CN114820529A - Method and system for detecting bubbles of rubber product - Google Patents
Method and system for detecting bubbles of rubber product Download PDFInfo
- Publication number
- CN114820529A CN114820529A CN202210462366.5A CN202210462366A CN114820529A CN 114820529 A CN114820529 A CN 114820529A CN 202210462366 A CN202210462366 A CN 202210462366A CN 114820529 A CN114820529 A CN 114820529A
- Authority
- CN
- China
- Prior art keywords
- suspected
- bubble area
- pixel point
- edge pixel
- bubble
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
- G06T7/0008—Industrial image inspection checking presence/absence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- 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
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- 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
- G06T2207/10004—Still image; Photographic image
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Geometry (AREA)
- Quality & Reliability (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The invention relates to the technical field of material testing and analysis, in particular to a method and a system for detecting bubbles of a rubber product. The invention carries out material testing and analysis processing on the visible light surface image, detects whether the rubber product has bubble defects and determines the position of the bubbles, effectively improves the detection speed and the detection accuracy of the bubble detection of the rubber product, and can be used for realizing the detection, the metering and the like of new materials.
Description
Technical Field
The invention relates to the technical field of material testing and analysis, in particular to a method and a system for detecting bubbles of a rubber product.
Background
When a rubber product is vulcanized, due to various reasons that the dispersion of an auxiliary agent in a rubber compound is uneven during mixing, the mixing temperature is low, the water is not completely volatilized, the unevenness of the surface of a die influences the flowability of a rubber material and the like, bubbles are often generated on the surface or in the middle of the rubber product, the bubbles are quality defects which often occur during rubber vulcanization, the appearance quality of the product is influenced, and even the inherent quality of the product is influenced. In order to ensure the quality of the rubber product, the bubble defect produced after the rubber product is vulcanized needs to be detected.
The patent document with publication number CN108414529A in the prior art discloses a method and a system for detecting air bubbles inside a rubber product, which particularly utilizes the principle that the thermal expansion coefficients of the air bubbles and a rubber substrate are different to cause discontinuity and concentration of a strain field on the surface of the rubber product after heating, and utilizes digital images of the rubber product before and after heating, measuring a first main strain field cloud picture on the surface of the heated rubber product by an image correlation method, detecting whether bubbles exist in the rubber product and judging the positions of the bubbles by the characteristic of uneven distribution of first main strain of each pixel point in the first main strain field cloud picture, however, when the rubber product is subjected to heat treatment, the heating may be uneven or the heating temperature may be inconsistent, the strain field distribution result on the surface of the rubber is easy to be inaccurate under the condition, and further, the bubble detection result of the rubber product is inaccurate.
Disclosure of Invention
In order to solve the problem that the existing rubber product bubble detection result is inaccurate, the invention aims to provide a rubber product bubble detection method and system.
The invention provides a method for detecting bubbles of a rubber product, which comprises the following steps:
acquiring a visible light surface image of the rubber product to be detected, and processing the visible light surface image to obtain a visible light edge image of the rubber product to be detected, so as to obtain each suspected bubble area in the visible light edge image;
screening each suspected bubble area in the visible light edge image of the rubber product to be detected, so as to determine each first suspected bubble area in the visible light edge image of the rubber product to be detected;
determining a target circle area corresponding to each first suspected bubble area according to each first suspected bubble area in the visible light edge image of the rubber product to be detected;
determining a morphological index value corresponding to each first suspected bubble area according to each first suspected bubble area and a target circle area corresponding to each first suspected bubble area in a visible light edge image of the rubber product to be detected;
screening each first suspected bubble area according to the morphological index value corresponding to each first suspected bubble area, so as to determine each second suspected bubble area in the visible light edge image of the rubber product to be detected;
determining the confidence coefficient that each second suspected bubble area is a real bubble area according to each second suspected bubble area in the visible light edge image of the rubber product to be detected;
and screening each second suspected bubble area according to the confidence degree that each second suspected bubble area is a real bubble area, so as to determine the bubble area in the visible light surface image of the rubber product to be detected.
Further, the step of determining the target circle area corresponding to each first suspected bubble area includes:
according to each first suspected bubble area in the visible light edge image of the rubber product to be detected, respectively performing vertical scanning movement and horizontal scanning movement in each first suspected bubble area by adopting a horizontal scanning line and a vertical scanning line, and determining the intersection point of each horizontal scanning line and each first suspected bubble area and the intersection point of each vertical scanning line and each first suspected bubble area;
determining the circle center coordinates of the target circle area corresponding to each first suspected bubble area according to the intersection point of each horizontal scanning line and each first suspected bubble area and the intersection point of each vertical scanning line and each first suspected bubble area;
determining each target edge pixel point of each first suspected bubble area according to each edge pixel point of each first suspected bubble area in the visible light edge image of the rubber product to be detected;
determining the radius of a target circle area corresponding to each first suspected bubble area according to the coordinate position of each target edge pixel point of each first suspected bubble area and the center coordinate of the target circle area corresponding to each first suspected bubble area;
and determining the target circle area corresponding to each first suspected bubble area according to the circle center coordinate and the radius of the target circle area corresponding to each first suspected bubble area.
Further, the step of determining each target edge pixel point of each first suspected bubble area includes:
determining an angle index value corresponding to each edge pixel point of each first suspected bubble area according to each edge pixel point of each first suspected bubble area in a visible light edge image of the rubber product to be detected;
determining the confidence degree that each edge pixel point of each first suspected bubble area is a target edge pixel point according to the angle index value corresponding to each edge pixel point of each first suspected bubble area;
and screening each edge pixel point of each first suspected bubble area according to the confidence degree that each edge pixel point of each first suspected bubble area is a target edge pixel point, so as to determine each target edge pixel point of each first suspected bubble area.
Further, the step of determining the angle index value corresponding to each edge pixel point of each first suspected bubble area includes:
according to each edge pixel point of each first suspected bubble area in a visible light edge image of the rubber product to be detected, taking each edge pixel point of each first suspected bubble area as a central pixel point, selecting K edge pixel points from two sides of the central pixel point, and further determining a related pixel point corresponding to each edge pixel point;
determining a fitting line segment corresponding to each edge pixel point of each first suspected bubble area according to the related pixel point corresponding to each edge pixel point of each first suspected bubble area;
and taking the included angle between the fitting line segment corresponding to each edge pixel point of each first suspected bubble area and the positive direction of the X axis as the angle index value corresponding to each edge pixel point of each first suspected bubble area.
Further, the step of determining the confidence that each edge pixel point of each first suspected bubble area is a target edge pixel point includes:
determining neighborhood pixel points corresponding to the edge pixel points of each first suspected bubble area according to the edge pixel points of each first suspected bubble area;
determining a maximum angle index value and a minimum angle index value corresponding to each edge pixel point of each first suspected bubble area according to the angle index value of each edge pixel point of each first suspected bubble area and the angle index value of a neighborhood pixel point corresponding to each edge pixel point;
determining the confidence degree that each edge pixel point of each first suspected bubble area is a target edge pixel point according to the maximum angle index value and the minimum angle index value corresponding to each edge pixel point of each first suspected bubble area, wherein the calculation formula is as follows:
wherein the content of the first and second substances,
the confidence coefficient, gamma, that the jth edge pixel point of the c-th first suspected bubble area is the target edge pixel point c,j,max The maximum angle index value, gamma, corresponding to the jth edge pixel point of the jth first suspected bubble area c,j,min And v is a super parameter, and is a minimum angle index value corresponding to the jth edge pixel point of the jth first suspected bubble area.
Further, a calculation formula for determining a morphological index value corresponding to each first suspected bubble area is as follows:
where ρ is c The morphological index value u corresponding to the c-th first suspected bubble area c The number of edge pixels, U, of the edge pixel point where each edge pixel point of the c-th first suspected bubble area coincides with each edge pixel point of the corresponding target circle area c The number of pixels of each edge pixel point of the c-th first suspected bubble area.
Further, the step of determining the confidence that each second suspected bubble area is a true bubble area includes:
determining a horizontal direction gradient amplitude and a vertical direction gradient amplitude corresponding to each edge pixel point of each second suspected bubble area according to each second suspected bubble area in the visible light edge image of the rubber product to be detected, and further determining a gradient direction corresponding to each edge pixel point of each second suspected bubble area;
determining a direction index corresponding to each edge pixel point of each second suspected bubble area according to the pixel coordinate of each edge pixel point of each second suspected bubble area and the circle center coordinate of the target circle area corresponding to each second suspected bubble area;
determining a direction difference index corresponding to each edge pixel point of each second suspected bubble area according to the gradient direction and the direction index corresponding to each edge pixel point of each second suspected bubble area;
and determining the confidence degree that each second suspected bubble area is a real bubble area according to the direction difference index corresponding to each edge pixel point of each second suspected bubble area.
Further, a calculation formula for determining the direction difference index corresponding to each edge pixel point of each second suspected bubble area is as follows:
wherein d is c,j Is a direction difference index theta corresponding to the jth edge pixel point of the jth second suspected bubble area c j The gradient direction corresponding to the jth edge pixel point of the jth second suspected bubble area,
and the direction index corresponding to the jth edge pixel point of the jth second suspected bubble area.
Further, the step of determining the confidence that each second suspected bubble area is a true bubble area includes:
judging whether the direction difference index corresponding to each edge pixel point of each second suspected bubble area is lower than a preset threshold value or not according to the direction difference index corresponding to each edge pixel point of each second suspected bubble area, and further counting the number of pixels of edge pixel points of which the direction difference index corresponding to each second suspected bubble area is lower than the preset threshold value;
and determining the confidence coefficient of each second suspected bubble area as a real bubble area according to the pixel number of each edge pixel point of each second suspected bubble area and the pixel number of the edge pixel point of which the corresponding direction difference index is lower than a preset threshold value.
The invention also provides a rubber product bubble detection system which comprises a processor and a memory, wherein the processor is used for processing the instructions stored in the memory so as to realize a rubber product bubble detection method.
The invention has the following beneficial effects:
the invention provides a method and a system for detecting bubbles of a rubber product, which can be used for realizing detection, metering, relevant standardization and the like of a new material, and particularly obtain a visible light edge image of the rubber product to be detected by a visible light means, perform material test and analysis on the visible light edge image, particularly determine each suspected bubble area in the visible light edge image according to the visible light edge image, and screen each suspected bubble area, thereby determining each first suspected bubble area in the visible light edge image of the rubber product to be detected. And screening the first suspected bubble areas through the obtained morphological index values corresponding to the first suspected bubble areas, so as to determine the second suspected bubble areas. And screening each second suspected bubble area according to the confidence coefficient that each obtained second suspected bubble area is a real bubble area, so as to determine the bubble area in the visible light surface image of the rubber product to be detected.
The invention mainly tests and analyzes the bubbles in the surface image of the rubber product by the material testing and analyzing technology, can accurately determine the bubble area on the surface of the rubber product, solves the problem of inaccurate bubble detection of the rubber product, improves the accuracy of the bubble detection of the rubber product, and can be used for realizing the detection, the metering and the like of new materials.
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 embodiments or the description of 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 detecting bubbles in a rubber product according to an embodiment of the present invention;
fig. 2 is a flowchart illustrating a process of determining a target circle region corresponding to each first suspected-bubble region in an embodiment of the present invention.
Detailed Description
To further explain the technical means and effects of the present invention adopted to achieve the predetermined objects, the following detailed description of the embodiments, structures, features and effects of the technical solutions according to the present invention will be given with reference to the accompanying drawings and preferred embodiments. In the following description, different references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
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 a method for detecting bubbles in a rubber product, which comprises the following steps as shown in FIG. 1:
(1) and acquiring a visible light surface image of the rubber product to be detected, processing the visible light surface image to obtain a visible light edge image of the rubber product to be detected, and further obtaining each suspected bubble area in the visible light edge image.
(1-1) acquiring a visible light surface image of the rubber product to be detected.
