CN115876804A - Visual detection method and system for defects of mask - Google Patents
Visual detection method and system for defects of mask Download PDFInfo
- Publication number
- CN115876804A CN115876804A CN202310132555.0A CN202310132555A CN115876804A CN 115876804 A CN115876804 A CN 115876804A CN 202310132555 A CN202310132555 A CN 202310132555A CN 115876804 A CN115876804 A CN 115876804A
- Authority
- CN
- China
- Prior art keywords
- mask
- defect
- evaluation
- score
- 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.)
- Granted
Links
Images
Classifications
-
- 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
- Image Analysis (AREA)
Abstract
The invention discloses a method and a system for visual inspection of mask defects, which relate to the technical field of visual inspection of mask defects and comprise the following steps: an image acquisition unit; constructing a unit and constructing a mask defect evaluation system; the first scoring unit is used for scoring the current mask according to a mask defect evaluation system; the second evaluation unit screens the masks for the first time according to the final mask defect scores, marks the mask three-dimensional model according to the mask defect scores and mask defect positions, and evaluates the mask defect degree to form a defect evaluation value; and the screening unit marks the position of the mask with the defect evaluation value lower than the threshold value, removes the mask from the production line by using the grabbing device, and performs secondary screening to finish defect detection. Can accomplish defect mark fast to the gauze mask, for conventional judgement mode based on machine vision, the evaluation flow is shorter, and the error is also less, can once only confirm that there are a plurality of defects in the gauze mask, and evaluation efficiency is also higher.
Description
Technical Field
The invention relates to the technical field of mask defect visual detection, in particular to a mask defect visual detection method and system.
Background
The mask mainly comprises: the surface body part: the noodle body is divided into an inner layer, a middle layer and an outer layer. The outer layer and the inner layer adopt PP non-woven fabrics, and the middle layer adopts melt-blown fabrics. The outer layer is used for preventing spray and dust, the middle layer is a core functional layer and is used for filtering spray, particles or bacteria, and the inner layer mainly absorbs moisture; nose bridge strip: the nose bridge strip is mainly used for a part of the attaching face and the nose bridge to play roles in sealing and fixing, and is made of three common materials, namely full plastic, plastic-coated metal cores (single core and double cores) and aluminum; ear belt: ear straps are used for wearing masks, and are generally flat and round.
Most gauze masks all adopt the non-woven fabrics to produce at present, melt and spout various defects such as cloth mask can produce ear area welding failure, the nose bridge strip is not installed or the installation is wrong, ear area, nose bridge strip length are different, even melt and spout cloth breakage in process of production, these defects can lead to the gauze mask to use well or be difficult to play due effect.
Therefore, in the production process of the melt-blown mask, the mask must be detected to reduce the probability of the occurrence of the above-mentioned defective products, and the existing detection method for the mask defects is usually realized based on machine vision in consideration of low detection efficiency and high cost of artificial defects and with the rapid development of image processing and pattern recognition technology.
Although the existing detection method based on machine vision is efficient, the defect judgment standard and the actual sale standard have certain difference during detection, and masks on a production line can be judged to be unqualified products as long as the defects exist, so that the probability of defective products is larger than expected if the defects of the masks are judged based on machine vision, and the economic benefit is reduced.
Disclosure of Invention
Technical problem to be solved
Aiming at the defects of the prior art, the invention provides a visual detection method and a system for defects of a mask, wherein an image acquisition unit is arranged; constructing a unit and constructing a mask defect evaluation system; the first scoring unit is used for scoring the current mask according to a mask defect evaluation system; the second evaluation unit is used for screening the mask for the first time according to the final mask defect score, marking the mask three-dimensional model according to the mask defect score and the mask defect position, and evaluating the mask defect degree to form a defect evaluation value; and the screening unit marks the position of the mask with the defect evaluation value lower than the threshold value, removes the mask from the production line by using the grabbing device, and performs secondary screening to finish defect detection. Can accomplish defect mark fast to the gauze mask, for conventional judgement mode based on machine vision, the evaluation flow is shorter, and the error is also less, can once only confirm that there are a plurality of defects in the gauze mask, and evaluation efficiency is also higher, has solved the problem that proposes in the background art.
(II) technical scheme
In order to realize the purpose, the invention is realized by the following technical scheme: a visual detection method for defects of a mask comprises the following steps: step one, collecting a mask image in multiple angles by adopting an image collecting device, establishing a mask three-dimensional model based on the mask image and the mask size, and marking each area of the mask on the mask three-dimensional model; step two, obtaining mask use evaluation according to a public channel, determining the influence degree of each regional defect of the mask on actual use, evaluating the defect degree, and constructing a mask defect evaluation system; comparing the acquired mask image with a standard mask image, scoring the current mask according to a mask defect evaluation system, and marking the defect score and a corresponding defect area on a three-dimensional model;
fourthly, screening the mask for the first time according to the final mask defect score, marking the mask three-dimensional model according to the mask defect score and the mask defect position, and evaluating the mask defect degree to form a defect evaluation value; and fifthly, marking the mask position with the defect evaluation value lower than the threshold value, removing the mask position from the production line by using a gripping device, and carrying out secondary screening to finish defect detection.
Further, the first step includes: step 101, arranging a plurality of rotating mechanisms and corresponding imaging devices on the mask along the advancing direction of a production line to form a plurality of multi-angle images on the mask; and 102, carrying out imaging quality evaluation on the obtained mask images, deleting the parts with the imaging quality lower than a threshold value, and establishing a mask image set.
Further, step 102 is followed by: 103, obtaining the sizes of the masks in the current batch, establishing a three-dimensional model of the masks, and rendering the three-dimensional model of the masks based on the current mask color; and 104, identifying the mask image set, comparing the mask image set with a standard mask image, determining each area of the current mask, and marking on the three-dimensional model of the mask.
Further, the step two includes the following steps: step 201, establishing a deep web crawler, performing deep retrieval on currently-disclosed network information, acquiring corresponding evaluation information of a user when each area of the mask is damaged, and establishing an evaluation information base; step 202, identifying evaluation information according to a semantic identification model established based on a natural language processing algorithm, and forming corresponding evaluation by a user when the mask is distinguished to have defects; classifying a plurality of different evaluations according to the trained SVM classifier, distinguishing different levels, and acquiring a plurality of evaluation level evaluation data sets after corresponding to the mask area.
Further, step 202 is followed by: step 203, establishing a mask scoring model based on a machine learning algorithm, training and testing, and scoring the evaluations in a plurality of evaluation data sets of different levels; and 204, constructing a mask defect evaluation system according to the defect degree, the defect area and the defect score, and uploading the mask defect evaluation system to a processing end.
Further, the third step includes: step 301, determining a corresponding mask defect standard image according to the defect score, the defect degree and the defect area to form a defect standard image library; step 302, after the mask passes through the production line, obtaining a mask image with imaging quality evaluation higher than a threshold value, comparing the mask image with images in a standard defect image library, and determining one or more masks with similarity exceeding the corresponding threshold value;
step 303, picking one or more images with the similarity exceeding a threshold, and scoring the defect of the current mask image as a final mask defect score when only one high-similarity image exists; when the high-similarity image is more than one, taking the accumulated score as the final mask defect score according to the sum of the plurality of defect scores; and 304, acquiring the final mask defect score and the mask defect position, and uploading to a processing end.
Further, the fourth step includes: step 401, obtaining a final mask defect score, comparing the final mask defect score with a corresponding threshold value, screening out a part of the final mask defect score larger than the threshold value, and determining the part as an unqualified mask;
step 402, aiming at the masks with a plurality of defects and final mask defect scores smaller than a threshold value, acquiring mask defect positions and corresponding defect positions, and marking the mask defect positions on a mask three-dimensional model;
and 403, constructing a mask wearing model according to the three-dimensional mask model, carrying out wearing simulation analysis, and determining the influence on mask wearing according to multiple linear regression analysis by changing the defect degree when the mask has defects in different areas to form area influence factors.
Further, after the standard service life of the mask is determined, the current mask is subjected to wearing simulation analysis, the time when the mask defect score reaches a threshold value is determined, and the service life of the mask with the defect is determined;
comparing the service life of the defective mask with that of the standard mask to form a service life ratio, and determining the defective mask as an unqualified mask if the value of the service life ratio is smaller than a corresponding threshold value;
and step 404, combining the region influence factor and the defect score of the region, and synthesizing to form a defect evaluation value QP.
Further, the association method of the defect evaluation value QP conforms to the following formula:
wherein the defect score of the noodle body part is M f Regional influence factor A m (ii) a Nose bridge strip Defect score B f The regional influence factor is A b (ii) a Ear band Defect score E f The regional influence factor is A e And C is a constant correction coefficient.
A mask defect visual inspection system, comprising:
the image acquisition unit acquires mask images in multiple angles by adopting an image acquisition device, establishes a three-dimensional mask model, and marks each area of the mask on the three-dimensional mask model;
the method comprises the steps of constructing a unit, obtaining mask use evaluation, determining the influence degree of defects of each area of the mask on actual use, evaluating the defect degree, and constructing a mask defect evaluation system;
the first scoring unit is used for comparing the acquired mask image with a standard mask image, scoring the current mask according to a mask defect evaluation system, and marking the defect score and a corresponding defect area on the three-dimensional model;
the second evaluation unit is used for screening the mask for the first time according to the final mask defect score, marking the mask three-dimensional model according to the mask defect score and the mask defect position, and evaluating the mask defect degree to form a defect evaluation value;
and the screening unit marks the position of the mask with the defect evaluation value lower than the threshold value, removes the mask from the production line by using the grabbing device, and performs secondary screening to finish defect detection.
(III) advantageous effects
The invention provides a visual detection method and a system for defects of a mask, which have the following beneficial effects:
by acquiring multi-angle high-quality mask images, the success rate of identifying mask defects based on vision is increased, errors are reduced, and by completing primary modeling and marking on the mask, the defects are displayed more visually when being judged, and the difficulty of quality inspection is reduced;
the method has the advantages that the method completes the construction of a mask defect evaluation system by collecting and distinguishing the use evaluation of the mask disclosed on the network, has an evaluation standard which is suitable for practical use when evaluating the mask defect, and the evaluation mark can also be directly hooked with the market sale, so that the matching degree with the market sale is higher; a mask defect evaluation system is established according to actual use, and the difference of different users in mask defect evaluation is eliminated.
Based on the construction of a mask defect evaluation system, a defect evaluation standard and a defect scoring judgment method based on similarity judgment are determined, the mask can be rapidly scored for defects, compared with a conventional judgment mode based on machine vision, the method is shorter in evaluation flow, smaller in error, capable of determining that a plurality of defects exist in the mask at one time, and higher in evaluation efficiency.
And finally, generating a final mask defect score, a defect evaluation value QP and a service life ratio which are used as standards for judging whether the mask with the defects is qualified or not, and further completing multi-round screening of the mask and finally completing mask defect detection.
Drawings
FIG. 1 is a schematic flow chart of a visual inspection method for defects of a mask according to the present invention;
FIG. 2 is a schematic view of a mask defect vision inspection system according to the present invention;
FIG. 3 is a schematic view of the screening process of the mask of the present invention;
in the figure: 10. an image acquisition unit; 20. a building unit; 30. a first scoring unit; 40. a second evaluation unit; 50. and a screening unit.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Example 1
Referring to fig. 1-3, the present invention provides a visual inspection method for mask defects, comprising the following steps:
step one, collecting a mask image in multiple angles by adopting an image collecting device, establishing a mask three-dimensional model based on the mask image and the mask size, and marking each area of the mask on the mask three-dimensional model; enabling the user to know intuitively;
the first step comprises the following steps:
step 101, arranging a plurality of rotating mechanisms and corresponding imaging devices on the mask along the advancing direction of a production line to form a plurality of multi-angle images on the mask; dead angles of photographing are reduced, so that all areas are covered;
102, carrying out imaging quality evaluation on the obtained mask images, deleting the parts with imaging quality lower than a threshold value, only leaving the parts with better imaging, avoiding the quality images from influencing the image recognition result, and establishing a mask image set; among them, the method for imaging quality evaluation belongs to the prior art, and no further disclosure is made here.
103, obtaining the sizes of the masks in the current batch, establishing a three-dimensional model of the masks, and rendering the three-dimensional model of the masks based on the current mask color, so that the three-dimensional model of the masks is closer to an actual product and is more authentic;
and 104, identifying the mask image set, comparing the mask image set with a standard mask image, determining each area of the current mask, and marking on the three-dimensional model of the mask.
During the use, combine the content in step 101 to 104, through obtaining the high-quality gauze mask image of multi-angle, can increase the success rate of discernment gauze mask defect, reduce the error, simultaneously, based on accomplish preliminary model building and mark to the gauze mask, also can show more directly perceived when the defect is judged, reduce the degree of difficulty of quality control.
Step two, obtaining mask use evaluation according to a public channel, determining the influence degree of each regional defect of the mask on actual use, evaluating the defect degree, and constructing a mask defect evaluation system; therefore, when the defect degree of the mask is evaluated, the adaptability to the actual use condition is better;
the second step comprises the following steps:
step 201, establishing a deep web crawler, performing deep retrieval on currently-disclosed network information, acquiring corresponding evaluation information of a user when each area of the mask is damaged, and establishing an evaluation information base; the authenticity is higher by taking the evaluation made by the user as a reference;
step 202, identifying evaluation information according to a semantic identification model established based on a natural language processing algorithm, and forming corresponding evaluation by a user when the mask is distinguished to have defects;
classifying a plurality of different evaluations according to the trained SVM classifier, distinguishing different levels, and acquiring a plurality of evaluation level evaluation data sets after corresponding to the mask area;
when the mask evaluation method is used, the evaluation of the mask on the public network in the use process is obtained by combining the contents in the steps 201 and 202, and the good evaluation, the medium evaluation and the bad evaluation in the evaluation are distinguished according to the semantics, so that the difficulty of scoring is reduced.
Step 203, establishing a mask scoring model based on a machine learning algorithm, training and testing, and scoring the evaluations in a plurality of evaluation data sets of different levels;
step 204, constructing a mask defect evaluation system according to the defect degree, the defect area and the defect score, and uploading the mask defect evaluation system to a processing end; when there is the defect in the gauze mask of production line, can form specific score according to the defect degree, above defect all is acquireed through the public channel moreover, and the authenticity is better, also more accords with reality.
When the mask defect evaluation system is used, the mask defect evaluation system is constructed by collecting and distinguishing the public mask use evaluation on the network, so that the mask defect evaluation system has a higher truth degree scoring standard when the mask defect is scored, and the mask defect evaluation system is established according to actual use, so that the difference of different people in mask defect evaluation is eliminated.
Comparing the acquired mask image with a standard mask image, scoring the current mask according to a mask defect evaluation system, and marking the defect score and a corresponding defect area on a three-dimensional model; the defect evaluation is visualized;
the third step comprises the following steps:
step 301, determining a corresponding mask defect standard image according to the defect score, the defect degree and the defect area to form a defect standard image library;
step 302, after the mask passes through the production line, obtaining a mask image with imaging quality evaluation higher than a threshold value, comparing the mask image with images in a standard defect image library, and determining one or more masks with similarity exceeding the corresponding threshold value;
according to the constructed mask defect evaluation system, a mask defect standard image can be determined more directly, the score of the mask defect standard image is determined, and if a mask image with high similarity exists, the score of the mask defect can be determined rapidly;
step 303, picking one or more images with the similarity exceeding a threshold value, and scoring the final mask defect by using the defect score of the current mask image when only one high-similarity image exists; when more than one high-similarity image exists, according to the sum of the defect scores, taking the accumulated score as the final mask defect score; according to the judgment of the similarity, the defect scoring of the mask is completed finally, the difficulty in judging the similarity is low, compared with machine vision identification, the judgment process is less, and the probability of generating errors is lower.
And 304, acquiring the final mask defect score and the mask defect position, and uploading to a processing end.
In use, in conjunction with the contents of steps 301 to 304: after the construction of the mask defect evaluation system is completed, the defect evaluation standard and the defect scoring judgment means based on similarity judgment are determined, so that the mask defect can be scored quickly, and compared with a conventional judgment mode based on machine vision, the method is shorter in evaluation flow and smaller in error.
Fourthly, screening the mask for the first time according to the final mask defect score, marking the mask three-dimensional model according to the mask defect score and the mask defect position, and evaluating the mask defect degree to form a defect evaluation value; thereby making secondary evaluation on the mask which is inconvenient to evaluate by defect scoring directly;
for example, the existing mask has more defects, but each defect is smaller, which does not necessarily affect the practical use, if the mask is abandoned directly because of the defect, the economic loss is more, but the whole public praise of the mask is reduced without any treatment, so that part of the mask needs to be screened.
The fourth step comprises the following steps:
step 401, obtaining a final mask defect score, comparing the final mask defect score with a corresponding threshold value, screening out a part of the final mask defect score larger than the threshold value, and determining the part as an unqualified mask; and after the unqualified mask is removed, the first screening is finished, and the user can reprocess the unqualified mask or use the unqualified mask as a waste product.
Step 402, aiming at the masks with a plurality of defects and final mask defect scores smaller than a threshold value, acquiring mask defect positions and corresponding defect positions, and marking the mask defect positions on a mask three-dimensional model; namely, when the existing mask has defects, the parameters are changed to map the mask defects on the mask three-dimensional model;
step 403, constructing a mask wearing model according to the three-dimensional mask model, carrying out wearing simulation analysis, and determining the influence on mask wearing according to multiple linear regression analysis by changing the defect degree when the mask has defects in different areas to form area influence factors; based on the determined region influence factors, the region influence factors are used as parameters for evaluating the current mask defect score;
as a further improvement, after the standard service life of the mask is determined (disposable use is not considered, the service life of the mask is the defect after the mask is used for a period of time, and the defect score reaches a threshold value), the wearing simulation analysis is carried out on the current mask, the time when the defect score of the mask reaches the threshold value is determined, and the service life of the defect mask is determined;
comparing the service life of the defective mask with that of the standard mask to form a service life ratio, and determining that the defective mask is unqualified if the value of the service life ratio is smaller than a corresponding threshold value; wherein, the standard service life of the mask can refer to the minimum value indicated in the national standard.
Step 404, combining the region influence factors and the defect scores of the regions, and forming a defect evaluation value after synthesis;
among them, explanation is made on the defect evaluation value QP formation by way of example: the mask mainly comprises:
1. the surface body part: the noodle body is divided into an inner layer, a middle layer and an outer layer. The outer layer and the inner layer adopt PP non-woven fabrics, and the middle layer adopts melt-blown fabrics. The outer layer is used for preventing spray and dust, the middle layer is a core functional layer and is used for filtering spray, particles or bacteria, and the inner layer mainly absorbs moisture;
2. nose bridge strip: the nose bridge strip is mainly used for the part of the attaching face and the nose bridge to play roles of sealing and fixing, and is made of three common materials, namely full plastic, plastic-coated metal cores (single core and double core) and aluminum;
3. ear belt: ear straps are used for wearing masks, and are generally flat and round.
The association method of the defect evaluation value QP conforms to the following formula:
wherein the defect score of the noodle body part is M f Regional influence factor A m (ii) a Nose bridge strip Defect score B f The regional influence factor is A b (ii) a Ear band Defect score E f The regional influence factor is A e And C is a constant correction coefficient.
When the mask defect detection method is used, the final mask defect score, the defect evaluation value QP and the service life ratio are generated by combining the contents in the steps 401 to 404, so that whether the defective mask is qualified or not can be judged and read, multiple rounds of screening of the mask are further completed, and mask defect detection is finally completed.
And fifthly, marking the mask position with the defect evaluation value lower than the threshold value, removing the mask position from the production line by using a gripping device, and carrying out secondary screening to finish defect detection.
With reference to the first step to the fifth step, the following effects are at least present in the present application:
by acquiring multi-angle high-quality mask images, the success rate of identifying mask defects based on vision is increased, errors are reduced, and by completing primary modeling and marking on the mask, the defects are displayed more visually when being judged, and the difficulty of quality inspection is reduced;
the method has the advantages that the method completes the construction of a mask defect evaluation system by collecting and distinguishing the use evaluation of the mask disclosed on the network, has an evaluation standard which is suitable for practical use when evaluating the mask defect, and the evaluation mark can also be directly hooked with the market sale, so that the matching degree with the market sale is higher; a mask defect evaluation system is established according to actual use, and differences of different users in mask defect evaluation are eliminated.
Based on the construction of a mask defect evaluation system, a defect evaluation standard and a defect scoring judgment method based on similarity judgment are determined, the mask can be rapidly scored for defects, compared with a conventional judgment mode based on machine vision, the method is shorter in evaluation flow, smaller in error, capable of determining that a plurality of defects exist in the mask at one time, and higher in evaluation efficiency.
And generating a final mask defect score, a defect evaluation value QP and a service life ratio as standards for judging whether the defective mask is qualified or not, thereby completing multi-round screening of the masks and finally completing mask defect detection.
Example 2
Referring to fig. 1-3, the present invention provides a visual inspection system for mask defects, comprising:
the image acquisition unit 10 acquires mask images at multiple angles by adopting an image acquisition device, establishes a three-dimensional mask model, and marks each area of the mask on the three-dimensional mask model;
the construction unit 20 is used for obtaining the mask use evaluation, determining the influence degree of each regional defect of the mask on the actual use, evaluating the defect degree and constructing a mask defect evaluation system;
the first scoring unit 30 compares the acquired mask image with a standard mask image, scores the current mask according to a mask defect evaluation system, and marks the defect score and the corresponding defect area on the three-dimensional model;
the second evaluation unit 40 is used for screening the mask for the first time according to the final mask defect score, marking the mask three-dimensional model according to the mask defect score and the mask defect position, and evaluating the mask defect degree to form a defect evaluation value;
and the screening unit 50 marks the position of the mask with the defect evaluation value lower than the threshold value, removes the mask from the production line by using a grabbing device, and performs secondary screening to complete defect detection.
The above embodiments may be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented in software, the above-described embodiments may be implemented in whole or in part in the form of a computer program product. The computer program product comprises one or more computer instructions or computer programs. The procedures or functions described in accordance with the embodiments of the present application are produced in whole or in part when the computer instructions or the computer program are loaded or executed on a computer. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored on a computer readable storage medium or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. that contains one or more collections of available media. The usable medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. The semiconductor medium may be a solid state disk.
Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
In the several embodiments provided in the present application, it should be understood that the disclosed system, apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the units is only some of the logical functions, and there may be other divisions when actually implemented, for example, a plurality of units or components may be combined or may be integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one position, or may be distributed on multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit.
The functions may be stored in a computer-readable storage medium if they are implemented in the form of software functional units and sold or used as separate products. Based on such understanding, the technical solution of the present application or portions thereof that substantially contribute to the prior art may be embodied in the form of a software product stored in a storage medium and including instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a read-only memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
The above description is only for the specific embodiments of the present application, but the scope of the present application is not limited thereto, and any person skilled in the art can easily think of the changes or substitutions within the technical scope of the present application, and shall be covered by the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
And finally: the above description is only a preferred embodiment of the present invention, and should not be taken as limiting the invention, and any modifications, equivalents, and improvements made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. A visual detection method for defects of a mask is characterized by comprising the following steps: the method comprises the following steps:
step one, collecting a mask image in multiple angles by adopting an image collecting device, establishing a mask three-dimensional model based on the mask image and the mask size, and marking each area of the mask on the mask three-dimensional model;
step two, obtaining mask use evaluation according to a public channel, determining the influence degree of each regional defect of the mask on actual use, evaluating the defect degree, and constructing a mask defect evaluation system;
comparing the acquired mask image with a standard mask image, scoring the current mask according to a mask defect evaluation system, and marking the defect score and a corresponding defect area on the three-dimensional model;
fourthly, screening the mask for the first time according to the final mask defect score, marking the mask three-dimensional model according to the mask defect score and the mask defect position, and evaluating the mask defect degree to form a defect evaluation value;
and fifthly, marking the mask position with the defect evaluation value lower than the threshold value, removing the mask position from the production line by using a gripping device, and carrying out secondary screening to finish defect detection.
2. The visual inspection method of a mask defect of claim 1, wherein: the first step comprises the following steps: step 101, arranging a plurality of rotating mechanisms and corresponding imaging devices on the mask along the advancing direction of a production line to form a plurality of multi-angle images on the mask; and 102, carrying out imaging quality evaluation on the obtained mask images, deleting the parts with the imaging quality lower than a threshold value, and establishing a mask image set.
3. The visual inspection method of a mask defect of claim 2, wherein: step 102 is followed by: 103, obtaining the sizes of the masks in the current batch, establishing a three-dimensional model of the masks, and rendering the three-dimensional model of the masks based on the current mask color; and 104, identifying the mask image set, comparing the mask image set with a standard mask image, determining each area of the current mask, and marking on the three-dimensional model of the mask.
4. The visual inspection method of a mask defect of claim 1, wherein: the second step comprises the following steps: step 201, establishing a deep web crawler, performing deep retrieval on currently-disclosed network information, acquiring corresponding evaluation information of a user when each area of the mask is damaged, and establishing an evaluation information base; step 202, identifying evaluation information according to a semantic identification model established based on a natural language processing algorithm, and forming corresponding evaluation by a user when the mask is distinguished to have defects; classifying a plurality of different evaluations according to the trained SVM classifier, distinguishing different levels, and acquiring a plurality of evaluation level evaluation data sets after corresponding to the mask area.
5. The visual inspection method of a mask defect of claim 4, wherein: step 202 is followed by: step 203, establishing a mask scoring model based on a machine learning algorithm, training and testing, and scoring the evaluations in a plurality of evaluation data sets of different levels; and 204, constructing a mask defect evaluation system according to the defect degree, the defect area and the defect score, and uploading the mask defect evaluation system to a processing end.
6. The visual inspection method of a mask defect of claim 1, wherein: the third step comprises the following steps: step 301, determining a corresponding mask defect standard image according to the defect score, the defect degree and the defect area to form a defect standard image library;
step 302, after the mask passes through a production line, obtaining a mask image with imaging quality evaluation higher than a threshold value, comparing the mask image with images in a standard defect image library, and determining one or more masks with similarity exceeding the corresponding threshold value;
step 303, picking one or more images with the similarity exceeding a threshold value, and scoring the final mask defect by using the defect score of the current mask image when only one high-similarity image exists; when the high-similarity image is more than one, taking the accumulated score as the final mask defect score according to the sum of the plurality of defect scores;
and 304, acquiring the final mask defect score and the mask defect position, and uploading to a processing end.
7. The visual inspection method of a mask defect of claim 1, wherein: the fourth step comprises: step 401, obtaining a final mask defect score, comparing the final mask defect score with a corresponding threshold value, screening out a part of the final mask defect score larger than the threshold value, and determining the part as an unqualified mask;
step 402, aiming at the masks with a plurality of defects and final mask defect scores smaller than a threshold value, acquiring mask defect positions and corresponding defect positions, and marking the mask defect positions on a mask three-dimensional model;
and step 403, constructing a mask wearing model according to the three-dimensional mask model, carrying out wearing simulation analysis, and determining the influence on mask wearing according to multiple linear regression analysis by changing the defect degree when different areas of the mask have defects so as to form area influence factors.
8. The visual inspection method of a mask defect of claim 7, wherein: after the standard service life of the mask is determined, carrying out wearing simulation analysis on the current mask, determining the time when the mask defect score reaches a threshold value, and determining the service life of the mask with the defect;
comparing the service life of the defective mask with that of the standard mask to form a service life ratio, and determining the defective mask as an unqualified mask if the value of the service life ratio is smaller than a corresponding threshold value;
and step 404, combining the area influence factor and the defect score of the area, and forming a defect evaluation value QP after synthesis.
9. The visual inspection method of a mask defect of claim 8, wherein: the association method of the defect evaluation value QP conforms to the following formula:
wherein the defect score of the noodle part is M f Regional influence factor A m (ii) a Nose bridge strip Defect score B f The regional influence factor is A b (ii) a Ear band Defect score E f The regional influence factor is A e And C is a constant correction coefficient.
10. A mask defect visual inspection system, characterized in that: the method comprises the following steps:
the image acquisition unit (10) acquires mask images in multiple angles by adopting an image acquisition device, establishes a three-dimensional mask model, and marks each area of the mask on the three-dimensional mask model;
the method comprises the steps that a construction unit (20) obtains use evaluation of the mask, determines the influence degree of defects of each area of the mask on actual use, evaluates the defect degree, and constructs a mask defect evaluation system;
the first scoring unit (30) compares the acquired mask image with a standard mask image, scores the current mask according to a mask defect evaluation system, and marks the defect score and the corresponding defect area on the three-dimensional model;
a second evaluation unit (40) for screening the mask for the first time according to the final mask defect score, marking the mask three-dimensional model according to the mask defect score and the mask defect position, and evaluating the mask defect degree to form a defect evaluation value; and the screening unit (50) marks the position of the mask with the defect evaluation value lower than the threshold value, removes the mask from the production line by using a grabbing device, and performs secondary screening to finish defect detection.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310132555.0A CN115876804B (en) | 2023-02-20 | 2023-02-20 | Mask defect visual detection method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310132555.0A CN115876804B (en) | 2023-02-20 | 2023-02-20 | Mask defect visual detection method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115876804A true CN115876804A (en) | 2023-03-31 |
| CN115876804B CN115876804B (en) | 2023-08-01 |
Family
ID=85761302
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310132555.0A Active CN115876804B (en) | 2023-02-20 | 2023-02-20 | Mask defect visual detection method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115876804B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116223529A (en) * | 2023-05-09 | 2023-06-06 | 张家港大裕橡胶制品有限公司 | Intelligent detection method and system for production of film-pressed gloves |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107014822A (en) * | 2017-02-28 | 2017-08-04 | 深圳市维图视技术有限公司 | A kind of non-woven fabrics gauze mask defective vision detecting system and method |
| CN109765238A (en) * | 2018-12-12 | 2019-05-17 | 弓立（厦门）医疗用品有限公司 | A kind of product quality detection method of mask fully-automatic production detection device |
| WO2020103324A1 (en) * | 2018-11-20 | 2020-05-28 | 深圳市维图视技术有限公司 | On-line mask detection system and method |
| CN111768404A (en) * | 2020-07-08 | 2020-10-13 | 北京滴普科技有限公司 | Mask appearance defect detection system, method and device and storage medium |
| CN112102813A (en) * | 2020-07-31 | 2020-12-18 | 南京航空航天大学 | Method for generating voice recognition test data based on context in user comment |
| CN112858310A (en) * | 2021-01-20 | 2021-05-28 | 河南体健医疗器械有限公司 | AI intelligent inspection mask system |
| CN113405994A (en) * | 2021-06-24 | 2021-09-17 | 深圳回收宝科技有限公司 | Defect detection method and defect detection system |
| CN114689512A (en) * | 2022-04-15 | 2022-07-01 | 中科芯集成电路有限公司 | Mask detection method integrating machine vision and deep learning |
| CN115266577A (en) * | 2021-04-29 | 2022-11-01 | 广州深路自动化科技有限公司 | Mask defect detection method, system and storage medium |
| CN115601355A (en) * | 2022-11-10 | 2023-01-13 | 四川启睿克科技有限公司（Cn） | Method and device for detecting and classifying product surface defects and storage medium |
-
2023
- 2023-02-20 CN CN202310132555.0A patent/CN115876804B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107014822A (en) * | 2017-02-28 | 2017-08-04 | 深圳市维图视技术有限公司 | A kind of non-woven fabrics gauze mask defective vision detecting system and method |
| WO2020103324A1 (en) * | 2018-11-20 | 2020-05-28 | 深圳市维图视技术有限公司 | On-line mask detection system and method |
| CN109765238A (en) * | 2018-12-12 | 2019-05-17 | 弓立（厦门）医疗用品有限公司 | A kind of product quality detection method of mask fully-automatic production detection device |
| CN111768404A (en) * | 2020-07-08 | 2020-10-13 | 北京滴普科技有限公司 | Mask appearance defect detection system, method and device and storage medium |
| CN112102813A (en) * | 2020-07-31 | 2020-12-18 | 南京航空航天大学 | Method for generating voice recognition test data based on context in user comment |
| CN112858310A (en) * | 2021-01-20 | 2021-05-28 | 河南体健医疗器械有限公司 | AI intelligent inspection mask system |
| CN115266577A (en) * | 2021-04-29 | 2022-11-01 | 广州深路自动化科技有限公司 | Mask defect detection method, system and storage medium |
| CN113405994A (en) * | 2021-06-24 | 2021-09-17 | 深圳回收宝科技有限公司 | Defect detection method and defect detection system |
| CN114689512A (en) * | 2022-04-15 | 2022-07-01 | 中科芯集成电路有限公司 | Mask detection method integrating machine vision and deep learning |
| CN115601355A (en) * | 2022-11-10 | 2023-01-13 | 四川启睿克科技有限公司（Cn） | Method and device for detecting and classifying product surface defects and storage medium |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116223529A (en) * | 2023-05-09 | 2023-06-06 | 张家港大裕橡胶制品有限公司 | Intelligent detection method and system for production of film-pressed gloves |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115876804B (en) | 2023-08-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| TW201732662A (en) | Method and apparatus for establishing data identification model | |
| CN111344703B (en) | User authentication device and method based on iris recognition | |
| CN107644213A (en) | Video person extraction method and device | |
| CN108896278A (en) | A kind of optical filter silk-screen defect inspection method, device and terminal device | |
| CN111507426B (en) | Non-reference image quality grading evaluation method and device based on visual fusion characteristics | |
| CN111914633B (en) | Face-changing video tampering detection method based on face characteristic time domain stability and application thereof | |
| CN115841624B (en) | Blast furnace gas flow distribution identification method based on infrared image | |
| CN115876804A (en) | Visual detection method and system for defects of mask | |
| CN103577838A (en) | Face recognition method and device | |
| CN109145708A (en) | A kind of people flow rate statistical method based on the fusion of RGB and D information | |
| CN106611160A (en) | CNN (Convolutional Neural Network) based image hair identification method and device | |
| CN107785061A (en) | Autism-spectrum disorder with children mood ability interfering system | |
| CN101853397A (en) | Bionic human face detection method based on human visual characteristics | |
| CN113065474A (en) | Behavior recognition method and device and computer equipment | |
| CN112613471B (en) | Face living body detection method, device and computer readable storage medium | |
| CN109583295A (en) | A kind of notch of switch machine automatic testing method based on convolutional neural networks | |
| Samsudin et al. | Evaluation and grading systems of facial paralysis for facial rehabilitation | |
| CN114565602A (en) | Image identification method and device based on multi-channel fusion and storage medium | |
| CN110321944A (en) | A kind of construction method of the deep neural network model based on contact net image quality evaluation | |
| CN110097603B (en) | Fashionable image dominant hue analysis method | |
| CN117152092B (en) | Full-reference image evaluation method, device, electronic equipment and computer storage medium | |
| CN109859199A (en) | A kind of method of the fresh water pipless pearl quality testing of SD-OCT image | |
| CN115841731A (en) | Infrared-monitoring park fire early warning method | |
| CN110287795A (en) | A kind of eye age detection method based on image analysis | |
| JP2005346222A (en) | Sweat gland pore removing device, sweat gland pore removing method and sweat gland pore removing program |
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 |