CN116805311B - Automobile part surface defect monitoring method based on robot vision - Google Patents
Automobile part surface defect monitoring method based on robot vision Download PDFInfo
- Publication number
- CN116805311B CN116805311B CN202311042712.5A CN202311042712A CN116805311B CN 116805311 B CN116805311 B CN 116805311B CN 202311042712 A CN202311042712 A CN 202311042712A CN 116805311 B CN116805311 B CN 116805311B
- Authority
- CN
- China
- Prior art keywords
- monitoring
- information
- grade
- target accessory
- target
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 176
- 230000007547 defect Effects 0.000 title claims abstract description 126
- 238000000034 method Methods 0.000 title claims abstract description 25
- 230000005477 standard model Effects 0.000 claims abstract description 12
- 238000007514 turning Methods 0.000 claims description 12
- 230000002950 deficient Effects 0.000 claims description 11
- 230000002159 abnormal effect Effects 0.000 claims description 10
- 238000004519 manufacturing process Methods 0.000 claims description 7
- 239000003086 colorant Substances 0.000 claims description 5
- 238000012795 verification Methods 0.000 claims description 5
- 238000001914 filtration Methods 0.000 claims description 3
- 238000003754 machining Methods 0.000 claims description 3
- 238000004458 analytical method Methods 0.000 abstract description 5
- 238000012545 processing Methods 0.000 description 11
- 238000001514 detection method Methods 0.000 description 5
- 238000011156 evaluation Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- 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
- G01N21/8851—Scan 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T19/00—Manipulating 3D models or images for computer graphics
- G06T19/20—Editing of 3D images, e.g. changing shapes or colours, aligning objects or positioning parts
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
-
- 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
- G01N21/8851—Scan 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
- G01N2021/8854—Grading and classifying of flaws
-
- 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
- G01N21/8851—Scan 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
- G01N2021/8887—Scan 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30164—Workpiece; Machine component
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30232—Surveillance
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computer Graphics (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- Signal Processing (AREA)
- Architecture (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The application discloses a method for monitoring surface defects of automobile parts based on robot vision, which relates to the technical field of defect monitoring, and comprises the steps of collecting multi-azimuth images of standard target parts, and establishing reference information according to the multi-azimuth images, wherein the reference information comprises a standard model and brightness characteristic information of the multi-azimuth images of the standard target parts; and setting monitoring grade information of the target accessory, wherein the monitoring grade information is an emergency grade set for the plurality of monitoring item information, and setting a deviation threshold value of the target accessory of the emergency grade. The application can carry out a division judgment on whether the defect can be recycled and reprocessed or can be used normally, has a better defect analysis function, can be used for judging the severity degree of the subsequent defect monitoring, greatly improves the efficiency of defect monitoring, and can also improve the accuracy of defect monitoring.
Description
Technical Field
The application relates to the technical field of defect monitoring, in particular to an automobile part surface defect monitoring method based on robot vision.
Background
Along with the development of society, the manufacturing of accessories in the automobile industry can be combined with the prior art to realize the defect monitoring of the surfaces of the automobile accessories, the existing accessory surface defect monitoring means mostly adopt manual cooperation devices to jointly detect, the defects on the surfaces of the riding accessories are difficult to quickly and accurately confirm, meanwhile, the defect degree is difficult to judge, whether the defects of the accessories can be reprocessed to be qualified or not cannot be judged, and therefore, the automobile accessory surface defect monitoring method based on robot vision is provided.
Disclosure of Invention
The application aims to provide a method for monitoring surface defects of automobile parts based on robot vision, which aims to solve the defects in the background technology.
In order to achieve the above object, the present application provides the following technical solutions: a method for monitoring surface defects of automobile parts based on robot vision comprises the following steps:
collecting multi-azimuth images of a standard target accessory, and establishing reference information according to the multi-azimuth images, wherein the reference information comprises a standard model and brightness characteristic information of the multi-azimuth images of the standard target accessory;
setting monitoring grade information of the target accessory, wherein the monitoring grade information is an emergency grade set for a plurality of monitoring item information, and setting a deviation threshold value of the target accessory of the emergency grade;
acquiring a multidirectional monitoring image of a target accessory based on robot vision, and establishing monitoring information according to the multidirectional monitoring image, wherein the monitoring information comprises a monitoring model and the brightness characteristic information of the multidirectional monitoring image of the target accessory;
and comparing the monitoring information with the reference information to obtain distinguishing characteristic information, and evaluating the distinguishing characteristic information based on the monitoring grade information to obtain the defect index of the target accessory.