In the embodiment, the visible light surface image of the rubber product to be detected can be obtained through the image acquisition device, the image acquisition device comprises a light source, a lens, an industrial camera, a detection table, rubber product conveying equipment and the like, and an implementer can deploy and install the image acquisition device according to the actual situation of the implementer. In this embodiment, the rubber product conveying device conveys the rubber product to be detected to the detection table, and it should be noted that after the rubber product conveying device finishes conveying, the position information of the rubber product on the detection table is kept fixed. The industrial camera is arranged right above the detection table, the shooting range of the industrial camera can cover the whole outline of the rubber product to be detected, and in addition, the industrial camera can avoid the influence of external factors on the image of the rubber product by shooting the rubber product to be detected through a overlooking visual angle. The light sources in the image acquisition equipment are distributed around the detection table, and the brightness intensity of the detection table can be ensured.
And (1-2) preprocessing the visible light surface image according to the visible light surface image of the rubber product to be detected to obtain the preprocessed visible light surface image of the rubber product to be detected.
And (3) carrying out denoising treatment and illumination equalization treatment on the visible light surface image according to the visible light surface image of the rubber product to be detected obtained in the step (1-1) so as to improve the image quality of the visible light surface image and reduce the influence of noise and illumination nonuniformity in the visible light surface image, thereby obtaining the visible light surface image of the rubber product to be detected after denoising treatment and illumination equalization treatment. The method for denoising and illumination equalization of the image has many ways, an implementer can select the method, the method mainly adopts Gaussian filtering and median filtering to denoise the visible light surface image, and meanwhile, the method carries out equalization on the visible light surface image through gamma conversion. The process of processing the visible light surface image by gaussian filtering, median filtering and gamma transformation is prior art and is not within the scope of the present invention, and will not be elaborated herein.
And (1-3) obtaining a visible light edge image of the rubber product to be detected according to the pretreated visible light surface image of the rubber product to be detected, and further obtaining each suspected bubble area in the visible light edge image.
And according to the pretreated visible light surface image of the rubber product to be detected, carrying out edge detection on the visible light surface image by using an edge detection operator to obtain a visible light edge image of the rubber product to be detected. Since rubber articles often suffer from quality defects during vulcanization, which may be bubble defects or other defects, there are regions of suspected bubbles in the visible edge image of the rubber article to be inspected. And obtaining each suspected bubble area in the visible light edge image according to the visible light edge image of the rubber product to be detected. The process of edge detection of an image by a visible light edge image edge detection operator is the prior art, is out of the protection scope of the invention, and is not elaborated herein.
(2) And screening each suspected bubble area in the visible light edge image of the rubber product to be detected, so as to determine each first suspected bubble area in the visible light edge image of the rubber product to be detected.
In this embodiment, each suspected bubble area in the visible light edge image of the rubber product to be detected is preliminarily detected through hough circle detection, each suspected bubble area after preliminary detection corresponds to a plurality of candidate circles, each suspected bubble area has a plurality of candidate circles corresponding to the suspected bubble area, the number of the candidate circles corresponding to each suspected bubble area is counted, the number of the candidate circles is marked as N, for example, the number of the candidate circles corresponding to the ith suspected bubble area in the visible light edge image of the rubber product to be detected is marked as N i Thereby obtaining a number sequence consisting of the number of candidate circles corresponding to each suspected bubble area, the number sequence N being { N ═ N 1 ,N 2 ,…,N n And n is the number of each suspected bubble area in the visible light edge image of the rubber product to be detected.
Through the candidate circle number sequence corresponding to each suspected bubble area, the rubber to be detectedAnd screening each suspected bubble area in the visible light edge image of the rubber product, and taking the suspected bubble area with the candidate circle number smaller than a preset candidate circle number threshold value as a first suspected bubble area, wherein the first suspected bubble area refers to each screened suspected bubble area, so as to obtain each first suspected bubble area in the visible light edge image of the rubber product to be detected. The embodiment marks the preset candidate circle number threshold as N T And N is T ＝15。
It should be noted that, according to priori knowledge, most of the shapes of the bubbles of the rubber product are approximate circles or circles, when hough circle detection is performed on an approximate circle region, the number of candidate circles corresponding to the approximate circle region is small, and the smaller the number of candidate circles corresponding to a certain suspected bubble region is, the higher the possibility that the suspected bubble region is a bubble of the rubber product is, in this embodiment, each suspected bubble region in a visible light edge image of the rubber product to be detected is screened based on the above contents, and the screened suspected bubble regions are further analyzed, so that each first suspected bubble region after screening is subsequently detected, and therefore, the accuracy of the bubble detection of the rubber product is improved.
(3) A flow chart of determining a target circle area corresponding to each first suspected bubble area according to each first suspected bubble area in a visible light edge image of a rubber product to be detected is shown in fig. 2, and the determining step includes:
and (3-1) according to each first suspected bubble area in the visible light edge image of the rubber product to be detected, respectively performing vertical scanning movement and horizontal scanning movement in each first suspected bubble area by adopting a horizontal scanning line and a vertical scanning line, and determining the intersection point of each horizontal scanning line and each first suspected bubble area and the intersection point of each vertical scanning line and each first suspected bubble area.
In this embodiment, the horizontal scanning lines are moved in a vertical scanning manner in each first suspected bubble area, and the vertical scanning lines are moved in a horizontal scanning manner in each first suspected bubble area, and the scanning lines intersect with the edge of each first suspected bubble area in the moving process, so that an intersection point of each horizontal scanning line and each first suspected bubble area and an intersection point of each vertical scanning line and each first suspected bubble area are obtained.
And (3-2) determining the circle center coordinates of the target circle area corresponding to each first suspected bubble area according to the intersection point of each horizontal scanning line and each first suspected bubble area and the intersection point of each vertical scanning line and each first suspected bubble area.
First, since it is known from step (2) that each first bubble pseudo area is approximately circular, in the present embodiment, only the case where the scanning line intersects the first bubble pseudo area by 2 intersections is considered, and the case where there are single or multiple intersections will be ignored.
In this embodiment, according to the coordinate positions of 2 intersection points of each horizontal scanning line and each first suspected bubble area, the central point abscissa between the 2 intersection points may be determined, and the central point abscissas corresponding to each horizontal scanning line are sorted, so as to obtain the central point abscissa set corresponding to each first suspected bubble area. According to the center point abscissa set corresponding to each first suspected bubble area, based on the symmetry principle, the center point abscissa having the highest frequency in the center point abscissa set corresponding to each first suspected bubble area is taken as the center of circle abscissa of the corresponding target circle area and is marked as x.
