐标满足如下关系:
X=W*(m-1)+x;
Y=H*(n-1)+y;
采用四分制或者十分制标准,对疵点类型进行评估和打分。根据单个疵点的分数,生成 织物的总分数、关于码长的平均分数等统计信息。以四分制为例,疵点类型及对应分数如下:
疵点类型 疵点分数
Ⅰ类 1分
Ⅱ类 2分
Ⅲ类 3分
Ⅳ类 4分
若疵点的数目为N,各疵点的类别为Ti,则织物的总分数P满足:
Figure PCTCN2018082661-appb-000001
若织物的总码长为L,则检测关于码长的平均分数为:
Figure PCTCN2018082661-appb-000002
采用数据库对织物每个检出疵点进行信息存储,包括但不限于疵点等级、疵点分数、疵点局部坐标、疵点全局坐标、疵点原始图存储路径、疵点检测图存储路径、织物整体检测分数、织物检测码长及单位码长平均分数等。
根据疵点检测结果,可生成所检测织物的检测报告,并可通过打印机打印出纸质的检测报告。
针对采样图片,若检测出存在疵点,则中控电脑1通过设定变频器2较低的频率值,降低验布机滚轮电机5的转速,从而降低织物的传输速度。中控电脑1控制疵点标记装置7横向平移到疵点全局坐标X处,并控制疵点标记装置7定位到所检测的疵点(X,Y)处进行疵点标记。疵点标记装置7亦可不平移,仅需定位疵点标记装置7到所检测疵点的全局纵坐标Y处,在织物的边缘进行疵点标记。标记方式包括但不限于荧光标记、标签纸标记等。
可选地,该系统可以在全自动和交互模式下运行。
在该实例中,本发明的疵点检测系统中的噪声清除子系统、疵点检测子系统和疵点自动标记子系统在织物传输过程的处理区域分布如图4所示。
通过噪声清除子系统,可以消除诸如纤维,小纱线,灰尘等的噪声因素。因此,可以得到干净的图像以用于后续处理。如果没有噪声过滤装置,则拍摄的图像可能会因污染而使检测结果不可靠,从而导致误报。噪声清除子系统确保了摄像头捕获的高品质保真图片,为疵点检测子系统提供了基础,降低了噪声因素对检测结果的影响。疵点检测子系统检测结果为疵点自动标记子系统提供了精准的坐标参考,使得疵点可以进行精确标记,为后续裁剪提供了重要指导及参考意义。
基于计算机视觉图像的自动检测方法可有效消除人工检测面临的检测效率低、漏检率高及不适合长时间连续作业等缺陷,在确保检测精度的前提下可大幅提升检测速度,从而提升整体织物检验效率。此外,由于基于计算机视觉图像的疵点检测过程中,噪声消除、疵点检测、打分及存储、疵点标记等均自动完成,从而一名操作员可管理多台自动验布系统,大大 降低工厂的人力成本投入。
根据本发明的织物疵点自动检测设备,可以包括存储器和控制器,其中,控制配置成执行如上所述的方法。
根据本发明,还提供一种计算机可读存储介质,该存储介质上存储有计算机程序,该计算机程序被执行时,执行如上文所述的方法。
描述于本申请实施例中所涉及到的模块或单元可以通过软件的方式实现,也可以通过硬件的方式来实现。本领域技术人员可以理解,上述实施方式的全部或部分步骤可以被实现为由CPU执行的计算机程序或者指令。在该计算机程序被CPU执行时,执行本发明提供的上述方法所限定的上述功能。程序可以存储于一种计算机可读存储介质中,该存储介质可以是只读存储器,磁盘或光盘等。
本发明是根据特定实施例进行描述的,但本领域的技术人员应明白在不脱离本发明范围时,可进行各种变化和等同替换。此外,为适应本发明技术的特定场合或材料,可对本发明进行诸多修改而不脱离其保护范围。因此,本发明并不限于在此公开的特定实施例,而包括所有落入到权利要求保护范围的实施例。

Claims (26)