In a preferred embodiment, the step of acquiring a multi-aspect image of the standard target assembly and creating the reference information based on the multi-aspect image comprises:
selecting a standard part shape of the target accessory as a standard target accessory;
setting standard light information, and acquiring images of a plurality of directions of a standard target accessory based on the standard light information to obtain a multi-direction image;
filtering the colors of the multi-azimuth image to obtain a target image, and storing the brightness value and the brightness contrast value of the target accessory image corresponding to the type of the target accessory as brightness characteristic information of the standard target accessory multi-azimuth image;
and constructing a three-dimensional model of the standard target accessory according to the multi-azimuth image, and marking line characteristic points of the target accessory according to the three-dimensional model, wherein the line characteristic points are turning points corresponding to the shape line change of the target accessory so as to distinguish the color of the three-dimensional model of the target accessory as the line characteristic point marking color, thereby obtaining the standard model.
In a preferred embodiment, the step of constructing a three-dimensional model of the standard target fitting from the multi-dimensional image comprises:
obtaining a contour map of a standard target fitting according to the multi-azimuth image, and paving lines on the surface of the contour map to obtain a net-shaped contour map;
and performing color filling on the grid positions of the mesh profile to obtain a three-dimensional model.
In a preferred embodiment, the step of formulating the monitoring level information of the target accessory includes:
collecting monitoring items corresponding to the types of the target accessories, wherein the monitoring items are the positions and types of the surface defects of the accessories, and monitoring information is obtained based on the corresponding monitoring items of the reference information;
and dividing and formulating an emergency grade according to the monitoring items, wherein the emergency grade comprises a first emergency grade and a second emergency grade.
In a preferred embodiment, the step of obtaining the monitoring information based on the reference information corresponding to the monitoring item includes:
acquiring index information required by defect verification according to a monitoring item based on reference information, wherein the information included in the index information is any one or combination of multiple information of line characteristic points, brightness values and brightness contrast values respectively;
the index information is associated with the monitoring item data as monitoring information.
In a preferred embodiment, the step of setting the deviation threshold of the emergency level target accessory includes:
acquiring index information corresponding to the emergency level, splitting information in the index information, and obtaining one or more pieces of split information;
and formulating deviation thresholds for the split information one by one, wherein the deviation thresholds comprise reasonable deviation thresholds and abnormal deviation thresholds.
In a preferred embodiment, the step of acquiring the multi-azimuth monitoring image of the target accessory based on the robot vision and establishing monitoring information according to the multi-azimuth monitoring image includes:
acquiring corresponding monitoring items according to the target accessory;
acquiring a multi-azimuth monitoring image of the target accessory based on robot vision, and obtaining a brightness value and a brightness contrast value of the multi-azimuth monitoring image;
constructing a contour map of the target accessory according to the multidirectional monitoring image, and paving the contour map on the surface of the contour map defect monitoring position through lines based on the defect monitoring position corresponding to the monitoring item to obtain a net-shaped contour map of the monitoring position;
and performing color filling on the grid positions of the monitoring position network contour map to obtain a monitoring three-dimensional model.
In a preferred embodiment, the step of comparing the monitoring information with the reference information to obtain the distinguishing characteristic information, and evaluating the distinguishing characteristic information based on the monitoring grade information to obtain the defect index of the target accessory includes:
comparing the monitoring three-dimensional model with the standard model based on the monitoring item of the target accessory to obtain distinguishing characteristic information;
obtaining emergency grade of the distinguishing characteristic information based on the monitoring grade information, taking the first emergency grade as a comparison optimal monitoring grade, and taking the second emergency grade as a post-setting monitoring grade;
obtaining distinguishing characteristic information meeting a deviation threshold value and an abnormal deviation threshold value respectively based on the emergency level, respectively obtaining a qualified target accessory and a defective target accessory, and marking the distinguishing characteristic information of the defective target accessory as a disqualified target accessory if the target accessory exists;
obtaining a defect index of the target accessory according to the distinguishing characteristic information;
the defect index that does not satisfy the threshold condition is used as the low quality target accessory machining information.
In a preferred embodiment, the method further includes the step of storing the defect index into a corresponding target accessory category as historical defect information, collecting the historical defect information to perform frequent classification, and determining whether the target accessory category is a high-risk manufacturing target accessory category, including:
storing the corresponding target types of the defect indexes as historical defect information, and obtaining the historical defect indexes of the target accessories according to the historical defect information;
and (3) dividing the frequent-hair grade, and obtaining a corresponding frequent-hair grade according to a frequent-hair grade value which is correspondingly set by the frequent-hair grade, wherein the frequent-hair grade comprises a high-risk grade, a dangerous grade, a general grade and a good grade.
In the technical scheme, the application has the technical effects and advantages that:
the application is convenient for comparing the distinguishing data in the follow-up monitoring of the target accessory, and greatly improves the follow-up efficiency of monitoring the surface defects of the target accessory; the defect analysis method has the advantages that whether the defect can be recycled and reprocessed or not or whether the defect can be normally used can be judged, the defect analysis function is good, the defect analysis method can be used for judging the severity degree of subsequent defect monitoring, the defect monitoring efficiency is greatly improved, and meanwhile the defect monitoring accuracy can be improved.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments described in the present application, and other drawings may be obtained according to these drawings for a person having ordinary skill in the art.
FIG. 1 is a flow chart of the method of the present application.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present application more apparent, the technical solutions of the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application, and it is apparent that the described embodiments are some embodiments of the present application, but not all embodiments of the present application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
Embodiment 1, please refer to fig. 1, the method for monitoring the surface defects of the automobile parts based on the robot vision according to the embodiment comprises the following steps:
s1, acquiring a multi-azimuth image of a standard target accessory, and establishing reference information according to the multi-azimuth image, wherein the reference information comprises a standard model and brightness characteristic information of the multi-azimuth image of the standard target accessory;
in one embodiment, the step S1 of collecting a multi-azimuth image of the standard target accessory and establishing the reference information according to the multi-azimuth image includes:
s11, selecting a standard part shape of the target accessory as a standard target accessory;
s12, setting standard light information, and acquiring images of a plurality of directions of a standard target accessory based on the standard light information to obtain a multi-direction image;
s13, filtering the colors of the multi-azimuth image to obtain a target image, and storing the brightness value and the brightness contrast value of the target accessory image corresponding to the type of the target accessory as brightness characteristic information of the standard target accessory multi-azimuth image;
s14, constructing a three-dimensional model of the standard target accessory according to the multi-azimuth image, and marking line feature points of the target accessory according to the three-dimensional model, wherein the line feature points are turning points corresponding to the shape line change of the target accessory, and distinguishing the color of the three-dimensional model of the target accessory as the line feature point marking color to obtain the standard model;
as described in the above steps S11-S14, the standard component shape of the target accessory is selected as the standard target accessory, the standard target accessory is subjected to multi-directional image acquisition under the set standard light information to obtain multi-directional images, wherein the standard light information is daily light processed in a factory, the multi-directional images of the standard target accessory are acquired under the standard light information, then the multi-directional images are acquired and the brightness contrast value is saved to represent the surface flatness of the target accessory, the three-dimensional model of the standard target accessory is constructed through the multi-directional images after the brightness value is acquired, and the line feature points corresponding to the standard target accessory on the three-dimensional model are marked, for example, the corner of an automobile accessory with lines is marked through the line feature points, and the line feature points corresponding to a plurality of different three-dimensional model colors on the three-dimensional model are marked to obtain the target model, so that the standard target accessory can be better acquired, the target accessory in production and processing can be conveniently monitored, and the target accessory has a better data reference function;
in one embodiment, the step S14 of constructing a three-dimensional model of the standard target accessory according to the multi-aspect image includes:
s141, obtaining a contour map of a standard target accessory according to the multi-azimuth image, and paving lines on the surface of the contour map to obtain a net-shaped contour map;
s142, performing color filling on grid positions of the mesh profile to obtain a three-dimensional model;
as described in the steps S141 and S142, a shape outline drawing of the standard target accessory is obtained according to the multi-azimuth image drawing, a net outline drawing is obtained by laying lines according to the outline drawing, then a net part of the net outline drawing is filled by colors to obtain a three-dimensional model of a similar entity type, the lines are laid in the outline drawing, change turning points of the lines appear according to turning changes of the outlines and are used as reference information for marking characteristic points of the lines, the change information of the standard target accessory can be better known through net layout, the comparison of the difference data in the follow-up monitoring of the target accessory is facilitated, and the efficiency of the follow-up monitoring of the surface defects of the target accessory is greatly improved;
s2, setting monitoring grade information of the target accessory, wherein the monitoring grade information is an emergency grade set for a plurality of monitoring item information, and setting a deviation threshold value of the target accessory of the emergency grade;
in one embodiment, the step S21 of formulating the monitoring level information of the target accessory includes:
s211, acquiring monitoring items corresponding to the types of the target accessories, wherein the monitoring items are the positions and the types of the surface defects of the accessories, and acquiring monitoring information based on the corresponding monitoring items of the reference information;
s212, dividing and formulating an emergency grade according to the monitoring items, wherein the emergency grade comprises a first emergency grade and a second emergency grade;