Similarly, the center of circle ordinate of the target circle area corresponding to each first suspected bubble area can be obtained and recorded as y by referring to the determination process of the center abscissa of the target circle area corresponding to each first suspected bubble area and according to the intersection point of each vertical scanning straight line and each first suspected bubble area. And obtaining the center coordinates (x, y) of the target circle area corresponding to each first suspected bubble area according to the center abscissa x and the center ordinate y of the target circle area corresponding to each first suspected bubble area.
And (3-3) determining each target edge pixel point of each first suspected bubble area according to each edge pixel point of each first suspected bubble area in the visible light edge image of the rubber product to be detected.
It should be noted that, in order to improve the accuracy of the radius of the target circle region corresponding to the first suspected bubble region, ensure the accuracy of the morphological index value corresponding to the subsequently obtained first suspected bubble region, and avoid the problem of reduction of detection precision caused by artificial random selection of subjectivity, the present embodiment needs to determine each target edge pixel point of each first suspected bubble region, and the steps include:
(3-3-1) according to each edge pixel point of each first suspected bubble area in the visible light edge image of the rubber product to be detected, determining an angle index value corresponding to each edge pixel point of each first suspected bubble area, wherein the step comprises the following steps:
(3-3-1-1) according to each edge pixel point of each first suspected bubble area in the visible light edge image of the rubber product to be detected, taking each edge pixel point of each first suspected bubble area as a central pixel point, selecting K edge pixel points from two sides of the central pixel point, and further determining related pixel points corresponding to each edge pixel point.
In this embodiment, for each edge pixel point of each first suspected bubble region, each edge pixel point is taken as a center edge pixel point, and K edge pixel points are respectively selected from two sides of the center edge pixel point, so as to obtain a related pixel point corresponding to each edge pixel point, where each edge pixel point has a related pixel point corresponding thereto, and the related pixel point refers to 2K +1 edge pixel points including the center edge pixel point itself.
And (3-3-1-2) determining a fitting line segment corresponding to each edge pixel point of each first suspected bubble area according to the related pixel point corresponding to each edge pixel point of each first suspected bubble area.
In this embodiment, 2K +1 relevant pixel points corresponding to each edge pixel point of each first suspected bubble region obtained in the step (3-3-1-1) are fit into a line segment, so as to determine a fit line segment corresponding to each edge pixel point of each first suspected bubble region, where each edge pixel point has a corresponding fit line segment.
(3-3-1-3) taking the included angle between the fitting line segment corresponding to each edge pixel point of each first suspected bubble area and the positive direction of the X axis as the angle index value corresponding to each edge pixel point of each first suspected bubble area.
And (3) according to the fitting line segment corresponding to each edge pixel point of each first suspected bubble area obtained in the step (3-3-1-2), obtaining an included angle between the fitting line segment corresponding to each edge pixel point and the positive direction of the X axis, and taking the included angle as an angle index value corresponding to the corresponding edge pixel point, so as to obtain the angle index value corresponding to each edge pixel point of each first suspected bubble area, and marking the angle index value as gamma.
(3-3-2) according to the angle index value corresponding to each edge pixel point of each first suspected bubble area, determining the confidence degree that each edge pixel point of each first suspected bubble area is a target edge pixel point, wherein the step comprises the following steps:
and (3-3-2-1) determining neighborhood pixel points corresponding to the edge pixel points of each first suspected bubble area according to the edge pixel points of each first suspected bubble area.
In this embodiment, M edge pixels are selected from both sides of each edge pixel, respectively, so as to obtain 2M neighborhood pixels, and the M edge pixels on both sides of the edge pixel are used as the neighborhood pixels corresponding to each edge pixel, and each edge pixel has 2M neighborhood pixels corresponding to the edge pixel.
(3-3-2-2) determining the maximum angle index value and the minimum angle index value corresponding to each edge pixel point of each first suspected bubble area according to the angle index value of each edge pixel point of each first suspected bubble area and the angle index value of the neighborhood pixel point corresponding to each edge pixel point.
Determining angle index values corresponding to the 2M neighborhood pixel points according to the 2M neighborhood pixel points corresponding to the edge pixel points of each first suspected bubble area obtained in the step (3-3-2-1), and determining angle index values in the 2M +1 angle index values according to the angle index value of each edge pixel point of each first suspected bubble area and the angle index value of each 2M neighborhood pixel point corresponding to each edge pixel pointThe maximum value and the minimum value are obtained, thereby obtaining the maximum angle index value and the minimum angle index value corresponding to each edge pixel point, and the maximum angle index value is recorded as gamma max The minimum angle index value is recorded as gamma min 。
(3-3-2-3) according to the maximum angle index value and the minimum angle index value corresponding to each edge pixel point of each first suspected bubble area, determining the confidence degree that each edge pixel point of each first suspected bubble area is a target edge pixel point.
In this embodiment, for example, the confidence that the jth edge pixel point of the c-th first suspected bubble area is determined as the target edge pixel point is determined, according to the maximum angle index value and the minimum angle index value corresponding to the jth edge pixel point of the c-th first suspected bubble area, the confidence that the jth edge pixel point of the c-th first suspected bubble area is the target edge pixel point is calculated, the target edge pixel point is a plurality of edge pixel points screened from each edge pixel point, the confidence that the edge pixel point is determined as the target edge pixel point is helpful for subsequently calculating the radius of the target circle area, and the confidence calculation formula is as follows:
wherein the content of the first and second substances,
the confidence coefficient, gamma, that the jth edge pixel point of the c-th first suspected bubble area is the target edge pixel point c,j,max The maximum angle index value, gamma, corresponding to the jth edge pixel point of the jth first suspected bubble area c,j,min And v is a super parameter, and is a minimum angle index value corresponding to the jth edge pixel point of the jth first suspected bubble area.
And referring to the determination process of the confidence degree that the jth edge pixel point of the c-th first suspected bubble area is the target edge pixel point, and obtaining the confidence degree that each edge pixel point of each first suspected bubble area is the target edge pixel point.
And (3-3-3) screening each edge pixel point of each first suspected bubble area according to the confidence degree that each edge pixel point of each first suspected bubble area is a target edge pixel point, so as to determine each target edge pixel point of each first suspected bubble area.
In this embodiment, the confidence levels of the edge pixel points of each first suspected bubble area as the target edge pixel points are arranged in descending order to obtain a confidence level sequence corresponding to each first suspected bubble area, and the confidence level sequence is marked as
From confidence series
And selecting edge pixel points corresponding to the previous Q confidences, and taking the edge pixel points corresponding to the previous Q confidences as target edge pixel points of the corresponding first suspected bubble areas, so as to obtain the target edge pixel points of the first suspected bubble areas.
And (3-4) determining the radius of the target circle area corresponding to each first suspected bubble area according to the coordinate position of each target edge pixel point of each first suspected bubble area and the center coordinate of the target circle area corresponding to each first suspected bubble area.
In this embodiment, the coordinate positions of Q target edge pixel points of each first suspected bubble region are determined through the Q target edge pixel points of each first suspected bubble region obtained in step (3-3-3), and a formula (x) is calculated based on the existing circular equation according to the coordinate positions of the Q target edge pixel points of each first suspected bubble region and the center coordinates of the target circle region corresponding to each first suspected bubble region obtained in step (3-2) q -x) 2 +(y q -y) 2 ＝r q 2 Wherein (x) q ,y q ) Is the coordinate position of the q-th target edge pixel point in each first suspected bubble area, (x, y) is the center coordinate of the target circle area corresponding to each first suspected bubble area, r q For the q-th target circle radius index corresponding to each first suspected bubble area, the target circle radius index corresponding to each first suspected bubble area can be obtained and is marked as r q Q is 1,2, …, Q. In this embodiment, the mean value of the target circle radius indexes corresponding to each first suspected bubble area is used as the radius of the target circle area, so as to obtain the radius of the target circle area corresponding to each first suspected bubble area.
And (3-5) determining the target circle area corresponding to each first suspected bubble area according to the circle center coordinate and the radius of the target circle area corresponding to each first suspected bubble area.
In this embodiment, the target circle area corresponding to each first suspected bubble area can be obtained according to the center coordinate and the radius of the target circle area corresponding to each first suspected bubble area, and the process of determining the target circle area through the center coordinate and the radius is the prior art and is not within the protection scope of the present invention, and is not described in detail herein.
(4) And determining the morphological index value corresponding to each first suspected bubble area according to each first suspected bubble area in the visible light edge image of the rubber product to be detected and the target circle area corresponding to each first suspected bubble area.
In this embodiment, taking calculation of a morphological index value corresponding to the c-th first suspected bubble area as an example, according to the c-th first suspected bubble area in the visible light edge image of the rubber product to be detected and the target circle area corresponding to the c-th first suspected bubble area, the pixel number of each edge pixel point of the c-th first suspected bubble area, and the pixel number of each edge pixel point of the c-th first suspected bubble area, which is overlapped with each edge pixel point of the target circle area corresponding thereto, are counted. According to the pixel number of each edge pixel point of the c-th first suspected bubble area and the pixel number of the edge pixel point of the c-th first suspected bubble area, which is coincided with each edge pixel point of the corresponding target circle area, calculating a morphological index value corresponding to the c-th first suspected bubble area, wherein the calculation formula is as follows:
where ρ is c The morphological index value u corresponding to the c-th first suspected bubble area c The number of edge pixels, U, of the edge pixel point where each edge pixel point of the c-th first suspected bubble area coincides with each edge pixel point of the corresponding target circle area c The number of pixels of each edge pixel point of the c-th first suspected bubble area.
In addition, the morphology index value corresponding to each first suspected-bubble area can be obtained by referring to the calculation process of the morphology index value corresponding to the c-th first suspected-bubble area. U in the above formula for calculating morphometric index value c The larger the size, the morphological index value ρ corresponding to the c-th first suspected bubble area c The larger the size, the more likely the c-th first pseudo-bubble area is to be a rubber product bubble area.
(5) And screening each first suspected bubble area according to the morphological index value corresponding to each first suspected bubble area, so as to determine each second suspected bubble area in the visible light edge image of the rubber product to be detected.
In this embodiment, the morphology index value corresponding to each first bubble pseudo area is normalized. According to the morphological index value corresponding to each first suspected bubble area after normalization processing, comparing the morphological index value corresponding to each first suspected bubble area after normalization processing with a preset morphological index threshold value, selecting each first suspected bubble area larger than the preset morphological index threshold value, and taking the first suspected bubble area larger than the preset morphological index threshold value as a second suspected bubble area, where the preset morphological index threshold value is set to be 0.6 in this embodiment.
(6) And determining the confidence degree that each second suspected bubble area is a real bubble area according to each second suspected bubble area in the visible light edge image of the rubber product to be detected.
It should be noted that, because the contrast ratio between the rubber product bubble area and the rubber product background area is relatively low, in order to accurately detect the rubber product bubble area and accurately extract the bubble area, in this embodiment, the confidence degree that each second suspected bubble area is the real bubble area is calculated, so as to screen each second suspected bubble area again, and the step of determining the confidence degree that each second suspected bubble area is the real bubble area includes:
(6-1) according to each second suspected bubble area in the visible light edge image of the rubber product to be detected, determining a horizontal direction gradient amplitude and a vertical direction gradient amplitude corresponding to each edge pixel point of each second suspected bubble area, and further determining a gradient direction corresponding to each edge pixel point of each second suspected bubble area.
In this embodiment, a graying process is performed on a visible light edge image of the rubber product to be detected to obtain a grayed visible light edge image of the rubber product to be detected, and according to each second suspected bubble region in the grayed visible light edge image of the rubber product to be detected, a horizontal direction gradient amplitude and a vertical direction gradient amplitude corresponding to each edge pixel point of each second suspected bubble region can be obtained, where the horizontal direction gradient amplitude is recorded as T x And the gradient amplitude in the vertical direction is recorded as T y . The gradient amplitude T in the horizontal direction corresponding to each edge pixel point of each second suspected bubble area x And vertical gradient amplitude T y And calculating the gradient direction corresponding to each edge pixel point of each second suspected bubble area.
In this embodiment, for example, the gradient direction corresponding to the jth edge pixel point of the c-th second suspected bubble area is calculated, according to the vertical gradient amplitude corresponding to the jth edge pixel point of the c-th second suspected bubble area and the horizontal gradient amplitude corresponding to the jth edge pixel point of the c-th second suspected bubble area, the gradient direction corresponding to the jth edge pixel point of the c-th second suspected bubble area is calculated, and the calculation formula is as follows:
wherein, theta c j A gradient direction T corresponding to the jth edge pixel point of the jth second suspected bubble area c,y j A vertical gradient amplitude value T corresponding to the jth edge pixel point of the jth second suspected bubble area c,x j For the horizontal direction gradient amplitude corresponding to the jth edge pixel point of the mth second suspected bubble area, arctan () is an arctangent function.
And obtaining the gradient direction corresponding to each edge pixel point of each second suspected bubble area by referring to the calculation process of the gradient direction corresponding to the jth edge pixel point of the jth second suspected bubble area.
And (6-2) determining a direction index corresponding to each edge pixel point of each second suspected bubble area according to the pixel coordinate of each edge pixel point of each second suspected bubble area and the center coordinate of the target circle area corresponding to each second suspected bubble area.
In this embodiment, for example, the direction index corresponding to the jth edge pixel point of the c-th second suspected bubble area is determined, and according to the pixel coordinate of the jth edge pixel point of the c-th second suspected bubble area and the center coordinate of the target circle area corresponding to the c-th second suspected bubble area, the direction index corresponding to the jth edge pixel point of the c-th second suspected bubble area is calculated, where the calculation formula is as follows:
wherein the content of the first and second substances,
a direction index y corresponding to the jth edge pixel point of the jth second suspected bubble area c Is the center ordinate, y, of the target circle area corresponding to the c-th second suspected bubble area c,j Is the pixel ordinate, x, of the jth edge pixel point of the c second suspected bubble area c Is the c second suspected bubble areaCenter abscissa, x, of target circle region corresponding to field c,j And the arctan () is an arctangent function which is the pixel abscissa of the jth edge pixel point of the c-th second suspected bubble area.
And obtaining the direction indexes corresponding to the edge pixel points of the second suspected bubble areas by referring to the determination process of the direction indexes corresponding to the jth edge pixel point of the jth second suspected bubble area.
And (6-3) determining a direction difference index corresponding to each edge pixel point of each second suspected bubble area according to the gradient direction and the direction index corresponding to each edge pixel point of each second suspected bubble area.
In this embodiment, taking the determination of the direction difference index corresponding to the jth edge pixel point of the c-th second suspected bubble area as an example, according to the gradient direction corresponding to the jth edge pixel point of the c-th second suspected bubble area obtained in step (6-1) and the direction index corresponding to the jth edge pixel point of the c-th second suspected bubble area obtained in step (6-2), the direction difference index corresponding to the jth edge pixel point of the c-th second suspected bubble area may be calculated, and a calculation formula thereof is as follows:
wherein d is c,j Is a direction difference index theta corresponding to the jth edge pixel point of the jth second suspected bubble area c j The gradient direction corresponding to the jth edge pixel point of the jth second suspected bubble area,
and the direction index corresponding to the jth edge pixel point of the jth second suspected bubble area.
And obtaining the direction difference index corresponding to each edge pixel point of each second suspected bubble area by referring to the determination process of the direction difference index corresponding to the jth edge pixel point of the jth second suspected bubble area, wherein each edge pixel point has the corresponding direction difference index.
(6-4) according to the direction difference index corresponding to each edge pixel point of each second suspected bubble area, determining the confidence degree that each second suspected bubble area is a real bubble area, wherein the step comprises the following steps:
(6-4-1) judging whether the direction difference index corresponding to each edge pixel point of each second suspected bubble area is lower than a preset threshold value according to the direction difference index corresponding to each edge pixel point of each second suspected bubble area, and further counting the number of pixels of the edge pixel points of which the direction difference index corresponding to each second suspected bubble area is lower than the preset threshold value.
In this embodiment, according to the direction difference index corresponding to each edge pixel point of each second suspected bubble area, it is first determined whether the direction difference index corresponding to each edge pixel point of each second suspected bubble area is lower than a preset threshold, and then the number of pixels of edge pixel points whose direction difference index is lower than the preset threshold is counted, and the number of pixels of edge pixel points lower than the preset threshold is recorded as a.
And (6-4-2) determining the confidence degree that each second suspected bubble area is a real bubble area according to the pixel number of each edge pixel point of each second suspected bubble area and the pixel number of the edge pixel point of which the corresponding direction difference index is lower than a preset threshold value.
In this embodiment, for example, the confidence that the c-th second suspected bubble area is the real bubble area is determined, and the confidence that the c-th second suspected bubble area is the real bubble area is calculated according to the number of pixels of each edge pixel of the c-th second suspected bubble area and the number of pixels of edge pixels of which the corresponding direction difference index is lower than the preset threshold, where the calculation formula is as follows:
wherein, ω is c Is the confidence that the c-th second suspected bubble area is the real bubble area, A c The number of pixels of each edge pixel point of the c-th second suspected bubble area，a c And tau is a hyper-parameter larger than 1, wherein tau is the number of the edge pixels lower than a preset threshold value in all the edge pixels of the c-th second suspected bubble area.
And referring to the determination process of the confidence degree that the c-th second suspected bubble area is the real bubble area, obtaining the confidence degree that each second suspected bubble area is the real bubble area.
(7) And screening each second suspected bubble area according to the confidence degree that each second suspected bubble area is a real bubble area, so as to determine the bubble area in the surface image of the rubber product to be detected.
In this embodiment, according to the confidence that each second suspected bubble area is a real bubble area, normalization processing is performed on the confidence that each second suspected bubble area is a real bubble area to ensure that the confidence value is [0,1], and the normalization processing is helpful for more intuitively analyzing whether each second suspected bubble area is a real bubble area or not in the following.
And screening each second suspected bubble area according to the confidence degree that each second suspected bubble area is a real bubble area, if the confidence degree that a certain second suspected bubble area is a real bubble area is greater than a preset confidence degree threshold value, determining that the second suspected bubble area is a bubble area in the surface image of the rubber product to be detected, otherwise, determining that the second suspected bubble area is not a bubble area in the surface image of the rubber product to be detected, wherein the preset confidence degree threshold value is set to be 0.6 in the embodiment. Based on this, in the present embodiment, the second suspected-bubble area after the screening is used as the final bubble area of the rubber product.
According to the bubble area in the surface image of the rubber product to be detected, marking the pixel value of the pixel point of the bubble area in the surface image of the rubber product to be detected as 1, and marking the pixel values of the pixel points of other areas as 0, thereby obtaining the binary image of the surface image of the rubber product to be detected. The binary image of the surface image of the rubber product to be detected is helpful for observing the bubble area of the rubber product to be detected more intuitively.
The embodiment also provides a rubber product bubble detection system which comprises a processor and a memory, wherein the processor is used for processing the instructions stored in the memory so as to realize a rubber product bubble detection method.
The method mainly utilizes a material testing and analyzing technology to screen each suspected bubble area according to the image characteristic information of the bubbles of the rubber product, thereby determining the bubble area in the rubber product and realizing the detection of the bubble defects of the rubber product.
The above-mentioned embodiments are only used for illustrating the technical solutions of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; such modifications and substitutions do not substantially depart from the spirit and scope of the embodiments of the present application and are intended to be included within the scope of the present application.
Claims (10)
1. A method for detecting bubbles in a rubber product is characterized by comprising the following steps:
acquiring a visible light surface image of the rubber product to be detected, and processing the visible light surface image to obtain a visible light edge image of the rubber product to be detected, so as to obtain each suspected bubble area in the visible light edge image;
screening each suspected bubble area in the visible light edge image of the rubber product to be detected, so as to determine each first suspected bubble area in the visible light edge image of the rubber product to be detected;
determining a target circle area corresponding to each first suspected bubble area according to each first suspected bubble area in the visible light edge image of the rubber product to be detected;
determining a morphological index value corresponding to each first suspected bubble area according to each first suspected bubble area and a target circle area corresponding to each first suspected bubble area in a visible light edge image of the rubber product to be detected;
screening each first suspected bubble area according to the morphological index value corresponding to each first suspected bubble area, so as to determine each second suspected bubble area in the visible light edge image of the rubber product to be detected;
determining the confidence coefficient that each second suspected bubble area is a real bubble area according to each second suspected bubble area in the visible light edge image of the rubber product to be detected;
and screening each second suspected bubble area according to the confidence degree that each second suspected bubble area is a real bubble area, so as to determine the bubble area in the visible light surface image of the rubber product to be detected.
2. The method of claim 1, wherein the step of determining the target circle area corresponding to each first suspected bubble area comprises:
according to each first suspected bubble area in the visible light edge image of the rubber product to be detected, respectively performing vertical scanning movement and horizontal scanning movement in each first suspected bubble area by adopting a horizontal scanning line and a vertical scanning line, and determining the intersection point of each horizontal scanning line and each first suspected bubble area and the intersection point of each vertical scanning line and each first suspected bubble area;
determining the circle center coordinates of the target circle area corresponding to each first suspected bubble area according to the intersection point of each horizontal scanning line and each first suspected bubble area and the intersection point of each vertical scanning line and each first suspected bubble area;
determining each target edge pixel point of each first suspected bubble area according to each edge pixel point of each first suspected bubble area in the visible light edge image of the rubber product to be detected;
determining the radius of a target circle area corresponding to each first suspected bubble area according to the coordinate position of each target edge pixel point of each first suspected bubble area and the center coordinate of the target circle area corresponding to each first suspected bubble area;
and determining the target circle area corresponding to each first suspected bubble area according to the circle center coordinate and the radius of the target circle area corresponding to each first suspected bubble area.
3. The method of claim 2, wherein the step of determining the target edge pixel point of each first suspected bubble area comprises:
determining an angle index value corresponding to each edge pixel point of each first suspected bubble area according to each edge pixel point of each first suspected bubble area in a visible light edge image of the rubber product to be detected;
determining the confidence degree that each edge pixel point of each first suspected bubble area is a target edge pixel point according to the angle index value corresponding to each edge pixel point of each first suspected bubble area;
and screening each edge pixel point of each first suspected bubble area according to the confidence degree that each edge pixel point of each first suspected bubble area is a target edge pixel point, so as to determine each target edge pixel point of each first suspected bubble area.
4. The method according to claim 3, wherein the step of determining the angle index value corresponding to each edge pixel point of each first suspected bubble area comprises:
according to each edge pixel point of each first suspected bubble area in a visible light edge image of the rubber product to be detected, taking each edge pixel point of each first suspected bubble area as a central pixel point, selecting K edge pixel points from two sides of the central pixel point, and further determining a related pixel point corresponding to each edge pixel point;
determining a fitting line segment corresponding to each edge pixel point of each first suspected bubble area according to the related pixel point corresponding to each edge pixel point of each first suspected bubble area;
and taking the included angle between the fitting line segment corresponding to each edge pixel point of each first suspected bubble area and the positive direction of the X axis as the angle index value corresponding to each edge pixel point of each first suspected bubble area.
5. The method for detecting bubbles in a rubber product according to claim 3 or 4, wherein the step of determining the confidence that each edge pixel point of each first suspected bubble area is a target edge pixel point comprises:
determining neighborhood pixel points corresponding to the edge pixel points of each first suspected bubble area according to the edge pixel points of each first suspected bubble area;
determining a maximum angle index value and a minimum angle index value corresponding to each edge pixel point of each first suspected bubble area according to the angle index value of each edge pixel point of each first suspected bubble area and the angle index value of a neighborhood pixel point corresponding to each edge pixel point;
determining the confidence degree that each edge pixel point of each first suspected bubble area is a target edge pixel point according to the maximum angle index value and the minimum angle index value corresponding to each edge pixel point of each first suspected bubble area, wherein the calculation formula is as follows:
wherein the content of the first and second substances,
the confidence coefficient, gamma, that the jth edge pixel point of the c-th first suspected bubble area is the target edge pixel point c,j,max A maximum angle index value gamma corresponding to the jth edge pixel point of the c-th first suspected bubble area c,j,min And v is a super parameter, and is a minimum angle index value corresponding to the jth edge pixel point of the jth first suspected bubble area.
6. The method of claim 1, wherein the formula for determining the morphometric index value corresponding to each of the first suspected bubble areas is as follows:
where ρ is c The morphological index value u corresponding to the c-th first suspected bubble area c The number of edge pixels, U, of the edge pixel point where each edge pixel point of the c-th first suspected bubble area coincides with each edge pixel point of the corresponding target circle area c The number of pixels of each edge pixel point of the c-th first suspected bubble area.
7. The method of claim 1, wherein the step of determining the confidence level that each second suspected bubble area is a true bubble area comprises:
determining a horizontal direction gradient amplitude and a vertical direction gradient amplitude corresponding to each edge pixel point of each second suspected bubble area according to each second suspected bubble area in the visible light edge image of the rubber product to be detected, and further determining a gradient direction corresponding to each edge pixel point of each second suspected bubble area;
determining a direction index corresponding to each edge pixel point of each second suspected bubble area according to the pixel coordinate of each edge pixel point of each second suspected bubble area and the circle center coordinate of the target circle area corresponding to each second suspected bubble area;
determining a direction difference index corresponding to each edge pixel point of each second suspected bubble area according to the gradient direction and the direction index corresponding to each edge pixel point of each second suspected bubble area;
and determining the confidence degree of each second suspected bubble area as a real bubble area according to the direction difference index corresponding to each edge pixel point of each second suspected bubble area.
8. The method for detecting bubbles in a rubber product according to claim 7, wherein a calculation formula for determining the direction difference index corresponding to each edge pixel point of each second suspected bubble area is as follows:
d c,j ＝1-exp(-|θ c j -θ′ c j |)
wherein d is c,j Is a direction difference index theta corresponding to the jth edge pixel point of the jth second suspected bubble area c j Is the gradient direction corresponding to the jth edge pixel point of the jth second suspected bubble area, theta' c j And the direction index corresponding to the jth edge pixel point of the jth second suspected bubble area.
9. The method according to claim 7, wherein the step of determining the confidence that each second suspected bubble area is a true bubble area comprises:
judging whether the direction difference index corresponding to each edge pixel point of each second suspected bubble area is lower than a preset threshold value or not according to the direction difference index corresponding to each edge pixel point of each second suspected bubble area, and further counting the number of pixels of edge pixel points of which the direction difference index corresponding to each second suspected bubble area is lower than the preset threshold value;
and determining the confidence degree of each second suspected bubble area as a real bubble area according to the pixel number of each edge pixel point of each second suspected bubble area and the pixel number of the edge pixel point of which the corresponding direction difference index is lower than a preset threshold value.
10. A rubber article bubble detection system comprising a processor and a memory, the processor being configured to process instructions stored in the memory to implement a rubber article bubble detection method as recited in any one of claims 1-9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210462366.5A CN114820529A (en) | 2022-04-28 | 2022-04-28 | Method and system for detecting bubbles of rubber product |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210462366.5A CN114820529A (en) | 2022-04-28 | 2022-04-28 | Method and system for detecting bubbles of rubber product |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114820529A true CN114820529A (en) | 2022-07-29 |
Family
ID=82508637
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210462366.5A Pending CN114820529A (en) | 2022-04-28 | 2022-04-28 | Method and system for detecting bubbles of rubber product |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114820529A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116228758A (en) * | 2023-05-08 | 2023-06-06 | 深圳市前海誉卓科技有限公司 | Internal bubble detection method for polarizer production |
| CN116423957A (en) * | 2023-04-17 | 2023-07-14 | 浙江顶善美集成家居股份有限公司 | Preparation method of folding waterproof multilayer honeycomb board |
-
2022
- 2022-04-28 CN CN202210462366.5A patent/CN114820529A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116423957A (en) * | 2023-04-17 | 2023-07-14 | 浙江顶善美集成家居股份有限公司 | Preparation method of folding waterproof multilayer honeycomb board |
| CN116423957B (en) * | 2023-04-17 | 2023-11-03 | 浙江顶善美集成家居股份有限公司 | Preparation method of folding waterproof multilayer honeycomb board |
| CN116228758A (en) * | 2023-05-08 | 2023-06-06 | 深圳市前海誉卓科技有限公司 | Internal bubble detection method for polarizer production |
| CN116228758B (en) * | 2023-05-08 | 2023-07-07 | 深圳市前海誉卓科技有限公司 | Internal bubble detection method for polarizer production |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114820529A (en) | Method and system for detecting bubbles of rubber product | |
| US8331650B2 (en) | Methods, systems and apparatus for defect detection | |
| CN115018853B (en) | Mechanical component defect detection method based on image processing | |
| CN111047568B (en) | Method and system for detecting and identifying steam leakage defect | |
| KR102168724B1 (en) | Method And Apparatus for Discriminating Normal and Abnormal by using Vision Inspection | |
| CN117197141B (en) | Method for detecting surface defects of automobile parts | |
| CN109584239B (en) | High-light object surface defect detection system and method based on reflected light | |
| CN115984272B (en) | Semitrailer axle defect identification method based on computer vision | |
| CN110189375A (en) | A kind of images steganalysis method based on monocular vision measurement | |
| CN115953409B (en) | Injection molding surface defect detection method based on image processing | |
| CN117408995B (en) | Power adapter appearance quality detection method based on multi-feature fusion | |
| CN111861979A (en) | Positioning method, positioning equipment and computer readable storage medium | |
| KR102242996B1 (en) | Method for atypical defects detect in automobile injection products | |
| CN117237747B (en) | Hardware defect classification and identification method based on artificial intelligence | |
| CN116563298B (en) | Cross line center sub-pixel detection method based on Gaussian fitting | |
| Zhang et al. | Defect detection of mobile phone screen based on improved difference image method | |
| CN113554688B (en) | O-shaped sealing ring size measurement method based on monocular vision | |
| CN113011417B (en) | Target matching method based on intersection ratio coverage rate loss and repositioning strategy | |
| CN114693626A (en) | Method and device for detecting chip surface defects and computer readable storage medium | |
| Zhang et al. | Automatic optical defect inspection and dimension measurement of drill bit | |
| CN107239754B (en) | Automobile logo identification method based on sparse sampling intensity profile and gradient distribution | |
| Karim et al. | Comparing the Main Approaches of Image Segmentation | |
| CN112763506A (en) | Flaw detection method and device with AOI and AI functions | |
| CN110441315A (en) | Electronic component test equipment and method | |
| US11270402B2 (en) | Vial content detection using illuminated background pattern |
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 |