CN116309848B - Welding positioning method of battery and protection board based on computer vision - Google Patents
Welding positioning method of battery and protection board based on computer vision Download PDFInfo
- Publication number
- CN116309848B CN116309848B CN202310545421.1A CN202310545421A CN116309848B CN 116309848 B CN116309848 B CN 116309848B CN 202310545421 A CN202310545421 A CN 202310545421A CN 116309848 B CN116309848 B CN 116309848B
- Authority
- CN
- China
- Prior art keywords
- suspected
- surface image
- tab
- battery
- edge point
- 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
- 238000003466 welding Methods 0.000 title claims abstract description 51
- 238000000034 method Methods 0.000 title claims abstract description 37
- 238000003708 edge detection Methods 0.000 claims description 7
- 230000001681 protective effect Effects 0.000 claims description 7
- 238000001514 detection method Methods 0.000 claims description 4
- 238000012545 processing Methods 0.000 abstract description 8
- 239000004973 liquid crystal related substance Substances 0.000 description 14
- 238000004364 calculation method Methods 0.000 description 5
- 239000011159 matrix material Substances 0.000 description 4
- 238000009826 distribution Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000007429 general method Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000003908 quality control method Methods 0.000 description 1
- 238000011179 visual inspection Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
-
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/762—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using clustering, e.g. of similar faces in social networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30148—Semiconductor; IC; Wafer
-
- 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
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
- Y02E60/10—Energy storage using batteries
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Multimedia (AREA)
- Geometry (AREA)
- Quality & Reliability (AREA)
- Connection Of Batteries Or Terminals (AREA)
Abstract
The invention relates to the technical field of image processing, in particular to a welding positioning method of a battery and a protection plate based on computer vision, which comprises the following steps: acquiring a battery surface image and a protection plate surface image, acquiring an intersection point set according to edge points and gradient directions, classifying the intersection point set to obtain a plurality of intersection point categories, acquiring the circle center of a tab and a suspected tab edge point according to the circle center probability of the intersection point categories, acquiring the suspected radius and the radius of the tab according to the suspected tab edge point and the circle center, further obtaining a distance deviation fluctuation degree, acquiring the tab probability of the suspected tab edge point according to the distance deviation fluctuation degree, acquiring the correspondence of the suspected tab edge points in the two images according to the tab probability, further obtaining the correspondence between the tab edge point and the tab edge point, and performing welding positioning to realize automatic welding. The invention reduces the calculated amount, ensures the integrity and the accuracy of the characteristic matching of the battery and the protection board, and ensures the accuracy of automatic welding.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to a welding positioning method of a battery and a protective plate based on computer vision.
Background
The welding of the battery and the protection plate is an indispensable part in the battery assembly manufacturing, the quality of the battery directly influences the safety and the performance of the whole battery assembly, and the problems of infirm welding or overhigh welding temperature and the like can influence the reliability of electric connection and even cause the safety problems of short circuit, electric leakage and the like of the battery assembly.
Therefore, in the welding process of the battery and the protection plate, the welding quality control and the process optimization need to be very important to avoid various welding defects, the welding positioning accuracy is one of the important conditions of welding success, and the manual visual inspection is a common method for obtaining the welding positioning quality, but the method has high labor cost and is greatly influenced by human factors, and the computer vision method with high precision and strong stability is gradually widely applied.
The battery and protection board welding positioning method based on computer vision comprises the basic steps of collecting images of the battery and the protection board, preprocessing the images, extracting features, matching features, estimating positions and welding, wherein the steps are used for extracting and matching all feature points of the images of the battery and the protection board, the calculated amount is large, the welding positioning characteristics of the battery and the protection board are not considered, the feature matching data amount is too large, the waste of data redundancy, resources and time and inaccuracy of feature matching are caused, the welding positioning is time-consuming, even inaccurate positioning is easily caused, and the welding of the battery and the protection board is influenced.
Disclosure of Invention
The invention provides a welding positioning method of a battery and a protective plate based on computer vision, which aims to solve the existing problems.
The welding positioning method of the battery and the protection board based on computer vision adopts the following technical scheme:
one embodiment of the invention provides a welding positioning method of a battery and a protection plate based on computer vision, which comprises the following steps:
collecting a battery surface image and a protection plate surface image;
performing edge detection on the battery surface image to obtain all edge points in the battery surface image and gradient directions of all edge points; acquiring gradient extension lines of each edge point according to the gradient direction of each edge point, and acquiring intersection points between every two gradient extension lines of all edge points to form an intersection point set; clustering the intersection point sets to obtain a plurality of intersection point categories; the circle center probability of each intersection point category is obtained; acquiring the intersection point category with the maximum circle center probability as a circle center category, and acquiring the circle center of the lug and the edge point of the suspected lug in the battery surface image according to the circle center category; taking the Euclidean distance from each suspected tab edge point to the circle center as the suspected radius corresponding to each suspected tab edge point, and acquiring the radius of the tab in the battery surface image according to all the suspected radii; acquiring the distance offset fluctuation degree of all suspected tab edge points in the battery surface image; acquiring the tab probability of each suspected tab edge point in the battery surface image according to the distance offset fluctuation degree and the suspected radius;
acquiring the tab probability of all suspected tab edge points and each suspected tab edge point in the surface image of the protection plate;
acquiring a suspected tab sequence of the battery surface image and a suspected tab sequence of the protection plate surface image; acquiring the distance between a suspected tab sequence of the battery surface image and each suspected tab edge point in the suspected tab sequence of the protection plate surface image according to the tab probability; acquiring a corresponding relation between a suspected tab edge point in the battery surface image and a suspected tab edge point in the protection plate surface image according to the distance;
acquiring the corresponding relation between the battery surface image and the tab edge point in the protection plate surface image according to the corresponding relation between the battery surface image and the suspected tab edge point in the protection plate surface image; and carrying out welding positioning according to the corresponding relation between the battery surface image and the tab edge point in the protection plate surface image, so as to realize automatic welding.
Preferably, the step of obtaining the gradient extension line of each edge point according to the gradient direction of each edge point includes the following specific steps:
a straight line passing through each edge point and in the gradient direction of each edge point in the battery surface image is acquired as a gradient extension line of each edge point.
Preferably, the obtaining the circle center probability of each intersection point category includes the following specific steps:
performing convex hull detection on all the intersections in each intersection category to obtain convex hull areas of each intersection category; and taking the ratio of the number of the intersections contained in each intersection category to the area of the convex hull area as the circle center probability of each intersection category.
Preferably, the step of obtaining the circle center of the tab and the edge point of the suspected tab in the battery surface image according to the circle center category includes the following specific steps:
taking the centroids of all the intersection points in the circle center category as the circle centers of the lugs in the battery surface image; and taking the edge point of the gradient extension line passing through the intersection point in the circle center category as a suspected polar edge point.
Preferably, the step of obtaining the radius of the tab in the battery surface image according to all the suspected radii includes the following specific steps:
rounding the suspected radiuses corresponding to all the suspected tab edge points in the battery surface image, forming a suspected radius sequence by the rounded suspected radiuses, and obtaining the numerical value with the largest occurrence frequency in the suspected radius sequence as the radius of the tab in the battery surface image.
Preferably, the step of obtaining the distance offset fluctuation degree of all the suspected tab edge points in the battery surface image includes the following specific steps:
，
wherein, the liquid crystal display device comprises a liquid crystal display device,the fluctuation degree of the distance offset of all the suspected tab edge points in the battery surface image is calculated; />Is the>The suspected radiuses corresponding to the edge points of the suspected tabs; />Radius of the lug in the battery surface image; />Is doubt in the battery surface imageNumber of pole-like edge points.
Preferably, the acquiring the tab probability of each suspected tab edge point in the battery surface image according to the distance offset fluctuation degree and the suspected radius includes the following specific steps:
，
wherein, the liquid crystal display device comprises a liquid crystal display device,is the>The possibility of the electrode lug of each suspected electrode lug edge point; />The fluctuation degree of the distance offset of all the suspected tab edge points in the battery surface image is calculated; />Is the>The suspected radiuses corresponding to the edge points of the suspected tabs; />Radius of the lug in the battery surface image; />Is an absolute value symbol; />Is a natural constant.
Preferably, the acquiring the suspected tab sequence of the battery surface image and the suspected tab sequence of the protection board surface image includes the following specific steps:
starting from the suspected electrode edge point with the highest possibility of the electrode lug in the battery surface image, acquiring all the suspected electrode edge points in a clockwise direction, and forming a suspected electrode lug sequence of the battery surface image by all the suspected electrode edge points; and similarly, acquiring a suspected tab sequence of the surface image of the protection plate.
Preferably, the step of obtaining the distance between the suspected tab sequence of the battery surface image and each suspected tab edge point in the suspected tab sequence of the protection plate surface image according to the tab probability comprises the following specific steps:
，
wherein, the liquid crystal display device comprises a liquid crystal display device,the suspected tab sequence which is the image of the battery surface is +.>The suspected tab edge points and the suspected tab sequence of the surface image of the protection board are +.>The distance between the edge points of the suspected electrode lugs; />The suspected tab sequence which is the image of the battery surface is +.>The possibility of the electrode lug of each suspected electrode lug edge point; />The suspected tab sequence which is the surface image of the protection boardThe possibility of the electrode lug of each suspected electrode lug edge point; />The suspected tab sequence which is the image of the battery surface is +.>Gray values of the edge points of the suspected electrode lugs; />The suspected tab sequence for protecting the plate surface image is +.>Gray values of the edge points of the suspected electrode lugs; />The number of suspected polar edge points in the battery surface image; />Is the number of suspected polar edge points in the protective plate surface image.
Preferably, the obtaining the correspondence between the battery surface image and the tab edge point in the protection plate surface image includes the following specific steps:
updating the corresponding relation of the suspected tab edge points in the battery surface image comprises the following steps: if one suspected tab edge point in the battery surface image corresponds to a plurality of suspected tab edge points in the protection plate surface image, the suspected tab edge point in the battery surface image is taken as a point to be matched, a plurality of corresponding suspected tab edge points in the protection plate surface image are respectively taken as candidate matching points, a candidate matching point with the smallest distance from the point to be matched is taken as a suspected tab edge point actually corresponding to the point to be matched in the protection plate surface image, and the corresponding relation between the rest candidate matching points and the point to be matched is relieved;
similarly, updating the corresponding relation of the suspected tab edge points in the surface image of the protection plate;
acquiring a tab edge point in a battery surface image, comprising: if one suspected tab edge point in the battery surface image does not correspond to any suspected tab edge point in the protection plate surface image, taking the suspected tab edge point in the battery surface image as a noise point, and removing the suspected tab edge point from a suspected tab sequence of the battery surface image; taking the rest suspected electrode lug edge points in the suspected electrode lug sequence of the battery surface image as electrode lug edge points in the battery surface image;
similarly, the tab edge points in the surface image of the protection plate are obtained, and the tab edge points in the surface image of the battery are in one-to-one correspondence with the tab edge points in the surface image of the protection plate.
The technical scheme of the invention has the beneficial effects that: according to the invention, circular fitting is carried out according to the gradient directions of all edge points in the battery surface image and the protection plate surface image, so that suspected tab edge points are obtained, tab characteristics of the battery and the protection plate during welding are only reserved, the data redundancy of the matching of the subsequent characteristic points is reduced, and the calculated amount is greatly reduced; according to the method, the likelihood of the tab of the suspected tab edge point is obtained according to the suspected radius of the suspected tab edge point and the fitted circle center, the likelihood of the tab is used as a weight, when the suspected tab edge point in the battery surface image and the suspected tab edge point in the protection board surface image are matched, the characteristic value is weighted when the sequence distance calculation is carried out, the distance between the edge points with large likelihood of the tab corresponding to the two images under the same gray value is ensured to be small, the probability of being identified is improved, omission of the tab characteristic point is avoided, the integrity and the accuracy of the matching of the battery and the protection board characteristics are ensured, the influence of noise on the matching result is avoided, the image processing efficiency is improved, the accurate position matching relation is obtained, and the follow-up automatic welding is more accurate.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart showing the steps of the welding positioning method of the battery and the protection plate based on computer vision.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following is a detailed description of the specific implementation, structure, characteristics and effects of the welding positioning method for the battery and the protection plate based on computer vision according to the invention, which are provided by the invention, with reference to the accompanying drawings and the preferred embodiment.
In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same.
Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following specifically describes a specific scheme of the welding positioning method of the battery and the protection board based on computer vision provided by the invention with reference to the accompanying drawings.
Referring to fig. 1, a flowchart of a method for welding and positioning a battery and a protection board based on computer vision according to an embodiment of the invention is shown, the method includes the following steps:
s001, acquiring a battery surface image and a protection plate surface image.
The battery is vertically placed on a welding workbench, a camera is erected above the welding workbench, and the RGB images of the battery surface and the RGB images of the surface of the protection board in the corresponding model on the welding workbench are shot through the camera.
In order to facilitate the subsequent calculation, the battery surface RGB image and the protection plate surface RGB image are converted into grayscale images, and are respectively recorded as a battery surface image and a protection plate surface image.
Thus, a battery surface image and a protection plate surface image are obtained.
S002, acquiring the suspected tab edge points of the battery surface image and the protection plate surface image, and the tab possibility of each suspected tab edge point.
It should be noted that, at present, a feature detection and matching method is generally used to weld and position a battery and a protection board, all feature points are extracted through a feature extraction algorithm and then matched to obtain matching pairs, all feature points need to be processed during feature matching, so that extremely large data redundancy is generated, the calculated amount in the processing process is excessively large, and time and resources are wasted.
The battery and the protection board are welded and positioned by a general method, the specific characteristics of the battery and the protection board are ignored, the processing method is lack of pertinence, and the processing result is inaccurate.
Because the tab is an important connection part when the battery and the protection plate are welded and the tab positions of the battery and the protection plate are circular, the embodiment of the invention screens out possible tab edge points through circular fitting, so that analysis is only carried out on the possible tab edge points when characteristic matching is carried out subsequently, the calculated amount is greatly reduced, specific characteristics are focused, and the accuracy of characteristic extraction is improved.
In the embodiment of the invention, edge detection is carried out on the battery surface image, and all edge points in the battery surface image and the gradient direction of each edge point are obtained.
The edge detection algorithm adopted by the embodiment of the invention is Canny edge detection, and in other embodiments, an operator can select the edge detection algorithm according to actual implementation conditions.
And similarly, acquiring all edge points in the surface image of the protection plate and the gradient direction of each edge point.
All edge points in the battery surface image and the protection plate surface image obtained by edge detection are all characteristic edge points in the image, and feature matching is directly carried out on all edge points in the battery surface image and the protection plate surface image, so that the calculation amount is large, the influence of other features is avoided, and the accuracy is low.
Because the tab is an important connection part when the battery and the protection plate are welded, and the battery tab and the protection plate tab have the same circular shape, edge points in the battery surface image and the protection plate surface image can be screened according to the characteristics, and possible tab edge points can be obtained, so that the characteristic matching can be carried out only on the possible tab edge points in the follow-up process, the calculated amount is reduced, and meanwhile, the matching accuracy is improved.
It should be further noted that, the gradient direction of the edge point is perpendicular to the edge direction of the current position on the edge where the edge is located, for the circular pattern, the gradient direction of the point on the edge of the circular pattern is perpendicular to the tangent line of the point, and the extension lines of the gradient directions of the point on the edge of the circular pattern all point to the circle center, i.e. the extension lines of the gradient directions of the point on the edge of the circular pattern all intersect at the circle center.
At this time, the circles in the battery surface image can be fitted according to the features of the circular patterns, so that the pixel points which are possibly the edges of the lugs are screened out.
In the embodiment of the invention, firstly, circular fitting is carried out on all edge points in the battery surface image, specifically:
and acquiring straight lines passing through each edge point in the battery surface image and in the gradient direction of each edge point as gradient extension lines of each edge point, and acquiring intersection points between every two gradient extension lines of all the edge points to form an intersection point set.
And carrying out mean shift clustering on all the intersections in the intersection set to obtain all the intersection categories.
Performing convex hull detection on all the intersections in each intersection category to obtain convex hull areas of each intersection category, and obtaining the circle center probability of each intersection category according to the convex hull areas:
，
wherein the method comprises the steps ofIs->Center probability of each intersection point category, +.>Is->The number of intersections included in the individual intersection categories;is->The areas of convex hull regions of the respective intersection categories; when->The greater the number of intersections included in the intersection categories, and the smaller the area of the convex hull region, the +.>The greater the density of the individual intersection categories, the greater the probability that the gradient extension lines of the edge points located on the tab edge in the battery surface image intersect at +.>The center of the tab is positioned at the first +.>The greater the likelihood of each intersection category.
Acquiring the intersection point category with the largest circle center probability as the circle center category, taking the mass centers of all intersection points in the circle center category as the circle centers of the lugs in the battery surface image, and marking as。
And taking the edge point of the gradient extension line passing through the intersection point in the circle center category as a suspected polar edge point.
It should be noted that, in the embodiment of the invention, through the intersection point of the gradient extension lines of each edge point, the circle center of the circle to be fitted is obtained according to the distribution of the intersection points, and all the feature points which are possibly the points on the edge of the circular tab are screened out according to the category of the intersection point where the circle center is located, so that the data volume is reduced and the integrity of the feature points is ensured compared with the feature matching of all the edge points in the battery surface image and the protection plate surface image.
Acquiring a suspected radius corresponding to each suspected tab edge point in the battery surface image:
，
wherein, the liquid crystal display device comprises a liquid crystal display device,is the>The suspected radiuses corresponding to the edge points of the suspected tabs; />The abscissa of the circle center of the lug in the battery surface image; />The ordinate of the circle center of the lug in the battery surface image; />Is the>The abscissa of each suspected polar edge point; />Is the>The ordinate of each suspected polar edge point.
Rounding the suspected radiuses corresponding to all the suspected tab edge points in the battery surface image, forming a suspected radius sequence by the rounded suspected radiuses, and obtaining the numerical value with the largest occurrence frequency in the suspected radius sequence as the radius of the tab in the battery surface image.
Therefore, the circle center and the radius of the lug in the battery surface image are obtained, circular fitting is realized, and the edge point of the suspected lug is obtained.
In the invention, the circle centers of the tabs are obtained through the distribution condition of the intersections between every two gradient extension lines of all edge points in the battery surface image, so as to obtain the edge points of the suspected tabs.
And obtaining the suspected radius corresponding to each suspected tab edge point by calculating the distance from each suspected tab edge point to the circle center of the tab, selecting the numerical value with the largest occurrence number in the suspected radius sequence as the radius of the tab in the battery surface image, and taking the distance relation between each suspected tab edge point and the circle center into account as much as possible so that most of the suspected tab edge points are on the fitted circle.
However, under the interference of noise and shooting angles, the electrode lugs in the battery surface image may not be perfect circles, so that part of electrode lug edge points are not on the fitted circles, and at the moment, the electrode lug possibility of each suspected electrode lug edge point needs to be obtained by combining the relation between the suspected electrode lug edge points and the fitted circles.
In the embodiment of the invention, the distance offset fluctuation degree of all suspected tab edge points in the battery surface image is firstly obtained:
，
wherein, the liquid crystal display device comprises a liquid crystal display device,the fluctuation degree of the distance offset of all the suspected tab edge points in the battery surface image is calculated; />Is the>The suspected radiuses corresponding to the edge points of the suspected tabs; />Radius of the lug in the battery surface image; />The number of suspected polar edge points in the battery surface image; when the suspected radius and the corresponding suspected radius of the edge point of the suspected tab areWhen the radius difference of the polar lugs is larger, the distance offset fluctuation degree is larger.
The electrode lug possibility of each suspected electrode lug edge point in the battery surface image is obtained according to the distance offset fluctuation degree:
，
wherein, the liquid crystal display device comprises a liquid crystal display device,is the>The possibility of the electrode lug of each suspected electrode lug edge point; />The fluctuation degree of the distance offset of all the suspected tab edge points in the battery surface image is calculated; />Is the>The suspected radiuses corresponding to the edge points of the suspected tabs; />Radius of the lug in the battery surface image; />Is an absolute value symbol; />Is a natural constant; the distance offset fluctuation degree reflects the average fluctuation level of the distance from the edge point of each suspected electrode lug to the circle center relative to the radius, and the radius of each suspected electrode lug is obtained by processing the suspected radius of the edge point of each suspected electrode lug by adopting a rounding method, so that the acceptable range of the distance from the edge point of each suspected electrode lug to the circle center is required to be given; if%>The difference between the distance from the edge point of each suspected electrode lug to the circle center and the radius is smaller than the distance deviation fluctuation degree, which can be considered as the +.>The probability of the edge points of each suspected tab on the fitted circle is high, and if the difference value and the distance deviation fluctuation degree are smaller, the first +.>The greater the likelihood that each suspected tab edge point is on the fitted circle; if%>The difference between the distance from the edge point of each suspected electrode lug to the circle center and the radius is larger than the distance deviation fluctuation degree, which can be considered as the +.>The probability of the edge points of each suspected electrode lug on the fitting circle is smaller, and the smaller the difference value is with the fluctuation degree of the distance offset, the +.>The more likely each suspected tab edge point is based on the less likely to fit to a circle.
And similarly, acquiring all suspected tab edge points in the surface image of the protection plate and the tab possibility of each suspected tab edge point.
Thus, the suspected tab edge points of the battery surface image and the protection plate surface image and the tab possibility of each suspected tab edge point are obtained.
S003, matching the tab edge points in the battery surface image and the protection plate surface image according to the tab possibility of the suspected tab edge points.
It should be noted that, the battery tab and the tab position of the protection plate have the same color characteristics, and at this time, the suspected tab edge points in the battery surface image and the suspected tab edge points in the protection plate surface image can be matched according to the gray sequence of all the suspected tab edge points.
However, the suspected tab edge points in the battery surface image and the protection plate surface image may also be noise points, and if the matching is directly performed, the matching result may be inaccurate under the influence of the noise points.
The likelihood of the edge point of the suspected electrode tab reflects the likelihood that the edge point of the suspected electrode tab is an actual edge point of the electrode tab, so that the likelihood of the electrode tab can be used as a weight, when the edge point of the suspected electrode tab in the battery surface image and the edge point of the suspected electrode tab in the protection board surface image are matched, the characteristic value is weighted when the sequence distance calculation is carried out, the distance between the edge points with large likelihood of the electrode tab corresponding to the two images under the same gray value is ensured to be small, the identified probability of the electrode tab is improved, omission of the characteristic point of the electrode tab is avoided, the integrity and the accuracy of the characteristic matching of the battery and the protection board are ensured, the influence of noise on the matching result is avoided, the image processing efficiency is improved, the accurate position matching relation is obtained, and the follow-up automatic welding is more accurate.
In the embodiment of the invention, starting from the suspected tab edge point with the highest possibility of the tab in the battery surface image, acquiring all the suspected tab edge points in a clockwise direction, forming a suspected tab sequence of the battery surface image by all the suspected tab edge points, and acquiring the suspected tab sequence of the protection board surface image in a similar way.
Obtaining the distance between the suspected tab sequence of the battery surface image and each suspected tab edge point in the suspected tab sequence of the protection plate surface image:
，
wherein, the liquid crystal display device comprises a liquid crystal display device,the suspected tab sequence which is the image of the battery surface is +.>The suspected electrode tab edge points and the suspected electrode tab sequence of the surface image of the protection board/>The distance between the edge points of the suspected electrode lugs; />The suspected tab sequence which is the image of the battery surface is +.>The possibility of the electrode lug of each suspected electrode lug edge point; />The suspected tab sequence which is the surface image of the protection boardThe possibility of the electrode lug of each suspected electrode lug edge point; />The suspected tab sequence which is the image of the battery surface is +.>Gray values of the edge points of the suspected electrode lugs; />The suspected tab sequence for protecting the plate surface image is +.>Gray values of the edge points of the suspected electrode lugs; when the gray levels of two suspected tab edge points in the battery surface image and the protection plate surface image are more similar and the possibility of the tabs of the two suspected tab edge points is higher, the distance between the two suspected tab edge points is smaller, and otherwise, the distance between the two suspected tab edge points is larger.
And constructing an accumulated distance matrix according to the distance between the suspected tab sequence of the battery surface image and each suspected tab edge point in the suspected tab sequence of the protection plate surface image through a DTW algorithm, searching a path from the lower left corner to the upper right corner of the matrix to enable elements on the path to be minimum, and obtaining a characteristic point matching pair according to the elements and the minimum path to obtain the corresponding relation between the suspected tab edge point in the battery surface image and the suspected tab edge point in the protection plate surface image.
It should be noted that, constructing the cumulative distance matrix, searching a path from the lower left corner to the upper right corner of the matrix, so that the sum of elements on the path is the smallest, and obtaining the feature point matching pair according to the element and the smallest path is a known technique in the DTW algorithm, which is not described in detail herein.
And if one suspected tab edge point in the battery surface image corresponds to a plurality of suspected tab edge points in the protection plate surface image, taking the suspected tab edge point in the battery surface image as a point to be matched, respectively taking the corresponding plurality of suspected tab edge points in the protection plate surface image as candidate matching points, taking the candidate matching point with the smallest distance with the point to be matched as the actually corresponding suspected tab edge point in the protection plate surface image, and releasing the corresponding relation between the rest candidate matching points and the point to be matched.
Similarly, if one suspected tab edge point in the surface image of the protection plate corresponds to a plurality of suspected tab edge points in the surface image of the battery, the suspected tab edge point in the surface image of the protection plate is taken as a point to be matched, the corresponding plurality of suspected tab edge points in the surface image of the battery are respectively taken as candidate matching points, the candidate matching point with the smallest distance with the point to be matched is taken as the suspected tab edge point actually corresponding to the point to be matched in the surface image of the battery, and the corresponding relation between the rest candidate matching points and the point to be matched is relieved.
If one suspected tab edge point in the battery surface image does not correspond to any suspected tab edge point in the protection plate surface image, the suspected tab edge point in the battery surface image is taken as a noise point and is removed from a suspected tab sequence of the battery surface image.
Similarly, if one suspected tab edge point in the surface image of the protection plate does not correspond to any suspected tab edge point in the surface image of the battery, the suspected tab edge point in the surface image of the protection plate is taken as a noise point and is removed from the suspected tab sequence of the surface image of the protection plate.
And taking the rest of the suspected tab edge points in the suspected tab sequence of the battery surface image as the tab edge points in the battery surface image, and taking the rest of the suspected tab edge points in the suspected tab sequence of the protective plate surface image as the tab edge points in the protective plate surface image.
At this time, the tab edge points in the battery surface image correspond to the tab edge points in the protection plate surface image one by one.
Therefore, the matching of the lug edge points of the battery surface image and the protection plate surface image is realized.
S004, positioning results are generated based on the corresponding relation of the tab edge points, and automatic welding of the battery and the protection plate is achieved.
According to the corresponding relation between the battery surface image and the tab edge point in the protection plate surface image, the positions of the battery and the protection plate are estimated and calculated through a mathematical model, a positioning result is generated, the positioning result is fed back to a control system, and the welding position and the technological parameters of the welding equipment are controlled, so that the automatic positioning and welding of the battery and the protection plate are realized.
Through the steps, the automatic positioning and welding of the battery and the protection plate are completed.
According to the embodiment of the invention, circular fitting is carried out according to the gradient directions of all edge points in the battery surface image and the protection plate surface image, so that suspected tab edge points are obtained, tab characteristics of the battery and the protection plate during welding are only reserved, the data redundancy of the matching of the subsequent characteristic points is reduced, and the calculated amount is greatly reduced; according to the method, the likelihood of the tab of the suspected tab edge point is obtained according to the suspected radius of the suspected tab edge point and the fitted circle center, the likelihood of the tab is used as a weight, when the suspected tab edge point in the battery surface image and the suspected tab edge point in the protection board surface image are matched, the characteristic value is weighted when the sequence distance calculation is carried out, the distance between the edge points with large likelihood of the tab corresponding to the two images under the same gray value is ensured to be small, the probability of being identified is improved, omission of the tab characteristic point is avoided, the integrity and the accuracy of the matching of the battery and the protection board characteristics are ensured, the influence of noise on the matching result is avoided, the image processing efficiency is improved, the accurate position matching relation is obtained, and the follow-up automatic welding is more accurate.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, alternatives, and improvements that fall within the spirit and scope of the invention.
Claims (6)
1. The welding positioning method of the battery and the protection board based on computer vision is characterized by comprising the following steps of:
collecting a battery surface image and a protection plate surface image;
performing edge detection on the battery surface image to obtain all edge points in the battery surface image and gradient directions of all edge points; acquiring gradient extension lines of each edge point according to the gradient direction of each edge point, and acquiring intersection points between every two gradient extension lines of all edge points to form an intersection point set; clustering the intersection point sets to obtain a plurality of intersection point categories; the circle center probability of each intersection point category is obtained; acquiring the intersection point category with the maximum circle center probability as a circle center category, and acquiring the circle center of the lug and the edge point of the suspected lug in the battery surface image according to the circle center category; taking the Euclidean distance from each suspected tab edge point to the circle center as the suspected radius corresponding to each suspected tab edge point, and acquiring the radius of the tab in the battery surface image according to all the suspected radii; acquiring the distance offset fluctuation degree of all suspected tab edge points in the battery surface image; acquiring the tab probability of each suspected tab edge point in the battery surface image according to the distance offset fluctuation degree and the suspected radius;
acquiring the tab probability of all suspected tab edge points and each suspected tab edge point in the surface image of the protection plate;
acquiring a suspected tab sequence of the battery surface image and a suspected tab sequence of the protection plate surface image; acquiring the distance between a suspected tab sequence of the battery surface image and each suspected tab edge point in the suspected tab sequence of the protection plate surface image according to the tab probability; acquiring a corresponding relation between a suspected tab edge point in the battery surface image and a suspected tab edge point in the protection plate surface image according to the distance;
acquiring the corresponding relation between the battery surface image and the tab edge point in the protection plate surface image according to the corresponding relation between the battery surface image and the suspected tab edge point in the protection plate surface image; welding and positioning are carried out according to the corresponding relation between the battery surface image and the tab edge point in the protection plate surface image, so that automatic welding is realized;
the method for acquiring the distance offset fluctuation degree of all suspected tab edge points in the battery surface image comprises the following specific steps:
；
the distance offset fluctuation degree is the distance offset fluctuation degree of all suspected tab edge points in the battery surface image; the method comprises the steps of obtaining a suspected radius corresponding to a first suspected tab edge point in a battery surface image; radius of the lug in the battery surface image; the number of suspected polar edge points in the battery surface image;
the method for acquiring the tab probability of each suspected tab edge point in the battery surface image according to the distance offset fluctuation degree and the suspected radius comprises the following specific steps:
；
the electrode lug possibility of the first suspected electrode lug edge point in the battery surface image is obtained; the fluctuation degree of the distance offset of all the suspected tab edge points in the battery surface image is calculated; the method comprises the steps of obtaining a suspected radius corresponding to a first suspected tab edge point in a battery surface image; radius of the lug in the battery surface image; is an absolute value symbol; is a natural constant;
the method for acquiring the suspected tab sequence of the battery surface image and the suspected tab sequence of the protection plate surface image comprises the following specific steps:
starting from the suspected electrode edge point with the highest possibility of the electrode lug in the battery surface image, acquiring all the suspected electrode edge points in a clockwise direction, and forming a suspected electrode lug sequence of the battery surface image by all the suspected electrode edge points; similarly, a suspected tab sequence of the surface image of the protection board is obtained;
the method for acquiring the distance between each suspected tab edge point in the suspected tab sequence of the battery surface image and the suspected tab sequence of the protection plate surface image according to the tab possibility comprises the following specific steps:
；
the distance between the edge point of the first suspected electrode tab in the suspected electrode tab sequence of the battery surface image and the edge point of the second suspected electrode tab in the suspected electrode tab sequence of the protective plate surface image is the distance; the potential of the tab is the edge point of the first suspected tab in the suspected tab sequence of the battery surface image; the possibility of the electrode lug of the edge point of the first suspected electrode lug in the suspected electrode lug sequence of the surface image of the protection plate; the gray value of the edge point of the first suspected electrode lug in the suspected electrode lug sequence of the battery surface image; the gray value of the edge point of the first suspected electrode lug in the suspected electrode lug sequence of the surface image of the protection plate; the number of suspected polar edge points in the battery surface image; is the number of suspected polar edge points in the protective plate surface image.
2. The welding positioning method for the battery and the protection plate based on computer vision according to claim 1, wherein the step of obtaining the gradient extension line of each edge point according to the gradient direction of each edge point comprises the following specific steps:
a straight line passing through each edge point and in the gradient direction of each edge point in the battery surface image is acquired as a gradient extension line of each edge point.
3. The welding positioning method of the battery and the protection board based on computer vision according to claim 1, wherein the obtaining the circle center probability of each intersection point category comprises the following specific steps:
performing convex hull detection on all the intersections in each intersection category to obtain convex hull areas of each intersection category; and taking the ratio of the number of the intersections contained in each intersection category to the area of the convex hull area as the circle center probability of each intersection category.
4. The welding positioning method for the battery and the protection plate based on computer vision according to claim 1, wherein the step of obtaining the circle center of the tab and the edge point of the suspected tab in the battery surface image according to the circle center category comprises the following specific steps:
taking the centroids of all the intersection points in the circle center category as the circle centers of the lugs in the battery surface image; and taking the edge point of the gradient extension line passing through the intersection point in the circle center category as a suspected polar edge point.
5. The welding positioning method of the battery and the protection plate based on computer vision according to claim 1, wherein the step of obtaining the radius of the tab in the battery surface image according to all suspected radii comprises the following specific steps:
rounding the suspected radiuses corresponding to all the suspected tab edge points in the battery surface image, forming a suspected radius sequence by the rounded suspected radiuses, and obtaining the numerical value with the largest occurrence frequency in the suspected radius sequence as the radius of the tab in the battery surface image.
6. The welding positioning method for a battery and a protection plate based on computer vision according to claim 1, wherein the obtaining of the correspondence between the battery surface image and the tab edge point in the protection plate surface image comprises the following specific steps:
updating the corresponding relation of the suspected tab edge points in the battery surface image comprises the following steps: if one suspected tab edge point in the battery surface image corresponds to a plurality of suspected tab edge points in the protection plate surface image, the suspected tab edge point in the battery surface image is taken as a point to be matched, a plurality of corresponding suspected tab edge points in the protection plate surface image are respectively taken as candidate matching points, a candidate matching point with the smallest distance from the point to be matched is taken as a suspected tab edge point actually corresponding to the point to be matched in the protection plate surface image, and the corresponding relation between the rest candidate matching points and the point to be matched is relieved;
similarly, updating the corresponding relation of the suspected tab edge points in the surface image of the protection plate;
acquiring a tab edge point in a battery surface image, comprising: if one suspected tab edge point in the battery surface image does not correspond to any suspected tab edge point in the protection plate surface image, taking the suspected tab edge point in the battery surface image as a noise point, and removing the suspected tab edge point from a suspected tab sequence of the battery surface image; taking the rest suspected electrode lug edge points in the suspected electrode lug sequence of the battery surface image as electrode lug edge points in the battery surface image;
similarly, the tab edge points in the surface image of the protection plate are obtained, and the tab edge points in the surface image of the battery are in one-to-one correspondence with the tab edge points in the surface image of the protection plate.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310545421.1A CN116309848B (en) | 2023-05-16 | 2023-05-16 | Welding positioning method of battery and protection board based on computer vision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310545421.1A CN116309848B (en) | 2023-05-16 | 2023-05-16 | Welding positioning method of battery and protection board based on computer vision |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116309848A CN116309848A (en) | 2023-06-23 |
| CN116309848B true CN116309848B (en) | 2023-08-01 |
Family
ID=86803449
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310545421.1A Active CN116309848B (en) | 2023-05-16 | 2023-05-16 | Welding positioning method of battery and protection board based on computer vision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116309848B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115272168A (en) * | 2022-05-19 | 2022-11-01 | 广州超音速自动化科技股份有限公司 | Pole piece quality detection method, system, equipment and storage medium |
| CN115837541A (en) * | 2022-11-15 | 2023-03-24 | 宁德时代新能源科技股份有限公司 | Battery cell pole positioning method and device, welding method and device, equipment and medium |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106815834B (en) * | 2016-12-25 | 2021-04-16 | 惠州市德赛电池有限公司 | Welding positioning method for battery and battery protection plate |
| US11890705B2 (en) * | 2018-11-13 | 2024-02-06 | Rivian Ip Holdings, Llc | Weld location detection in a battery cell pack |
| CN109697732B (en) * | 2018-11-27 | 2023-10-13 | 惠州市德赛电池有限公司 | Method for positioning resistance welding area |
-
2023
- 2023-05-16 CN CN202310545421.1A patent/CN116309848B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115272168A (en) * | 2022-05-19 | 2022-11-01 | 广州超音速自动化科技股份有限公司 | Pole piece quality detection method, system, equipment and storage medium |
| CN115837541A (en) * | 2022-11-15 | 2023-03-24 | 宁德时代新能源科技股份有限公司 | Battery cell pole positioning method and device, welding method and device, equipment and medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116309848A (en) | 2023-06-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112949564B (en) | Pointer type instrument automatic reading method based on deep learning | |
| CN114897864B (en) | Workpiece detection and defect judgment method based on digital-analog information | |
| CN109615609A (en) | A kind of solder joint flaw detection method based on deep learning | |
| CN113205063A (en) | Visual identification and positioning method for defects of power transmission conductor | |
| CN110378227B (en) | Method, device and equipment for correcting sample labeling data and storage medium | |
| CN109949227A (en) | Image split-joint method, system and electronic equipment | |
| CN105741281B (en) | Method for detecting image edge based on neighborhood dispersion | |
| CN115601347A (en) | Steel plate surface defect detection method based on gray texture analysis | |
| CN115861307B (en) | Fascia gun power supply driving plate welding fault detection method based on artificial intelligence | |
| CN112819748B (en) | Training method and device for strip steel surface defect recognition model | |
| CN115082477B (en) | Semiconductor wafer processing quality detection method based on light reflection removing effect | |
| CN112669295A (en) | Lithium battery pole piece defect detection method based on secondary threshold segmentation theory | |
| CN113554582A (en) | Defect detection method, device and system for functional hole in cover plate of electronic equipment | |
| CN110619623A (en) | Automatic identification method for heating of joint of power transformation equipment | |
| CN112001244A (en) | Computer-aided iris comparison method and device | |
| CN115018846A (en) | AI intelligent camera-based multi-target crack defect detection method and device | |
| CN112884746A (en) | Character defect intelligent detection algorithm based on edge shape matching | |
| CN115082922A (en) | Water meter digital picture processing method and system based on deep learning | |
| CN113705564B (en) | Pointer type instrument identification reading method | |
| CN117058411B (en) | Method, device, medium and equipment for identifying edge appearance flaws of battery | |
| CN116309848B (en) | Welding positioning method of battery and protection board based on computer vision | |
| CN108428247B (en) | Method and system for detecting direction of soldering tin point | |
| CN116664579B (en) | Battery pack welding quality image detection method | |
| CN113781413A (en) | Electrolytic capacitor positioning method based on Hough gradient method | |
| CN116363136B (en) | On-line screening method and system for automatic production of motor vehicle parts |
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 |