as described in the above steps S211-S212, the type acquisition monitoring item is obtained for the target accessory, wherein the monitoring item is a position of the target accessory, which needs defect monitoring, and affects the parameters of the target accessory, so that the type item is used as the monitoring item, and then the monitoring item reference information is set to the monitoring information for defect monitoring, so that the defect of the monitoring item can be monitored and verified according to the monitoring information, the pertinence of defect verification is greatly improved, and meanwhile, the verification efficiency and accuracy of defect are also improved, emergency grades are formulated according to the classification of the monitoring item, for example, a corner position and angle of the target accessory are used as a first emergency grade, the surface leveling state of the target accessory is used as a second emergency grade, when the first emergency grade is used as a problem, the use of the target accessory is affected seriously, and when the second emergency grade is in a problem, the use of the target accessory is affected in a controllable range, or the defect can be subsequently reworked and corrected, whether the defect can be recycled or can be used in an emergency grade, and the defect can be well classified, and the defect can be normally used is judged, and the defect can be analyzed;
in one embodiment, the step S211 of obtaining the monitoring information based on the reference information corresponding to the monitoring item includes:
s2111, acquiring index information required by defect verification according to a monitoring item based on reference information, wherein the information included in the index information is any one or combination of multiple information of line characteristic points, brightness values and brightness contrast values;
s2112, associating the index information with the monitoring item data as monitoring information;
as described in the above steps S2111-S2112, when the defect of the target accessory is monitored, the index information corresponding to the defect monitoring needs to be invoked when the defect is monitored, wherein the index information includes any one or more of the line feature point, the brightness value and the brightness contrast value, for example, the defect to be monitored is the turning angle of one corner of the target accessory, and therefore the required index information is the turning angle of the corner line and the line; the defects to be monitored are the arc-shaped surface of the target accessory at one position, then the required index information is the radian of an arc-shaped surface line and the flatness of the surface, one defect is monitored and corresponds to one index information, one defect is monitored and corresponds to a detection item, the detection item is related to the index information, namely the detection information, the related index information can be immediately called through the detected detection item, the defect monitoring efficiency is greatly improved, and meanwhile, the defect monitoring accuracy can also be improved;
in one embodiment, the step S22 of setting the deviation threshold of the target accessory of the emergency level includes:
s221, acquiring index information corresponding to the emergency level, and splitting information in the index information to obtain one or more pieces of split information;
s222, formulating deviation thresholds for the split information one by one, wherein the deviation thresholds comprise reasonable deviation thresholds and abnormal deviation thresholds;
as described in the above steps S221-S222, splitting the information monitored by the defect in the index information to obtain one or more splitting information, and making and dividing the splitting information into a reasonable deviation threshold and an abnormal deviation threshold according to the splitting information, wherein the splitting data in the reasonable deviation threshold is in a qualified state, so that the target accessory can be directly processed or used in the next stage, when the splitting data is in the abnormal deviation threshold, the splitting data is represented as an abnormal target accessory, normal use cannot be performed, and processing or use of the next link cannot be normally performed, for example, whether a turning angle has a defect needs to be monitored, therefore, the turning angle of a line feature point and the position of a line turning need to be obtained, wherein the position of the line is a distance between other turning feature points, and therefore, whether the defect exists in one turning angle is split into two pieces of information as the splitting information according to the making deviation threshold, and whether the target accessory is qualified or not can be directly obtained when the splitting information is monitored, so that the efficiency of monitoring the target accessory is improved and whether the target accessory is qualified or not is judged;
s3, acquiring a multi-azimuth monitoring image of the target accessory based on robot vision, and establishing monitoring information according to the multi-azimuth monitoring image, wherein the monitoring information comprises a monitoring model and brightness characteristic information of the multi-azimuth monitoring image of the target accessory;
in one embodiment, the step S3 of collecting the multi-azimuth monitoring image of the target accessory based on the robot vision and establishing the monitoring information according to the multi-azimuth monitoring image includes:
s31, acquiring corresponding monitoring items according to the target accessory;
s32, acquiring a multi-azimuth monitoring image of the target accessory based on robot vision, and obtaining a brightness value and a brightness contrast value of the multi-azimuth monitoring image;
s33, constructing a contour map of the target accessory according to the multi-azimuth monitoring image, and paving the defect monitoring positions corresponding to the monitoring items on the surface of the defect monitoring positions of the contour map through lines to obtain a net-shaped contour map of the monitoring positions;
s34, performing color filling on grid positions of the monitoring position network contour map to obtain a monitoring three-dimensional model;
in addition, since the monitoring items corresponding to the types of the target accessories are not all parts to be monitored, only the positions corresponding to the monitoring items are subjected to grid processing according to the monitoring items of the target accessories to obtain a monitoring three-dimensional model, so that the construction efficiency of the three-dimensional model of the target accessories can be improved, the positions which are not needed are not constructed, and the defect monitoring and identifying efficiency of the target accessories is improved;
