CN114841982A - Air conditioner filter element quality monitoring method based on image processing - Google Patents

Air conditioner filter element quality monitoring method based on image processing Download PDF

Info

Publication number
CN114841982A
CN114841982A CN202210553566.1A CN202210553566A CN114841982A CN 114841982 A CN114841982 A CN 114841982A CN 202210553566 A CN202210553566 A CN 202210553566A CN 114841982 A CN114841982 A CN 114841982A
Authority
CN
China
Prior art keywords
woven fabric
abnormal
image
target non
activated carbon
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
CN202210553566.1A
Other languages
Chinese (zh)
Other versions
CN114841982B (en
Inventor
王丽丽
陆辉
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.)
Zhejiang Zongchi Environmental Technology Co ltd
Original Assignee
Jiangsu Fengshen Air Conditioning Group 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 Jiangsu Fengshen Air Conditioning Group Co ltd filed Critical Jiangsu Fengshen Air Conditioning Group Co ltd
Priority to CN202210553566.1A priority Critical patent/CN114841982B/en
Publication of CN114841982A publication Critical patent/CN114841982A/en
Application granted granted Critical
Publication of CN114841982B publication Critical patent/CN114841982B/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
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F24—HEATING; RANGES; VENTILATING
    • F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
    • F24F11/00—Control or safety arrangements
    • F24F11/30—Control or safety arrangements for purposes related to the operation of the system, e.g. for safety or monitoring
    • F24F11/32—Responding to malfunctions or emergencies
    • F24F11/39—Monitoring filter performance
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes
    • G06T7/66—Analysis of geometric attributes of image moments or centre of gravity
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/70—Determining position or orientation of objects or cameras
    • G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
    • G06V10/507—Summing image-intensity values; Histogram projection analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/762—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using clustering, e.g. of similar faces in social networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10016—Video; Image sequence
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10024—Color image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; Learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30108—Industrial image inspection
    • 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
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A50/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE in human health protection, e.g. against extreme weather
    • Y02A50/20—Air quality improvement or preservation, e.g. vehicle emission control or emission reduction by using catalytic converters
    • 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
    • Y02B—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO BUILDINGS, e.g. HOUSING, HOUSE APPLIANCES OR RELATED END-USER APPLICATIONS
    • Y02B30/00—Energy efficient heating, ventilation or air conditioning [HVAC]
    • Y02B30/70—Efficient control or regulation technologies, e.g. for control of refrigerant flow, motor or heating

Abstract

The invention relates to an air conditioner filter element quality monitoring method based on image processing, and belongs to the technical field of air conditioner filter element quality monitoring. The method comprises the following steps: obtaining fiber sparsity corresponding to each abnormal region corresponding to each target non-woven fabric surface image according to the gray value of each abnormal pixel point in each abnormal region and the distribution concentration degree of the abnormal pixel points in each abnormal region; obtaining position indexes of the different regions corresponding to the surface images of the target non-woven fabrics according to the positions of the different regions in the corresponding surface images of the target non-woven fabrics; according to the position index and the fiber sparsity, obtaining the influence degree of each abnormal region corresponding to each target non-woven fabric surface image on the quality of the air-conditioning filter element; and monitoring the quality of the air conditioner filter element according to the influence degree of each abnormal area on the quality of the air conditioner filter element. The invention can improve the accuracy of monitoring the quality of the air conditioner filter element.

