CN115760782B - Machine vision-based in-mold labeling offset defect identification method - Google Patents
Machine vision-based in-mold labeling offset defect identification method Download PDFInfo
- Publication number
- CN115760782B CN115760782B CN202211459884.8A CN202211459884A CN115760782B CN 115760782 B CN115760782 B CN 115760782B CN 202211459884 A CN202211459884 A CN 202211459884A CN 115760782 B CN115760782 B CN 115760782B
- Authority
- CN
- China
- Prior art keywords
- circle
- edge
- image
- pixel
- gradient
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Abstract
The invention discloses a machine vision-based in-mold labeling offset defect identification method, which comprises the following steps of: s1, after calibration and correction of an industrial camera, acquiring an image of a labeled product in a mold, making a circle with a diameter smaller than that of a circle B, and dividing the image into a label part and an outer edge part; s2, acquiring an interested region by using threshold segmentation and morphological processing, and extracting sub-pixel edges on the interested region by using a Canny operator and a Zernike moment to realize edge extraction of a target circle A and a target circle C; s3, performing weight type edge fitting, namely performing edge fitting on the circle A and the circle C by using a weight function to obtain parameters of the radius and the circle center of the circle A and the circle C, and calculating the minimum distance g between the circle B and the circle C min ,g min And judging that the labels are offset defective products when the labels are smaller than the qualified spacing. The invention can solve the problem of detection of offset defects of the in-mold labeling of the product, and enables the detection of machine vision defects to have flexibility.
Description
Technical Field
The invention belongs to the technical field of machine vision, and particularly relates to a machine vision-based in-mold labeling offset defect identification method.
Background
The actual production process of the in-mold labeling product at the present stage shows the characteristic of typical flexible manufacturing: the shape of the product produced by each injection molding machine in the workshop is consistent, but the labels, the colors and the batches are different, and the products can be replaced irregularly according to the requirements of customers. In order to improve the production efficiency and fully utilize the rapid detection capability of the vision system, products on the conveyor belts of a plurality of injection molding machines are assembled on a total conveyor belt, and a set of machine vision system is used for detection in a concentrated manner. The label offset is the defect of highest detection difficulty in the in-mold labeling product, the narrow gap is difficult to directly judge by means of human eyes, a great amount of time and energy are consumed by means of a complex measuring tool, and the accuracy is poor, so that the detection task is difficult to finish in batch production.
Currently, there are few studies on visual inspection of in-mold label offset defects in injection molded articles, and the problem is also difficult to solve: firstly, due to the flexible production condition, the detected products have the diversity of labels and colors, and the traditional image segmentation algorithm cannot play a role; secondly, the structure of some products is compact, the information is complex, the extraction of the bias characteristic is difficult, and great challenges are brought to the detection of the bias defect of the label of the product.
Disclosure of Invention
The invention mainly aims to overcome the defects and shortcomings of the prior art and provides a machine vision-based in-mold labeling offset defect identification method.
In order to achieve the above purpose, the present invention adopts the following technical scheme:
the in-mold labeling offset defect identification method based on machine vision is characterized in that the outer circle of a labeling product is set as a circle A, the convex line circle of the labeling product is set as a circle B, and a label is regarded as a standard circle and is set as a circle C; the method comprises the following steps:
s1, after calibration and correction of an industrial camera, acquiring an image of a labeled product in a mold, making a circle with a diameter smaller than that of a circle B, and dividing the image into a label part and an outer edge part;
s2, acquiring an interested region by using threshold segmentation and morphological processing, and extracting sub-pixel edges on the interested region by using a Canny operator and a Zernike moment to realize edge extraction of a target circle A and a target circle C;
s3, performing weight type edge fitting, namely performing edge fitting on the circle A and the circle C by using a weight function to obtain parameters of the radius and the circle center of the circle A and the circle C, and calculating the minimum distance g between the circle B and the circle C min ,g min And judging that the labels are offset defective products when the labels are smaller than the qualified spacing.
Further, when the industrial camera collects the image of the in-mold labeling product, the annular light source is inclined at a certain angle so as to make the characteristics of the region of interest highlighted.
Further, the step S1 specifically includes:
and (3) using a positioning algorithm to obtain positioning coordinates (x, y) of the in-mold labeling product, taking the coordinates as a circle center to make a circle with the diameter smaller than that of the circle B, and dividing the image into a circle C and an outer edge part thereof.
