CN116152249A - Intelligent digital printing quality detection method - Google Patents

Intelligent digital printing quality detection method Download PDF

Info

Publication number
CN116152249A
CN116152249A CN202310424973.7A CN202310424973A CN116152249A CN 116152249 A CN116152249 A CN 116152249A CN 202310424973 A CN202310424973 A CN 202310424973A CN 116152249 A CN116152249 A CN 116152249A
Authority
CN
China
Prior art keywords
detected
ssim
image block
target
value
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.)
Granted
Application number
CN202310424973.7A
Other languages
Chinese (zh)
Other versions
CN116152249B (en
Inventor
周长坤
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jining Leader Printing Co ltd
Original Assignee
Jining Leader Printing Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Jining Leader Printing Co ltd filed Critical Jining Leader Printing Co ltd
Priority to CN202310424973.7A priority Critical patent/CN116152249B/en
Publication of CN116152249A publication Critical patent/CN116152249A/en
Application granted granted Critical
Publication of CN116152249B publication Critical patent/CN116152249B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • G06T7/001—Industrial image inspection using an image reference approach
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/90—Determination of colour characteristics
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30108—Industrial image inspection
    • G06T2207/30144—Printing quality
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/30—Computing systems specially adapted for manufacturing

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Quality & Reliability (AREA)
  • Image Processing (AREA)

Abstract

The invention relates to the technical field of image data processing, in particular to an intelligent digital printing quality detection method, which comprises the following steps: obtaining a template image and an image to be detected, performing blocking processing on the template image and the image to be detected, performing initial improvement on an original SSIM value by combining hue and saturation, obtaining an initial improved SSIM value of each image block to be detected, obtaining a difference row sequence of each row and a difference column sequence of each column in the image to be detected, adjusting the initial improved SSIM values of all the image blocks to be detected by combining the distribution condition of the difference row sequence and the difference column sequence, obtaining improved SSIM values of all the image blocks to be detected, and obtaining a suspected defect image block according to the improved SSIM values of all the image blocks to be detected, thereby finishing the quality detection of the printed matter image. The invention improves the detection accuracy of the defects and the integrity of the edges of the defects, eliminates noise interference and improves the accuracy of the subsequent quality detection of the printed image.

Description

Intelligent digital printing quality detection method
Technical Field
The invention relates to the technical field of image data processing, in particular to an intelligent digital printing quality detection method.
Background
In the complete digital printing process, deviation occurs in the quality of paper in the printing pretreatment, the printing pressure in the printing process is uneven, defects such as scratches, omission marks, color spots and the like on the surface of a printed matter can be caused by improper operations such as post-printing glazing and laminating, the printing quality can be influenced, and the printing quality is not accepted by customers. Therefore, detection of print quality is an indispensable operation therein.
In the prior art, similarity measurement is generally carried out on the acquired printed matter image and template image, and then quality detection is carried out on the printed matter image. In the prior art, an SSIM algorithm is often adopted to measure the similarity between a digital printing image and a template image.
Because of the randomness of the size and the position of the printing defect, when the tiny defect occurs, the result is larger when the SSIM algorithm is used for carrying out similarity measurement on the printed matter. Meanwhile, noise is inevitably generated when the printed matter images are acquired and transmitted, and the detail of the printed matter images is lost due to the fact that the noise removal processing is directly carried out, so that the accuracy of subsequent quality detection is affected. But noise points can influence similarity measurement at the same time, and the influence of micro defects and noise on printing quality is different, so that an SSIM algorithm needs to be adjusted by combining the shapes and distribution characteristics of the printing defects and the noise points, and the accuracy of printing quality detection is further improved.
Disclosure of Invention
The invention provides an intelligent digital printing quality detection method for solving the existing problems.
The intelligent digital printing quality detection method adopts the following technical scheme:
the invention provides an intelligent digital printing quality detection method, which comprises the following steps:
collecting a template image and an image to be detected, dividing the template image into a plurality of template image blocks with the size equal to a preset size, and dividing the image to be detected into a plurality of image blocks to be detected with the size equal to the preset size;
marking any image block to be detected as a target image block to be detected, and marking the template image blocks at the same positions in the template image as target template image blocks; calculating an SSIM value between the target image block to be detected and the target template image block, and marking the SSIM value as an original SSIM value of the target image block to be detected;
according to the tone value and the saturation value of the center points of the target to-be-detected image block and the target template image block, improving the original SSIM value of the target to-be-detected image block to obtain an initial improved SSIM value of the target to-be-detected image block; obtaining initial improved SSIM values of all image blocks to be detected;
marking a sequence formed by initial improved SSIM values of all image blocks to be detected in the same row in the image to be detected according to the sequence as an SSIM row sequence; marking a sequence formed by initial improved SSIM values of all image blocks to be detected in the same column in the image to be detected according to the sequence as an SSIM column sequence; obtaining a difference value row sequence of each SSIM row sequence and a difference value column sequence of each SSIM column sequence;
obtaining the defect edge possibility of the target image block to be detected according to the distribution characteristics of the difference line sequence of the line of the target image block to be detected and the difference column sequence of the column; the difference value between the initial improved SSIM value of the target image block to be detected and the defect edge possibility of the target image block to be detected is recorded as the improved SSIM value of the target image block to be detected, and the improved SSIM values of all the image blocks to be detected are obtained;
marking the image blocks to be detected with the improved SSIM value smaller than the threshold value as suspected defect image blocks, and marking all the suspected defect image blocks to finish the quality detection of the printed matter image.
Further, the obtaining the initial improved SSIM value of the target image block to be detected includes the following specific steps:
the calculation formula of the initial improved SSIM value of the target image block to be detected is as follows:
Figure SMS_1
in the method, in the process of the invention,
Figure SMS_3
an initial modified SSIM value representing a target image block to be detected,
Figure SMS_5
the original SSIM value representing the target image block to be detected,
Figure SMS_7
representing the hue value of the center point of the target image block to be detected,
Figure SMS_4
representing the hue value of the center point of the target template image block,
Figure SMS_6
a saturation value representing the center point of the target image block to be detected,
Figure SMS_8
a saturation value representing the center point of the target template image block,
Figure SMS_9
the representation takes the absolute value of the value,
Figure SMS_2
an exponential function based on a natural constant is represented.
Further, the method for obtaining the difference value row sequence of each SSIM row sequence and the difference value column sequence of each SSIM column sequence comprises the following specific steps:
acquiring a first-order differential sequence of each SSIM row sequence and each SSIM column sequence, and recording a sequence formed by a first numerical value of each SSIM row sequence and the first-order differential sequence of each SSIM row sequence as a differential value row sequence of each SSIM row sequence, wherein the ith data in the differential value row sequence is used as an SSIM row differential value of the ith image block to be detected in the corresponding row; and marking a sequence formed by the first numerical value of each SSIM sequence and the first differential sequence of each SSIM sequence as a differential sequence of each SSIM sequence, wherein the jth data in the differential sequence is used as the SSIM sequence differential value of the jth image block to be detected in the corresponding sequence.
Further, calculating an SSIM value between the target image block to be detected and the target template image block, including the following specific steps:
and calculating the SSIM value of an R channel between the target image block to be detected and the target template image block according to an SSIM algorithm, and similarly, obtaining the SSIM value of a G channel and the SSIM value of a B channel between the target image block to be detected and the target template image block, and recording the average value of the SSIM values of the three channels as the SSIM value between the target image block to be detected and the target template image block.
Further, obtaining the defect edge possibility of the target image block to be detected, which comprises the following specific steps:
the calculation formula of the defect edge probability of the target image block to be detected is as follows:
Figure SMS_10
in the method, in the process of the invention,
Figure SMS_12
representing the defect edge likelihood of the target image block to be detected,
Figure SMS_16
representing the maximum number of consecutive target to-be-detected image blocks in the same row that are to the left of the target to-be-detected image block and that have the same sign as the SSIM row difference value of the target to-be-detected image block,
Figure SMS_20
representing the maximum number of consecutive target to-be-detected image blocks in the same row that are located to the right of the target to-be-detected image block and that have the same sign as the SSIM row difference value of the target to-be-detected image block,
Figure SMS_14
SSIM line difference values representing image blocks to be detected located on the left side of the target image block to be detected in the same line,
Figure SMS_18
an SSIM row difference value of the image block to be detected, which is positioned on the right side of the target image block to be detected in the same row, is represented;
Figure SMS_21
representing the maximum number of consecutive target to-be-detected image blocks in the same column above the target to-be-detected image block and having the same sign as the SSIM column difference value of the target to-be-detected image block,
Figure SMS_23
successive objects representing the same column below the target image block to be detected and having the same sign as the SSIM column difference of the target image block to be detectedThe maximum number of image blocks to be detected,
Figure SMS_11
the SSIM column difference value representing the image block to be detected above the target image block to be detected in the same column,
Figure SMS_15
the SSIM column difference value representing the image block to be detected below the target image block to be detected in the same column,
Figure SMS_19
the representation takes the absolute value of the value,
Figure SMS_22
indicating that the maximum value is taken,
Figure SMS_13
representing a Sigmoid function, M and N respectively represent the number of rows and columns of the image to be detected, A is a preset side length,
Figure SMS_17
representing an upward rounding.
The technical scheme of the invention has the beneficial effects that: when the similarity of an image to be detected and a template image is measured by the SSIM algorithm, color characteristics, detail information and noise influence in the image to be detected are not considered, edges of defects cannot be accurately acquired, noise is wrongly identified as the defects, and accordingly detection of color defects such as missing marks, misprints and color spots in the image to be detected is inaccurate.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flow chart of steps of an intelligent digital print quality detection method according to the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the present invention to achieve the preset purpose, the following detailed description refers to specific embodiments, structures, features and effects of an intelligent digital print quality detection method according to the present invention with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following specifically describes a specific scheme of the intelligent digital printing quality detection method provided by the invention with reference to the accompanying drawings.
Referring to fig. 1, a flowchart illustrating steps of a method for detecting intelligent digital print quality according to an embodiment of the present invention is shown, the method includes the following steps:
s001, acquiring a template image and an image to be detected, and performing blocking processing on the template image and the image to be detected.
In this embodiment, quality detection is performed on a digital print image, and similarity measurement is required between an acquired print image and a template image, so that a camera is placed above a digital printer, the printed digital print image is acquired in real time, the acquired image is recorded as an image to be detected, and the print image without printing defects is recorded as a template image, wherein the image to be detected and the template image are both RGB images.
The size of the template image and the size of the image to be detected are M multiplied by N, wherein M represents the number of rows of the image to be detected, N represents the number of columns of the image to be detected, and the template image and the image to be detected are expanded into the size by filling pixel points with gray values of 0 at the right and the lower sides of the image
Figure SMS_24
Wherein A is a preset side length,
Figure SMS_25
representing an upward rounding; dividing a template image into a plurality of image blocks with the size equal to a preset size, marking the image blocks as template image blocks, dividing an image to be detected into a plurality of image blocks with the size equal to the preset size, marking the image blocks as the image blocks to be detected, wherein the preset size is A multiplied by A.
In the embodiment of the present invention, if the preset side length a is 11, the preset dimension a×a is 11×11, and in other embodiments, the operator may set the preset side length and the preset dimension according to the actual implementation situation.
S002, carrying out initial improvement on the original SSIM value by combining the hue and the saturation to obtain an initial improved SSIM value of each image block to be detected, obtaining a difference row sequence of each row and a difference column sequence of each column in the image to be detected, and adjusting the initial improved SSIM values of all the image blocks to be detected by combining the distribution condition of the difference row sequence and the difference column sequence to obtain improved SSIM values of all the image blocks to be detected.
1. And carrying out initial improvement on the original SSIM value by combining the hue and the saturation to obtain initial improved SSIM values of all the image blocks to be detected.
It should be noted that, when a micro printing defect occurs, directly using the SSIM algorithm to measure the similarity between the image to be detected and the template image may result in a larger SSIM value, which may further result in a larger error of quality detection. For printed matters, more defects are missing printing, misprinting, color spots and the like, so that the color characteristics of an image to be detected and a template image are different, and an SSIM algorithm only considers the brightness, contrast and structure of the printed matter image and does not consider the color tone and saturation greatly influencing the color characteristics of the printed matter image, so that the obtained original SSIM value is not combined with the color characteristic difference of the image to be detected and the template image, and the detection result of the defects such as missing printing, misprinting, color spots and the like is inaccurate. The present embodiment performs initial improvement on the original SSIM value in combination with hue and saturation, and obtains an initial improved SSIM value for each image block to be detected.
In this embodiment, for any one image block to be detected in the image to be detected, the template image block at the same position in the template image is recorded as the template image block of the image block to be detected; and marking any one image block to be detected as a target image block to be detected, and marking a template image block of the target image block to be detected as a target template image block.
And calculating the SSIM value of an R channel between the target image block to be detected and the target template image block according to an SSIM algorithm, and similarly, obtaining the SSIM value of a G channel and the SSIM value of a B channel between the target image block to be detected and the target template image block, and recording the average value of the SSIM values of the three channels as the SSIM value between the target image block to be detected and the target template image block.
And converting the target to-be-detected image block and the target template image block into corresponding HSV images by taking the pixel point at the central position of the target to-be-detected image block as the central point of the target to-be-detected image block, taking the pixel point at the central position of the target template image block as the central point of the target template image block, and obtaining the tone value and the saturation value of the central point of the target to-be-detected image block and the tone value and the saturation value of the central point of the target template image block.
The calculation formula of the initial improved SSIM value of the target image block to be detected is as follows:
Figure SMS_26
in the method, in the process of the invention,
Figure SMS_28
an initial modified SSIM value representing a target image block to be detected,
Figure SMS_30
the original SSIM value representing the target image block to be detected,
Figure SMS_32
representing the hue value of the center point of the target image block to be detected,
Figure SMS_29
representing the hue value of the center point of the target template image block,
Figure SMS_31
a saturation value representing the center point of the target image block to be detected,
Figure SMS_33
a saturation value representing the center point of the target template image block,
Figure SMS_34
the representation takes the absolute value of the value,
Figure SMS_27
an exponential function based on a natural constant is represented.
Figure SMS_35
Representing the difference of the hue value of the center point of the target image block to be detected and the center point of the target template image block,
Figure SMS_36
representing the difference of saturation values of the center point of the target to-be-detected image block and the center point of the target template image block, wherein the larger the difference of hue values and the difference of saturation values of the target to-be-detected image block and the target template image block, the target to-be-detected image block and the target template image block areThe larger the difference of the color characteristics of the target template image block is, the initial improved SSIM value of the target image block to be detected
Figure SMS_37
The smaller.
An initial improved SSIM value is obtained for all image blocks to be detected.
Considering that most defects of the printed matter image are color defects, combining the difference of hue values and saturation values of the image to be detected and the template image to obtain the color characteristic difference of the image to be detected and the template image, further improving the original SSIM values of the image to be detected and the template image according to the color characteristic difference, and obtaining an initial improved SSIM value, wherein the detection result of color defects such as missing printing, misprinting, color spots and the like, which are frequently caused by the printed matter image, is more accurate and is more suitable for detecting the quality of the printed matter.
2. And acquiring a difference line sequence of each line and a difference column sequence of each column in the image to be detected, and adjusting initial improved SSIM values of all the image blocks to be detected by combining the distribution conditions of the difference line sequence and the difference column sequence to obtain improved SSIM values of all the image blocks to be detected.
It should be noted that, the defect on the print image has a range characteristic, and for a defect on a certain position on the print image, the possibility that a plurality of image blocks to be detected around the defect have printing defects is greater, so that initial improved SSIM values of the plurality of image blocks to be detected are similar and smaller than those of the normal image blocks to be detected. Considering that the subsequent processing operations on different defects on the printed matter image are different, the integrity of defect detection of the printed matter image needs to be ensured, and the edge of the defect needs to be completely acquired, but for tiny defects, for example, when tiny missing printing occurs, the missing printing edge is close to the edge of a normal area on a template image, the SSIM value of an image block to be detected at the missing printing edge is larger, but in order to ensure that the defect is completely extracted, the initial improved SSIM value of the image block to be detected at the missing printing edge needs to be reduced and adjusted, so that the subsequent complete detection of the defect is facilitated. Meanwhile, when the printed matter image is acquired and transmitted, the printed matter image can generate noise, but for the image block to be detected where the noise is located, compared with the surrounding normal image blocks to be detected, the initial improvement SSIM is smaller, but the influence of the noise and the printing defect on the image quality is different, the initial improvement SSIM of the image block to be detected is required to be adjusted, and the value of the initial improvement SSIM of the image block to be detected where the noise is located is increased.
It should be further noted that, because the defect of the print image has a range, the damage degree at the center of the defect is the largest, the damage degree is diffused from the center to the periphery, and gradually decreases, particularly when the defect is embodied on the initial improved SSIM value of the image block to be detected, the difference between the corresponding image block to be detected at the center of the defect and the template image block is larger, the defect degree is gradually diffused outwards, the difference between the defect and the template image block is further decreased, and the corresponding initial improved SSIM value is decreased. However, for the image block to be detected, where the noise point is located, the surrounding image blocks to be detected are not different from the template image block, so that the initial improvement SSIM value of the surrounding image block to be detected is larger, and the surrounding image block to be detected does not have a gradient characteristic. Considering that the influence of noise and defects on the image quality of a printed matter is different, the initial improvement SSIM of the image block to be detected needs to be adjusted by combining the two-dimensional characteristics of the printing defects and the noise, and the initial improvement SSIM value of the image block to be detected at the edge of the defects is reduced.
In this embodiment, a sequence formed by the initial improved SSIM values of all image blocks to be detected located in the same line in the image to be detected in order is denoted as an SSIM line sequence; and marking a sequence formed by the initial improved SSIM values of all the image blocks to be detected in the same column in the image to be detected according to the sequence as an SSIM column sequence.
Acquiring a first-order differential sequence of each SSIM row sequence and each SSIM column sequence, and recording a sequence formed by a first numerical value of each SSIM row sequence and the first-order differential sequence of each SSIM row sequence as a differential value row sequence of each SSIM row sequence, wherein the ith data in the differential value row sequence is used as an SSIM row differential value of the ith image block to be detected in the corresponding row; and marking a sequence formed by the first numerical value of each SSIM sequence and the first differential sequence of each SSIM sequence as a differential sequence of each SSIM sequence, wherein the jth data in the differential sequence is used as the SSIM sequence differential value of the jth image block to be detected in the corresponding sequence.
Marking any image block to be detected as a target image block to be detected, marking a template image block of the target image block to be detected as a target template image block, and calculating the defect edge probability of the target image block to be detected by the following formula:
Figure SMS_38
in the method, in the process of the invention,
Figure SMS_40
representing the defect edge likelihood of the target image block to be detected,
Figure SMS_43
representing the maximum number of consecutive target to-be-detected image blocks in the same row that are to the left of the target to-be-detected image block and that have the same sign as the SSIM row difference value of the target to-be-detected image block,
Figure SMS_47
representing the maximum number of consecutive target to-be-detected image blocks in the same row that are located to the right of the target to-be-detected image block and that have the same sign as the SSIM row difference value of the target to-be-detected image block,
Figure SMS_42
SSIM line difference values representing image blocks to be detected located on the left side of the target image block to be detected in the same line,
Figure SMS_45
an SSIM row difference value of the image block to be detected, which is positioned on the right side of the target image block to be detected in the same row, is represented;
Figure SMS_49
representing a plurality of consecutive target to-be-detected image blocks in the same column above the target to-be-detected image block and having the same sign as the SSIM column difference value of the target to-be-detected image blockThe maximum number of times that the device can be used,
Figure SMS_50
representing the maximum number of consecutive target to-be-detected image blocks in the same column below the target to-be-detected image block and having the same sign as the SSIM column difference value of the target to-be-detected image block,
Figure SMS_39
the SSIM column difference value representing the image block to be detected above the target image block to be detected in the same column,
Figure SMS_44
the SSIM column difference value representing the image block to be detected below the target image block to be detected in the same column,
Figure SMS_48
the representation takes the absolute value of the value,
Figure SMS_51
indicating that the maximum value is taken,
Figure SMS_41
representing a Sigmoid function, M and N respectively represent the number of rows and columns of the image to be detected, A is a preset side length,
Figure SMS_46
representing an upward rounding.
Figure SMS_52
Representing the length of the SSIM line sequence and also representing the length of the SSIM line difference for normalization;
Figure SMS_53
the length of the SSIM column sequence is represented, as well as the length of the SSIM column difference for normalization.
Figure SMS_56
Variation trend of SSIM line difference value of image blocks to be detected, which are positioned around target image blocks to be detected in same lineIf the target to-be-detected image block is at the edge of the defect, the damage degree of the defect is gradually reduced due to the diffusion from the center to the periphery, the initial improvement SSIM value is gradually increased, and the initial improvement SSIM value of the to-be-detected image blocks around the target to-be-detected image block presents an increasing trend, namely, the more target to-be-detected image blocks which are positioned around the target to-be-detected image block in the same row and have the same sign as the SSIM row difference value of the target to-be-detected image block
Figure SMS_59
The larger the probability that the target image block to be detected is at the edge of the defect is, the greater the probability that the target image block to be detected is at the edge of the defect is
Figure SMS_61
The larger; when the target to-be-detected image block is located at the noise position, the initial improved SSIM values of the to-be-detected image blocks around the target to-be-detected image block are similar and larger, and the less the target to-be-detected image blocks which are located around the target to-be-detected image block in the same row and have the same sign as the SSIM row difference value of the target to-be-detected image block, namely
Figure SMS_55
The smaller the target image block to be detected is, the greater the possibility that the target image block to be detected is positioned at noise is, the smaller the possibility that the target image block to be detected is positioned at the edge of the defect is, the defect edge possibility of the target image block to be detected is
Figure SMS_58
The smaller; in the same way, the processing method comprises the steps of,
Figure SMS_60
the larger the probability that the target image block to be detected is at the edge of the defect is, the greater the probability that the target image block to be detected is at the edge of the defect is
Figure SMS_62
The larger the size of the container,
Figure SMS_54
the smaller the target to-be-detected imageThe greater the likelihood that the block is at noise, the less the likelihood that the target image block to be detected is at the edge of the defect, the likelihood that the target image block to be detected is at the edge of the defect
Figure SMS_57
The smaller.
Figure SMS_64
Representing the difference degree of the SSIM line differences of the image blocks to be detected which are positioned around the target image block to be detected in the same line, if the target image block to be detected is positioned at the edge of the defect, the initial improved SSIM value of the image block to be detected on one side of the target image block to be detected is smaller, and the initial improved SSIM value of the image block to be detected on the other side is larger, so that the greater the difference degree of the SSIM line differences of the image blocks to be detected around the target image block to be detected is, namely
Figure SMS_68
The larger the probability that the target image block to be detected is at the edge of the defect is, the greater the probability that the target image block to be detected is at the edge of the defect is
Figure SMS_69
The larger; if the target to-be-detected image block is positioned at the noise, the initial improved SSIM values of the to-be-detected image blocks around the target to-be-detected image block are similar, the smaller the difference degree of SSIM line difference values of the to-be-detected image blocks around the target to-be-detected image block is, namely
Figure SMS_65
The larger the target image block to be detected is, the larger the possibility that the target image block to be detected is positioned at noise is, the smaller the possibility that the target image block to be detected is positioned at the edge of the defect is, and the defect edge possibility of the target image block to be detected is
Figure SMS_67
The smaller; in the same way, the processing method comprises the steps of,
Figure SMS_70
the larger the target image block to be detectedThe greater the likelihood of being at the edge of the defect, the greater the likelihood of the defective edge of the target image block to be detected
Figure SMS_71
The larger the size of the container,
Figure SMS_63
the smaller the target image block to be detected is, the greater the possibility that the target image block to be detected is positioned at noise is, the smaller the possibility that the target image block to be detected is positioned at the edge of the defect is, the defect edge possibility of the target image block to be detected is
Figure SMS_66
The smaller.
For example, the difference line sequence of the SSIM line sequence of the line where the target image block to be detected is {0.2, -0.3, -0.4, -0.6, -0.2,0.7}, where the SSIM line difference of the target image block to be detected is-0.6, the sign of the SSIM line difference of the target image block to be detected is negative, at this time
Figure SMS_72
，
Figure SMS_73
。
And recording the difference value between the initial improved SSIM value of the target image block to be detected and the defect edge possibility of the target image block to be detected as the improved SSIM value of the target image block to be detected.
Improved SSIM values are obtained for all image blocks to be detected.
The initial improved SSIM value of the image block to be detected is adjusted through the two-dimensional characteristics of the printing defect and the noise, the initial improved SSIM value of the image block to be detected at the edge of the defect is reduced, the initial improved SSIM value of the image block to be detected at the noise position is improved, the acquired edge integrity of the defect is ensured, the interference of the noise on the defect detection is eliminated, and the accuracy of the subsequent detection on the quality of the printing image is further improved.
S003, obtaining suspected defect image blocks according to the improved SSIM values of all the image blocks to be detected, and finishing quality detection of the printed matter images.
Marking the image blocks to be detected with the improved SSIM value smaller than the threshold value as suspected defect image blocks, and marking all the suspected defect image blocks to finish the quality detection of the printed matter image.
In this embodiment, the threshold is 0.6, and in other embodiments, the practitioner may set the threshold empirically.
When the similarity of an image to be detected and a template image is measured by the SSIM algorithm, color characteristics, detail information and noise influence in the image to be detected are not considered, edges of defects cannot be accurately acquired, noise is wrongly identified as the defects, and accordingly detection of color defects such as missing marks, misprints and color spots in the image to be detected is inaccurate.
It should be noted that: the sequence of the embodiments of the present invention is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments.

Claims (5)

1. An intelligent digital printing quality detection method is characterized by comprising the following steps:
collecting a template image and an image to be detected, dividing the template image into a plurality of template image blocks with the size equal to a preset size, and dividing the image to be detected into a plurality of image blocks to be detected with the size equal to the preset size;
marking any image block to be detected as a target image block to be detected, and marking the template image blocks at the same positions in the template image as target template image blocks; calculating an SSIM value between the target image block to be detected and the target template image block, and marking the SSIM value as an original SSIM value of the target image block to be detected;
according to the tone value and the saturation value of the center points of the target to-be-detected image block and the target template image block, improving the original SSIM value of the target to-be-detected image block to obtain an initial improved SSIM value of the target to-be-detected image block; obtaining initial improved SSIM values of all image blocks to be detected;
marking a sequence formed by initial improved SSIM values of all image blocks to be detected in the same row in the image to be detected according to the sequence as an SSIM row sequence; marking a sequence formed by initial improved SSIM values of all image blocks to be detected in the same column in the image to be detected according to the sequence as an SSIM column sequence; obtaining a difference value row sequence of each SSIM row sequence and a difference value column sequence of each SSIM column sequence;
obtaining the defect edge possibility of the target image block to be detected according to the distribution characteristics of the difference line sequence of the line of the target image block to be detected and the difference column sequence of the column; the difference value between the initial improved SSIM value of the target image block to be detected and the defect edge possibility of the target image block to be detected is recorded as the improved SSIM value of the target image block to be detected, and the improved SSIM values of all the image blocks to be detected are obtained;
marking the image blocks to be detected with the improved SSIM value smaller than the threshold value as suspected defect image blocks, and marking all the suspected defect image blocks to finish the quality detection of the printed matter image.
2. The intelligent digital print quality inspection method according to claim 1, wherein the obtaining of the initial improved SSIM value of the target image block to be inspected comprises the following specific steps:
the calculation formula of the initial improved SSIM value of the target image block to be detected is as follows:
Figure QLYQS_1
in the method, in the process of the invention,
Figure QLYQS_3
an initial improved SSIM value representing a target image block to be detected,/->
Figure QLYQS_5
Original SSIM value representing the target image block to be detected,/->
Figure QLYQS_7
Tone value representing the center point of the target image block to be detected,/->
Figure QLYQS_4
Tone value representing the center point of the target template image block,/->
Figure QLYQS_6
Saturation value representing the center point of the target image block to be detected,/or->
Figure QLYQS_8
Saturation value representing the center point of the target template image block,/->
Figure QLYQS_9
Representing absolute value>
Figure QLYQS_2
An exponential function based on a natural constant is represented.
3. The method for detecting the quality of intelligent digital printing according to claim 1, wherein the steps of obtaining the difference line sequence of each SSIM line sequence and the difference column sequence of each SSIM column sequence comprise the following specific steps:
acquiring a first-order differential sequence of each SSIM row sequence and each SSIM column sequence, and recording a sequence formed by a first numerical value of each SSIM row sequence and the first-order differential sequence of each SSIM row sequence as a differential value row sequence of each SSIM row sequence, wherein the ith data in the differential value row sequence is used as an SSIM row differential value of the ith image block to be detected in the corresponding row; and marking a sequence formed by the first numerical value of each SSIM sequence and the first differential sequence of each SSIM sequence as a differential sequence of each SSIM sequence, wherein the jth data in the differential sequence is used as the SSIM sequence differential value of the jth image block to be detected in the corresponding sequence.
4. The intelligent digital printing quality detection method according to claim 1, wherein calculating the SSIM value between the target image block to be detected and the target template image block comprises the following specific steps:
and calculating the SSIM value of an R channel between the target image block to be detected and the target template image block according to an SSIM algorithm, and similarly, obtaining the SSIM value of a G channel and the SSIM value of a B channel between the target image block to be detected and the target template image block, and recording the average value of the SSIM values of the three channels as the SSIM value between the target image block to be detected and the target template image block.
5. A method for intelligent digital print quality inspection according to claim 3, wherein obtaining the defect edge probability of the target image block to be inspected comprises the following specific steps:
the calculation formula of the defect edge probability of the target image block to be detected is as follows:
Figure QLYQS_10
in the method, in the process of the invention,
Figure QLYQS_12
representing the defect edge probability of the target image block to be detected,/->
Figure QLYQS_16
Representing the maximum number of consecutive target to-be-detected image blocks in the same row, located to the left of the target to-be-detected image block and having the same sign as the SSIM row difference value of the target to-be-detected image block,/>
Figure QLYQS_20
Representing the maximum number of consecutive target to-be-detected image blocks in the same row, located to the right of the target to-be-detected image block and having the same sign as the SSIM row difference value of the target to-be-detected image block,/>
Figure QLYQS_14
SSIM line difference value representing the image block to be detected located on the left side of the target image block to be detected in the same line,/->
Figure QLYQS_18
An SSIM row difference value of the image block to be detected, which is positioned on the right side of the target image block to be detected in the same row, is represented; />
Figure QLYQS_21
Representing the maximum number of consecutive target to-be-detected image blocks in the same column above the target to-be-detected image block and having the same sign as the SSIM column difference of the target to-be-detected image block,/>
Figure QLYQS_23
Representing the maximum number of consecutive target to-be-detected image blocks in the same column below the target to-be-detected image block and having the same sign as the SSIM column difference of the target to-be-detected image block,/the maximum number of consecutive target to-be-detected image blocks is equal to the SSIM column difference of the target to-be-detected image block>
Figure QLYQS_11
SSIM column difference values representing the image blocks to be detected located above the target image block to be detected in the same column,/>
Figure QLYQS_15
SSIM column difference values representing the image blocks to be detected located below the target image block to be detected in the same column,/>
Figure QLYQS_19
Representing absolute value>
Figure QLYQS_22
Indicating that the maximum value is taken>
Figure QLYQS_13
Representing a Sigmoid function, M and N respectively represent the number of rows and columns of an image to be detected, A is a preset side length, and +.>
Figure QLYQS_17
Representing an upward rounding. />
CN202310424973.7A 2023-04-20 2023-04-20 Intelligent digital printing quality detection method Active CN116152249B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310424973.7A CN116152249B (en) 2023-04-20 2023-04-20 Intelligent digital printing quality detection method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310424973.7A CN116152249B (en) 2023-04-20 2023-04-20 Intelligent digital printing quality detection method

Publications (2)

Publication Number Publication Date
CN116152249A true CN116152249A (en) 2023-05-23
CN116152249B CN116152249B (en) 2023-07-07

Family

ID=86362164

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310424973.7A Active CN116152249B (en) 2023-04-20 2023-04-20 Intelligent digital printing quality detection method

Country Status (1)

Country Link
CN (1) CN116152249B (en)

Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110069894A1 (en) * 2009-09-24 2011-03-24 Marie Vans Image defect detection
US20140022460A1 (en) * 2011-04-12 2014-01-23 Dolby Laboratories Licensing Corporation Quality Assessment for Images that Have Extended Dynamic Ranges or Wide Color Gamuts
CN104361593A (en) * 2014-11-14 2015-02-18 南京大学 Color image quality evaluation method based on HVSs and quaternions
JP2017164935A (en) * 2016-03-15 2017-09-21 株式会社リコー Printed matter inspection device, printed matter inspection method, and program
CN108010024A (en) * 2017-12-11 2018-05-08 宁波大学 It is a kind of blind with reference to tone mapping graph image quality evaluation method
CN110782453A (en) * 2019-11-06 2020-02-11 西安工程大学 Printed matter quality detection method based on image feature matching
CN112508826A (en) * 2020-11-16 2021-03-16 哈尔滨工业大学（深圳） Printed matter defect detection method based on feature registration and gradient shape matching fusion
CN113706464A (en) * 2021-07-22 2021-11-26 西安交通大学 Printed matter appearance quality detection method and system
CN113793274A (en) * 2021-08-26 2021-12-14 中国科学院沈阳自动化研究所 Highlight image restoration method based on tone
CN114842011A (en) * 2022-07-04 2022-08-02 启东万惠机械制造有限公司 Bearing wear detection method and system based on image processing
CN115272341A (en) * 2022-09-29 2022-11-01 华联机械集团有限公司 Packaging machine defect product detection method based on machine vision

Patent Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110069894A1 (en) * 2009-09-24 2011-03-24 Marie Vans Image defect detection
US20140022460A1 (en) * 2011-04-12 2014-01-23 Dolby Laboratories Licensing Corporation Quality Assessment for Images that Have Extended Dynamic Ranges or Wide Color Gamuts
CN104361593A (en) * 2014-11-14 2015-02-18 南京大学 Color image quality evaluation method based on HVSs and quaternions
JP2017164935A (en) * 2016-03-15 2017-09-21 株式会社リコー Printed matter inspection device, printed matter inspection method, and program
CN108010024A (en) * 2017-12-11 2018-05-08 宁波大学 It is a kind of blind with reference to tone mapping graph image quality evaluation method
CN110782453A (en) * 2019-11-06 2020-02-11 西安工程大学 Printed matter quality detection method based on image feature matching
CN112508826A (en) * 2020-11-16 2021-03-16 哈尔滨工业大学（深圳） Printed matter defect detection method based on feature registration and gradient shape matching fusion
CN113706464A (en) * 2021-07-22 2021-11-26 西安交通大学 Printed matter appearance quality detection method and system
CN113793274A (en) * 2021-08-26 2021-12-14 中国科学院沈阳自动化研究所 Highlight image restoration method based on tone
CN114842011A (en) * 2022-07-04 2022-08-02 启东万惠机械制造有限公司 Bearing wear detection method and system based on image processing
CN115272341A (en) * 2022-09-29 2022-11-01 华联机械集团有限公司 Packaging machine defect product detection method based on machine vision

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
曾台英;邵雪;汪祖辉;: "基于灰平衡的印刷质量控制和评价", 包装学报, no. 04 *
李雨;王振东;孙威;林冠男;: "海洋文化印刷品中图像的改进增强算法", 盐城工学院学报(自然科学版), no. 03 *

Also Published As

Publication number Publication date
CN116152249B (en) 2023-07-07

Similar Documents

Publication Publication Date Title
CN108918526B (en) Notch defect detection method for flexible IC packaging substrate circuit
CN108918536B (en) Tire mold surface character defect detection method, device, equipment and storage medium
CN113344857B (en) Defect detection network training method, defect detection method and storage medium
CN113554629A (en) Strip steel red rust defect detection method based on artificial intelligence
CN114494259A (en) Cloth defect detection method based on artificial intelligence
CN107389693B (en) Automatic detection method for defects of printed matter based on machine vision
CN111242896A (en) Color printing label defect detection and quality rating method
CN109886954B (en) Printed matter defect detection method
CN113888536B (en) Printed matter double image detection method and system based on computer vision
CN114723705A (en) Cloth flaw detection method based on image processing
CN113160222A (en) Production data identification method for industrial information image
CN116630304B (en) Lithium battery mold processing detection method and system based on artificial intelligence
CN114266764A (en) Character integrity detection method and device for printed label
CN115078365A (en) Soft package printing quality defect detection method
CN115880280A (en) Detection method for quality of steel structure weld joint
CN117152129B (en) Visual detection method and system for surface defects of battery cover plate
CN115115603A (en) Automobile accessory flywheel surface detection method based on artificial intelligence
CN117541588B (en) Printing defect detection method for paper product
CN116152249B (en) Intelligent digital printing quality detection method
CN116402822B (en) Concrete structure image detection method and device, electronic equipment and storage medium
CN116934761A (en) Self-adaptive detection method for defects of latex gloves
CN116109577A (en) Printing label defect detection system and method
CN115272343A (en) Method for detecting printing quality of packaging printed matter
CN113223098B (en) Preprocessing optimization method for image color classification
CN115457563A (en) Zero-missing-detection and low-error-identification ship water gauge reading method

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