  1. 一种织物疵点自动检测方法,包括:
    通过噪声清除设备在织物传输过程中对待检测织物区域做预处理,消除检测过程中的噪声干扰;
    根据验布机的滚轮电机的转动信息,触发摄像头对待检测织物进行图像采集;以及
    基于采集的图像,自动识别待检测织物上的疵点。
  2. 如权利要求1所述的方法,还包括,当识别出疵点时,调节验布机的滚轮电机的转速以降低织物传输速度,通过疵点标记装置对所述织物上的检测出的疵点位置进行标记。
  3. 如权利要求2所述的方法,还包括,当疵点位置标记完成时,提高验布机的滚轮电机的转速,以恢复织物的传输速度。
  4. 如权利要求1所述的方法,其中,自动识别待检测织物上的疵点还包括,对跨接两个或多个相邻图像的疵点进行疵点图像拼接。
  5. 如权利要求2所述的方法,还包括获取疵点在整体织物中的全局坐标位置,包括:获取疵点在采样图片上的局部坐标、疵点区域大小、疵点的长度和疵点的方向,根据验布机的滚轮电机的转动信息获得疵点在整体织物中的全局纵坐标,根据采样图片所对应的摄像头的编号获得疵点在整体织物中的全局横坐标,并根据疵点在整体织物中的全局坐标位置对疵点位置进行准确标记。
  6. 如权利要求5所述的方法,其中,对疵点位置进行标记还包括:根据所述疵点在整体织物中的全局坐标位置,控制疵点标记装置横向平移到疵点上方进行疵点标记,并且疵点标记装置的纵向移动速度与织物传输速度相匹配。
  7. 如权利要求5所述的方法,其中,对疵点位置进行标记还包括:根据所述疵点在整体织物中的全局坐标位置,控制疵点标记装置定位到所识别出的疵点对应的织物边缘,并在织物边缘进行标记,而不平移所述疵点标记装置。
  8. 如权利要求1所述的方法,其中,所述噪声清除设备包括吹风装置和清扫装置。
  9. 如权利要求8所述的方法,还包括,根据待检测织物特性调节所述吹风装置的出风速度和强弱,织物特性包括但不限于织物颜色、织物花型、织物纤维程度和织物结构; 并且通过调节所述清扫装置来控制所述清扫装置的清扫面与织物表面接触的松紧程度或者间隙高低,清扫面包括但不限于纤维组织、毛刷、海绵等。
  10. 如权利要求1所述的方法,还包括,根据疵点特性对疵点进行分级和打分,疵点特性包括但不限于疵点长度、疵点区域大小、疵点方向、疵点出现的单位码长密度等。
  11. 如权利要求10所述的方法,还包括对识别出的织物的疵点进行信息存储,存储的疵点信息包括但不限于疵点等级、疵点分数、疵点局部坐标、疵点全局坐标、疵点原始图像的存储路径、疵点检测图像的存储路径、织物整体检测分数、织物检测码长以及单位码长平均分数等。
  12. 如权利要求1所述的方法,还包括自动生成并输出待检测织物的检测报告。
  13. 一种织物疵点自动检测系统,包括:
    噪声清除设备,配置成在织物传输过程中对待检测织物区域做预处理,消除检测过程中的噪声干扰;
    摄像头,配置成根据验布机的滚轮电机的转动信息被触发以对待检测织物进行图像采集;以及
    疵点识别装置,配置成基于采集的图像自动识别待检测织物上的疵点。
  14. 如权利要求13所述的系统,还包括疵点标记装置,当所述疵点识别装置识别出疵点时,验布机的滚轮电机的转速被降低,以降低织物传输速度,通过所述疵点标记装置对所述织物上的检测出的疵点位置进行标记。
  15. 如权利要求14所述的系统,当疵点位置标记完成时,验布机的滚轮电机的转速被提高,以恢复织物的高速传输。
  16. 如权利要求13所述的系统,其中,所述疵点识别装置还配置成对跨接两个或多个相邻图像的疵点进行疵点图像拼接。
  17. 如权利要求14所述的系统,其中,所述疵点识别装置还配置成获取疵点在整体织物中的全局坐标位置,通过获取疵点在采样图片上的局部坐标、疵点区域大小、疵点的长度和疵点的方向,根据验布机的滚轮电机的转动信息获得疵点在整体织物中的全局纵坐标,并根据采样图片所对应的摄像头的编号获得疵点在整体织物中的全局横坐标,并且疵点标记装置根据疵点在整体织物中的全局坐标位置对疵点位置进行准确标记。
  18. 如权利要求17所述的系统,其中,所述疵点标记装置根据所述疵点在整体织物中的全局坐标位置横向平移到疵点上方进行疵点标记,并且疵点标记装置的纵向移动速度与织物传输速度相匹配。
  19. 如权利要求17所述的系统,其中,所述疵点标记装置根据所述疵点在整体织物中的全局坐标位置定位到所识别出的疵点对应的织物边缘,并在织物边缘进行标记,同时所述疵点标记装置不发生平移。
  20. 如权利要求13所述的系统,其中,所述噪声清除装置包括吹风装置和清扫装置。
  21. 如权利要求20所述的系统,其中,所述吹风装置的出风速度和强弱是根据待检测织物特性调节的,织物特性包括但不限于织物颜色、织物花型、织物纤维程度和织物结构;并且所述清扫装置的清扫面与织物表面接触的松紧程度或者间隙高低是通过调节所述清扫装置来控制的,清扫面包括但不限于纤维组织、毛刷、海绵等。
  22. 如权利要求13所述的系统,还包括疵点分级装置,配置成根据疵点特性对疵点进行分级和打分,疵点特性包括但不限于疵点长度、疵点区域大小、疵点方向、疵点出现的单位码长密度等。
  23. 如权利要求13所述的系统,还包括数据库,配置成对识别出的织物的疵点进行信息存储,存储的疵点信息包括但不限于疵点等级、疵点分数、疵点局部坐标、疵点全局坐标、疵点原始图像的存储路径、疵点检测图像的存储路径、织物整体检测分数、织物检测码长以及单位码长平均分数等。
  24. 如权利要求13所述的系统,还包括检测报告装置,配置成自动生成并输出待检测织物的检测报告。
  25. 一种织物疵点自动检测设备,包括存储器和控制器,其中,所述控制配置成执行如权利要求1-12中任一项所述的方法。
  26. 一种计算机可读存储介质,该存储介质上存储有计算机程序,该计算机程序被执行时,执行如权利要求1-12中任一项所述的方法。
PCT/CN2018/082661 2017-05-24 2018-04-11 一种织物疵点自动检测方法、系统和计算机可读存储介质 WO2018214661A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US16/611,183 US10942133B2 (en) 2017-05-24 2018-04-11 Method and system for automatically detecting fabric defect, and computer readable storage medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201710373868.X 2017-05-24
CN201710373868.XA CN108931531B (zh) 2017-05-24 2017-05-24 一种织物疵点自动检测方法、系统和计算机可读存储介质

Publications (1)

Publication Number Publication Date
WO2018214661A1 true WO2018214661A1 (zh) 2018-11-29

Family

ID=64395287

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/082661 WO2018214661A1 (zh) 2017-05-24 2018-04-11 一种织物疵点自动检测方法、系统和计算机可读存储介质

Country Status (3)

Country Link
US (1) US10942133B2 (zh)
CN (1) CN108931531B (zh)
WO (1) WO2018214661A1 (zh)

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110632089A (zh) * 2019-09-05 2019-12-31 浙江翼晟科技有限公司 自动验布系统及其验布方法
CN111337506A (zh) * 2020-03-30 2020-06-26 河南科技学院 一种用于服装品质检验的智能装置
WO2020142727A1 (en) * 2019-01-03 2020-07-09 Palleon Pharmaceuticals Inc. Methods and compositions for treating cancer with immune cells
CN111462064A (zh) * 2020-03-26 2020-07-28 西安工程大学 一种基于大脑皮层细胞方向选择性的织物疵点检测方法
EP3792391A1 (en) * 2019-09-12 2021-03-17 Manteco S.p.A. System for the control and management of fabrics
CN114820556A (zh) * 2022-05-13 2022-07-29 东北大学秦皇岛分校 一种基于自定义算法的食品包装瑕疵样本的生成方法
CN115110230A (zh) * 2022-02-25 2022-09-27 耿艳 一种基于物联网大数据技术的纺织布匹漂白装置

Families Citing this family (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110057832A (zh) * 2019-04-24 2019-07-26 拓卡奔马机电科技有限公司 一种铺布机上的布料瑕疵检测控制系统
CN110389130A (zh) * 2019-07-04 2019-10-29 盎古(上海)科技有限公司 应用于织物的智能检测系统
CN110702690B (zh) * 2019-09-29 2022-05-20 浙江宝仁和中科技有限公司 一种布面疵点标记剔除系统
CN111812031A (zh) * 2020-06-02 2020-10-23 维库(厦门)信息技术有限公司 一种基于物联网检测系统的检测方法
CN111754470A (zh) * 2020-06-11 2020-10-09 厦门雨程户外运动用品有限公司 自动验布方法、装置、自动验布机和存储介质
CN113030102A (zh) * 2021-03-02 2021-06-25 成都小淞科技有限公司 基于机器视觉的漆面瑕疵检查系统
CN113335981A (zh) * 2021-05-18 2021-09-03 佛山市赛邦智能装备有限公司 一种无纺布专用高速复卷分切检测设备及其检测方法
CN113281342B (zh) * 2021-05-21 2024-04-26 绍兴随手智能科技有限公司 一种机器视觉与光谱仪融合的布匹瑕疵检测方法
CN113313690B (zh) * 2021-05-31 2022-08-09 齐鲁工业大学 一种皮革纤维mct序列切片图像的同根区域智能检测方法
CN113376168A (zh) * 2021-06-09 2021-09-10 深圳技术大学 一种检修一体化设备
CN113588671B (zh) * 2021-09-28 2021-12-10 江苏祥顺布业有限公司 一种低功耗的验布机智能控制方法及系统
CN114393895B (zh) * 2022-01-21 2024-05-17 山东晶创新材料科技有限公司 一种基于聚丙烯长丝非织造布的复合防水卷材tpo制备方法
CN115049644B (zh) * 2022-08-12 2022-11-01 山东三微新材料有限公司 一种基于铝管表面瑕疵识别的温度控制方法及系统
CN117288761B (zh) * 2023-11-27 2024-02-06 天津市海迅科技发展有限公司 一种基于测试材料的瑕疵检测分类评估方法及系统

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN203688459U (zh) * 2013-10-25 2014-07-02 厦门大学 基于图像识别的织物质量在线检测系统
CN204882397U (zh) * 2015-06-29 2015-12-16 广东溢达纺织有限公司 不停机自动标识纺织面料疵点的系统
CN205786382U (zh) * 2016-05-21 2016-12-07 晋江市池店镇泉捷鞋机设计服务部 一种预除尘双向验布机

Family Cites Families (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4124300A (en) * 1976-02-23 1978-11-07 Greenwood Mills, Inc. Method for automatic fabric inspection
US5087822A (en) * 1990-06-22 1992-02-11 Alcan International Limited Illumination system with incident beams from near and far dark field for high speed surface inspection of rolled aluminum sheet
US5870203A (en) * 1996-03-15 1999-02-09 Sony Corporation Adaptive lighting control apparatus for illuminating a variable-speed web for inspection
US20010012381A1 (en) * 1997-09-26 2001-08-09 Hamed Sari-Sarraf Vision-based, on-loom fabric inspection system
FI114741B (fi) * 2001-05-11 2004-12-15 Metso Automation Oy Lämpökuvaukseen perustuva laadun- ja kunnonvalvontamenetelmä
DE102004012727A1 (de) * 2004-03-16 2005-10-06 Voith Paper Patent Gmbh J-Linie
US20090060316A1 (en) * 2006-02-22 2009-03-05 Hannu Ruuska Method for Monitoring a Rapidly-Moving Paper Web and Corresponding System
DE102006057215B4 (de) * 2006-12-01 2022-08-11 Trützschler GmbH & Co Kommanditgesellschaft Vorrichtung an einer Spinnereivorbereitungsanlage zum Erkennen von Fremdstoffen in Fasergut
JP4440281B2 (ja) * 2007-04-02 2010-03-24 株式会社プレックス シート状物品の検査方法および装置
CN101587082B (zh) * 2009-06-24 2011-08-03 天津工业大学 一种应用于织物疵点检测的快速三维重建方法
CN103207186A (zh) * 2013-04-08 2013-07-17 北京经纬纺机新技术有限公司 自动验布机疵点检测识别方法及其系统
CN204575549U (zh) * 2015-03-13 2015-08-19 赵晓明 一种可预除尘的纺织品缺陷检测装置
JP6479706B2 (ja) * 2016-04-11 2019-03-06 株式会社PSM International 長尺シート材の厚み計測方法および厚み計測システム

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN203688459U (zh) * 2013-10-25 2014-07-02 厦门大学 基于图像识别的织物质量在线检测系统
CN204882397U (zh) * 2015-06-29 2015-12-16 广东溢达纺织有限公司 不停机自动标识纺织面料疵点的系统
CN205786382U (zh) * 2016-05-21 2016-12-07 晋江市池店镇泉捷鞋机设计服务部 一种预除尘双向验布机

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2020142727A1 (en) * 2019-01-03 2020-07-09 Palleon Pharmaceuticals Inc. Methods and compositions for treating cancer with immune cells
CN110632089A (zh) * 2019-09-05 2019-12-31 浙江翼晟科技有限公司 自动验布系统及其验布方法
EP3792391A1 (en) * 2019-09-12 2021-03-17 Manteco S.p.A. System for the control and management of fabrics
CN111462064A (zh) * 2020-03-26 2020-07-28 西安工程大学 一种基于大脑皮层细胞方向选择性的织物疵点检测方法
CN111462064B (zh) * 2020-03-26 2023-05-12 西安工程大学 一种基于大脑皮层细胞方向选择性的织物疵点检测方法
CN111337506A (zh) * 2020-03-30 2020-06-26 河南科技学院 一种用于服装品质检验的智能装置
CN115110230A (zh) * 2022-02-25 2022-09-27 耿艳 一种基于物联网大数据技术的纺织布匹漂白装置
CN115110230B (zh) * 2022-02-25 2023-11-21 耿艳 一种基于物联网大数据技术的纺织布匹漂白装置
CN114820556A (zh) * 2022-05-13 2022-07-29 东北大学秦皇岛分校 一种基于自定义算法的食品包装瑕疵样本的生成方法
CN114820556B (zh) * 2022-05-13 2024-05-17 东北大学秦皇岛分校 一种基于自定义算法的食品包装瑕疵样本的生成方法

Also Published As

Publication number Publication date
CN108931531B (zh) 2021-10-22
US20200158659A1 (en) 2020-05-21
US10942133B2 (en) 2021-03-09
CN108931531A (zh) 2018-12-04

Similar Documents

Publication Publication Date Title
WO2018214661A1 (zh) 一种织物疵点自动检测方法、系统和计算机可读存储介质
CN109550712B (zh) 一种化纤丝尾丝外观缺陷检测系统及方法
CN109454006B (zh) 基于化纤丝锭绊丝缺陷在线检测和分级的装置的检测分级方法
WO2021248554A1 (zh) 高速高精度锂离子电池极片的毛刺检测方法及检测系统
JP3822242B2 (ja) スライド及び試料の調製品質を評価するための方法及び装置
JP5151019B2 (ja) 疵検出装置及び疵検出方法
US6687396B1 (en) Optical member inspection apparatus, image-processing apparatus, image-processing method, and computer readable medium
CN107449782A (zh) 手机外壳侧边表面缺陷数字化定量检测系统
JPH0328749A (ja) オンライン式構造検査方式
WO2007062563A1 (fr) Procede d'examen automatique en ligne pour des detections de defauts de surface d'acier pendant le pretraitement de toles d'acier
CN106802215A (zh) 一种水管漏水检测装置及检测方法
CN110889823A (zh) 一种SiC缺陷的检测方法和系统
CN111415349B (zh) 一种基于图像处理技术的涤纶长丝毛丝检测方法
CN107024478A (zh) 一种零件加工自动缺陷识别系统
CN110084787B (zh) 一种基于机器视觉的合成革涂胶在线监测方法
CN106599890A (zh) 变电站巡检机器人数字式仪表识别算法
CN107156902A (zh) 一种烟支外观质量在线检测装置
CN111724445A (zh) 一种大视野小尺寸识别码的识别方法及系统
JP2007003243A (ja) 長尺物の外観検査装置
CN113781458A (zh) 一种基于人工智能的识别方法
CN102609699A (zh) 一种激光扫描铸件工件号的识别装置和方法
CN105699386B (zh) 一种采用接触式图像传感器的自动验布标记方法
CN208432555U (zh) 一种机组式印刷品自动检测设备
CN113727022B (zh) 巡检图像的采集方法及装置、电子设备、存储介质
JP2012202957A (ja) 欠陥位置情報生成装置、欠陥確認システム及び欠陥位置情報生成方法

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: 18806632

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: 18806632

Country of ref document: EP

Kind code of ref document: A1