Further, in step S2, the region of interest is acquired specifically as follows:
the circle C and the outer edge part are respectively divided by threshold values, and the expression of threshold value division is as follows:
wherein f (x, y), g (x, y) are the gray values of the image before and after segmentation, T 1 ,T 2 Is a threshold value;
and obtaining an image after threshold segmentation, performing a closed operation by using a circular structural element with a proper radius, eliminating interference of low gray noise in the region on the premise of not causing the deficiency of the target circular edge, and taking the two regions after the closed operation as the interested region extracted by the target circular edge.
Further, during threshold segmentation, different thresholds are selected for different labeling products for segmentation.
Further, extracting the subpixel edges using the Canny operator and the Zernike moments includes:
pixel-level edge positioning is performed on the region of interest by using a Canny operator;
on the pixel-level edges detected by the Canny operator, a Zernike moment sub-pixel edge localization algorithm is used.
Further, performing pixel-level edge positioning on the region of interest by using a Canny operator specifically includes:
step one, performing Gaussian smoothing filtering on an image;
step two, calculating a gradient amplitude diagram and a gradient direction; the calculation formula of the gradient amplitude is:
the calculation formula of the gradient angle is as follows:
wherein G is x And G y The gradients in the x-direction and y-direction, respectively;
step three, non-maximum suppression is applied to the gradient amplitude image;
step four, detecting pixel-level edges by using double-threshold processing and edge connection; the processing method of the double threshold value specifically comprises the following steps:
selecting proper high threshold and low threshold, if the gradient value of a certain pixel is higher than the high threshold, reserving; discarding if the gradient value of a certain pixel is below a low threshold; if the gradient value of a certain pixel is between the high threshold value and the low threshold value, if the gradient of the pixel in the 8-connected region of the pixel is higher than the high threshold value, the gradient is reserved, and if the gradient is not present, the gradient is abandoned.
Further, the edge positioning algorithm using the Zernike moment sub-pixel specifically comprises the following steps:
step one, calculating Zernike moment of an image; the calculation formula of the Zernike moment of the gray image n-order m times is as follows:
wherein f (x, y) represents the gray value of the image point (x, y), V * mn (ρ, θ) is a Zernike polynomial V mn (ρ, θ) complex conjugation;
step two, calculating 4 edge parameters, namely background gray level h, step height k, vertical distance l from the center of the disc to the edge and included angle omega between the vertical line and the x axis, through the rotation invariance of the Zernike moment;
and thirdly, positioning the edges of the sub-pixels according to the edge parameters.
Further, in step S3, the edge fitting is specifically:
calculating the circle center and radius of the fitting circle by using a least square circle fitting algorithm and calculating the minimum value of algebraic distance from each edge point of the target circle to the fitting circle; the least squares method is calculated as follows:
wherein (alpha, beta) is the center of a fitting circle, R is the radius of the fitting circle, (X) i ,Y i ) Is an edge point;
introducing a Tukey weight function, wherein the function introduces a weight omega for each sub-pixel edge point, and the function expression is as follows:
wherein delta is the distance from the edge point to the fitting circle; τ is a clipping factor, derived from the standard deviation of the edge points, which are ignored when fitting circles when δ is greater than τ, and whose weight varies smoothly between 1 and 0 when δ is less than τ.
Further, in step S3, g min Is calculated as follows:
the center of circle A (x) A ,y A ) And radius r A Center of circle C (x) C ,y C ) And radius r C G is calculated according to the following formula min :
g min =r A -r C -l-s
Wherein l is the processing distance from circle A to circle B, and s is the center distance of circle A to circle C.
Compared with the prior art, the invention has the following advantages and beneficial effects:
1. the method provided by the invention aims at different in-mold labeling products, can accurately and effectively judge whether the labeling is biased, and realizes the requirement of flexible detection.
2. Unlike conventional methods for processing the whole image, the invention proposes a divide-and-conquer idea; for the situation that the target spacing is small or local overlap is caused, the whole treatment can lead to the connection of the regions between the targets; the method is used for dividing two target circles with small spacing or partially overlapped, then independently carrying out operations such as edge extraction, edge fitting and the like on the separated target circles, and finally integrating.
Drawings
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is a schematic diagram of the structure of an in-mold labeling article;
FIG. 3 is a graph of circle A, circle C position parameters and g min Is a relationship of (2);
FIG. 4 is a diagram of a target circle edge extraction process;
fig. 5 is a graph of the results after edge fitting of the target circle.