s4, comparing the monitoring information with the reference information to obtain distinguishing characteristic information, and evaluating the distinguishing characteristic information based on the monitoring grade information to obtain a defect index of the target accessory;
in one embodiment, the step S4 of comparing the monitoring information with the reference information to obtain the distinguishing characteristic information, and evaluating the distinguishing characteristic information based on the monitoring level information to obtain the defect index of the target accessory includes:
s41, comparing the monitored three-dimensional model with a standard model based on a monitoring item of the target accessory to obtain distinguishing characteristic information;
s42, obtaining emergency grade of the distinguishing characteristic information based on the monitoring grade information, taking the first emergency grade as a comparison optimal monitoring grade, and taking the second emergency grade as a post-setting monitoring grade;
s43, obtaining distinguishing characteristic information meeting a deviation threshold value and an abnormal deviation threshold value respectively based on the emergency level, respectively obtaining a qualified target accessory and a defective target accessory, and marking the distinguishing characteristic information of the defective target accessory as a non-qualified target accessory if the target accessory exists;
s44, obtaining a defect index of the target accessory according to the distinguishing characteristic information, wherein the defect index is calculated according to the following formula:
wherein->For defect index->For the sum of the distinguishing characteristic information data of the qualified target fitting in the case of the first emergency level, +.>For the difference characteristic information data sum of the defective target fittings in the case of the first emergency level, +.>For the first emergency level evaluation factor, +.>For the sum of the distinguishing characteristic information data of the qualified target fitting in the case of the second emergency level, +.>For the second emergency level, the difference characteristic information data sum of the defective target fitting, +.>For the second emergency level evaluation factor, it is to be noted that +.>Numerical value sum +.>The greater the number of (2), i.e. +.>The larger the value of (c) representing a defect of the target fitting, the larger the defect representing a processing quality of the corresponding target fitting, the worse;
s45, taking the defect index which does not meet the threshold condition as low-quality target accessory machining information;
as described in the above steps S41-S45, comparing the obtained monitored three-dimensional model of the target accessory with the standard model to obtain distinguishing characteristic information, for example, taking the difference value of the standard model corresponding to the monitored three-dimensional labyrinth of the current target accessory as the distinguishing characteristic information, obtaining a grade corresponding to the detection item, and performing preferential selection analysis on the distinguishing characteristic information according to the emergency grade; the method comprises the steps that data meeting a deviation threshold value is preferentially judged according to a first emergency level to obtain a qualified target accessory, distinguishing characteristic information meeting an abnormal deviation threshold value is obtained to obtain a defective target accessory, then distinguishing characteristic information of a second emergency level is judged to obtain the qualified target accessory and the defective target accessory respectively, when the defective target accessory is judged to exist in the same target accessory, the fact that the target accessory is a disqualified accessory is represented, the processing is needed, in addition, the processing quality condition of the target accessory can be known according to the defect index of the target accessory, the defect degree of the target accessory can be well known, and meanwhile, whether the target accessory can be recycled and reprocessed to obtain the qualified accessory can be also achieved;
s5, storing the defect index into a corresponding target accessory type to serve as historical defect information, collecting the historical defect information to divide the frequent grades, and determining whether the target accessory type is a high-risk manufacturing target accessory type or not;
in one embodiment, the step S5 of storing the defect index into the corresponding target accessory category as the historical defect information, collecting the historical defect information to perform frequent classification, and determining whether the target accessory category is a high-risk manufacturing target accessory category includes:
s51, storing the target types corresponding to the defect indexes as historical defect information, and obtaining the historical defect indexes of the target fittings according to the historical defect information, wherein the calculation formula of the historical defect indexes is as follows:
the method comprises the steps of carrying out a first treatment on the surface of the In (I)>For historical defect index->For the number of times of occurrence of defect index +.>For the evaluation coefficient of the processing difficulty of the target fitting, it should be noted that +.>And->The greater the number of ++>The larger the number, the more critical the manufacturing of the target fitting, and the more likely the defect;
s52, dividing the frequent-hair grade, and obtaining a corresponding frequent-hair grade according to a frequent-hair grade value which is correspondingly set by the frequent-hair grade, wherein the frequent-hair grade comprises a high-risk grade, a dangerous grade, a general grade and a good grade;
as described in the above steps S51-S52, the defect index is stored to correspond to the type of the target accessory, and then is used as a history defect index, a history defect index is obtained according to the defect index calculation, the processing dangerous condition of the type of the target accessory can be obtained, frequent grades are classified, wherein the frequent grades include a high-risk grade, a dangerous grade, a general grade and a good grade, each grade under the frequent grade corresponds to a frequent grade value, the corresponding frequent grade is obtained according to the frequent grade value satisfied by the history defect index, and further the processing dangerous degree of the target accessory is obtained, the processing defect degree of the target accessory can be known, and the defect degree can be conveniently obtained in the subsequent quality condition of the same processing of the target accessory.
The foregoing is merely illustrative of the present application, and the present application is not limited thereto, and any person skilled in the art will readily recognize that variations or substitutions are within the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (7)
1. The method for monitoring the surface defects of the automobile spare and accessory parts based on the robot vision is characterized by comprising the following steps of:
collecting multi-azimuth images of a standard target accessory, and establishing reference information according to the multi-azimuth images, wherein the reference information comprises a standard model and brightness characteristic information of the multi-azimuth images of the standard target accessory;
the step of collecting the multi-azimuth image of the standard target accessory and establishing the reference information according to the multi-azimuth image comprises the following steps:
selecting a standard part shape of the target accessory as a standard target accessory;
setting standard light information, and acquiring images of a plurality of directions of a standard target accessory based on the standard light information to obtain a multi-direction image;
filtering the colors of the multi-azimuth image to obtain a target image, and storing the brightness value and the brightness contrast value of the target accessory image corresponding to the type of the target accessory as brightness characteristic information of the standard target accessory multi-azimuth image;
constructing a three-dimensional model of the standard target accessory according to the multi-azimuth image, and marking line characteristic points of the target accessory according to the three-dimensional model, wherein the line characteristic points are turning points corresponding to the shape line change of the target accessory, so as to distinguish the color of the three-dimensional model of the target accessory as the line characteristic point marking color, and obtaining the standard model;
the step of constructing a three-dimensional model of the standard target accessory according to the multi-azimuth image comprises the following steps:
obtaining a contour map of a standard target fitting according to the multi-azimuth image, and paving lines on the surface of the contour map to obtain a net-shaped contour map;
color filling is carried out on the grid positions of the mesh profile to obtain a three-dimensional model;
setting monitoring grade information of the target accessory, wherein the monitoring grade information is an emergency grade set for a plurality of monitoring item information, and setting a deviation threshold value of the target accessory of the emergency grade;
acquiring a multidirectional monitoring image of a target accessory based on robot vision, and establishing monitoring information according to the multidirectional monitoring image, wherein the monitoring information comprises a monitoring model and the brightness characteristic information of the multidirectional monitoring image of the target accessory;
and comparing the monitoring information with the reference information to obtain distinguishing characteristic information, and evaluating the distinguishing characteristic information based on the monitoring grade information to obtain the defect index of the target accessory.
2. The method for monitoring the surface defects of the automobile spare and accessory parts based on the robot vision according to claim 1, wherein the method comprises the following steps of: the step of formulating the monitoring grade information of the target accessory comprises the following steps:
collecting monitoring items corresponding to the types of the target accessories, wherein the monitoring items are the positions and types of the surface defects of the accessories, and monitoring information is obtained based on the corresponding monitoring items of the reference information;
and dividing and formulating an emergency grade according to the monitoring items, wherein the emergency grade comprises a first emergency grade and a second emergency grade.
3. The method for monitoring the surface defects of the automobile spare and accessory parts based on the robot vision according to claim 2, wherein the method comprises the following steps of: the step of obtaining the monitoring information based on the corresponding monitoring item of the reference information comprises the following steps:
acquiring index information required by defect verification according to a monitoring item based on reference information, wherein the information included in the index information is any one or combination of multiple information of line characteristic points, brightness values and brightness contrast values respectively;
the index information is associated with the monitoring item data as monitoring information.
4. A method for monitoring surface defects of automobile parts based on robot vision according to claim 3, wherein: the step of setting the deviation threshold of the urgent grade target accessory comprises the following steps:
acquiring index information corresponding to the emergency level, splitting information in the index information, and obtaining one or more pieces of split information;
and formulating deviation thresholds for the split information one by one, wherein the deviation thresholds comprise reasonable deviation thresholds and abnormal deviation thresholds.
5. The method for monitoring the surface defects of the automobile spare and accessory parts based on the robot vision according to claim 1, wherein the method comprises the following steps of: the robot vision-based multi-azimuth monitoring image of the target accessory is collected, and the step of establishing monitoring information according to the multi-azimuth monitoring image comprises the following steps:
acquiring corresponding monitoring items according to the target accessory;
acquiring a multi-azimuth monitoring image of the target accessory based on robot vision, and obtaining a brightness value and a brightness contrast value of the multi-azimuth monitoring image;
constructing a contour map of the target accessory according to the multidirectional monitoring image, and paving the contour map on the surface of the contour map defect monitoring position through lines based on the defect monitoring position corresponding to the monitoring item to obtain a net-shaped contour map of the monitoring position;
and performing color filling on the grid positions of the monitoring position network contour map to obtain a monitoring three-dimensional model.
6. The method for monitoring the surface defects of the automobile spare and accessory parts based on the robot vision according to claim 1, wherein the method comprises the following steps of: the step of comparing the monitoring information with the reference information to obtain the distinguishing characteristic information, and evaluating the distinguishing characteristic information based on the monitoring grade information to obtain the defect index of the target accessory comprises the following steps:
comparing the monitoring three-dimensional model with the standard model based on the monitoring item of the target accessory to obtain distinguishing characteristic information;
obtaining emergency grade of the distinguishing characteristic information based on the monitoring grade information, taking the first emergency grade as a comparison optimal monitoring grade, and taking the second emergency grade as a post-setting monitoring grade;
obtaining distinguishing characteristic information meeting a deviation threshold value and an abnormal deviation threshold value respectively based on the emergency level, respectively obtaining a qualified target accessory and a defective target accessory, and marking the distinguishing characteristic information of the defective target accessory as a disqualified target accessory if the target accessory exists;
obtaining a defect index of the target accessory according to the distinguishing characteristic information;
the defect index that does not satisfy the threshold condition is used as the low quality target accessory machining information.
7. The method for monitoring the surface defects of the automobile spare and accessory parts based on the robot vision according to claim 1, wherein the method comprises the following steps of: the method also comprises the steps of storing the defect index into the corresponding target accessory type as historical defect information, collecting the historical defect information to carry out frequent grade division, and determining whether the target accessory type is a high-risk manufacturing target accessory type or not, and comprises the following steps:
storing the corresponding target types of the defect indexes as historical defect information, and obtaining the historical defect indexes of the target accessories according to the historical defect information;
and (3) dividing the frequent-hair grade, and obtaining a corresponding frequent-hair grade according to a frequent-hair grade value which is correspondingly set by the frequent-hair grade, wherein the frequent-hair grade comprises a high-risk grade, a dangerous grade, a general grade and a good grade.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311042712.5A CN116805311B (en) | 2023-08-18 | 2023-08-18 | Automobile part surface defect monitoring method based on robot vision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311042712.5A CN116805311B (en) | 2023-08-18 | 2023-08-18 | Automobile part surface defect monitoring method based on robot vision |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116805311A CN116805311A (en) | 2023-09-26 |
| CN116805311B true CN116805311B (en) | 2023-11-07 |
Family
ID=88079547
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311042712.5A Active CN116805311B (en) | 2023-08-18 | 2023-08-18 | Automobile part surface defect monitoring method based on robot vision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116805311B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117236138B (en) * | 2023-11-08 | 2024-01-30 | 无锡学院 | Digital twinning-based robot motion control and state monitoring method and system |
| CN117252486B (en) * | 2023-11-14 | 2024-02-02 | 长春师范大学 | Automobile part defect detection method and system based on Internet of things |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103941569A (en) * | 2014-04-30 | 2014-07-23 | 天津工业大学 | Method for LED multi-angle holographic three-dimensional data representation |
| CN111461210A (en) * | 2020-03-31 | 2020-07-28 | 天津智惠未来科技有限责任公司 | Unmanned aerial vehicle-based wind power inspection blade defect grade determination method |
| CN112013789A (en) * | 2020-10-27 | 2020-12-01 | 山东海德智能科技有限公司 | High-precision part deviation detection system based on 3D vision algorithm |
| CN114998217A (en) * | 2022-05-10 | 2022-09-02 | 济宁海富光学科技有限公司 | Method for determining defect grade of glass substrate, computer device and storage medium |
| CN115047142A (en) * | 2022-05-24 | 2022-09-13 | 中国铁道科学研究院集团有限公司铁道建筑研究所 | Method and system for analyzing tunnel lining quality |
| WO2023068055A1 (en) * | 2021-10-20 | 2023-04-27 | 株式会社神戸製鋼所 | Method and device for monitoring welding, and method and device for laminate molding |
| CN116309564A (en) * | 2023-05-17 | 2023-06-23 | 厦门微图软件科技有限公司 | Method and system for detecting appearance defects of battery cells based on artificial intelligent image recognition |
-
2023
- 2023-08-18 CN CN202311042712.5A patent/CN116805311B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103941569A (en) * | 2014-04-30 | 2014-07-23 | 天津工业大学 | Method for LED multi-angle holographic three-dimensional data representation |
| CN111461210A (en) * | 2020-03-31 | 2020-07-28 | 天津智惠未来科技有限责任公司 | Unmanned aerial vehicle-based wind power inspection blade defect grade determination method |
| CN112013789A (en) * | 2020-10-27 | 2020-12-01 | 山东海德智能科技有限公司 | High-precision part deviation detection system based on 3D vision algorithm |
| WO2023068055A1 (en) * | 2021-10-20 | 2023-04-27 | 株式会社神戸製鋼所 | Method and device for monitoring welding, and method and device for laminate molding |
| CN114998217A (en) * | 2022-05-10 | 2022-09-02 | 济宁海富光学科技有限公司 | Method for determining defect grade of glass substrate, computer device and storage medium |
| CN115047142A (en) * | 2022-05-24 | 2022-09-13 | 中国铁道科学研究院集团有限公司铁道建筑研究所 | Method and system for analyzing tunnel lining quality |
| CN116309564A (en) * | 2023-05-17 | 2023-06-23 | 厦门微图软件科技有限公司 | Method and system for detecting appearance defects of battery cells based on artificial intelligent image recognition |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116805311A (en) | 2023-09-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116805311B (en) | Automobile part surface defect monitoring method based on robot vision | |
| CN109685760B (en) | MATLAB-based SLM powder bed powder laying image convex hull depression defect detection method | |
| CN103714077A (en) | Method and device for retrieving objects and method and device for verifying retrieval | |
| CN107966444B (en) | Textile flaw detection method based on template | |
| CN115690108A (en) | Aluminum alloy rod production quality evaluation method based on image processing | |
| CN105550938B (en) | Method for testing abnormal value of county-area cultivated land quality evaluation result | |
| CN111127571A (en) | Small sample defect classification method and device | |
| CN115144399B (en) | Assembly quality detection method and device based on machine vision | |
| CN112784821A (en) | Building site behavior safety detection and identification method and system based on YOLOv5 | |
| CN105302123A (en) | Online data monitoring method | |
| CN108921840A (en) | Display screen peripheral circuit detection method, device, electronic equipment and storage medium | |
| CN110288624A (en) | Detection method, device and the relevant device of straightway in a kind of image | |
| CN112183264B (en) | Method for judging someone remains under crane boom based on spatial relationship learning | |
| CN117036353B (en) | Temperature-resistant foam coating detection method for new energy battery | |
| CN113033401A (en) | Human activity change recognition and supervision method for ecological protection red line | |
| CN115240146A (en) | Intelligent machine tool assembly acceptance method based on computer vision | |
| CN116883446B (en) | Real-time monitoring system for grinding degree of vehicle-mounted camera lens | |
| KR20160087600A (en) | Apparatus for inspecting defect and method thereof | |
| CN116993733B (en) | Earphone sleeve appearance quality detection method and system | |
| CN113778091A (en) | Method for inspecting equipment of wind power plant booster station | |
| CN112836967A (en) | New energy automobile battery safety risk assessment system | |
| CN105095897B (en) | A kind of digit recognition method based on gradient image and Similarity-Weighted | |
| Powell et al. | Automated road distress detection | |
| WO2000028309A1 (en) | Method for inspecting inferiority in shape | |
| US20190325606A1 (en) | Inspection apparatus |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |