CN116441200A - Online intelligent detection system and method for code spraying defects - Google Patents
Online intelligent detection system and method for code spraying defects Download PDFInfo
- Publication number
- CN116441200A CN116441200A CN202310172312.XA CN202310172312A CN116441200A CN 116441200 A CN116441200 A CN 116441200A CN 202310172312 A CN202310172312 A CN 202310172312A CN 116441200 A CN116441200 A CN 116441200A
- Authority
- CN
- China
- Prior art keywords
- character
- code spraying
- image
- characters
- defect
- 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.)
- Pending
Links
- 238000005507 spraying Methods 0.000 title claims abstract description 167
- 230000007547 defect Effects 0.000 title claims abstract description 104
- 238000001514 detection method Methods 0.000 title claims abstract description 53
- 238000000034 method Methods 0.000 title claims abstract description 30
- 238000012545 processing Methods 0.000 claims abstract description 40
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 37
- 238000004458 analytical method Methods 0.000 claims abstract description 21
- 230000002950 deficient Effects 0.000 claims abstract description 14
- 238000007405 data analysis Methods 0.000 claims abstract description 11
- 238000007689 inspection Methods 0.000 claims abstract description 9
- 238000004891 communication Methods 0.000 claims abstract description 4
- 230000006872 improvement Effects 0.000 claims description 21
- 230000011218 segmentation Effects 0.000 claims description 18
- 238000012423 maintenance Methods 0.000 claims description 16
- 230000003449 preventive effect Effects 0.000 claims description 15
- 238000007781 pre-processing Methods 0.000 claims description 12
- 238000012549 training Methods 0.000 claims description 11
- 230000002265 prevention Effects 0.000 claims description 8
- 239000007921 spray Substances 0.000 claims description 7
- 238000001914 filtration Methods 0.000 claims description 6
- 238000009499 grossing Methods 0.000 claims description 6
- 230000005611 electricity Effects 0.000 claims description 5
- 238000012986 modification Methods 0.000 claims description 5
- 230000004048 modification Effects 0.000 claims description 5
- 230000003068 static effect Effects 0.000 claims description 5
- 230000003993 interaction Effects 0.000 claims description 4
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 claims description 3
- 238000013527 convolutional neural network Methods 0.000 claims description 3
- 238000013135 deep learning Methods 0.000 claims description 3
- 238000009826 distribution Methods 0.000 claims description 3
- 230000002708 enhancing effect Effects 0.000 claims description 3
- 238000000605 extraction Methods 0.000 claims description 3
- 238000010030 laminating Methods 0.000 claims description 3
- 239000011159 matrix material Substances 0.000 claims description 3
- 230000009467 reduction Effects 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 238000012163 sequencing technique Methods 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 239000011248 coating agent Substances 0.000 claims description 2
- 238000000576 coating method Methods 0.000 claims description 2
- 230000008439 repair process Effects 0.000 claims description 2
- 238000004806 packaging method and process Methods 0.000 abstract description 9
- 238000004519 manufacturing process Methods 0.000 abstract description 6
- 230000008569 process Effects 0.000 description 9
- 241000208125 Nicotiana Species 0.000 description 8
- 235000002637 Nicotiana tabacum Nutrition 0.000 description 8
- 235000019504 cigarettes Nutrition 0.000 description 8
- 239000000463 material Substances 0.000 description 4
- 238000009434 installation Methods 0.000 description 3
- 238000003908 quality control method Methods 0.000 description 3
- 230000007613 environmental effect Effects 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000012797 qualification Methods 0.000 description 2
- 239000000779 smoke Substances 0.000 description 2
- 229920000298 Cellophane Polymers 0.000 description 1
- 241001292396 Cirrhitidae Species 0.000 description 1
- 229920004449 Halon® Polymers 0.000 description 1
- 229910000831 Steel Inorganic materials 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- PXBRQCKWGAHEHS-UHFFFAOYSA-N dichlorodifluoromethane Chemical compound FC(F)(Cl)Cl PXBRQCKWGAHEHS-UHFFFAOYSA-N 0.000 description 1
- 239000011521 glass Substances 0.000 description 1
- 238000010409 ironing Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000001179 sorption measurement Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 239000010959 steel Substances 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
- B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
- B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
- B07C5/36—Sorting apparatus characterised by the means used for distribution
- B07C5/361—Processing or control devices therefor, e.g. escort memory
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
- B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
- B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
- B07C5/34—Sorting according to other particular properties
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
- B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
- B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
- B07C5/36—Sorting apparatus characterised by the means used for distribution
- B07C5/361—Processing or control devices therefor, e.g. escort memory
- B07C5/362—Separating or distributor mechanisms
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/50—Constructional details
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N25/00—Circuitry of solid-state image sensors [SSIS]; Control thereof
- H04N25/70—SSIS architectures; Circuits associated therewith
- H04N25/76—Addressed sensors, e.g. MOS or CMOS sensors
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Signal Processing (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Multimedia (AREA)
- Quality & Reliability (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The invention discloses an online intelligent detection system and method for code spraying defects, comprising an industrial personal computer, an electric control unit, an image acquisition unit, an image processing unit and a data acquisition analysis unit which are in communication connection with the industrial personal computer. The system can accurately identify products with quality defects and accurately reject the products with the quality defects through online shooting and image algorithm processing, can realize full-automatic detection and rejection of the quality defects of the sprayed codes, and can actively prevent and control influencing factors and time periods of the quality defects easily through big data analysis, and trend inspection early warning is carried out on the quality problems which do not occur yet, so that the passive rejection is changed into active risk early warning. The invention can greatly improve the detection rate of the code-spraying defective products, stabilize the product packaging quality, improve the stability of production operation equipment, and simultaneously reduce the labor intensity of manual frequent spot check.
Description
Technical Field
The invention relates to the field of code spraying quality detection, in particular to an online intelligent detection system and method for a code spraying defect of transparent paper.
Background
In the cigarette production process, part of products cancel the traditional steel seal number for anti-counterfeiting and flow direction tracking, and an online code spraying device is adopted to spray the ink number on the transparent paper material of the GD packaging machine in real time, so that the operation of the code spraying machine is easily influenced by various aspects such as ink quality, maintenance, site environment and the like, the quality problems such as missing code spraying, wrong code spraying, light code spraying color, code spraying position deviation and the like can be caused in the actual production process, and the quality control is unreliable by the whole manual work due to the fact that an online code spraying detection system is not provided, and the quality control capability is not at all for the occasional code spraying quality problem, so that the product quality is seriously influenced.
At present, the industry has a method for checking the outer packaging paper, and the method for checking the outer packaging paper after the material is unfolded and sprayed with codes is unreliable because the processes of guide roller conveying, negative pressure adsorption conveying, packaging folding, side seal ironing and the like are required from the package paper unfolding to package forming.
Disclosure of Invention
In order to overcome the problems, the application provides an online intelligent detection system and an online intelligent detection method for the code spraying defects of transparent paper, which can accurately identify products with quality defects and accurately reject the products with quality defects through online shooting and image algorithm processing, realize full-automatic detection rejection of the code spraying quality defects, simultaneously judge preventive trend of the frequently-sent code spraying quality problems, improve the detection rate of the code spraying defect products, stabilize the product packaging quality, and reduce the labor intensity of manual frequent spot check.
The technical scheme adopted for solving the technical problems is as follows:
the intelligent on-line detection system for code spraying defect includes industrial computer and communication connection with the same
The electronic control unit is used for generating a detection phase trigger signal and a defect product rejection signal;
the image acquisition unit comprises a camera and a light source, and is arranged behind the transparent paper coating station and before the laminating station;
the image processing unit is used for judging the code spraying defects of the real-time acquired images based on the defect recognition algorithm and feeding the defective products back to the electronic control unit for removal;
the data acquisition analysis unit is used for carrying out data analysis on the generated code spraying defects to obtain a quality influence factor and a time period of the frequently-transmitted code spraying so as to actively prevent and control the frequently-transmitted code spraying; the method is used for carrying out trend inspection early warning on the code spraying defects which do not occur based on an intelligent prediction model, and analyzing by combining code spraying quality influence factors and equipment operation data so as to carry out preventive improvement or maintenance.
As an improvement of the technical scheme, the electronic control unit generates a detection phase trigger signal of the camera through an original machine phase angle, wherein the original machine phase angle refers to a rotation angle of a main encoder of GD packaging machine equipment, and reflects the position of a corresponding product in the product operation process.
As an improvement of the technical scheme, the camera adopts a CMOS high-speed industrial camera.
As an improvement of the above technical solution, the defect recognition algorithm includes an image preprocessing algorithm and a defect determination algorithm;
the image preprocessing algorithm is used for carrying out image basic processing, image depth processing, character denoising segmentation and single character processing on the acquired basic image, and accurately segmenting the code-spraying character region so as to realize image noise reduction and character enhancement, so that preparation is made for next character recognition judgment; the specific method of the image preprocessing algorithm comprises the following steps:
(1) image basic processing: the method comprises the steps of reading an acquired image, acquiring a single-channel picture of an RGB image, and carrying out corresponding rotation and preliminary code spraying area extraction;
(2) image depth processing:
firstly, converting an extracted preliminary code spraying area image into a gray image, enhancing image contrast, and smoothing the image by using a mean smoothing filter according to an output result, so as to obtain a dynamic gray threshold value of the position of each character;
then, dynamically dividing the picture character region through a global dynamic gray threshold value, so as to obtain an image code-spraying character region;
finally, the obtained code-spraying character area is processed: firstly, screening the width and the area of a code spraying area by utilizing a characteristic distribution histogram, removing meaningless areas, secondly, corroding an image to achieve the purposes of increasing dark parts and reducing bright parts, thirdly, self-adaptively carrying out image rotation by utilizing an affine transformation matrix according to the character angle of the obtained image characters, and adjusting the image characters to be horizontally displayed;
(3) character denoising segmentation:
firstly, performing binarization segmentation on an image subjected to advanced treatment, extracting a current character string region, performing connected domain analysis on characters in the region, and filtering meaningless miscellaneous points through the area;
then, each character lattice point of the lattice characters is connected by a closed operation, so that the characters can be connected completely and are not divided into a plurality of parts; the adhesion between the characters is broken by utilizing open operation, so that each character can be independently segmented, and the subsequent character recognition is not influenced;
finally, by carrying out circumscribed rectangle on each character area to obtain a plurality of independent rectangle areas, and by identifying rectangle areas far exceeding single characters, carrying out length re-segmentation on the average width of the characters to prevent the occurrence of characters which are adhered too closely and are not segmented, so as to finish the de-noising segmentation of the code-spraying character strings;
(4) single character processing:
and (3) carrying out single-region processing again on the single character region after denoising and segmentation to realize further denoising of the character region, respectively carrying out connected region analysis and noise point filtering on each character region, merging and connecting the characters in each rectangular region again, sequencing the character regions, counting the final number of the characters, and finally obtaining clear character blocks of a plurality of different regions.
The defect determination algorithm is a Halcon-based basic image digital algorithm.
The defect judgment algorithm comprises the following specific methods:
(1) defining the correct code spraying character on duty:
automatically generating a correct code-spraying character string of the shift through the actual meaning of the code-spraying;
(2) training a character classification template:
firstly, establishing a character model library, carrying out actual image acquisition and image preprocessing by adjusting an inkjet printer to spray different characters of 0-9 and A-M to obtain single segmentation pictures of different characters, and then training and updating the character library by utilizing a large number of inkjet character images of different shifts and dates to finally obtain an inkjet character library template;
(3) character recognition judgment:
comparing and judging single characters segmented by the preprocessed image with different characters in a character library template, judging and identifying each code spraying character, finally outputting a section of character string from the obtained identification result, comparing the character string with a defined correct code spraying character string, and if the character string is identical to the defined correct code spraying character string, judging that the character string is qualified, and if the character string is not identical to the defined correct code spraying character string, judging that the character string is a code spraying defect.
As an improvement of the technical scheme, the intelligent prediction model is a code spraying defect prediction model established by training based on a YOLOv5 deep learning convolutional neural network algorithm.
The establishment process of the code spraying defect prediction model comprises the following steps:
classifying original code spraying defect pictures acquired by equipment, classifying the original code spraying defect pictures into 11 types according to code spraying quality influence factors of code spraying defect tobacco packages, namely material static electricity, transparent paper replacement, environmental light source influence, shutdown of a code spraying machine, equipment shutdown, no smoke of the equipment, shaking of the transparent paper, loosening of the transparent paper, flip-chip of the tobacco packages, damage of the tobacco packages and dirt entrainment, respectively carrying out feature calibration on each type of code spraying defect pictures, and obtaining a code spraying defect prediction model of the equipment through Python training;
the application of the code spraying defect prediction model comprises the following steps:
(1) data analysis is carried out on the generated code spraying defects to obtain the quality influence factors and the time period of the frequently-transmitted code spraying so as to actively prevent and control:
and (3) classifying and judging the code spraying defect tobacco bale pictures which occur in the current period by utilizing a prediction model, and counting the occurrence frequency and the number of the code spraying defects and the code spraying quality influence factors of the code spraying defects in different machine stations, shifts and time periods based on the classification judgment:
judging the running state of the equipment code spraying;
reminding an operator of the frequent reason and the frequent time period of the current removing of the cigarette packet with the code spraying defect so as to actively maintain and adjust and reduce the generation of unqualified products;
(2) trending detection and early warning are carried out on the code spraying defects which do not occur based on an intelligent prediction model, and analysis is carried out by combining code spraying quality influence factors and equipment operation data so as to carry out preventive improvement or maintenance:
according to the judgment and statistics result of the prediction model, combining the code spraying quality influence factors which cause the code spraying defect to have a large number of cigarette packets and high frequency, and carrying out early warning prompt on the code spraying defect with the highest probability of occurrence;
based on the equipment operation data, the frequent reasons are improved or maintained in a preventive manner, and the code spraying qualification rate of the equipment is continuously improved.
As an improvement of the above technical solution, the preventive improvement and maintenance includes:
(1) adjusting a bracket of the code spraying equipment;
(2) the maintenance of the ink jet numbering machine and the spray gun;
(3) adjustment of the static electricity eliminating device;
(4) setting and optimizing detection parameters of the industrial personal computer;
(5) adjusting OPC trigger parameters of equipment;
(6) technical improvements or modifications are made to the main cause of the problem.
The online intelligent detection method of the code spraying defect comprises the following steps of
The electric control unit sets a detection phase triggering angle;
the industrial personal computer converts the detection phase trigger signal to an image acquisition unit for image acquisition;
the industrial personal computer performs display interaction on the acquired images, and performs image processing based on a defect recognition algorithm; when the image product is judged to be a defective product, the industrial personal computer sends a signal to the electronic control unit, and the electronic control unit generates a defective product rejection signal to reject the defective product at a rejection port;
the industrial personal computer transmits the graphic processing result and the equipment operation data to the data acquisition analysis unit, and performs data analysis on the generated code spraying defects to obtain code spraying quality influence factors and frequent time periods for active prevention and control; and carrying out trend inspection early warning on the code spraying defects which do not occur based on an intelligent prediction model, and analyzing by combining the code spraying quality influence factors and equipment operation data so as to carry out preventive improvement or maintenance.
The invention has the beneficial effects that:
the method and the device can realize the online code spraying quality detection of each package of the coated transparent paper finished product, integrate an original machine control system to accurately trigger and reject, intelligently analyze code spraying quality problem influence factors and make trend prediction by combining image processing results and operation data, improve and optimize key nodes, continuously reduce the generation of code spraying quality problems, and improve the code spraying quality control capability of the transparent paper.
The wisdom prevention and control of this application mainly falls into three stages:
the first stage: detection rejection
Confirming the unique installation position of the detection assembly, enabling the detection assembly to effectively detect unqualified cigarette packets based on an efficient and reliable defect recognition algorithm, and enabling the unqualified cigarette packets to be triggered and removed accurately through linkage of an electric control unit so as to prevent unqualified cigarettes from flowing into a next process;
and a second stage: active prevention and control
The image acquisition data and the equipment operation data are transmitted into a data acquisition analysis unit in real time to carry out big data analysis, the influence factors and frequent time periods of unqualified products are confirmed, active prevention and control are carried out on the time periods of easy occurrence of quality defects, trend inspection early warning is carried out on the quality problems which do not occur yet, and passive rejection is changed into active risk early warning;
and a third stage: improved optimization
And by combining with the data statistics analysis result, the nodes with the frequency quality problem are improved and optimized, and the production equipment parameters are subjected to self-adaptive dynamic regulation and control, so that the production system has intelligence, and finally, under the continuous cyclic optimization, the generation of unqualified products is reduced, and the normal operation of the equipment is ensured.
The invention can greatly improve the detection rate of the code-spraying defective products, stabilize the product packaging quality, improve the stability of production operation equipment, and simultaneously reduce the labor intensity of manual frequent spot check.
Drawings
The invention is further described with reference to the drawings and the specific embodiments below:
FIG. 1 is a block diagram of an online intelligent detection system for code spraying defects;
FIG. 2 is a flow chart of the method for online intelligent detection of code spraying defects.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Example 1
Referring to fig. 1, the on-line intelligent detection system for the code spraying defect comprises an industrial personal computer, an electric control unit, an image acquisition unit, an image processing unit and a data acquisition analysis unit, wherein the electric control unit, the image acquisition unit and the image processing unit are in communication connection with the industrial personal computer.
Referring to fig. 2, the basic flow of the intelligent detection system is as follows:
the G.D electric control cabinet generates signals when corresponding to the detection phase, the signals trigger a camera and a light source of the image acquisition unit to shoot through the industrial personal computer, the shot images are uploaded to the industrial personal computer, the industrial personal computer performs display interaction by utilizing a VC interface, performs image processing by utilizing a defect recognition algorithm, and products which are unqualified in processing judgment are fed back to the electric control unit and are accurately removed at a corresponding original machine removing station. Meanwhile, in the running process of the system, all equipment running data are uploaded to a data acquisition analysis unit in real time for statistical analysis, and analysis results and defect reasons are fed back and early-warned to operation and repair personnel and management and quality inspection personnel respectively, so that the real-time condition of the code spraying quality of each machine is mastered at any time, and the factors of frequently occurring code spraying quality problems are eliminated.
In this embodiment, the electronic control unit is mainly used for generating a detection phase trigger signal and a defect product rejection signal. The electronic control unit generates a detection phase trigger signal of the camera through an original machine phase angle, wherein the original machine phase angle refers to a rotation angle of a main encoder of the GD packaging machine equipment, and reflects the position of a corresponding product in the product operation process.
In this embodiment, the image acquisition unit includes a camera, a bracket, and a light source, which are mounted after the cellophane wrapping station and before the laminating station. The camera adopts a CMOS high-speed industrial camera, the installation of the camera does not affect the operation, maintenance and operation of the equipment, the installation and the timely detection of the camera at the position of the coated product of the transparent paper are required, each finished product is required to be clearly detected, the detection and the image processing calculation are completed before the rejection opening is rejected, and the effective rejection can be realized.
In this embodiment, the image processing unit is mainly used for judging the code spraying defect of the real-time collected image based on the defect recognition algorithm, and feeding back the defective product to the electronic control unit for removal.
Because of poor field detection environment, the noise of the basic image is large, the processing difficulty of the basic image is high, the variety of the quality defects of the code spraying is large, and various conditions of deformation can occur in the spraying process in the operation process, so that a visual algorithm is relatively complex. For this reason, the defect recognition algorithm adopted in the present embodiment includes an image preprocessing algorithm and a defect determination algorithm:
the image preprocessing algorithm is used for carrying out image basic processing, image depth processing, character denoising segmentation and single character processing on the acquired basic image, and accurately segmenting the code-spraying character region so as to realize image noise reduction and character enhancement, and is ready for next character recognition and judgment.
In a preferred embodiment, the specific method of the image preprocessing algorithm is as follows:
(1) image basic processing: the method comprises the steps of reading an acquired image, acquiring a single-channel picture of an RGB image, and carrying out corresponding rotation and preliminary code spraying area extraction;
(2) image depth processing:
firstly, converting an extracted preliminary code spraying area image into a gray image, enhancing image contrast, and smoothing the image by using a mean smoothing filter according to an output result, so as to obtain a dynamic gray threshold value of the position of each character;
then, dynamically dividing the picture character region through a global dynamic gray threshold value, so as to obtain an image code-spraying character region;
finally, the obtained code-spraying character area is processed: firstly, screening the width and the area of a code spraying area by utilizing a characteristic distribution histogram, removing meaningless areas, secondly, corroding an image to achieve the purposes of increasing dark parts and reducing bright parts, thirdly, self-adaptively carrying out image rotation by utilizing an affine transformation matrix according to the character angle of the obtained image characters, and adjusting the image characters to be horizontally displayed;
(3) character denoising segmentation:
firstly, performing binarization segmentation on an image subjected to advanced treatment, extracting a current character string region, performing connected domain analysis on characters in the region, and filtering meaningless miscellaneous points through the area;
then, each character lattice point of the lattice characters is connected by a closed operation, so that the characters can be connected completely and are not divided into a plurality of parts; the adhesion between the characters is broken by utilizing open operation, so that each character can be independently segmented, and the subsequent character recognition is not influenced;
finally, by carrying out circumscribed rectangle on each character area to obtain a plurality of independent rectangle areas, and by identifying rectangle areas far exceeding single characters, carrying out length re-segmentation on the average width of the characters to prevent the occurrence of characters which are adhered too closely and are not segmented, so as to finish the de-noising segmentation of the code-spraying character strings;
(4) single character processing:
and (3) carrying out single-region processing again on the single character region after denoising and segmentation to realize further denoising of the character region, respectively carrying out connected region analysis and noise point filtering on each character region, merging and connecting the characters in each rectangular region again, sequencing the character regions, counting the final number of the characters, and finally obtaining clear character blocks of a plurality of different regions.
The defect decision algorithm is a halon-based base image digital algorithm.
In a preferred embodiment, the specific method of the defect determination algorithm is:
(1) defining the correct code spraying character on duty:
automatically generating and defining correct code spraying character strings of the shift through the actual meaning of the code spraying;
(2) training a character classification template:
firstly, establishing a character model library, carrying out actual image acquisition and image preprocessing by adjusting an inkjet printer to spray different characters of 0-9 and A-M to obtain single segmentation pictures of different characters, and then training and updating the character library by utilizing a large number of inkjet character images of different shifts and dates to finally obtain an inkjet character library template;
(3) character recognition judgment:
comparing and judging single characters segmented by the preprocessed image with different characters in a character library template, judging and identifying each code spraying character, finally outputting a section of character string from the obtained identification result, comparing the character string with a defined correct code spraying character string, and if the character string is identical to the defined correct code spraying character string, judging that the character string is qualified, and if the character string is not identical to the defined correct code spraying character string, judging that the character string is a code spraying defect.
In this embodiment, the data analysis unit may perform data analysis on the occurring code spraying defect to obtain a code spraying quality influence factor and a frequent time period for active prevention and control; and carrying out trend inspection early warning on the code spraying defects which do not occur based on an intelligent prediction model, and analyzing by combining the code spraying quality influence factors and equipment operation data so as to carry out preventive improvement or maintenance.
And accessing the data acquisition platform into a workshop wrapping equipment big data platform, establishing a code spraying defect prediction model based on a YOLOv5 deep learning convolutional neural network algorithm, programming and writing the algorithm through PHP language, and training and reasoning through Python. The establishment process of the code spraying defect prediction model comprises the following steps:
the method comprises the steps of classifying original code spraying defect pictures collected by equipment, classifying the original code spraying defect pictures into 11 types according to code spraying quality influence factors of code spraying defect tobacco packages, namely static electricity of materials, transparent paper replacement, influence of an environmental light source, shutdown of a code spraying machine, equipment shutdown, no smoke of the equipment, shaking of the transparent paper, loosening of the transparent paper, flip-chip of the tobacco packages, damage of the tobacco packages and dirt entrainment, respectively carrying out feature calibration on each type of code spraying defect pictures, and obtaining a code spraying defect prediction model of the equipment through Python training.
The specific application of the code spraying defect prediction model is as follows:
(1) data analysis is carried out on the generated code spraying defects to obtain the quality influence factors and the time period of the frequently-transmitted code spraying so as to actively prevent and control:
and (3) classifying and judging the code spraying defect tobacco bale pictures which occur in the current period by utilizing a prediction model, and counting the occurrence frequency and the number of the code spraying defects and the code spraying quality influence factors of the code spraying defects in different machine stations, shifts and time periods based on the classification judgment:
judging the running state of the equipment code spraying;
reminding an operator of the frequent reason and the frequent time period of the current removing of the cigarette packet with the code spraying defect so as to actively maintain and adjust and reduce the generation of unqualified products;
(2) trending detection and early warning are carried out on the code spraying defects which do not occur based on an intelligent prediction model, and analysis is carried out by combining code spraying quality influence factors and equipment operation data so as to carry out preventive improvement or maintenance:
according to the judgment and statistics result of the prediction model, combining the code spraying quality influence factors which cause the code spraying defect to have a large number of cigarette packets and high frequency, and carrying out early warning prompt on the code spraying defect with the highest probability of occurrence;
based on the equipment operation data, the frequent reasons are improved or maintained in a preventive manner, and the code spraying qualification rate of the equipment is continuously improved.
For example, aiming at unstable vehicle speed, glass paper shaking, encoder slipping and other possible influence factors, intelligent analysis and code spraying quality risk early warning are carried out by combining system operation statistical data, and according to analysis results, preventive improvement and maintenance are carried out, so that key influence factors are subjected to self-adaptive dynamic adjustment, the generation of unqualified products is reduced, and normal operation of equipment is ensured.
Wherein the preventive improvement or maintenance may also be:
(1) adjusting a bracket of the code spraying equipment;
(2) the maintenance of the ink jet numbering machine and the spray gun;
(3) adjustment of the static electricity eliminating device;
(4) setting and optimizing detection parameters of the industrial personal computer;
(5) adjusting OPC trigger parameters of equipment;
(6) technical improvements or modifications are made to the main cause of the problem.
Example 2
Referring to FIG. 2, the online intelligent detection method of the code spraying defect comprises the following steps of
The electric control unit sets a detection phase triggering angle;
the industrial personal computer converts the detection phase trigger signal to an image acquisition unit for image acquisition;
the industrial personal computer performs display interaction on the acquired images, and performs image processing based on a defect recognition algorithm; when the image product is judged to be a defective product, the industrial personal computer sends a signal to the electronic control unit, and the electronic control unit generates a defective product rejection signal to reject the defective product at a rejection port;
the industrial personal computer transmits the graphic processing result and the equipment operation data to the data acquisition analysis unit, and performs data analysis on the generated code spraying defects to obtain code spraying quality influence factors and frequent time periods for active prevention and control; and carrying out trend inspection early warning on the code spraying defects which do not occur based on an intelligent prediction model, and analyzing by combining the code spraying quality influence factors and equipment operation data so as to carry out preventive improvement or maintenance.
It should be noted that: the foregoing description is only illustrative of the preferred embodiments of the present invention, and although the present invention has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that modifications may be made to the embodiments described, or equivalents may be substituted for elements thereof, and any modifications, equivalents, improvements or changes may be made without departing from the spirit and principles of the present invention.
Claims (8)
1. Spout online wisdom detecting system of sign indicating number defect, its characterized in that: comprises an industrial personal computer and a communication connection with the industrial personal computer
The electronic control unit is used for generating a detection phase trigger signal and a defect product rejection signal;
the image acquisition unit comprises a camera and a light source, and is arranged behind the transparent paper coating station and before the laminating station;
the image processing unit is used for judging the code spraying defects of the real-time acquired images based on the defect recognition algorithm and feeding the defective products back to the electronic control unit for removal;
the data acquisition analysis unit is used for carrying out data analysis on the generated code spraying defects to obtain code spraying quality influence factors and frequent time periods for active prevention and control; the method is used for carrying out trend inspection early warning on the code spraying defects which do not occur based on an intelligent prediction model, and analyzing by combining code spraying quality influence factors and equipment operation data so as to carry out preventive improvement or maintenance.
2. The online intelligent detection system for code spraying defects according to claim 1, wherein:
the electric control unit generates a detection phase trigger signal of the camera by using the rotation angle of the equipment main encoder.
3. The online intelligent detection system for code spraying defects according to claim 1, wherein:
the defect identification algorithm comprises an image preprocessing algorithm and a defect judging algorithm;
the image preprocessing algorithm is used for carrying out image basic processing, image depth processing, character denoising segmentation and single character processing on the acquired basic image, and accurately segmenting the code-spraying character region so as to realize image noise reduction and character enhancement, and is ready for next character recognition and judgment.
4. The online intelligent detection system for code spraying defects according to claim 3, wherein:
the specific method of the image preprocessing algorithm comprises the following steps:
(1) image basic processing: the method comprises the steps of reading an acquired image, acquiring a single-channel picture of an RGB image, and carrying out corresponding rotation and preliminary code spraying area extraction;
(2) image depth processing:
firstly, converting an extracted preliminary code spraying area image into a gray image, enhancing image contrast, and smoothing the image by using a mean smoothing filter according to an output result, so as to obtain a dynamic gray threshold value of the position of each character;
then, dynamically dividing the picture character region through a global dynamic gray threshold value, so as to obtain an image code-spraying character region;
finally, the obtained code-spraying character area is processed: firstly, screening the width and the area of a code spraying area by utilizing a characteristic distribution histogram, removing meaningless areas, secondly, corroding an image to achieve the purposes of increasing dark parts and reducing bright parts, thirdly, self-adaptively carrying out image rotation by utilizing an affine transformation matrix according to the character angle of the obtained image characters, and adjusting the image characters to be horizontally displayed;
(3) character denoising segmentation:
firstly, performing binarization segmentation on an image subjected to advanced treatment, extracting a current character string region, performing connected domain analysis on characters in the region, and filtering meaningless miscellaneous points through the area;
then, each character lattice point of the lattice characters is connected by a closed operation, so that the characters can be connected completely and are not divided into a plurality of parts; the adhesion between the characters is broken by utilizing open operation, so that each character can be independently segmented, and the subsequent character recognition is not influenced;
finally, by carrying out circumscribed rectangle on each character area to obtain a plurality of independent rectangle areas, and by identifying rectangle areas far exceeding single characters, carrying out length re-segmentation on the average width of the characters to prevent the occurrence of characters which are adhered too closely and are not segmented, so as to finish the de-noising segmentation of the code-spraying character strings;
(4) single character processing:
and (3) carrying out single-region processing again on the single character region after denoising and segmentation to realize further denoising of the character region, respectively carrying out connected region analysis and noise point filtering on each character region, merging and connecting the characters in each rectangular region again, sequencing the character regions, counting the final number of the characters, and finally obtaining clear character blocks of a plurality of different regions.
5. The online intelligent detection system for code spraying defects according to claim 3, wherein:
the defect judgment algorithm comprises the following specific methods:
(1) defining the correct code spraying character on duty:
automatically generating and defining correct code spraying character strings of the shift through the actual meaning of the code spraying;
(2) training a character classification template:
firstly, establishing a character model library, carrying out actual image acquisition and image preprocessing by adjusting an inkjet printer to spray different characters of 0-9 and A-M to obtain single segmentation pictures of different characters, and then training and updating the character library by utilizing a large number of inkjet character images of different shifts and dates to finally obtain an inkjet character library template;
(3) character recognition judgment:
comparing and judging single characters segmented by the preprocessed image with different characters in a character library template, judging and identifying each code spraying character, finally outputting a section of character string from the obtained identification result, comparing the character string with a defined correct code spraying character string, and if the character string is identical to the defined correct code spraying character string, judging that the character string is qualified, and if the character string is not identical to the defined correct code spraying character string, judging that the character string is a code spraying defect.
6. The online intelligent detection system for code spraying defects according to claim 1, wherein:
the intelligent prediction model is a code spraying defect prediction model established by training based on a YOLOv5 deep learning convolutional neural network algorithm.
7. The online intelligent detection system for code spraying defects according to claim 1, wherein:
the preventive improvement and repair includes:
(1) adjusting a bracket of the code spraying equipment;
(2) the maintenance of the ink jet numbering machine and the spray gun;
(3) adjustment of the static electricity eliminating device;
(4) setting and optimizing detection parameters of the industrial personal computer;
(5) adjusting OPC trigger parameters of equipment;
(6) technical improvements or modifications are made to the main cause of the problem.
8. The online intelligent detection method for the code spraying defects is characterized by comprising the following steps of: the method is applied to the code spraying defect online intelligent detection system as claimed in any one of claims 1 to 7, and comprises the following steps of
The electric control unit sets a detection phase triggering angle;
the industrial personal computer converts the detection phase trigger signal to an image acquisition unit for image acquisition;
the industrial personal computer performs display interaction on the acquired images, and performs image processing based on a defect recognition algorithm; when the image product is judged to be a defective product, the industrial personal computer sends a signal to the electronic control unit, and the electronic control unit generates a defective product rejection signal to reject the defective product at a rejection port;
the industrial personal computer transmits the graphic processing result and the equipment operation data to the data acquisition analysis unit, and performs data analysis on the generated code spraying defects to obtain code spraying quality influence factors and frequent time periods for active prevention and control; and carrying out trend inspection early warning on the code spraying defects which do not occur based on an intelligent prediction model, and analyzing by combining the code spraying quality influence factors and equipment operation data so as to carry out preventive improvement or maintenance.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310172312.XA CN116441200A (en) | 2023-02-27 | 2023-02-27 | Online intelligent detection system and method for code spraying defects |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310172312.XA CN116441200A (en) | 2023-02-27 | 2023-02-27 | Online intelligent detection system and method for code spraying defects |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116441200A true CN116441200A (en) | 2023-07-18 |
Family
ID=87120951
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310172312.XA Pending CN116441200A (en) | 2023-02-27 | 2023-02-27 | Online intelligent detection system and method for code spraying defects |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116441200A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117358615A (en) * | 2023-10-08 | 2024-01-09 | 南京三隆包装有限公司 | Automatic code-spraying printing defect detection method and system |
| CN117734330A (en) * | 2023-12-22 | 2024-03-22 | 深圳日上光电有限公司 | COB lamp strip code spraying control method and system |
| CN118150596A (en) * | 2024-05-10 | 2024-06-07 | 东莞市卡的智能科技有限公司 | RFID smart card defect detection method and system |
-
2023
- 2023-02-27 CN CN202310172312.XA patent/CN116441200A/en active Pending
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117358615A (en) * | 2023-10-08 | 2024-01-09 | 南京三隆包装有限公司 | Automatic code-spraying printing defect detection method and system |
| CN117358615B (en) * | 2023-10-08 | 2024-04-30 | 南京三隆包装有限公司 | Automatic code-spraying printing defect detection method and system |
| CN117734330A (en) * | 2023-12-22 | 2024-03-22 | 深圳日上光电有限公司 | COB lamp strip code spraying control method and system |
| CN118150596A (en) * | 2024-05-10 | 2024-06-07 | 东莞市卡的智能科技有限公司 | RFID smart card defect detection method and system |
| CN118150596B (en) * | 2024-05-10 | 2024-07-12 | 东莞市卡的智能科技有限公司 | RFID smart card defect detection method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116441200A (en) | Online intelligent detection system and method for code spraying defects | |
| CN104992449B (en) | Information identification and surface defect online test method based on machine vision | |
| CN109724990A (en) | The quick positioning and detection method in coding region in a kind of packing box label | |
| CN108562589A (en) | A method of magnetic circuit material surface defect is detected | |
| CN113487570B (en) | High-temperature continuous casting billet surface defect detection method based on improved yolov5x network model | |
| CN114266751B (en) | Product packaging bag coding defect detection method and system based on AI technology | |
| CN107392931A (en) | Bar tobacco board sorter and method | |
| CN113077450A (en) | Cherry grading detection method and system based on deep convolutional neural network | |
| CN118506299B (en) | Rail train body abnormality detection method, device and storage medium | |
| CN110310275A (en) | A kind of chain conveyor defect inspection method based on image procossing | |
| CN116188459B (en) | Line laser rapid identification method and system for belt tearing detection | |
| CN111968082A (en) | Product packaging defect detection and identification method based on machine vision | |
| CN111539927A (en) | Detection process and algorithm of automobile plastic assembly fastening buckle lack-assembly detection device | |
| CN110723582B (en) | Web die-cutting connecting line two-dimensional code jet printing and detecting device | |
| CN116152261B (en) | Visual inspection system for quality of printed product | |
| CN115311250A (en) | Cigarette appearance detection method based on deep learning and space-time feature fusion | |
| CN114359155A (en) | Film laminating method and system | |
| Sa et al. | Packaging defect detection system based on machine vision and deep learning | |
| CN111570313A (en) | Condom packaging box code detection system and detection method based on machine vision | |
| JPH09311031A (en) | Method and apparatus for inspection of quality of punched and worked product | |
| CN116482125A (en) | Cloth defect processing method based on machine vision | |
| Yin et al. | Research on feature extraction of local binary pattern of SLM powder bed gray image | |
| CN110674754A (en) | Online intermittent hollow filter stick visual defect detection and identification system | |
| CN118150596B (en) | RFID smart card defect detection method and system | |
| CN117593528B (en) | Rail vehicle bolt loosening detection method based on machine vision |
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 |