Detailed Description
The present invention will be described in further detail with reference to examples and drawings, but embodiments of the present invention are not limited thereto.
Examples
In the embodiment, a labeling plastic product is taken as an example, wherein the outer circle of the product is set as a circle A, the convex line circle of the product is set as a circle B, and the distance between the circle A and the circle B is set as a fixed value l; the label is a non-standard circle, but the main body is a circle, so the label is regarded as a standard circle, and the circle is referred to as a circle C.
As shown in fig. 1 and 2, the method for identifying offset defects of in-mold labeling based on machine vision comprises the following steps:
s1, after calibration and correction of an industrial camera, acquiring an image of a labeled product in a mold, making a circle with a diameter smaller than that of a circle B, and dividing the image into a label part and an outer edge part; when the industrial camera collects the image of the in-mold labeling product, the annular light source is inclined at a certain angle to highlight the characteristics of the region of interest, and in the embodiment, 70-degree annular light source is adopted for inclined lighting.
And (3) obtaining positioning coordinates (x, y) of the in-mold labeling product by using a positioning algorithm, for example, obtaining a detection frame of the in-mold labeling product by using a YOLO algorithm, taking the central point of the frame as the positioning coordinates (x, y), and simultaneously removing the background outside the frame. A circle with the diameter smaller than that of the circle B is made by taking the coordinate as the center of the circle, and the image is divided into a circle C and the outer edge part thereof.
S2, acquiring an interested region by using threshold segmentation and morphological processing, and extracting sub-pixel edges on the interested region by using a Canny operator and a Zernike moment to realize edge extraction of a target circle A and a target circle C; as shown in fig. 4, a process diagram of the target circle edge extraction is shown.
In this embodiment, the region of interest acquisition is specifically:
the circle C and the outer edge part are respectively divided by using thresholds, and different labeling products are respectively divided by using different thresholds; the expression of the threshold segmentation is as follows:
wherein f (x, y), g (x, y) are the gray values of the image before and after segmentation, T 1 ,T 2 Is a threshold value;
and obtaining an image after threshold segmentation, performing a closed operation by using a circular structural element with a proper radius, eliminating interference of low gray noise in the region on the premise of not causing the deficiency of the target circular edge, and taking the two regions after the closed operation as the interested region extracted by the target circular edge.
In this embodiment, extracting the subpixel edges using the Canny operator and the Zernike moments includes:
pixel-level edge localization using Canny operators at the region of interest, specifically includes:
step one, performing Gaussian smoothing filtering on an image;
step two, calculating a gradient amplitude diagram and a gradient direction; the calculation formula of the gradient amplitude is:
the calculation formula of the gradient angle is as follows:
wherein G is x And G y Respectively denoted as x-directionAnd a gradient in the y-direction;
step three, non-maximum suppression is applied to the gradient amplitude image;
step four, detecting pixel-level edges by using double-threshold processing and edge connection; the processing method of the double threshold value specifically comprises the following steps:
selecting proper high threshold and low threshold (the high threshold is approximately 2-3 times of the low threshold), and if the gradient value of a certain pixel is higher than the high threshold, reserving; discarding if the gradient value of a certain pixel is below a low threshold; if the gradient value of a certain pixel is between the high threshold value and the low threshold value, if the gradient of the pixel in the 8-connected region of the pixel is higher than the high threshold value, the gradient is reserved, and if the gradient is not present, the gradient is abandoned.
On the pixel-level edge detected by the Canny operator, a Zernike moment sub-pixel edge positioning algorithm is used, and specifically:
step one, calculating Zernike moment of an image; the calculation formula of the Zernike moment of the gray image n-order m times is as follows:
wherein f (x, y) represents the gray value of the image point (x, y), V * mn (ρ, θ) is a Zernike polynomial V mn (ρ, θ) complex conjugation;
step two, calculating 4 edge parameters, namely background gray level h, step height k, vertical distance l from the center of the disc to the edge and included angle omega between the vertical line and the x axis, through the rotation invariance of the Zernike moment;
and thirdly, positioning the edges of the sub-pixels according to the edge parameters.
S3, performing weight type edge fitting, namely performing edge fitting on the circle A and the circle C by using a weight function to obtain parameters of the radius and the circle center of the circle A and the circle C, and calculating the minimum distance g between the circle B and the circle C min ,g min And judging that the labels are offset defective products when the labels are smaller than the qualified spacing.
In this embodiment, the edge fitting is specifically:
calculating the circle center and radius of the fitting circle by using a least square circle fitting algorithm and calculating the minimum value of algebraic distance from each edge point of the target circle to the fitting circle; the least squares method is calculated as follows:
wherein (alpha, beta) is the center of a fitting circle, R is the radius of the fitting circle, (X) i ,Y i ) Is an edge point;
introducing a Tukey weight function, wherein the function introduces a weight omega for each sub-pixel edge point, and the function expression is as follows:
wherein delta is the distance from the edge point to the fitting circle; τ is a clipping factor, derived from the standard deviation of the edge points, which are ignored when fitting circles when δ is greater than τ, and whose weight varies smoothly between 1 and 0 when δ is less than τ.
Comparing the fitted circle with the target circle, as shown in fig. 5, the fitted circle is a result diagram of the edge of the target circle, and the ideal fitting effect can be seen from the diagram.
In the present embodiment, g as shown in FIG. 3 min Is calculated as follows:
the center of circle A (x) A ,y A ) And radius r A Center of circle C (x) C ,y C ) And radius r C G is calculated according to the following formula min :
g min =r A -r C -l-s
Wherein l is the processing distance from circle A to circle B, and s is the center distance of circle A to circle C.
In this example g min And judging that the defective products with the label offset exist when the label offset is smaller than 3+/-0.3 mm.
It should also be noted that in this specification, terms such as "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (9)
1. The in-mold labeling offset defect identification method based on machine vision is characterized in that the outer circle of a labeling product is set as a circle A, the convex line circle of the labeling product is set as a circle B, and a label is regarded as a standard circle and is set as a circle C; the method comprises the following steps:
s1, after calibration and correction of an industrial camera, acquiring an image of a labeled product in a mold, making a circle with a diameter smaller than that of a circle B, and dividing the image into a label part and an outer edge part;
s2, acquiring an interested region by using threshold segmentation and morphological processing, and extracting sub-pixel edges on the interested region by using a Canny operator and a Zernike moment to realize edge extraction of a target circle A and a target circle C;
s3, performing weight type edge fitting, namely performing edge fitting on the circle A and the circle C by using a weight function to obtain parameters of the radius and the circle center of the circle A and the circle C, and calculating the minimum distance g between the circle B and the circle C min ,g min Judging that the label is not good when the qualified space is smaller than the qualified space; g min Is calculated as follows:
acquisition circle ACenter of circle (x) A ,y A ) And radius r A Center of circle C (x) C ,y C ) And radius r C G is calculated according to the following formula min :
g min =r A -r C -l-s
Wherein l is the processing distance from circle A to circle B, and s is the center distance of circle A to circle C.
2. The machine vision based in-mold labeling offset defect identification method of claim 1, wherein the industrial camera captures the in-mold labeling product image using an annular light source tilt angle to highlight the region of interest feature.
3. The method for identifying offset defects of in-mold labeling based on machine vision according to claim 1, wherein step S1 is specifically:
and (3) using a positioning algorithm to obtain positioning coordinates (x, y) of the in-mold labeling product, taking the coordinates as a circle center to make a circle with the diameter smaller than that of the circle B, and dividing the image into a circle C and an outer edge part thereof.
4. The method for identifying offset defects of in-mold labeling based on machine vision according to claim 1, wherein in step S2, the region of interest is obtained specifically as follows:
the circle C and the outer edge part are respectively divided by threshold values, and the expression of threshold value division is as follows:
wherein f (x, y), g (x, y) are the gray values of the image before and after segmentation, T 1 ,T 2 Is a threshold value;
and obtaining an image after threshold segmentation, performing a closed operation by using a circular structural element with a proper radius, eliminating interference of low gray noise in the region on the premise of not causing the deficiency of the target circular edge, and taking the two regions after the closed operation as the interested region extracted by the target circular edge.
5. The machine vision based in-mold labeling offset defect identification method of claim 4, wherein different thresholds are selected for different labeling products for segmentation.
6. The machine vision based in-mold labeling offset defect identification method of claim 1, wherein extracting sub-pixel edges using Canny operators and Zernike moments comprises:
pixel-level edge positioning is performed on the region of interest by using a Canny operator;
on the pixel-level edges detected by the Canny operator, a Zernike moment sub-pixel edge localization algorithm is used.
7. The machine vision based in-mold labeling offset defect identification method of claim 6, wherein performing pixel-level edge localization using Canny operator on the region of interest specifically comprises:
step one, performing Gaussian smoothing filtering on an image;
step two, calculating a gradient amplitude diagram and a gradient direction; the calculation formula of the gradient amplitude is:
the calculation formula of the gradient angle is as follows:
wherein G is x And G y The gradients in the x-direction and y-direction, respectively;
step three, non-maximum suppression is applied to the gradient amplitude image;
step four, detecting pixel-level edges by using double-threshold processing and edge connection; the processing method of the double threshold value specifically comprises the following steps:
selecting a high threshold value and a low threshold value, and if the gradient value of a certain pixel is higher than the high threshold value, reserving; discarding if the gradient value of a certain pixel is below a low threshold; if the gradient value of a certain pixel is between the high threshold value and the low threshold value, if the gradient of the pixel in the 8-connected region of the pixel is higher than the high threshold value, the gradient is reserved, and if the gradient is not present, the gradient is abandoned.
8. The method for identifying offset defects of in-mold labeling based on machine vision according to claim 6, wherein the Zernike moment subpixel edge positioning algorithm is specifically:
step one, calculating Zernike moment of an image; the calculation formula of the Zernike moment of the gray image n-order m times is as follows:
wherein f (x, y) represents the gray value of the image point (x, y), V * mn (ρ, θ) is a Zernike polynomial V mn (ρ, θ) complex conjugation;
step two, calculating 4 edge parameters, namely background gray level h, step height k, vertical distance l from the center of the disc to the edge and included angle omega between the vertical line and the x axis, through the rotation invariance of the Zernike moment;
and thirdly, positioning the edges of the sub-pixels according to the edge parameters.
9. The method for identifying offset defects of in-mold labeling based on machine vision according to claim 1, wherein in step S3, the edge fitting is specifically:
calculating the circle center and radius of the fitting circle by using a least square circle fitting algorithm and calculating the minimum value of algebraic distance from each edge point of the target circle to the fitting circle; the least squares method is calculated as follows:
wherein (alpha, beta) is the center of a fitting circle, R is the radius of the fitting circle, (X) i ,Y i ) Is an edge point;
introducing a Tukey weight function, wherein the function introduces a weight omega for each sub-pixel edge point, and the function expression is as follows:
wherein delta is the distance from the edge point to the fitting circle; τ is a clipping factor, derived from the standard deviation of the edge points, which are ignored when fitting circles when δ is greater than τ, and whose weight varies smoothly between 1 and 0 when δ is less than τ.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211459884.8A CN115760782B (en) | 2022-11-16 | 2022-11-16 | Machine vision-based in-mold labeling offset defect identification method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211459884.8A CN115760782B (en) | 2022-11-16 | 2022-11-16 | Machine vision-based in-mold labeling offset defect identification method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN115760782A CN115760782A (en) | 2023-03-07 |
CN115760782B true CN115760782B (en) | 2023-06-16 |
Family
ID=85334185
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202211459884.8A Active CN115760782B (en) | 2022-11-16 | 2022-11-16 | Machine vision-based in-mold labeling offset defect identification method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115760782B (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116309510B (en) * | 2023-03-29 | 2024-03-22 | 清华大学 | Numerical control machining surface defect positioning method and device |
CN116612116A (en) * | 2023-07-19 | 2023-08-18 | 天津伍嘉联创科技发展股份有限公司 | Crystal appearance defect detection method based on deep learning image segmentation |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106127253A (en) * | 2016-06-27 | 2016-11-16 | 北京航空航天大学 | A kind of method for detecting infrared puniness target utilizing sample characteristics learning classification |
CN109658391A (en) * | 2018-12-04 | 2019-04-19 | 东北大学 | A kind of radius of circle measurement method being fitted based on contour mergence and convex closure |
CN109900711A (en) * | 2019-04-02 | 2019-06-18 | 天津工业大学 | Workpiece, defect detection method based on machine vision |
CN111879241A (en) * | 2020-06-24 | 2020-11-03 | 西安交通大学 | Mobile phone battery size measuring method based on machine vision |
CN112884708A (en) * | 2021-01-15 | 2021-06-01 | 深圳市悦创进科技有限公司 | Method for detecting burrs of circular injection molding piece |
CN113129322A (en) * | 2021-04-22 | 2021-07-16 | 中煤科工集团重庆研究院有限公司 | Sub-pixel edge detection method |
CN113610917A (en) * | 2021-08-09 | 2021-11-05 | 河南工业大学 | Circular array target center image point positioning method based on blanking points |
CN114239378A (en) * | 2021-11-09 | 2022-03-25 | 华南理工大学 | Injection molding product size prediction method based on custom LightGBM model loss |
CN115018833A (en) * | 2022-08-05 | 2022-09-06 | 山东鲁芯之光半导体制造有限公司 | Processing defect detection method of semiconductor device |
-
2022
- 2022-11-16 CN CN202211459884.8A patent/CN115760782B/en active Active
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106127253A (en) * | 2016-06-27 | 2016-11-16 | 北京航空航天大学 | A kind of method for detecting infrared puniness target utilizing sample characteristics learning classification |
CN109658391A (en) * | 2018-12-04 | 2019-04-19 | 东北大学 | A kind of radius of circle measurement method being fitted based on contour mergence and convex closure |
CN109900711A (en) * | 2019-04-02 | 2019-06-18 | 天津工业大学 | Workpiece, defect detection method based on machine vision |
CN111879241A (en) * | 2020-06-24 | 2020-11-03 | 西安交通大学 | Mobile phone battery size measuring method based on machine vision |
CN112884708A (en) * | 2021-01-15 | 2021-06-01 | 深圳市悦创进科技有限公司 | Method for detecting burrs of circular injection molding piece |
CN113129322A (en) * | 2021-04-22 | 2021-07-16 | 中煤科工集团重庆研究院有限公司 | Sub-pixel edge detection method |
CN113610917A (en) * | 2021-08-09 | 2021-11-05 | 河南工业大学 | Circular array target center image point positioning method based on blanking points |
CN114239378A (en) * | 2021-11-09 | 2022-03-25 | 华南理工大学 | Injection molding product size prediction method based on custom LightGBM model loss |
CN115018833A (en) * | 2022-08-05 | 2022-09-06 | 山东鲁芯之光半导体制造有限公司 | Processing defect detection method of semiconductor device |
Also Published As
Publication number | Publication date |
---|---|
CN115760782A (en) | 2023-03-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN115760782B (en) | Machine vision-based in-mold labeling offset defect identification method | |
CN109141232B (en) | Online detection method for disc castings based on machine vision | |
CN106952257B (en) | A kind of curved surface label open defect detection method based on template matching and similarity calculation | |
WO2020248439A1 (en) | Crown cap surface defect online inspection method employing image processing | |
WO2022027949A1 (en) | Machine vision-based detecting method and system for glass bottle bottom defects | |
CN102495076B (en) | Method for detecting defects of metal zipper teeth of zipper on basis of machine vision | |
CN103604808B (en) | A kind of bottle cap defective vision detection method | |
CN104680519B (en) | Seven-piece puzzle recognition methods based on profile and color | |
CN107590837A (en) | A kind of vision positioning intelligent precise puts together machines people and its camera vision scaling method | |
CN108896574B (en) | Bottled liquor impurity detection method and system based on machine vision | |
CN109900711A (en) | Workpiece, defect detection method based on machine vision | |
CN104574389A (en) | Battery piece chromatism selection control method based on color machine vision | |
CN107894423B (en) | Automatic detection equipment and method for vehicle body surface quality loss and intelligent vehicle detection system | |
CN101799434A (en) | Printing image defect detection method | |
CN104050446A (en) | Meter pointer image identification method based on pointer width character | |
CN102680487A (en) | System and method for monitoring painting quality of components, in particular of motor-vehicle bodies | |
CN106824816A (en) | A kind of PE based on machine vision bottles of detection and method for sorting | |
CN110766684A (en) | Stator surface defect detection system and detection method based on machine vision | |
CN104297255A (en) | Visual inspection method and system device for paper cup defects | |
CN111080638B (en) | Method for detecting dirt at bottom of molded bottle | |
CN106018422A (en) | Matching-based visual outline defect inspection system and method for specially-shaped stamping parts | |
CN107036542A (en) | A kind of ring gear internal-and external diameter appearance detecting method and device | |
CN109324056B (en) | Sewing thread trace measuring method based on machine vision | |
CN110567965A (en) | Smartphone glass cover plate edge visual defect detection method | |
CN106560910B (en) | Tube core defect detection method and device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |