CN117031052B - Single printed matter front and back vision detection control system - Google Patents
Single printed matter front and back vision detection control system Download PDFInfo
- Publication number
- CN117031052B CN117031052B CN202311298675.4A CN202311298675A CN117031052B CN 117031052 B CN117031052 B CN 117031052B CN 202311298675 A CN202311298675 A CN 202311298675A CN 117031052 B CN117031052 B CN 117031052B
- Authority
- CN
- China
- Prior art keywords
- printed matter
- matching
- samples
- defect
- image
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 107
- 230000007547 defect Effects 0.000 claims abstract description 134
- 238000007689 inspection Methods 0.000 claims abstract description 11
- 238000011179 visual inspection Methods 0.000 claims abstract description 11
- 230000000007 visual effect Effects 0.000 claims description 31
- 238000000034 method Methods 0.000 claims description 20
- 230000007246 mechanism Effects 0.000 claims description 19
- 230000007306 turnover Effects 0.000 claims description 18
- 230000005477 standard model Effects 0.000 claims description 11
- 230000002950 deficient Effects 0.000 claims description 9
- 238000007781 pre-processing Methods 0.000 claims description 8
- 238000005286 illumination Methods 0.000 claims description 6
- 238000003708 edge detection Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 5
- 230000005540 biological transmission Effects 0.000 claims description 4
- 238000012937 correction Methods 0.000 claims description 4
- 230000000877 morphologic effect Effects 0.000 claims description 4
- 238000001179 sorption measurement Methods 0.000 claims description 4
- 230000001804 emulsifying effect Effects 0.000 claims description 3
- 238000003706 image smoothing Methods 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 3
- 230000001360 synchronised effect Effects 0.000 abstract description 2
- 230000001965 increasing effect Effects 0.000 description 8
- 230000008569 process Effects 0.000 description 7
- 238000004364 calculation method Methods 0.000 description 4
- 238000005457 optimization Methods 0.000 description 4
- 238000000605 extraction Methods 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 239000002761 deinking Substances 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 235000002566 Capsicum Nutrition 0.000 description 1
- 239000006002 Pepper Substances 0.000 description 1
- 235000016761 Piper aduncum Nutrition 0.000 description 1
- 235000017804 Piper guineense Nutrition 0.000 description 1
- 244000203593 Piper nigrum Species 0.000 description 1
- 235000008184 Piper nigrum Nutrition 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 150000003839 salts Chemical class 0.000 description 1
- 230000008961 swelling Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N35/00—Automatic analysis not limited to methods or materials provided for in any single one of groups G01N1/00 - G01N33/00; Handling materials therefor
- G01N35/00584—Control arrangements for automatic analysers
-
- 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
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N35/00—Automatic analysis not limited to methods or materials provided for in any single one of groups G01N1/00 - G01N33/00; Handling materials therefor
- G01N35/0099—Automatic analysis not limited to methods or materials provided for in any single one of groups G01N1/00 - G01N33/00; Handling materials therefor comprising robots or similar manipulators
Landscapes
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Engineering & Computer Science (AREA)
- Robotics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Image Processing (AREA)
Abstract
The invention relates to the technical field of machine vision, in particular to a front and back vision detection control system for single printed matter, which comprises the following components: a first vision inspection system and a second vision inspection system, the first vision inspection system comprising: the first image acquisition module and the first control module are used for matching the first image characteristic information with a preset first printed matter defect database, judging whether the matching is successful or not, and outputting a first control instruction and a second control instruction if the matching is successful, wherein the first control instruction is used for controlling taking out unqualified printed matters, and the second control instruction is used for controlling the second visual inspection system to be in a rest state; the second vision inspection system includes: the second image acquisition module, the second control module and the second printed matter defect database are used for realizing synchronous update of defect sample data between the first printed matter defect database and the second printed matter defect database through a cross-database trigger; the invention can optimize the detection flow, improve the detection efficiency and improve the detection efficiency.
Description
Technical Field
The invention relates to the technical field of machine vision, in particular to a front and back vision detection control system for single printed matter.
Background
In the prior art, the need of printing data such as bar codes, two-dimensional codes or special patterns on product packages is increasing, and in order to ensure the integrity and usability of various printed data printed on the product packages, quality detection is required to be carried out on the printed product packages, namely printed matters, in the mass production printing process of the product packages, unqualified printed matters such as incomplete printed data, fuzzy patterns and the like are timely detected, so that the quality of the finally sold printed matters is ensured.
For the single-sheet double-sided printed package printed matter, the on-line detection method generally comprises the steps of finishing front detection of single-sheet printed matter in batches through a visual detection system, manually turning over paper, and then carrying out back detection of the single-sheet printed matter in batches.
The patent application number is CN202221588132.7, has disclosed a single double-sided printed matter on-line measuring system, including the front that follows the printed matter and transmits the direction and detects cylinder, turn over cylinder and reverse side that sets gradually; at least one group of gripper mechanisms are respectively arranged on the front detection roller, the turnover roller and the back detection roller; a paper head positioning mechanism is arranged above the turn-over roller, and a paper tail adsorption mechanism and a paper swing mechanism are arranged on one side, close to the reverse side detection roller, of the upper side of the turn-over roller; visual detection systems are respectively arranged above the front detection roller and the back detection roller. The detection of the front printing quality of the printed matter is finished by the visual detection system at the front detection roller, the detection of the back printing quality of the printed matter is finished by the visual detection system at the back detection roller, and the paper is short in transmission path, less in connection times and high in detection speed in the process of transmitting the paper from the front detection roller to the back detection roller, so that the detection method is particularly suitable for high-speed online detection of the front and the back of the printed matter.
The print is required to be qualified by meeting the requirements on both sides, that is, the print is not qualified as long as one side is not qualified. However, the two visual detection systems of the online detection system for the single double-sided printed matter independently work, and the defective printed matter detected on the first surface also needs to be detected on the second surface, so that an invalid workflow is increased, the pressure of the visual detection system for collecting and processing data is increased, the detection speed is influenced, and the detection efficiency is lower.
Disclosure of Invention
In order to solve the problems, the invention provides a control system for detecting the front and back sides of a single printed matter, which can optimize the detection flow, reduce the calculation complexity, improve the detection efficiency and reduce the power consumption of the system; the defect detection is preferably adopted, so that whether the printed matter is qualified or not can be detected more specifically, and the defective printed matter can be found more quickly; when no matching is found through the defect database, and then the matching is compared with the standard model for detection, whether the printed matter is qualified or not can be judged more comprehensively, and missing detection is prevented; meanwhile, new defect samples are generated according to the image characteristic information and updated to a printed matter defect database, the types of the defect samples in the printed matter defect database are increased, and the efficiency of detecting the same defects of subsequent products is improved.
In order to achieve the above purpose, the invention adopts the following technical scheme:
the front and back visual detection control system for the single printed matter comprises a front detection roller, a turnover roller and a back detection roller which are sequentially arranged along the transmission direction of the printed matter, wherein the printed matter is transmitted and handed over by the rollers, and a first visual detection system and a second visual detection system are respectively arranged above the front detection roller and the back detection roller; it is characterized in that the method comprises the steps of,
the first vision inspection system includes:
the first image acquisition module is used for acquiring a front image of the printed matter to be detected;
the first control module is used for executing the following steps:
preprocessing the collected front image of the printed matter to be detected, extracting first image characteristic information based on the preprocessed front image, matching the first image characteristic information with a preset first printed matter defect database, judging whether the matching is successful,
if the matching is successful, outputting a first control instruction, a second control instruction and a warning signal of unqualified printed matters, wherein the first control instruction is used for controlling the unqualified printed matters to be taken out, and the second control instruction is used for controlling the second visual detection system to be in a rest state;
if the matching is unsuccessful, comparing and calculating the first image characteristic information with a preset printed matter front standard model, and generating a first difference value;
determining whether the first difference value exceeds a first set threshold,
if yes, outputting a first control instruction, a second control instruction and a warning signal of unqualified printed matter, and generating a new defect sample based on the first image characteristic information and updating the new defect sample into the first printed matter defect database;
the second vision inspection system includes:
the second image acquisition module is used for acquiring a reverse image of the printed matter to be detected;
the second control module is used for executing the following steps:
preprocessing the collected reverse image of the to-be-detected printed matter, extracting second image characteristic information based on the preprocessed reverse image, matching the second image characteristic information with a preset second printed matter defect database, judging whether the matching is successful or not,
if the matching is successful, outputting a third control instruction and a warning signal of unqualified printed matters, wherein the third control instruction is used for controlling the unqualified printed matters to be taken out;
if the matching is unsuccessful, comparing and calculating the second image characteristic information with a preset standard model of the reverse side of the printed matter, and generating a second difference value;
determining whether the second difference value exceeds a second set threshold,
if yes, outputting a third control instruction and a warning signal of unqualified printed matter, and generating a new defect sample based on the second image characteristic information and updating the new defect sample into the second printed matter defect database;
the defect sample data is synchronously updated between the first printed matter defect database and the second printed matter defect database through a cross-database trigger;
wherein the defect sample comprises: overprinting inaccuracy samples, dot gain samples, plate pasting samples, dirty plate samples, plate dropping samples, ghost samples, deinking samples, ink piling samples, ink emulsifying samples, dirt sticking samples, napping samples, character water generating samples, color cast samples, dark and bright color samples, paper tearing samples, fold samples, hole samples and foreign object interference samples.
Further, the matching of the first image feature information with a preset first printed matter defect database is performed, and whether the matching is successful or not is judged, and the method comprises the following steps:
s1, inquiring in a first printed matter defect database by taking first image characteristic information as an inquiry condition;
s2, calculating a good suffix table and a bad character table in a first printed matter defect database through a Boyer-Moore algorithm, wherein the good suffix table stores the same character string appearing on the rightmost side in each substring, and the bad character table stores the rightmost position of each character in a query mode; gradually sliding the query pattern to the rightmost position from the tail of the characteristic information in the defect sample for comparison, and returning to the matched position if the query pattern is completely matched, wherein the matching is completed; otherwise, executing S3;
s3, continuing to move the query mode rightward in the sample, selecting a proper moving distance according to the bad character table and the good suffix table, sliding the query mode to the next position, and repeatedly executing S2 until the sample boundary is reached;
s4, judging according to the matching result of the S2 and the S3, wherein the judging specifically comprises the following steps: if the matching is completed, determining that the query pattern is successfully matched with the current defect sample; if the matching is not completed and the query mode cannot be matched with the current defect sample, the query mode is moved backwards to the designated position in the good suffix table until all the defect samples are queried or the matched position is found;
s5, if all the defect samples are inquired and the matched positions are not found, determining that the inquiry mode is not successfully matched with the first printed matter defect database.
Further, the S2 includes:
creating a hash table of characters in the defect sample and the query pattern, mapping each character to a unique hash value;
starting from the end of the characteristic information of the defect sample, sliding the query pattern to the rightmost position;
calculating a hash value of the query pattern;
comparing the hash value of the query pattern with the hash value in the defect sample corresponding to the current position, and if the hash value is matched, then:
comparing the characters in the query pattern and the defect sample one by one, and if the characters are completely matched, returning to the matched position and indicating that the matching is completed.
Further, between S2 and S3, the method includes:
the same character is selected for matching in the query mode and the sample string through the Galil rule, and the specific steps are as follows:
finding the first position of the same character as the current failed match character in the rightmost query pattern, selecting a second position for movement based on whether the first position has the prefix of the same defective suffix in the sample string,
if the prefix of the same suffix exists, aligning the query pattern and the sample string at a second position, and skipping the comparison of the invalid portion;
otherwise, S3 is performed.
Further, eight groups of gripper mechanisms are respectively arranged on the front detection roller, the turnover roller and the back detection roller.
Further, the preprocessing includes: image denoising, image enhancement, image graying, image smoothing, edge detection, morphological processing, illumination correction, and size normalization.
Further, the first visual inspection system and the second visual inspection system each include a camera, a lens, and a light source.
Further, the method further comprises the following steps: and the manipulator is used for taking out unqualified printed matters according to the first control instruction or the third control instruction.
Further, a paper head positioning mechanism is arranged above the turn-over roller, and a paper tail adsorption mechanism and a paper swing and feed mechanism are arranged on one side, close to the reverse side detection roller, of the upper side of the turn-over roller.
Further, the swing radius of the paper swing mechanism is 1/3 of the radius of the turn-over roller.
The invention has the beneficial effects that:
1. when the first visual detection system detects that the front side of the printed matter is unqualified, the second visual detection system outputs a second control instruction for controlling the second visual detection system to enter a rest state, in other words, once the first visual detection system detects that the front side of the printed matter is unqualified, the back side of the printed matter is not detected. Therefore, the invention can optimize the detection flow to reduce the calculation complexity; meanwhile, the optimization scheme can effectively reduce the power consumption of the system and improve the detection efficiency.
2. The invention preferentially adopts defect detection, can detect whether the printed matter is qualified or not more specifically, and can find the defective printed matter more quickly. When no matching is found through the defect database, and then the matching is compared with the standard model for detection, whether the printed matter is qualified or not can be judged more comprehensively, and missing detection is prevented; meanwhile, new defect samples are generated according to the image characteristic information and updated to a printed matter defect database, the types of the defect samples in the printed matter defect database are increased, and the efficiency of detecting the same defects of subsequent products is improved.
3. The invention can improve the speed of detecting the defects of the printed matter through the Boyer-Moore algorithm, avoid a large number of invalid comparisons and improve the matching efficiency of the image characteristic information and the defect database of the printed matter. The defect matching strategy is optimized by calculating the hash value and the Galil rule through the hash table, so that the matching speed can be optimized, invalid comparison is reduced, and the matching time is shortened, thereby further improving the defect matching efficiency.
4. The invention is provided with more defect samples, can meet the detection of various defects and improves the detection efficiency.
Drawings
Fig. 1 is a block diagram of the structure of the visual inspection control system for the front and back sides of a single printed matter of the present invention.
Detailed Description
Referring to fig. 1, the invention relates to a control system for detecting front and back surfaces of a single printed matter, which comprises a front surface detection roller, a turnover roller and a back surface detection roller which are sequentially arranged along the transmission direction of the printed matter, wherein the printed matter is transmitted and handed over by each roller, and a first visual detection system and a second visual detection system are respectively arranged above the front surface detection roller and the back surface detection roller;
the first vision inspection system includes:
the first image acquisition module is used for acquiring a front image of the printed matter to be detected;
the first control module is used for executing the following steps:
preprocessing the collected front image of the printed matter to be detected, extracting first image characteristic information based on the preprocessed front image, matching the first image characteristic information with a preset first printed matter defect database, judging whether the matching is successful,
if the matching is successful, outputting a first control instruction, a second control instruction and a warning signal of unqualified printed matters, wherein the first control instruction is used for controlling the unqualified printed matters to be taken out, and the second control instruction is used for controlling the second visual detection system to be in a rest state; the object defined here is that the first visual inspection system outputs a second control instruction for controlling the second visual inspection system to enter a rest state when the front side failure of the printed matter is detected, in other words, the invention does not detect the back side of the printed matter once the front side failure of the printed matter is detected. Therefore, the invention can optimize the detection flow to reduce the calculation complexity; meanwhile, the optimization scheme can effectively reduce the power consumption of the system and improve the detection efficiency.
If the matching is unsuccessful, comparing and calculating the first image characteristic information with a preset printed matter front standard model, and generating a first difference value;
determining whether the first difference value exceeds a first set threshold,
if yes, outputting a first control instruction, a second control instruction and a warning signal of unqualified printed matter, and generating a new defect sample based on the first image characteristic information and updating the new defect sample into the first printed matter defect database; the purpose of this definition is to increase the variety of defect samples in the printed matter defect database and to increase the efficiency of detecting the same defects in subsequent products. The first control instruction is used for controlling the taking-out of unqualified printed matters, and the second control instruction is used for controlling the second visual detection system to be in a rest state; the object defined here is that the first visual inspection system outputs a second control instruction for controlling the second visual inspection system to enter a rest state when the front side failure of the printed matter is detected, in other words, the invention does not detect the back side of the printed matter once the front side failure of the printed matter is detected. Therefore, the invention can optimize the detection flow to reduce the calculation complexity; meanwhile, the optimization scheme can effectively reduce the power consumption of the system and improve the detection efficiency.
The second vision inspection system includes:
the second image acquisition module is used for acquiring a reverse image of the printed matter to be detected;
the second control module is used for executing the following steps:
preprocessing the collected reverse image of the to-be-detected printed matter, extracting second image characteristic information based on the preprocessed reverse image, matching the second image characteristic information with a preset second printed matter defect database, judging whether the matching is successful or not,
if the matching is successful, outputting a third control instruction and a warning signal of unqualified printed matters, wherein the third control instruction is used for controlling the unqualified printed matters to be taken out;
if the matching is unsuccessful, comparing and calculating the second image characteristic information with a preset standard model of the reverse side of the printed matter, and generating a second difference value;
determining whether the second difference value exceeds a second set threshold,
if yes, outputting a third control instruction and a warning signal of unqualified printed matter, and generating a new defect sample based on the second image characteristic information and updating the new defect sample into the second printed matter defect database;
the defect sample data is synchronously updated between the first printed matter defect database and the second printed matter defect database through a cross-database trigger;
wherein the defect sample comprises: overprinting inaccuracy samples, dot gain samples, plate pasting samples, dirty plate samples, plate dropping samples, ghost samples, deinking samples, ink piling samples, ink emulsifying samples, dirt sticking samples, napping samples, character water generating samples, color cast samples, dark and bright color samples, paper tearing samples, fold samples, hole samples and foreign object interference samples. The invention is provided with more defect samples, can meet the detection of various defects and improves the detection efficiency.
In the scheme, the defect detection is preferentially adopted, so that whether the printed matter is qualified or not can be detected more specifically, and the defective printed matter can be found more quickly. When no matching is found through the defect database, the matching is compared with the standard model for detection, so that whether the printed matter is qualified or not can be comprehensively judged, and missing detection is prevented.
Further, the matching of the first image feature information with a preset first printed matter defect database, and judging whether the matching is successful, includes the following steps:
s1, inquiring in a first printed matter defect database by taking first image characteristic information as an inquiry condition;
s2, calculating a good suffix table and a bad character table in a first printed matter defect database through a Boyer-Moore algorithm, wherein the good suffix table stores the same character string appearing on the rightmost side in each substring, and the bad character table stores the rightmost position of each character in a query mode; gradually sliding the query pattern to the rightmost position from the tail of the characteristic information in the defect sample for comparison, and returning to the matched position if the query pattern is completely matched, wherein the matching is completed; otherwise, executing S3;
s3, continuing to move the query mode rightward in the sample, selecting a proper moving distance according to the bad character table and the good suffix table, sliding the query mode to the next position, and repeatedly executing S2 until the sample boundary is reached;
s4, judging according to the matching result of the S2 and the S3, wherein the judging specifically comprises the following steps: if the matching is completed, determining that the query pattern is successfully matched with the current defect sample; if the matching is not completed and the query mode cannot be matched with the current defect sample, the query mode is moved backwards to the designated position in the good suffix table until all the defect samples are queried or the matched position is found;
s5, if all the defect samples are inquired and the matched positions are not found, determining that the inquiry mode is not successfully matched with the first printed matter defect database.
In the scheme, the speed of detecting the defects of the printed matter can be improved through the Boyer-Moore algorithm, a large number of invalid comparisons are avoided, and the matching efficiency of the image characteristic information and the defect database of the printed matter is improved.
Further, the step S2 includes:
creating a hash table of characters in the defect sample and the query pattern, mapping each character to a unique hash value;
starting from the end of the characteristic information of the defect sample, sliding the query pattern to the rightmost position;
calculating a hash value of the query pattern;
comparing the hash value of the query pattern with the hash value in the defect sample corresponding to the current position, and if the hash value is matched, then:
comparing the characters in the query pattern and the defect sample one by one, and if the characters are completely matched, returning to the matched position and indicating that the matching is completed.
According to the embodiment, the hash table is created on the defect sample and the characters in the query mode, each character is mapped to the unique hash value, the quick matching of the character strings can be realized, and the matching efficiency is improved. In addition, the hash value of the query mode is utilized to accelerate the matching process, the hash value is compared with the hash value of the current position in the defect sample, if the hash value is not matched, the query mode is directly slid to the right by one sub-string length, the current comparison process is skipped, invalid comparison is avoided, and the matching efficiency can be further improved.
The query mode is slid to the rightmost position from the tail of the characteristic information of the defect sample, so that the sliding times can be effectively reduced, and meanwhile, the number of times of unmatched characters can be increased, and the probability of missed judgment is reduced.
The hash value of the query mode is calculated, so that the matching process can be accelerated, whether a plurality of matching positions exist or not can be judged rapidly, and the matching accuracy is improved.
And comparing the hash value of the query mode with the hash value in the missing sample corresponding to the current position, if the hash value is matched, comparing the characters in the query mode and the missing sample one by one, and if the character comparison is completely matched, returning to the matched position and indicating that the matching is completed, so that invalid comparison and wasted time are avoided, and the matching effect is accelerated.
Further, between S2 and S3, the method includes:
the same character is selected for matching in the query mode and the sample string through the Galil rule, and the specific steps are as follows:
finding the first position of the same character as the current failed match character in the rightmost query pattern, selecting a second position for movement based on whether the first position has the prefix of the same defective suffix in the sample string,
if the prefix of the same suffix exists, aligning the query pattern and the sample string at a second position, and skipping the comparison of the invalid portion;
otherwise, S3 is performed.
In the above scheme, between S2 and S3, the matching process can be further optimized by selecting the same character in the query pattern and the sample string for matching by the Galil rule. The method comprises the following steps: find the first position of the same character as the failed match character in the rightmost query pattern. The second location of the movement is selected based on whether the first location has a prefix of the same defective suffix in the sample string. If the prefix of the same suffix exists, the query pattern and the sample string are aligned at a second location, skipping the comparison of the invalid portion. Therefore, invalid parts which are not matched with the query mode can be quickly skipped, and the matching speed is improved. If the prefixes of the same suffixes do not exist, the step S3 is continued to be executed, and the comparison is performed by the gradual sliding query mode.
By using the Galil rule, partial matching information can be better utilized, and matching efficiency is improved. When there is a character that fails to match, the location of the movement is determined by rules to reduce invalid comparison operations and to find the location of the match faster. Therefore, the algorithm can be optimized, the matching time is shortened, and the matching effect is improved.
It should be noted that, the method for matching the second image feature information with the preset second printed matter defect database is the same as the method for matching the first image feature information with the preset first printed matter defect database, and the working principle is the same, which is not described here again.
Further, the preprocessing includes: image denoising, image enhancement, image graying, image smoothing, edge detection, morphological processing, illumination correction, and size normalization. The purpose defined here is to remove some of the unwanted information of the original image of the print, enhancing the useful information and thus improving the accuracy and efficiency of the matching. In particular, the method comprises the steps of,
denoising an image: interference factors in the image, such as salt and pepper noise, gaussian noise and the like, are removed, so that noise in the image can be reduced, and a subsequent algorithm is more accurate and stable.
Image enhancement: the characteristics of the image are enhanced by increasing the contrast, brightness and the like of the image, so that the matching result is more obvious, and the matching precision and robustness are improved.
Graying the image: converting an image from an RGB color space to a gray scale image can reduce the amount of computation and facilitate the extraction and matching of subsequent features.
Smoothing of the image: the image is smoothed by using a Gaussian filter and other methods, so that noise and details of the image can be removed, and subsequent feature extraction and matching are easier and more accurate.
Edge detection: the image is subjected to edge detection through a Sobel or Canny algorithm, so that the edge information of the image can be extracted, and the subsequent feature extraction and matching are facilitated.
Morphological treatment: such as corrosion, swelling, etc., can change the shape and size of the image, help remove unnecessary interference and noise, and enhance the characteristics of the useful information.
And (3) illumination correction: the illumination and brightness of the image are corrected, so that the matching error caused by illumination influence can be reduced, and the matching precision is improved.
Size standardization: the images are subjected to size standardization processing, so that images with different sizes have the same proportion in comparison, and matching is performed more accurately.
Further, the first visual detection system and the second visual detection system each include a camera, a lens, and a light source.
Further, the method further comprises the following steps: and the manipulator is used for taking out unqualified printed matters according to the first control instruction or the third control instruction.
Further, eight groups of gripper mechanisms are respectively arranged on the front detection roller, the turnover roller and the back detection roller. The paper head positioning mechanism is arranged above the turnover roller, and the paper tail adsorption mechanism and the paper swing mechanism are arranged on one side, close to the reverse side detection roller, of the upper side of the turnover roller. The swing radius of the paper swing mechanism is 1/3 of the radius of the turn-over roller.
Compared with the prior art, the embodiment has the advantages that:
the embodiment can automatically collect, process and compare the image characteristic information of the printed matter to be detected and match with a preset defect database. Whether the difference of the characteristic information exceeds a set threshold value or not is judged, so that the qualification of the printed matter can be automatically judged, and a corresponding control instruction and a corresponding warning signal are output; the image characteristic information is extracted and matched with a standard model or a defect database, so that whether the printed matter has defects can be accurately judged; when the matching is unsuccessful, a new defect sample can be generated according to the image characteristic information and updated into a printed matter defect database, so that the detection accuracy is further improved. The first printed matter defect database and the second printed matter defect database can realize synchronous updating of data through the cross-database trigger. This ensures that the data of both databases remain consistent throughout. The quality of the printed matter can be rapidly and accurately judged through the visual detection system, and the unqualified matter can be timely controlled to be taken out; this helps to reduce the production of rejects, improves production efficiency and yield, and reduces the risk and cost of rejects.
In this embodiment, when the first visual detection system detects that the front side of the printed matter is not qualified, the second visual detection system outputs a second control instruction for controlling the second visual detection system to enter a rest state, in other words, once the front side of the printed matter is detected to be unqualified, the reverse side of the printed matter is not detected any more. Therefore, the detection flow can be optimized to reduce the computational complexity; meanwhile, the optimization scheme can effectively reduce the power consumption of the system and improve the detection efficiency.
According to the embodiment, the defect detection is preferentially adopted, whether the printed matter is qualified or not can be detected more specifically, and the defective printed matter can be found more quickly. When no matching is found through the defect database, and then the matching is compared with the standard model for detection, whether the printed matter is qualified or not can be judged more comprehensively, and missing detection is prevented; meanwhile, new defect samples are generated according to the image characteristic information and updated to a printed matter defect database, the types of the defect samples in the printed matter defect database are increased, and the efficiency of detecting the same defects of subsequent products is improved.
The method and the device can improve the speed of detecting the defects of the printed matter through a Boyer-Moore algorithm, avoid a large number of invalid comparisons, and improve the matching efficiency of the image characteristic information and the defect database of the printed matter. The defect matching strategy is optimized by calculating the hash value and the Galil rule through the hash table, so that the matching speed can be optimized, invalid comparison is reduced, and the matching time is shortened, thereby further improving the defect matching efficiency.
The embodiment is provided with more defect samples, can meet the detection of various defects, and improves the detection efficiency.
The above embodiments are merely illustrative of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention, and various modifications and improvements made by those skilled in the art to the technical solution of the present invention should fall within the scope of protection defined by the claims of the present invention without departing from the spirit of the design of the present invention.
Claims (4)
1. The front and back visual detection control system for the single printed matter comprises a front detection roller, a turnover roller and a back detection roller which are sequentially arranged along the transmission direction of the printed matter, wherein the printed matter is transmitted and handed over by the rollers, and a first visual detection system and a second visual detection system are respectively arranged above the front detection roller and the back detection roller; it is characterized in that the method comprises the steps of,
the first vision inspection system includes:
the first image acquisition module is used for acquiring a front image of the printed matter to be detected;
the first control module is used for executing the following steps:
preprocessing the collected front image of the printed matter to be detected, extracting first image characteristic information based on the preprocessed front image, matching the first image characteristic information with a preset first printed matter defect database, judging whether the matching is successful,
if the matching is successful, outputting a first control instruction, a second control instruction and a warning signal of unqualified printed matters, wherein the first control instruction is used for controlling the unqualified printed matters to be taken out, and the second control instruction is used for controlling the second visual detection system to be in a rest state;
if the matching is unsuccessful, comparing and calculating the first image characteristic information with a preset printed matter front standard model, and generating a first difference value;
determining whether the first difference value exceeds a first set threshold,
if yes, outputting a first control instruction, a second control instruction and a warning signal of unqualified printed matter, and generating a new defect sample based on the first image characteristic information and updating the new defect sample into the first printed matter defect database;
the second vision inspection system includes:
the second image acquisition module is used for acquiring a reverse image of the printed matter to be detected;
the second control module is used for executing the following steps:
preprocessing the collected reverse image of the to-be-detected printed matter, extracting second image characteristic information based on the preprocessed reverse image, matching the second image characteristic information with a preset second printed matter defect database, judging whether the matching is successful or not,
if the matching is successful, outputting a third control instruction and a warning signal of unqualified printed matters, wherein the third control instruction is used for controlling the unqualified printed matters to be taken out;
if the matching is unsuccessful, comparing and calculating the second image characteristic information with a preset standard model of the reverse side of the printed matter, and generating a second difference value;
determining whether the second difference value exceeds a second set threshold,
if yes, outputting a third control instruction and a warning signal of unqualified printed matter, and generating a new defect sample based on the second image characteristic information and updating the new defect sample into the second printed matter defect database;
the defect sample data is synchronously updated between the first printed matter defect database and the second printed matter defect database through a cross-database trigger;
wherein the defect sample comprises: overprinting inaccuracy samples, dot gain samples, burnt-in samples, dirty-in samples, plate-drop samples, ghost samples, deinked samples, ink piling samples, ink emulsifying samples, sticky dirt samples, napping samples, character water-out samples, color cast samples, dark and bright color samples, paper tearing samples, fold samples, hole samples and foreign matter interference samples;
the step of matching the first image characteristic information with a preset first printed matter defect database and judging whether the matching is successful or not comprises the following steps:
s1, inquiring in a first printed matter defect database by taking first image characteristic information as an inquiry condition;
s2, calculating a good suffix table and a bad character table in a first printed matter defect database through a Boyer-Moore algorithm, wherein the good suffix table stores the same character string appearing on the rightmost side in each substring, and the bad character table stores the rightmost position of each character in a query mode; gradually sliding the query pattern to the rightmost position from the tail of the characteristic information in the defect sample for comparison, and returning to the matched position if the query pattern is completely matched, wherein the matching is completed; otherwise, executing S3;
s3, continuing to move the query mode rightward in the sample, selecting a proper moving distance according to the bad character table and the good suffix table, sliding the query mode to the next position, and repeatedly executing S2 until the sample boundary is reached;
s4, judging according to the matching result of the S2 and the S3, wherein the judging specifically comprises the following steps: if the matching is completed, determining that the query pattern is successfully matched with the current defect sample; if the matching is not completed and the query mode cannot be matched with the current defect sample, the query mode is moved backwards to the designated position in the good suffix table until all the defect samples are queried or the matched position is found;
s5, if all the defect samples are inquired and the matched positions are not found, determining that the inquiry mode is not successfully matched with the first printed matter defect database;
the step S2 comprises the following steps:
creating a hash table of characters in the defect sample and the query pattern, mapping each character to a unique hash value;
starting from the end of the characteristic information of the defect sample, sliding the query pattern to the rightmost position;
calculating a hash value of the query pattern;
comparing the hash value of the query pattern with the hash value in the defect sample corresponding to the current position, and if the hash value is matched, then:
comparing the characters in the query mode and the defect sample one by one, and if the characters are completely matched, returning to the matched position and indicating that the matching is completed;
the steps between S2 and S3 include:
the same character is selected for matching in the query mode and the sample string through the Galil rule, and the specific steps are as follows:
finding the first position of the same character as the current failed match character in the rightmost query pattern, selecting a second position for movement based on whether the first position has the prefix of the same defective suffix in the sample string,
if the prefix of the same suffix exists, aligning the query pattern and the sample string at a second position, and skipping the comparison of the invalid portion;
otherwise, executing S3;
the pretreatment comprises the following steps: image denoising, image enhancement, image graying, image smoothing, edge detection, morphological processing, illumination correction and size standardization;
the first visual detection system and the second visual detection system comprise a camera, a lens and a light source;
the single printed matter positive and negative visual detection control system still includes: and the manipulator is used for taking out unqualified printed matters according to the first control instruction or the third control instruction.
2. The visual inspection control system for the obverse and reverse sides of a single printed matter as claimed in claim 1 wherein eight gripper mechanisms are provided on each of said obverse, reverse and reverse side inspection rollers.
3. The visual inspection control system for the front and back sides of a single printed matter according to claim 2, wherein a paper head positioning mechanism is arranged above the turnover roller, and a paper tail adsorption mechanism and a paper swing mechanism are arranged on one side, close to the back side detection roller, above the turnover roller.
4. A single sheet printed matter obverse and reverse side visual inspection control system as claimed in claim 3 wherein a swing radius of said sheet swing mechanism is 1/3 of a radius of said turn-over roller.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311298675.4A CN117031052B (en) | 2023-10-09 | 2023-10-09 | Single printed matter front and back vision detection control system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311298675.4A CN117031052B (en) | 2023-10-09 | 2023-10-09 | Single printed matter front and back vision detection control system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN117031052A CN117031052A (en) | 2023-11-10 |
CN117031052B true CN117031052B (en) | 2024-01-09 |
Family
ID=88632264
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202311298675.4A Active CN117031052B (en) | 2023-10-09 | 2023-10-09 | Single printed matter front and back vision detection control system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN117031052B (en) |
Citations (18)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101685006A (en) * | 2009-06-25 | 2010-03-31 | 上海交通大学 | Automatic detection system for vision of contact part feature machine |
CN103226106A (en) * | 2013-03-15 | 2013-07-31 | 江南大学 | Machine vision based bearing dust cap quality monitoring system |
CN105600533A (en) * | 2015-11-16 | 2016-05-25 | 河北天星恒科技有限公司 | Metal printed matter testing machine |
CN106144693A (en) * | 2015-04-21 | 2016-11-23 | 征图新视(江苏)科技有限公司 | Positive back face printing detection equipment |
CN109308700A (en) * | 2017-07-27 | 2019-02-05 | 南京敏光视觉智能科技有限公司 | A kind of visual identity defect inspection method based on printed matter character |
CN109374630A (en) * | 2018-09-28 | 2019-02-22 | 新兴铸管股份有限公司 | Ductile iron pipe casting flaw intelligent detecting method |
CN110502612A (en) * | 2019-08-08 | 2019-11-26 | 南京逸杰软件科技有限公司 | A kind of Traffic information demonstration safety detection method based on blacklist intelligent recognition |
CN110806978A (en) * | 2019-10-31 | 2020-02-18 | 吉林亿联银行股份有限公司 | Defect management method and device for third-party component |
CN111650220A (en) * | 2020-07-15 | 2020-09-11 | 博科视(苏州)技术有限公司 | Vision-based image-text defect detection method |
CN111693548A (en) * | 2019-03-12 | 2020-09-22 | 株式会社斯库林集团 | Inspection apparatus and inspection method |
CN111912846A (en) * | 2020-07-13 | 2020-11-10 | 苏州亚朴智能科技有限公司 | Machine vision-based surface defect and edge burr detection method |
CN112577970A (en) * | 2019-09-30 | 2021-03-30 | 深圳中科飞测科技股份有限公司 | Detection method, alignment method of detection equipment and detection equipment |
CN112748128A (en) * | 2020-12-24 | 2021-05-04 | 凌云光技术股份有限公司 | Front and back synchronous detection system |
CN113070240A (en) * | 2021-03-25 | 2021-07-06 | 南京工业大学 | Copper plate surface defect detection and automatic classification method based on machine vision and deep learning |
CN217766133U (en) * | 2022-06-23 | 2022-11-08 | 潍坊华田精工设备有限公司 | Single double-sided printed matter on-line detection system |
CN218567234U (en) * | 2021-05-07 | 2023-03-03 | 新乡市新机迪元科技有限公司 | Reel printing quality on-line measuring device |
CN115855951A (en) * | 2022-11-26 | 2023-03-28 | 苏州三立检测技术有限公司 | Cloth cover flaw detection method and system |
CN219574806U (en) * | 2022-12-13 | 2023-08-22 | 东信和平科技股份有限公司 | Automatic checking system for large printed matter of intelligent card |
-
2023
- 2023-10-09 CN CN202311298675.4A patent/CN117031052B/en active Active
Patent Citations (18)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101685006A (en) * | 2009-06-25 | 2010-03-31 | 上海交通大学 | Automatic detection system for vision of contact part feature machine |
CN103226106A (en) * | 2013-03-15 | 2013-07-31 | 江南大学 | Machine vision based bearing dust cap quality monitoring system |
CN106144693A (en) * | 2015-04-21 | 2016-11-23 | 征图新视(江苏)科技有限公司 | Positive back face printing detection equipment |
CN105600533A (en) * | 2015-11-16 | 2016-05-25 | 河北天星恒科技有限公司 | Metal printed matter testing machine |
CN109308700A (en) * | 2017-07-27 | 2019-02-05 | 南京敏光视觉智能科技有限公司 | A kind of visual identity defect inspection method based on printed matter character |
CN109374630A (en) * | 2018-09-28 | 2019-02-22 | 新兴铸管股份有限公司 | Ductile iron pipe casting flaw intelligent detecting method |
CN111693548A (en) * | 2019-03-12 | 2020-09-22 | 株式会社斯库林集团 | Inspection apparatus and inspection method |
CN110502612A (en) * | 2019-08-08 | 2019-11-26 | 南京逸杰软件科技有限公司 | A kind of Traffic information demonstration safety detection method based on blacklist intelligent recognition |
CN112577970A (en) * | 2019-09-30 | 2021-03-30 | 深圳中科飞测科技股份有限公司 | Detection method, alignment method of detection equipment and detection equipment |
CN110806978A (en) * | 2019-10-31 | 2020-02-18 | 吉林亿联银行股份有限公司 | Defect management method and device for third-party component |
CN111912846A (en) * | 2020-07-13 | 2020-11-10 | 苏州亚朴智能科技有限公司 | Machine vision-based surface defect and edge burr detection method |
CN111650220A (en) * | 2020-07-15 | 2020-09-11 | 博科视(苏州)技术有限公司 | Vision-based image-text defect detection method |
CN112748128A (en) * | 2020-12-24 | 2021-05-04 | 凌云光技术股份有限公司 | Front and back synchronous detection system |
CN113070240A (en) * | 2021-03-25 | 2021-07-06 | 南京工业大学 | Copper plate surface defect detection and automatic classification method based on machine vision and deep learning |
CN218567234U (en) * | 2021-05-07 | 2023-03-03 | 新乡市新机迪元科技有限公司 | Reel printing quality on-line measuring device |
CN217766133U (en) * | 2022-06-23 | 2022-11-08 | 潍坊华田精工设备有限公司 | Single double-sided printed matter on-line detection system |
CN115855951A (en) * | 2022-11-26 | 2023-03-28 | 苏州三立检测技术有限公司 | Cloth cover flaw detection method and system |
CN219574806U (en) * | 2022-12-13 | 2023-08-22 | 东信和平科技股份有限公司 | Automatic checking system for large printed matter of intelligent card |
Non-Patent Citations (1)
Title |
---|
李达辉等.数字图像处理核心技术及应用.电子科技大学出版社,2019,第151页. * |
Also Published As
Publication number | Publication date |
---|---|
CN117031052A (en) | 2023-11-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111650220B (en) | Vision-based image-text defect detection method | |
CN108548820B (en) | Cosmetic paper label defect detection method | |
CN113239930B (en) | Glass paper defect identification method, system, device and storage medium | |
CN106570511B (en) | Keyboard character defect online detection system and method based on computer vision | |
CN109767445B (en) | High-precision PCB defect intelligent detection method | |
CN107084992A (en) | A kind of capsule detection method and system based on machine vision | |
CN109142383A (en) | One kind being based on morphologic character defect inspection method and device | |
CN113034488B (en) | Visual inspection method for ink-jet printed matter | |
CN110853013A (en) | Method for detecting spraying quality of outer wall of white household appliance | |
CN114549504A (en) | Appearance quality detection method based on machine vision | |
CN117031052B (en) | Single printed matter front and back vision detection control system | |
CN112712058A (en) | Character recognition and extraction method | |
CN110533660B (en) | Method for detecting silk-screen defect of electronic product shell | |
CN114581368A (en) | Bar welding method and device based on binocular vision | |
CN113936291A (en) | Aluminum template quality inspection and recovery method based on machine vision | |
CN116482104B (en) | Thermal transfer film detection method | |
CN116309375B (en) | Method for detecting double-sided defects of solid wood plate and determining intelligent processing coordinates | |
CN112247675A (en) | System and method for detecting service life of cutter based on big data autonomous learning | |
CN113538323A (en) | Defect detection method and system based on deep learning | |
CN117761060A (en) | Visual detection system and detection method thereof | |
CN115457563A (en) | Zero-missing-detection and low-error-identification ship water gauge reading method | |
CN111091086B (en) | Method for improving identification rate of single characteristic information of logistics surface by utilizing machine vision technology | |
CN109624502B (en) | Efficient low-loss printed product production system and method | |
CN116429782B (en) | Saw chain defect detection method based on residual error network and knowledge coding | |
CN117282687A (en) | Automatic mark picking and supplementing control system for visual inspection of printed matter |
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 |