Description

Air conditioner filter element quality monitoring method based on image processing
Technical Field
The invention relates to the technical field of quality monitoring of air conditioner filter elements, in particular to an air conditioner filter element quality monitoring method based on image processing.
Background
The air-conditioning filter element is made of the following raw materials, namely, activated carbon non-woven fabric which is generally made of two non-woven fabrics which are compounded, wherein tiny granular activated carbon is sandwiched between the two non-woven fabrics, and the activated carbon non-woven fabric is processed and folded to be made into the air-conditioning filter element; the non-woven fabric is produced by adopting polyester fiber and polyester fiber materials and is manufactured by a needling process; the non-woven fabric is a fabric formed without spinning woven fabric, so that the phenomenon of uneven thickness of partial positions is inevitable in the production process; if the non-woven fabrics thickness that uses when making air conditioner filter core is inhomogeneous, can produce very probably to tear in the position of comparison thin, and the activated carbon non-woven fabrics is through folding, and the position of tearing is more difficult to detect out, consequently can make air conditioner filter core's filter effect difficult to guarantee to the product quality who leads to producing is uneven.
The existing quality monitoring method for the air conditioner filter element generally carries out quality monitoring based on a manual mode, the mode for carrying out quality monitoring based on manual work has strong subjectivity and wastes more resources, the limitation of the monitoring mode is strong, only obvious holes or other obvious defects can be detected, and the phenomenon that the quality monitoring of the air conditioner filter element is not accurate enough exists.
Disclosure of Invention
The invention provides an air conditioner filter element quality monitoring method based on image processing, which is used for solving the problem that the quality of an air conditioner filter element cannot be accurately monitored in the prior art, and adopts the following technical scheme:
in a first aspect, an embodiment of the present invention provides an air conditioner filter element quality monitoring method based on image processing, including the following steps:
acquiring continuous multiframe activated carbon non-woven fabric surface gray level images for processing an air conditioner filter element;
obtaining the fiber uniformity degree corresponding to the surface gray level image of each activated carbon non-woven fabric according to the gray level histogram corresponding to the surface gray level image of each activated carbon non-woven fabric; screening the gray level images of the surfaces of the activated carbon non-woven fabrics according to the fiber uniformity degree to obtain the surface images of the target non-woven fabrics corresponding to the air conditioner filter element;
obtaining a Gaussian mixture model corresponding to the surface image of each target non-woven fabric according to the corresponding gray level histogram of the surface image of each target non-woven fabric; obtaining abnormal pixel points corresponding to the surface images of the target non-woven fabrics according to the Gaussian mixture model;
clustering the abnormal pixel points to obtain different abnormal areas corresponding to the surface images of the target non-woven fabrics; obtaining characteristic values corresponding to the principal component directions of the abnormal regions corresponding to the surface images of the target non-woven fabrics according to the coordinates of the pixel points in the abnormal regions;
obtaining the distribution concentration degree of abnormal pixel points in each abnormal area corresponding to the surface image of each target non-woven fabric according to the characteristic value of the main component direction; obtaining fiber sparsity corresponding to each abnormal region corresponding to each target non-woven fabric surface image according to the gray value of each abnormal pixel point in each abnormal region and the distribution concentration degree of the abnormal pixel points in each abnormal region;
obtaining position indexes of the different regions corresponding to the surface images of the target non-woven fabrics according to the positions of the different regions in the corresponding surface images of the target non-woven fabrics;
according to the position index and the fiber sparsity, obtaining the influence degree of each abnormal region corresponding to each target non-woven fabric surface image on the quality of the air-conditioning filter element; and monitoring the quality of the air conditioner filter element according to the influence degree of each abnormal area on the quality of the air conditioner filter element.
Has the advantages that: the characteristic values corresponding to the principal component directions of the abnormal regions corresponding to the surface images of the target non-woven fabrics are used as the basis for obtaining the distribution concentration degree of the abnormal pixel points in the abnormal regions corresponding to the surface images of the target non-woven fabrics, and the gray value of each abnormal pixel point in each abnormal region and the distribution concentration degree of the abnormal pixel points in each abnormal region are used as the basis for obtaining the fiber sparsity degree corresponding to each abnormal region corresponding to the surface images of the target non-woven fabrics; taking the position of each abnormal area in the corresponding target non-woven fabric surface image as a basis for obtaining the position index of each abnormal area corresponding to each target non-woven fabric surface image; taking the position index and the fiber sparsity of each abnormal region corresponding to each target non-woven fabric surface image as a basis for obtaining the influence degree of each abnormal region corresponding to each target non-woven fabric surface image on the quality of the air-conditioning filter element; the influence degree of each abnormal area on the quality of the air-conditioning filter element is used as a basis for monitoring the quality of the air-conditioning filter element; compared with a manual quality monitoring mode, the quality monitoring method and the quality monitoring device can improve the accuracy of quality monitoring of the air conditioner filter element.
Preferably, the fiber uniformity degree corresponding to the gray level image on the surface of each activated carbon non-woven fabric is obtained according to the gray level histogram corresponding to the gray level image on the surface of each activated carbon non-woven fabric; the method for screening the gray level images of the surfaces of the activated carbon non-woven fabrics according to the fiber uniformity degree to obtain the surface images of the target non-woven fabrics corresponding to the air conditioner filter element comprises the following steps:
obtaining a gray value corresponding to the maximum ordinate on the gray histogram corresponding to the gray image on the surface of each activated carbon non-woven fabric according to the gray histogram corresponding to the gray image on the surface of each activated carbon non-woven fabric;
obtaining the probability of each gray value on the gray histogram on the surface gray image of the corresponding activated carbon non-woven fabric according to each gray value on the gray histogram corresponding to the surface gray image of the activated carbon non-woven fabric and the vertical coordinate corresponding to each gray value;
obtaining the fiber uniformity degree corresponding to the surface gray level image of the activated carbon non-woven fabric according to the difference between each gray level value on the gray level histogram corresponding to the surface gray level image of each activated carbon non-woven fabric and the gray level value corresponding to the corresponding maximum ordinate and the probability of each gray level value on the gray level histogram appearing on the corresponding surface gray level image of the activated carbon non-woven fabric;
and judging whether the fiber uniformity is smaller than a preset fiber uniformity threshold value, if so, rejecting the surface gray level image of the activated carbon non-woven fabric corresponding to the fiber uniformity, and recording the surface gray level image of the activated carbon non-woven fabric remaining after rejection as a target non-woven fabric surface image.
Preferably, a Gaussian mixture model corresponding to the surface image of each target non-woven fabric is obtained according to the gray level histogram corresponding to the surface image of each target non-woven fabric; the method for obtaining the abnormal pixel points corresponding to the surface images of the target non-woven fabrics according to the Gaussian mixture model comprises the following steps:
fitting by using an EM (effective electromagnetic) algorithm to obtain a one-dimensional Gaussian mixture model corresponding to each target non-woven fabric surface image;
obtaining expectation and variance of the one-dimensional Gaussian mixture model corresponding to the surface image of each target non-woven fabric according to the one-dimensional Gaussian mixture model;
obtaining a preset segmentation threshold corresponding to the surface image of each target non-woven fabric according to the expectation and the variance;
and judging whether the gray value of each pixel point on the surface image of each target non-woven fabric is greater than a preset segmentation threshold value or not, if so, extracting the corresponding pixel point, and marking the pixel point extracted from the surface image of each target non-woven fabric as each abnormal pixel point corresponding to the surface image of each target non-woven fabric.
Preferably, clustering is carried out on the abnormal pixel points to obtain abnormal areas corresponding to the surface images of the target non-woven fabrics; the method for obtaining the characteristic values corresponding to the principal component directions of the abnormal regions corresponding to the surface images of the target non-woven fabrics according to the coordinates of the pixel points in the abnormal regions comprises the following steps:
carrying out density clustering on each abnormal pixel point by using a mean shift clustering algorithm to obtain a plurality of categories corresponding to the surface images of the target non-woven fabrics; obtaining connected domains corresponding to all classes corresponding to the surface images of the target non-woven fabrics according to the coordinates of all abnormal pixel points in all classes corresponding to the surface images of the target non-woven fabrics; recording the connected domains as different connected domains corresponding to the surface images of the target non-woven fabrics;
and obtaining a first principal component direction and a second principal component direction corresponding to each abnormal connected domain by using a PCA algorithm, and obtaining a characteristic value corresponding to the first principal component direction and a characteristic value corresponding to the second principal component direction.
Preferably, the concentration of the distribution of the abnormal pixel points in each abnormal region corresponding to each target nonwoven fabric surface image is calculated according to the following formula:
Figure BDA0003653996480000041
wherein G is a,b F1 representing the degree of concentration of abnormal pixel points in the b-th abnormal region corresponding to the a-th target nonwoven fabric surface image a,b F2, which is a feature value corresponding to the first principal component direction corresponding to the b-th abnormal region corresponding to the a-th target nonwoven fabric surface image a,b And the characteristic value corresponding to the second principal component direction corresponding to the b-th abnormal area corresponding to the a-th target non-woven fabric surface image.
Preferably, the fiber sparsity corresponding to each abnormal region corresponding to each target nonwoven fabric surface image is calculated according to the following formula:
Figure BDA0003653996480000042
wherein S is a,b The fiber sparsity G corresponding to the b-th abnormal region corresponding to the a-th target non-woven fabric surface image a,b The concentration degree of the abnormal pixel points in the b-th abnormal region corresponding to the surface image of the a-th target nonwoven fabric F a,b The number of abnormal pixel points in the b-th abnormal area corresponding to the surface image of the a-th target non-woven fabric, K a,b,f The gray value corresponding to the f-th abnormal pixel point in the b-th abnormal area corresponding to the a-th target non-woven fabric surface image is obtained.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions and advantages of the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a flow chart of a method for monitoring the quality of an air conditioner filter element based on image processing according to the present invention;
FIG. 2 is a schematic view of an image capture device of the present invention;
FIG. 3 is a schematic representation of a surface image of a target nonwoven fabric of the present invention;
fig. 4 is a schematic view of a folded activated carbon nonwoven fabric of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be described clearly and completely with reference to the accompanying 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, rather than all embodiments, and all other embodiments obtained by those skilled in the art based on the embodiments of the present invention belong to the protection scope of the embodiments of the present invention.
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 embodiment provides an air conditioner filter element quality monitoring method based on image processing, which is described in detail as follows:
as shown in fig. 1, the method for monitoring the quality of the air conditioner filter element based on image processing comprises the following steps:
and S001, acquiring continuous multiframe activated carbon non-woven fabric surface gray level images for processing the air conditioner filter element.
In this embodiment, since the quality of the raw material activated carbon non-woven fabric for manufacturing the air conditioner filter element directly affects the quality of the produced air conditioner filter element, the raw material activated carbon non-woven fabric needs to be analyzed before the air conditioner filter element is processed, and the analysis result is used as a basis for subsequently monitoring the quality of the air conditioner filter element; in this embodiment, a fluorescent screen is placed below the activated carbon non-woven fabric in a direction parallel to the whole activated carbon non-woven fabric, the light of the fluorescent screen is white, and a camera is placed above the fluorescent screen and the activated carbon non-woven fabric, as shown in fig. 2, 1 is the advancing direction of the non-woven fabric; in this embodiment, according to the width of the activated carbon non-woven fabric in fig. 3, the focal length of the camera is adjusted in combination with the principle of camera imaging, so that the visual field range of the camera is the width of the activated carbon non-woven fabric; then determining the sampling frequency of the camera according to the moving speed of the activated carbon non-woven fabric, and recording the time for the camera to start to collect; in the embodiment, the image acquisition equipment based on the arrangement is used for acquiring the image of the activated carbon non-woven fabric to obtain continuous multiframe surface images of the activated carbon non-woven fabric of the air conditioner filter element to be processed; the appearance of the activated carbon non-woven fabric has no obvious processing defects such as hole tearing, and the surface image of the activated carbon non-woven fabric is an image which is not processed and folded by a machine; the surface image of the activated carbon non-woven fabric is an RGB image.
In this embodiment, the collected surface images of the respective activated carbon nonwoven fabrics are subjected to graying processing, and the surface images of the activated carbon nonwoven fabrics subjected to graying processing are referred to as activated carbon nonwoven fabric surface grayscale images.
S002, obtaining the fiber uniformity degree corresponding to the gray level image on the surface of each activated carbon non-woven fabric according to the gray level histogram corresponding to the gray level image on the surface of each activated carbon non-woven fabric; and screening the gray level images of the surfaces of the activated carbon non-woven fabrics according to the fiber uniformity degree to obtain the surface images of the target non-woven fabrics corresponding to the air conditioner filter element.
In this embodiment, since the fiber distribution on the surface of all the activated carbon non-woven fabrics is not uniform, in order to improve the detection efficiency, the degree of uniformity of the fiber corresponding to the gray level image on the surface of each activated carbon non-woven fabric needs to be analyzed and determined, and when the fiber distribution is determined to be non-uniform, the gray level image on the surface of the corresponding activated carbon non-woven fabric needs to be analyzed next. The thickness of the activated carbon non-woven fabric for processing the air conditioner filter element has certain requirements, so that the activated carbon non-woven fabric is not too thin, namely the whole brightness of the gray level image on the surface of the activated carbon non-woven fabric is not generated; therefore, under normal conditions, the gray value distribution on the gray histogram corresponding to the gray image on the surface of the activated carbon non-woven fabric is concentrated; however, when the thickness distribution of the fibers on the surface of the activated carbon nonwoven fabric is not uniform, the gray scale value range of the gray scale histogram of the gray scale image of the surface of the corresponding activated carbon nonwoven fabric is increased, so in this embodiment, the degree of uniformity of the fibers is determined for each surface image of the activated carbon nonwoven fabric according to the distribution of the gray scale values on the gray scale histogram of the surface gray scale image of each activated carbon nonwoven fabric. And screening the surface images of the activated carbon non-woven fabrics according to the judgment result.
In the embodiment, a gray level histogram corresponding to the gray level image on the surface of the activated carbon non-woven fabric is obtained according to the gray level value of each pixel point on the gray level image on the surface of the activated carbon non-woven fabric; obtaining a gray value corresponding to the maximum ordinate on a gray histogram corresponding to the gray image on the surface of the activated carbon non-woven fabric according to the gray histogram corresponding to the gray image on the surface of the activated carbon non-woven fabric; obtaining the probability of each gray value on the gray histogram corresponding to the gray image on the surface of the activated carbon non-woven fabric appearing on the gray image on the surface of the activated carbon non-woven fabric according to each gray value on the gray histogram corresponding to the gray image on the surface of the activated carbon non-woven fabric and the ordinate corresponding to each gray value; obtaining the fiber uniformity degree corresponding to the surface gray level image of each activated carbon non-woven fabric according to the difference between each gray level value on the gray level histogram corresponding to the surface gray level image of the activated carbon non-woven fabric and the gray level value corresponding to the maximum ordinate and the probability of each gray level value on the gray level histogram corresponding to the surface gray level image of each activated carbon non-woven fabric appearing on the corresponding surface gray level image of the activated carbon non-woven fabric; calculating the fiber uniformity degree corresponding to the gray level image on the surface of each activated carbon non-woven fabric according to the following formula:
Figure BDA0003653996480000061
wherein D is i The degree of uniformity, P, of the fibers corresponding to the gray level image on the surface of the ith activated carbon non-woven fabric i,j Is the ith activated carbon non-woven fabric watchProbability that the jth gray value on the gray histogram corresponding to the surface gray image appears on the corresponding gray image on the surface of the activated carbon non-woven fabric, H i,j The gray value is the jth gray value on the gray histogram corresponding to the ith activated carbon non-woven fabric surface gray image, H0 is the gray value corresponding to the maximum ordinate on the gray histogram corresponding to the ith activated carbon non-woven fabric surface gray image, and M is the number of gray values with the frequency not equal to 0 on the gray histogram corresponding to the ith activated carbon non-woven fabric surface gray image; degree D of fiber uniformity corresponding to surface gray level image of ith activated carbon non-woven fabric i The larger the value of (A) is, the lower the probability of the phenomenon of uneven fiber distribution on the gray level image on the surface of the activated carbon non-woven fabric is, and the gray level image on the surface of the activated carbon non-woven fabric does not need to be further detected; degree D of fiber uniformity corresponding to surface gray level image of ith activated carbon non-woven fabric i The smaller the value of (A) is, the higher the probability that the phenomenon of uneven fiber distribution appears on the gray level image of the surface of the activated carbon non-woven fabric is, and the gray level image of the surface of the activated carbon non-woven fabric needs to be further detected.
In the embodiment, the fiber uniformity degree corresponding to the gray level image on the surface of each activated carbon non-woven fabric can be obtained according to the process; and then judging whether the fiber uniformity degree corresponding to the surface gray level image of each activated carbon non-woven fabric is smaller than a preset fiber uniformity degree threshold value, if so, rejecting the surface gray level image of the activated carbon non-woven fabric corresponding to the fiber uniformity degree, and recording the remaining surface gray level image of the activated carbon non-woven fabric after rejection as a target non-woven fabric surface image.
In this embodiment, the threshold value of the degree of uniformity of the preset fibers is set to 0.6; as another embodiment, a preset fiber uniformity threshold may be set according to actual conditions.
S003, obtaining a Gaussian mixture model corresponding to the surface image of each target non-woven fabric according to the gray histogram corresponding to the surface image of each target non-woven fabric; and obtaining different abnormal pixel points corresponding to the surface images of the target non-woven fabrics according to the Gaussian mixture model.
In the embodiment, a Gaussian mixture model corresponding to each target non-woven fabric surface image is obtained by analyzing the gray level histogram corresponding to each target non-woven fabric surface image and fitting; obtaining abnormal pixel points corresponding to the surface images of the target non-woven fabrics according to the Gaussian mixture model; and taking each abnormal pixel point as a basis for subsequently obtaining each abnormal area corresponding to each target non-woven fabric surface image.
In the embodiment, according to each gray value on the gray histogram corresponding to each target non-woven fabric surface image and the occurrence frequency of each gray value, a one-dimensional Gaussian mixture model corresponding to each target non-woven fabric surface image is obtained by using EM (effective electromagnetic tomography) algorithm fitting, and the number of sub-Gaussian models in the one-dimensional Gaussian mixture model is K; the EM algorithm is the prior art, and therefore, the embodiment is not described in detail; in this embodiment, since the activated carbon non-woven fabric used for processing the air conditioner filter element has no obvious appearance defect, two obvious peak values do not appear in the surface image of each target non-woven fabric, and therefore the number of the sub-gaussian models in each one-dimensional gaussian mixture model in this embodiment is set to 1, that is, K is 1; obtaining expectation and variance of a one-dimensional Gaussian mixture model corresponding to the surface image of each target non-woven fabric; since most of the Gaussian distribution corresponding to the surface image of each target nonwoven fabric is (mu) distributed a -3σ a ,μ a +3σ a ) In the range of, wherein a Expectation of Gaussian mixture model corresponding to the a-th target nonwoven fabric surface image, sigma a The variance of a Gaussian mixture model corresponding to the a-th target non-woven fabric table image; because the number of the high-brightness pixel points in the surface image of each target non-woven fabric only accounts for a very small number of all the pixel points in the surface image of the corresponding target non-woven fabric, the mu corresponding to the surface image of each target non-woven fabric a +3σ a Rounding down, i.e.
Figure BDA0003653996480000081
The above-mentioned
Figure BDA0003653996480000082
To get the symbol downwards, the surface image of each target nonwoven fabric is mapped to μ a +3σ a The values after rounding down are recorded asA preset segmentation threshold corresponding to the target non-woven fabric surface image; and judging whether the gray value of each pixel point on the surface image of each target non-woven fabric is greater than a preset segmentation threshold value, if so, extracting the corresponding pixel point, and marking the pixel point extracted from the surface image of each target non-woven fabric as each abnormal pixel point corresponding to the surface image of each target non-woven fabric.
Step S004, clustering the abnormal pixel points to obtain different abnormal areas corresponding to the surface images of the target non-woven fabrics; and obtaining characteristic values corresponding to the principal component directions of the abnormal regions corresponding to the surface images of the target non-woven fabrics according to the coordinates of the pixel points in the abnormal regions.
In this embodiment, the characteristic values corresponding to the principal component directions of the different regions corresponding to the surface images of the target non-woven fabrics are obtained by analyzing the different abnormal pixel points corresponding to the surface images of the target non-woven fabrics; and analyzing and calculating the distribution concentration degree of each abnormal pixel point in each abnormal area corresponding to each target non-woven fabric surface image by using the obtained characteristic value corresponding to the principal component direction.
In the embodiment, the coordinates of each abnormal pixel point corresponding to each target non-woven fabric surface image are obtained, and then the density clustering is performed on each abnormal pixel point corresponding to each target non-woven fabric surface image by using a mean shift clustering algorithm to obtain a plurality of categories corresponding to each target non-woven fabric surface image; the mean shift clustering algorithm is the prior art, so the embodiment is not described in detail; performing convex hull detection on each abnormal pixel point in each category according to the coordinate of each abnormal pixel point in each category corresponding to each target non-woven fabric surface image to obtain a connected domain corresponding to each category corresponding to each target non-woven fabric surface image; and recording the connected domains corresponding to the types corresponding to the surface images of the target non-woven fabrics as the abnormal connected domains corresponding to the surface images of the target non-woven fabrics.
In the embodiment, the principal component directions of the coordinates of the abnormal pixel points in the abnormal connected domains corresponding to the surface images of the target non-woven fabrics are obtained by using a PCA algorithm, and each abnormal connected domain can obtain 2 principal component directions because the coordinates are 2-dimensional data, each principal component direction is a 2-dimensional unit vector, and each principal component direction corresponds to a characteristic value; in this embodiment, the obtained principal component direction with the largest eigenvalue is taken as a first principal component direction, and the obtained principal component direction with the smallest eigenvalue is taken as a second principal component direction; in this embodiment, the first principal component direction and the second principal component direction of each abnormal connected domain corresponding to each target nonwoven fabric surface image can be obtained through the above process, and the feature value corresponding to the first principal component direction and the feature value corresponding to the second principal component direction are obtained.
Step S005, obtaining the distribution concentration degree of abnormal pixel points in each abnormal area corresponding to each target non-woven fabric surface image according to the characteristic value of the principal component direction; and obtaining the fiber sparsity corresponding to each abnormal area corresponding to each target non-woven fabric surface image according to the gray value of each abnormal pixel point in each abnormal area and the distribution concentration degree of the abnormal pixel points in each abnormal area.
In this embodiment, since the activated carbon non-woven fabric is not woven, and the arrangement of the fibers is disordered, under normal conditions, abnormal pixel points are also easily generated in gaps between fibers of the activated carbon non-woven fabric, but when the distribution of the abnormal pixel points in a certain abnormal region is concentrated, it is indicated that the gaps between the fibers in the abnormal region are generally large, that is, the thickness of the activated carbon non-woven fabric corresponding to the abnormal region is relatively thin relative to other regions on the activated carbon non-woven fabric, and the larger the gray value of the abnormal pixel points in the abnormal region is, the thinner the thickness of the activated carbon non-woven fabric corresponding to the abnormal region is; therefore, in this embodiment, the distribution concentration degree of the abnormal pixel points in each abnormal region corresponding to each target nonwoven fabric surface image is obtained by analyzing the characteristic value corresponding to the first principal component direction and the characteristic value corresponding to the second principal component direction of each abnormal connected domain corresponding to each target nonwoven fabric surface image; analyzing the gray value of each abnormal pixel point in each abnormal region corresponding to each target non-woven fabric surface image and the distribution concentration degree of the abnormal pixel points in each corresponding abnormal region to obtain the fiber sparsity corresponding to each abnormal region corresponding to each target non-woven fabric surface image; obtaining fiber sparsity corresponding to each abnormal area corresponding to each target non-woven fabric surface image; and taking the obtained fiber sparsity corresponding to each abnormal area as a basis for subsequent analysis and calculation of the influence degree of each abnormal area on the quality of the air-conditioning filter element.
In this embodiment, when the difference between the characteristic value corresponding to the first principal component direction of a certain abnormal connected domain and the characteristic value corresponding to the second principal component direction is large, it indicates that the distribution of the abnormal pixel points in the abnormal connected domain is long, and at this time, the distribution concentration degree of the abnormal pixel points in the abnormal connected domain needs to be measured according to the characteristic value corresponding to the second principal component direction corresponding to the abnormal connected domain; the smaller the characteristic value corresponding to the second principal component direction corresponding to the abnormal connected domain is, the higher the distribution concentration degree of the abnormal pixel points in the abnormal connected domain is; in this embodiment, the distribution concentration degree of abnormal pixel points in each abnormal region corresponding to each target nonwoven fabric surface image is obtained according to the feature value corresponding to the first principal component direction and the feature value corresponding to the second principal component direction corresponding to each abnormal region corresponding to each target nonwoven fabric surface image; and calculating the distribution concentration degree of the abnormal pixel points in each abnormal area corresponding to the surface image of each target non-woven fabric according to the following formula:
Figure BDA0003653996480000101
wherein G is a,b F1 representing the degree of concentration of abnormal pixel points in the b-th abnormal region corresponding to the a-th target nonwoven fabric surface image a,b F2, which is a feature value corresponding to the first principal component direction corresponding to the b-th abnormal region corresponding to the a-th target nonwoven fabric surface image a,b The characteristic value corresponding to the second principal component direction corresponding to the b-th abnormal area corresponding to the a-th target non-woven fabric surface image; the concentration degree G of the abnormal pixel points in the b-th abnormal area corresponding to the surface image of the a-th target non-woven fabric a,b The larger the value of (a) indicates that the more concentrated the distribution of abnormal pixel points in the b-th abnormal region corresponding to the a-th target nonwoven fabric surface image isThe more concentrated distribution of abnormal pixel points in the b-th abnormal area corresponding to the surface image of the a target non-woven fabric indicates that gaps among fibers in the abnormal area are generally larger, namely, the thickness of the activated carbon non-woven fabric corresponding to the abnormal area is relatively thinner relative to other areas on the activated carbon non-woven fabric.
In the embodiment, fine fluff exists at the fiber edge of the activated carbon non-woven fabric, and when the distance between the fibers on the activated carbon non-woven fabric is relatively short, the fine fluff at the fiber edge has a large shielding effect on gaps between the fibers, so that the gray value of gap pixel points on the activated carbon non-woven fabric is small; when the distance between the fibers on the activated carbon non-woven fabric is long, the shielding effect of fine fluff on the edges of the fibers on gaps between the fibers is small, and white light of a fluorescent screen below the activated carbon non-woven fabric can penetrate through the gaps between the fibers, so that the gray value of the pixel points in the gaps on the activated carbon non-woven fabric is increased; therefore, the fiber sparsity corresponding to each abnormal region corresponding to each target non-woven fabric surface image also needs to be analyzed; in this embodiment, the fiber sparsity corresponding to each abnormal region corresponding to each target non-woven fabric surface image is obtained according to the gray value of each abnormal pixel in each abnormal region corresponding to each target non-woven fabric surface image and the distribution concentration degree of the abnormal pixel in each corresponding abnormal region; calculating the fiber sparsity corresponding to each abnormal area corresponding to each target non-woven fabric surface image according to the following formula:
Figure BDA0003653996480000102
wherein S is a,b The fiber sparsity G corresponding to the b-th abnormal region corresponding to the a-th target non-woven fabric surface image a,b The concentration degree of the abnormal pixel points in the b-th abnormal region corresponding to the surface image of the a-th target nonwoven fabric F a,b The number of abnormal pixel points in the b-th abnormal area corresponding to the surface image of the a-th target non-woven fabric, K a,b,f The b-th difference corresponding to the surface image of the a-th target non-woven fabricThe gray value of the f-th abnormal pixel point in the normal region; s a,b The larger the value of (b) indicates that the thickness of the activated carbon nonwoven fabric corresponding to the abnormal region is thinner.
And S006, obtaining the position indexes of the abnormal areas corresponding to the surface images of the target non-woven fabrics according to the positions of the abnormal areas in the corresponding surface images of the target non-woven fabrics.
In this embodiment, in the folding process of the activated carbon non-woven fabric, the tip pair of the folding machine will exert a force on the outer edge of the activated carbon non-woven fabric, and the folding process is as shown in fig. 4; therefore, when the position of a certain abnormal region is closer to the width edge of the activated carbon non-woven fabric corresponding to the surface image of the target non-woven fabric, the thickness of the activated carbon non-woven fabric corresponding to the abnormal region is thinner, and the tearing is more easily generated when the folding is carried out by using a folding machine; therefore, the influence degree of each abnormal region corresponding to each target non-woven fabric surface image on the quality of the air-conditioning filter element is measured according to the fiber sparsity of the abnormal region and the position index of each abnormal region; in the embodiment, the position indexes of the abnormal regions corresponding to the surface images of the target non-woven fabrics are obtained by analyzing the positions of the abnormal regions in the corresponding surface images of the target non-woven fabrics; and taking the obtained position indexes of the abnormal regions as a basis for subsequently analyzing the influence degree of the abnormal regions corresponding to the surface images of the target non-woven fabrics on the quality of the air-conditioning filter element.
In this embodiment, a coordinate system corresponding to each target nonwoven fabric surface image is constructed, as shown in fig. 3; recording a coordinate system corresponding to the constructed surface image of each target non-woven fabric as a target coordinate system corresponding to the surface image of each target non-woven fabric; according to the resolution of the surface image of each target non-woven fabric, the number of pixel points of the surface image of each target non-woven fabric in the X-axis direction on a corresponding target coordinate system is obtained, and the maximum abscissa value of the surface image of each target non-woven fabric in the X-axis direction is obtained; obtaining a central abscissa value of each target non-woven fabric surface image in the X-axis direction according to the maximum abscissa value of each target non-woven fabric surface image in the X-axis direction; obtaining distribution position indexes of the abnormal regions corresponding to the surface images of the target non-woven fabrics according to the abscissa value of each abnormal pixel point in each abnormal region in the surface image of the target non-woven fabrics in the corresponding target coordinate system and the maximum abscissa value of the surface image of the target non-woven fabrics in the X-axis direction; calculating the position indexes of the abnormal areas corresponding to the surface images of the target non-woven fabrics according to the following formula:
Figure BDA0003653996480000111
wherein, W a,b A position index corresponding to the b-th abnormal region corresponding to the a-th target nonwoven fabric surface image, F a,b The number X of abnormal pixel points in the b-th abnormal area corresponding to the surface image of the a-th target non-woven fabric a,b,f The X is the abscissa value of the f-th abnormal pixel point in the corresponding target coordinate system in the b-th abnormal area corresponding to the surface image of the a-th target non-woven fabric a,max The a is the maximum abscissa value of the surface image of the target non-woven fabric in the X-axis direction in a corresponding target coordinate system; x a,max L-1, wherein L is the number of pixels of the surface image of the ith target non-woven fabric in the X-axis direction on the corresponding target coordinate system;
Figure BDA0003653996480000121
the distance between the f-th abnormal pixel point in the b-th abnormal area corresponding to the a-th target non-woven fabric surface image and the corresponding abscissa value and the corresponding central abscissa value; w a,b The larger the pixel point is, the closer the f-th abnormal pixel point in the b-th abnormal area corresponding to the surface image of the a-th target non-woven fabric is to the width edge of the activated carbon non-woven fabric.
Step S007, obtaining the influence degree of each abnormal region corresponding to each target non-woven fabric surface image on the quality of the air-conditioning filter element according to the position index and the fiber sparsity; and monitoring the quality of the air conditioner filter element according to the influence degree of each abnormal area on the quality of the air conditioner filter element.
In this embodiment, the influence degree of each abnormal region corresponding to each target non-woven fabric surface image on the quality of the air-conditioning filter element is obtained by analyzing the position index and the fiber sparsity corresponding to each abnormal region corresponding to each target non-woven fabric surface image; and then monitoring the quality of the air-conditioning filter element according to the influence degree of each abnormal area corresponding to each target non-woven fabric surface image on the quality of the air-conditioning filter element.
In this embodiment, the position indexes corresponding to the abnormal regions corresponding to the surface images of the target non-woven fabrics are multiplied by the corresponding fiber sparsity to obtain the cracking tendency corresponding to the abnormal regions corresponding to the surface images of the target non-woven fabrics; the higher the fiber sparsity of a certain abnormal region is, the thinner the thickness of the activated carbon non-woven fabric region corresponding to the abnormal region is, the poorer the filtering effect is, the higher the cracking tendency of the produced air conditioner filter element is, the greater the influence on the performance and quality of the air conditioner filter element is, namely, the higher the possibility of tearing in the subsequent processing process is.
In the embodiment, according to the cracking tendency degree and the corresponding fiber sparsity of each abnormal region corresponding to each target non-woven fabric surface image, the influence degree of each abnormal region corresponding to each target non-woven fabric surface image on the quality of the air-conditioning filter element is obtained; and calculating the influence degree of each abnormal area corresponding to each target non-woven fabric surface image on the quality of the air-conditioning filter element according to the following formula:
Figure BDA0003653996480000131
wherein Z is a,b The influence degree of the b-th abnormal area corresponding to the a-th target non-woven fabric surface image on the quality of the air conditioner filter element, S a,b The fiber sparsity, U, corresponding to the b-th abnormal region corresponding to the a-th target non-woven fabric surface image a,b The cracking tendency degree corresponding to the b-th abnormal area corresponding to the a-th target non-woven fabric surface image; z a,b The larger the value of (a) is, the larger the influence of the b-th abnormal area corresponding to the surface image of the a-th target non-woven fabric on the quality of the air conditioner filter element is, and the worse the quality of the air conditioner filter element is.
In this embodiment, whether the influence degree of each abnormal region corresponding to each target non-woven fabric surface image on the quality of the air-conditioning filter element is greater than a preset influence degree threshold is judged, and if yes, a maximum vertical coordinate and a minimum vertical coordinate on a target coordinate system in the corresponding abnormal region are obtained; according to the number of times of shooting by a camera, the relative positions of the maximum ordinate and the minimum ordinate of the abnormal area which is larger than the preset influence degree threshold value in the surface image of each target non-woven fabric on the whole activated carbon non-woven fabric are obtained by combining the moving speed of the activated carbon non-woven fabric, and then the produced filter element number corresponding to the abnormal area which is larger than the preset influence degree threshold value in the surface image of each target non-woven fabric is obtained by combining the produced filter element specifications such as folding number, folding height and the like, and after the production is finished, the filter element corresponding to the number is removed, so that the product quality is ensured.
Has the advantages that: in this embodiment, the feature values corresponding to the principal component directions of the abnormal regions corresponding to the surface images of the target nonwoven fabrics are used as the basis for obtaining the distribution concentration degree of the abnormal pixel points in the abnormal regions corresponding to the surface images of the target nonwoven fabrics, and the gray value of each abnormal pixel point in each abnormal region and the distribution concentration degree of the abnormal pixel points in each abnormal region are used as the basis for obtaining the fiber sparsity degree corresponding to each abnormal region corresponding to the surface images of the target nonwoven fabrics; taking the position of each abnormal area in the corresponding target non-woven fabric surface image as a basis for obtaining the position index of each abnormal area corresponding to each target non-woven fabric surface image; taking the position index and the fiber sparsity of each abnormal region corresponding to each target non-woven fabric surface image as a basis for obtaining the influence degree of each abnormal region corresponding to each target non-woven fabric surface image on the quality of the air-conditioning filter element; the influence degree of each abnormal area on the quality of the air-conditioning filter element is used as a basis for monitoring the quality of the air-conditioning filter element; compared with a manual quality monitoring mode, the quality monitoring accuracy of the air conditioner filter element can be improved.
It should be noted that the order of the above-mentioned embodiments of the present invention is merely for description and does not represent the merits of the embodiments, and in some cases, actions or steps recited in the claims may be executed in an order different from the order of the embodiments and still achieve desirable results.

Claims (6)

1. An air conditioner filter element quality monitoring method based on image processing is characterized by comprising the following steps:
acquiring continuous multiframe activated carbon non-woven fabric surface gray level images for processing an air conditioner filter element;
obtaining the fiber uniformity degree corresponding to the surface gray level image of each activated carbon non-woven fabric according to the gray level histogram corresponding to the surface gray level image of each activated carbon non-woven fabric; screening the gray level images of the surfaces of the activated carbon non-woven fabrics according to the fiber uniformity degree to obtain the surface images of the target non-woven fabrics corresponding to the air conditioner filter element;
obtaining a Gaussian mixture model corresponding to the surface image of each target non-woven fabric according to the gray level histogram corresponding to the surface image of each target non-woven fabric; obtaining abnormal pixel points corresponding to the surface images of the target non-woven fabrics according to the Gaussian mixture model;
clustering the abnormal pixel points to obtain different abnormal areas corresponding to the surface images of the target non-woven fabrics; obtaining characteristic values corresponding to the principal component directions of the abnormal regions corresponding to the surface images of the target non-woven fabrics according to the coordinates of the pixel points in the abnormal regions;
obtaining the distribution concentration degree of abnormal pixel points in each abnormal area corresponding to the surface image of each target non-woven fabric according to the characteristic value of the main component direction; obtaining fiber sparsity corresponding to each abnormal region corresponding to each target non-woven fabric surface image according to the gray value of each abnormal pixel point in each abnormal region and the distribution concentration degree of the abnormal pixel points in each abnormal region;
obtaining position indexes of the different regions corresponding to the surface images of the target non-woven fabrics according to the positions of the different regions in the corresponding surface images of the target non-woven fabrics;
according to the position index and the fiber sparsity, obtaining the influence degree of each abnormal region corresponding to each target non-woven fabric surface image on the quality of the air-conditioning filter element; and monitoring the quality of the air conditioner filter element according to the influence degree of each abnormal area on the quality of the air conditioner filter element.
2. The method for monitoring the quality of the air conditioner filter element based on the image processing as claimed in claim 1, wherein the fiber uniformity degree corresponding to the gray level image on the surface of each activated carbon non-woven fabric is obtained according to the gray level histogram corresponding to the gray level image on the surface of each activated carbon non-woven fabric; the method for screening the gray level images of the surfaces of the activated carbon non-woven fabrics according to the fiber uniformity degree to obtain the surface images of the target non-woven fabrics corresponding to the air conditioner filter element comprises the following steps:
obtaining a gray value corresponding to the maximum ordinate on the gray histogram corresponding to the gray image on the surface of each activated carbon non-woven fabric according to the gray histogram corresponding to the gray image on the surface of each activated carbon non-woven fabric;
obtaining the probability of each gray value on the gray histogram on the surface gray image of the corresponding activated carbon non-woven fabric according to each gray value on the gray histogram corresponding to the surface gray image of the activated carbon non-woven fabric and the vertical coordinate corresponding to each gray value;
obtaining the fiber uniformity degree corresponding to the surface gray level image of the activated carbon non-woven fabric according to the difference between each gray level value on the gray level histogram corresponding to the surface gray level image of each activated carbon non-woven fabric and the gray level value corresponding to the corresponding maximum ordinate and the probability of each gray level value on the gray level histogram appearing on the corresponding surface gray level image of the activated carbon non-woven fabric;
and judging whether the fiber uniformity is smaller than a preset fiber uniformity threshold value, if so, rejecting the surface gray level image of the activated carbon non-woven fabric corresponding to the fiber uniformity, and recording the surface gray level image of the activated carbon non-woven fabric remaining after rejection as a target non-woven fabric surface image.
3. The method for monitoring the quality of the air conditioner filter element based on the image processing as claimed in claim 1, wherein the Gaussian mixture model corresponding to the surface image of each target non-woven fabric is obtained according to the gray level histogram corresponding to the surface image of each target non-woven fabric; the method for obtaining the abnormal pixel points corresponding to the surface images of the target non-woven fabrics according to the Gaussian mixture model comprises the following steps:
fitting by using an EM (effective electromagnetic) algorithm to obtain a one-dimensional Gaussian mixture model corresponding to each target non-woven fabric surface image;
obtaining expectation and variance of the one-dimensional Gaussian mixture model corresponding to the surface image of each target non-woven fabric according to the one-dimensional Gaussian mixture model;
obtaining a preset segmentation threshold corresponding to the surface image of each target non-woven fabric according to the expectation and the variance;
and judging whether the gray value of each pixel point on the surface image of each target non-woven fabric is greater than a preset segmentation threshold value or not, if so, extracting the corresponding pixel point, and marking the pixel point extracted from the surface image of each target non-woven fabric as each abnormal pixel point corresponding to the surface image of each target non-woven fabric.
4. The method for monitoring the quality of the air conditioner filter element based on the image processing as claimed in claim 1, wherein the abnormal pixel points are clustered to obtain abnormal regions corresponding to the surface images of the target non-woven fabrics; the method for obtaining the characteristic values corresponding to the principal component directions of the abnormal regions corresponding to the surface images of the target non-woven fabrics according to the coordinates of the pixel points in the abnormal regions comprises the following steps:
carrying out density clustering on each abnormal pixel point by using a mean shift clustering algorithm to obtain a plurality of categories corresponding to the surface images of the target non-woven fabrics; obtaining connected domains corresponding to all classes corresponding to the surface images of the target non-woven fabrics according to the coordinates of all abnormal pixel points in all classes corresponding to the surface images of the target non-woven fabrics; recording the connected domains as different connected domains corresponding to the surface images of the target non-woven fabrics;
and obtaining a first principal component direction and a second principal component direction corresponding to each abnormal connected domain by utilizing a PCA algorithm, and obtaining a characteristic value corresponding to the first principal component direction and a characteristic value corresponding to the second principal component direction.
5. The image processing-based quality monitoring method for the air conditioner filter element according to claim 4, wherein the distribution concentration degree of the abnormal pixel points in each abnormal area corresponding to each target non-woven fabric surface image is calculated according to the following formula:
Figure FDA0003653996470000031
wherein G is a,b F1 representing the degree of concentration of abnormal pixel points in the b-th abnormal region corresponding to the a-th target nonwoven fabric surface image a,b F2, which is a feature value corresponding to the first principal component direction corresponding to the b-th abnormal region corresponding to the a-th target nonwoven fabric surface image a,b And the characteristic value corresponding to the second principal component direction corresponding to the b-th abnormal area corresponding to the a-th target non-woven fabric surface image.
6. The image processing-based quality monitoring method for the air conditioner filter element according to claim 4, wherein the fiber sparsity corresponding to each abnormal region corresponding to each target non-woven fabric surface image is calculated according to the following formula:
Figure FDA0003653996470000032
wherein S is a,b The fiber sparsity G corresponding to the b-th abnormal region corresponding to the a-th target non-woven fabric surface image a,b The concentration degree of the abnormal pixel points in the b-th abnormal region corresponding to the surface image of the a-th target nonwoven fabric F a,b The number of abnormal pixel points in the b-th abnormal area corresponding to the surface image of the a-th target non-woven fabric, K a,b,f The gray value corresponding to the f-th abnormal pixel point in the b-th abnormal area corresponding to the a-th target non-woven fabric surface image is obtained.
CN202210553566.1A 2022-05-20 2022-05-20 Air conditioner filter element quality monitoring method based on image processing Active CN114841982B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210553566.1A CN114841982B (en) 2022-05-20 2022-05-20 Air conditioner filter element quality monitoring method based on image processing

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210553566.1A CN114841982B (en) 2022-05-20 2022-05-20 Air conditioner filter element quality monitoring method based on image processing

Publications (2)

Publication Number Publication Date
CN114841982A true CN114841982A (en) 2022-08-02
CN114841982B CN114841982B (en) 2024-01-09

Family

ID=82572783

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210553566.1A Active CN114841982B (en) 2022-05-20 2022-05-20 Air conditioner filter element quality monitoring method based on image processing

Country Status (1)

Country Link
CN (1) CN114841982B (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115293304A (en) * 2022-10-10 2022-11-04 普瑞奇科技（北京）股份有限公司 Intelligent monitoring method and system for filter element production
CN116237205A (en) * 2023-01-30 2023-06-09 南通市金月亮新材料科技发展有限公司 Door and window spraying adjustment method and system
CN116493295A (en) * 2023-06-28 2023-07-28 中建材佳星玻璃(黑龙江)有限公司 Glass production early warning method and system
CN116843692A (en) * 2023-09-04 2023-10-03 青岛冠宝林活性炭有限公司 Regenerated active carbon state detection method based on artificial intelligence

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102289677A (en) * 2011-08-02 2011-12-21 东华大学 Method for analyzing image based on principal component analysis and method applicable to detection of defects of fabric
US20130329098A1 (en) * 2012-05-31 2013-12-12 Apple Inc. Systems and Methods for Defective Pixel Correction with Neighboring Pixels
WO2022038575A1 (en) * 2020-08-21 2022-02-24 Abb Schweiz Ag Surface defect detection system

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102289677A (en) * 2011-08-02 2011-12-21 东华大学 Method for analyzing image based on principal component analysis and method applicable to detection of defects of fabric
US20130329098A1 (en) * 2012-05-31 2013-12-12 Apple Inc. Systems and Methods for Defective Pixel Correction with Neighboring Pixels
WO2022038575A1 (en) * 2020-08-21 2022-02-24 Abb Schweiz Ag Surface defect detection system

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
杨学志;左海琴;陈远;吴克伟;谢昭;: "PCA-NLM的纺织品缺陷检测", 中国图象图形学报, no. 12 *
汤勃;孔建益;伍世虔;: "机器视觉表面缺陷检测综述", 中国图象图形学报, no. 12 *

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115293304A (en) * 2022-10-10 2022-11-04 普瑞奇科技（北京）股份有限公司 Intelligent monitoring method and system for filter element production
CN115293304B (en) * 2022-10-10 2022-12-16 普瑞奇科技（北京）股份有限公司 Intelligent monitoring method and system for filter element production
CN116237205A (en) * 2023-01-30 2023-06-09 南通市金月亮新材料科技发展有限公司 Door and window spraying adjustment method and system
CN116237205B (en) * 2023-01-30 2023-11-24 南通市金月亮新材料科技发展有限公司 Door and window spraying adjustment method and system
CN116493295A (en) * 2023-06-28 2023-07-28 中建材佳星玻璃(黑龙江)有限公司 Glass production early warning method and system
CN116493295B (en) * 2023-06-28 2023-09-08 中建材佳星玻璃(黑龙江)有限公司 Glass production early warning method and system
CN116843692A (en) * 2023-09-04 2023-10-03 青岛冠宝林活性炭有限公司 Regenerated active carbon state detection method based on artificial intelligence
CN116843692B (en) * 2023-09-04 2023-12-01 青岛冠宝林活性炭有限公司 Regenerated active carbon state detection method based on artificial intelligence

Also Published As

Publication number Publication date
CN114841982B (en) 2024-01-09

Similar Documents

Publication Publication Date Title
CN114841982A (en) Air conditioner filter element quality monitoring method based on image processing
CN109934802B (en) Cloth defect detection method based on Fourier transform and image morphology
CN114549522B (en) Textile quality detection method based on target detection
DE60115314T2 (en) Method for the evaluation of images for defect detection
EP3580720B1 (en) Real-time, full web image processing method and system for web manufacturing supervision
CN109685766A (en) A kind of Fabric Defect detection method based on region fusion feature
CN115330784B (en) Cloth surface defect detection method
CN113554080A (en) Non-woven fabric defect detection and classification method and system based on machine vision
CN116777941B (en) Profile contour detection method and system based on machine vision
Çelik et al. Fabric defect detection using linear filtering and morphological operations
CN115131348B (en) Method and system for detecting textile surface defects
WO2006113979A1 (en) Method for identifying guignardia citricarpa
CN115330795B (en) Cloth burr defect detection method
CN115760696A (en) Surface scratch detection method
CN113706566A (en) Perfuming spray performance detection method based on edge detection
CN113610843A (en) Real-time defect identification system and method for optical fiber braided layer
Oni et al. Patterned fabric defect detection and classification (fddc) techniques: A review
CN116777917A (en) Defect detection method and system for optical cable production
Ozkaya et al. Faults detection with image processing methods in textile sector
US10388011B2 (en) Real-time, full web image processing method and system for web manufacturing supervision
CN109584224A (en) A method of it is analyzed and is shown for the radioscopic image to casting
EP3459045B1 (en) Real-time, full web image processing method and system for web manufacturing supervision
Geinitz et al. Detection of filament misalignment in carbon fiber production using a stereovision line scan camera system
Rippel et al. Automated Segmentation of Profiled Fibers in cross-sectional Micrographs for Quality Control
Zhao et al. An edge enhancement method of radiographic weld image using bidimensional empirical mode decomposition

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
TA01 Transfer of patent application right

Effective date of registration: 20231220

Address after: Factory Building No. 1, No. 37 Wangyun West Road, Taozhu Street, Zhuji City, Shaoxing City, Zhejiang Province, 311899

Applicant after: Zhejiang Zongchi Environmental Technology Co.,Ltd.

Address before: 226299 No.2, xiaohuaxiao Industrial Zone, west of Qidong City, Nantong City, Jiangsu Province

Applicant before: JIANGSU FENGSHEN AIR CONDITIONING GROUP Co.,Ltd.

TA01 Transfer of patent application right
GR01 Patent grant
GR01 Patent grant