CN115526860A - Battery piece defect detection method and device, computer equipment and readable storage medium - Google Patents

Battery piece defect detection method and device, computer equipment and readable storage medium Download PDF

Info

Publication number
CN115526860A
CN115526860A CN202211201816.1A CN202211201816A CN115526860A CN 115526860 A CN115526860 A CN 115526860A CN 202211201816 A CN202211201816 A CN 202211201816A CN 115526860 A CN115526860 A CN 115526860A
Authority
CN
China
Prior art keywords
defect
local
battery piece
target
determining
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202211201816.1A
Other languages
Chinese (zh)
Inventor
曹建伟
朱亮
傅林坚
刘华
文灿华
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhejiang Qiushi Semiconductor Equipment Co Ltd
Original Assignee
Zhejiang Qiushi Semiconductor Equipment Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Zhejiang Qiushi Semiconductor Equipment Co Ltd filed Critical Zhejiang Qiushi Semiconductor Equipment Co Ltd
Priority to CN202211201816.1A priority Critical patent/CN115526860A/en
Publication of CN115526860A publication Critical patent/CN115526860A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
    • 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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/70—Determining position or orientation of objects or cameras
    • G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/766—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using regression, e.g. by projecting features on hyperplanes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
    • G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
    • 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
    • G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/84—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using probabilistic graphical models from image or video features, e.g. Markov models or Bayesian networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/84—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using probabilistic graphical models from image or video features, e.g. Markov models or Bayesian networks
    • G06V10/85—Markov-related models; Markov random fields
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20084—Artificial neural networks [ANN]
    • 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

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Computing Systems (AREA)
  • Medical Informatics (AREA)
  • General Health & Medical Sciences (AREA)
  • Databases & Information Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Multimedia (AREA)
  • Probability & Statistics with Applications (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Quality & Reliability (AREA)
  • Geometry (AREA)
  • Secondary Cells (AREA)

Abstract

The application relates to a battery piece defect detection method and device, computer equipment and a readable storage medium. The method comprises the following steps: acquiring a battery piece image to be detected; inputting the battery piece image to be detected into a first defect detection model to obtain a main defect, wherein the main defect comprises a continuous defect area; inputting the image of the battery piece to be detected into a second defect detection model to obtain a local defect, wherein the local defect comprises at least one defect source point, and the defect source point comprises a defect end point or a defect intersection point; determining a target defect based on the primary defect and the local defect. According to the battery piece defect detection method provided by the embodiment of the application, on one hand, omission or errors of a main defect detection result are complemented or corrected through local defects; on the other hand, the local defects can more accurately express the positions of induced defects, so that an important reference basis is provided for analyzing the causes of the defects of the battery piece, and the accuracy and the detection efficiency of the defect detection of the battery piece are effectively improved.

Description

Battery piece defect detection method and device, computer equipment and readable storage medium
Technical Field
The present disclosure relates to the field of battery cell detection technologies, and in particular, to a battery cell defect detection method and apparatus, a computer device, and a readable storage medium.
Background
In recent years, as the overuse of non-renewable energy sources such as petroleum and coal and the environmental problems caused by the overuse of the non-renewable energy sources become more severe, new energy technology is emphasized and rapidly developed, while solar photovoltaic power generation technology is a typical representative of the new energy technology, and the wide application of the technology can well alleviate the energy problems encountered at present. However, the production process and flow of the photovoltaic cell are complex, and each production link may damage the cell to cause different cell defects.
In the conventional battery piece defect detection technology, due to the fact that different defect sizes and shapes of battery pieces are different, a data set for training a defect detection model is difficult to cover all defects, and a local area of the defect is actually the defect, so that a final detection result is possibly positioned in the local area of the defect, a regression effect of a detection algorithm is poor, and the accuracy of a defect detection result is low. Therefore, there is a need in the art for a detection technique that can improve the accuracy of detecting defects in a battery cell.
Disclosure of Invention
In view of the foregoing, it is desirable to provide a method, an apparatus, a computer device and a readable storage medium for detecting defects of a battery cell, which can improve the accuracy of detecting defects of the battery cell.
In a first aspect, the application provides a battery piece defect detection method. The method comprises the following steps:
acquiring a battery piece image to be detected;
inputting the battery piece image to be detected into a first defect detection model to obtain a main defect, wherein the main defect comprises a continuous defect area;
inputting the battery piece image to be detected into a second defect detection model to obtain a local defect, wherein the local defect comprises at least one defect source point, and the defect source point comprises a defect endpoint or a defect intersection point;
determining a target defect based on the primary defect and the local defect.
In one embodiment, the determining the target defect based on the main defect and the local defect comprises:
determining a main defect area of the main defect and a local defect area of the local defect;
determining the target defect based on the coincidence degree of the main defect region and the local defect region.
In one embodiment, the determining the target defect based on the coincidence degree of the main defect region and the local defect region includes:
if the contact ratio is larger than zero, determining the target defect based on the main defect area and the local defect;
and if the contact ratio is not greater than zero, determining the target defect based on the main defect.
In one embodiment, the determining the target defect based on the local defect comprises:
if the local defect is oblique hidden crack and the local defect comprises at least two oblique hidden crack source points, judging the position of the oblique hidden crack source points in the image of the cell to be detected;
if the oblique hidden crack source points are all internal end points, determining the main defect area, the oblique hidden crack and all the internal end points as the target defects;
and if the oblique hidden crack source points comprise at least one boundary endpoint, determining the main defect area, the oblique hidden crack and all the boundary endpoints as the target defects.
In one embodiment, the determining the target defect based on the local defect comprises:
and if the local defects comprise cross hidden cracks and oblique hidden cracks, determining the main defect region, the cross hidden cracks and all cross hidden crack source points as the target defects.
In one embodiment, after the determining the target defect based on the main defect and the local defect, the method further includes:
determining the size ratio of the image of the battery piece to be detected and the battery piece to be detected;
and determining the actual distribution of the target defects on the battery piece to be detected based on the position information of the target defects in the battery piece image to be detected and the size proportion.
In one embodiment, after determining the target defect based on the main defect and the local defect, the method further includes:
determining a comparison result of the target defect and a preset defect;
determining unusable defective battery pieces based on the comparison result, and removing the defective battery pieces.
In a second aspect, the application further provides a device for detecting defects of the battery piece. The device comprises:
the image acquisition module is used for acquiring an image of the battery piece to be detected;
the first defect detection module is used for inputting the battery piece image to be detected into a first defect detection model to obtain a main defect, and the main defect comprises a continuous defect area;
the second defect detection module is used for inputting the battery piece image to be detected into a second defect detection model to obtain a local defect, wherein the local defect comprises at least one defect point, and the defect point comprises a defect endpoint or a defect intersection point;
a defect determination module to determine a target defect based on the primary defect and the local defect.
In a third aspect, the present application further provides a computer device, which includes a memory and a processor, where the memory stores a computer program, and the processor implements the steps of the method described in any one of the above first aspects when executing the computer program.
In a fourth aspect, the present application also provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the method of any of the first aspects described above.
According to the battery piece defect detection method, the battery piece defect detection device, the computer equipment and the readable storage medium, the main defect and the local defect are obtained through the first detection model and the second detection model, and the target defect is determined based on the main defect and the local defect, so that on one hand, the internal relation between the main defect characteristic and the local defect characteristic can be fully considered, and the omission or the error of the main defect detection result can be complemented or corrected through the local defect; on the other hand, the local defects can more accurately express the positions of induced defects, important reference basis is provided for analyzing the causes of the defects of the battery piece, the links causing the defects in the production flow can be more quickly positioned, and further loss possibly caused in the generation of the battery piece is reduced. According to the battery piece defect detection method, the association between the main defect characteristics and the local defects is fully considered, so that the accuracy and the detection efficiency of the battery piece defect detection are effectively improved.
The details of one or more embodiments of the application are set forth in the accompanying drawings and the description below to provide a more thorough understanding of the application.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the application and together with the description serve to explain the application and not to limit the application. In the drawings:
FIG. 1 is a schematic flow chart illustrating a method for detecting defects in a battery cell according to an embodiment;
FIG. 2 is a schematic diagram of an embodiment of a full cell sheet cut into 6 cell sheet pieces;
FIG. 3 is a schematic diagram of a cell under test in one embodiment;
FIG. 4 is a schematic diagram of a cell slice with global and local defects in one embodiment;
FIG. 5 is a schematic diagram of a crack defect in a cell sheet under oblique conditions;
FIG. 6 is a schematic diagram of a bevel latent crack defect of a cell plate according to another embodiment;
FIG. 7 is a schematic diagram of a cross-hatch defect in one embodiment;
FIG. 8 is a schematic diagram of a defect distribution of a cell in one embodiment;
FIG. 9 is a schematic diagram illustrating a process for detecting a battery cell in one embodiment;
FIG. 10 is a block diagram of a device for detecting defects in a battery cell according to an embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
Unless defined otherwise, technical or scientific terms used herein shall have the same general meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The use of the terms "a" and "an" and "the" and similar referents in the context of this application do not denote a limitation of quantity, either in the singular or the plural. The terms "comprises," "comprising," "has," "having," and any variations thereof, as referred to in this application, are intended to cover non-exclusive inclusions; for example, a process, method, and system, article, or apparatus that comprises a list of steps or modules (elements) is not limited to the listed steps or modules, but may include other steps or modules (elements) not listed or inherent to such process, method, article, or apparatus. Reference in this application to "connected," "coupled," and the like is not intended to be limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. Reference to "a plurality" in this application means two or more. "and/or" describes the association relationship of the associated object, indicating that there may be three relationships, for example, "a and/or B" may indicate: a exists alone, A and B exist simultaneously, and B exists alone. In general, the character "/" indicates a relationship in which the objects associated before and after are an "or". Reference in the present application to the terms "first," "second," "third," etc., merely distinguish between similar objects and do not denote a particular order or importance to the objects.
The terms "module," "unit," and the like as used hereinafter are combinations of software and/or hardware that can achieve the intended functionality. Although the means described in the embodiments below are preferably implemented in hardware, an implementation in software, or a combination of software and hardware is also possible and contemplated.
In an embodiment, as shown in fig. 1, a method for detecting a defect of a battery piece is provided, and this method is applied to a terminal for example, it may be understood that the method may also be applied to a server, and may also be applied to a system including a terminal and a server, and is implemented through interaction between the terminal and the server. In this embodiment, the method includes the steps of:
s201: and acquiring an image of the battery piece to be detected.
The cell slice in the application refers to a solar cell slice such as monocrystalline silicon or polycrystalline silicon applied to photovoltaic power generation. In this embodiment of the application, the image of the battery piece to be detected may include an image of the whole battery piece, and may also include an image of a battery piece obtained by cutting the whole battery piece. Specifically, in some embodiments, the cell to be detected may include a whole collected image of the 210 × 210mm cell. If the whole cell piece is averagely cut into 6 pieces of 210 × 35mm cell pieces, the cell piece to be detected can also comprise the collected image of at least one cell piece. As shown in fig. 2, the whole battery piece may be divided into 1-6 battery pieces, and an image of each battery piece is obtained as the image of the battery piece to be detected, and fig. 3 is an image of the battery piece to be detected obtained in one embodiment, which is an image of the battery piece after being divided. As can be appreciated, after determining the defects of each battery piece, the defect distribution of the whole battery piece can be obtained.
In the embodiment of the application, based on the characteristic that the battery piece can emit light when being irradiated by some light sources, and because the light emitting characteristics of the defect position are greatly different from those of the normal position when the battery piece is irradiated by the light sources, the defect condition of the battery piece can be determined by acquiring the image of the battery piece irradiated by the light sources. The acquiring of the image of the battery piece to be detected comprises the steps of irradiating the battery piece by using a light source and then scanning by using camera equipment to acquire the image of the battery piece to be detected. The light source comprises an LED light source or a laser light source, and the image pickup equipment can comprise a TDI industrial line scan digital camera and the like. The battery piece is irradiated by an external light source, so that the battery piece is not influenced, and the TDI industrial linear array camera has high sensitivity and good imaging quality.
In some embodiments of the application, a space of about 0.5 m can be reserved on a production line of the battery piece for installing an LED light source, a TDI industrial linear array camera and an industrial personal computer, an external encoder is installed on a motor of a conveyor belt of the production line, the encoder can generate a trigger signal according to a preset absolute distance of movement of the conveyor belt and send the trigger signal to the TDI camera, and the TDI camera receives the trigger signal and then obtains an image of the battery piece to be detected. The installation space of about 0.5 m is reserved, so that the defect detection model in the embodiment of the application can obtain the buffer time of about 1 second, the reliability of defect detection is enhanced, and the normal operation of a production line cannot be influenced even if the accidental detection delay occurs. On the other hand, an external encoder is arranged on the conveyor belt motor, and a trigger signal can be generated according to the preset absolute distance of the movement of the conveyor belt, so that the conveyor belt can obtain consistent battery piece images at different transmission speeds, and the image stretching or compression caused by the fixed line frequency trigger mode when the speed of the conveyor belt changes is avoided.
S203: and inputting the image of the battery piece to be detected into a first defect detection model to obtain a main defect, wherein the main defect comprises a continuous defect area.
In the embodiment of the application, the first defect detection model can be used for identifying and matching defects of the input battery piece to be detected and outputting the main defects based on the characteristics of the defects. According to different training modes of the first defect detection model, the obtaining of the main defect may include marking the main defect in the image of the battery piece to be detected, or may include outputting information of the main defect, or may be outputting the main defect in a combination of the two modes. In an embodiment of the present application, the main defect includes a continuous defect region, where the continuous defect region includes a region having continuous defect characteristics formed by the same defect, and the defect characteristics include a defect line or a defect plane formed by the defect. Specifically, the continuous defect region may include a whole continuous hidden crack defect line or a whole continuous black spot defect surface. In some embodiments, the continuous defect region may further include a circumscribed rectangular frame of the continuous defect region. In other embodiments, the primary defect may further include a primary defect type and/or a primary defect source point. The main defect type may include oblique hidden crack, cross hidden crack, etc., and the main defect source point may include a defect end point or a defect intersection point.
In the embodiment of the application, after the image of the battery piece to be detected is obtained, a training data set of a first defect detection model is established, and the first defect detection model is obtained based on training of the training data set. In some embodiments, after the battery piece image to be detected is obtained, a qualified sample image and a defective sample image can be further obtained, and data enhancement operations such as rotation, mirroring, turning, gray scale conversion and the like are performed on the sample image. The defects of the battery piece are correspondingly distinguished and the defect types are determined, and the different defect types also correspond to different defect source points, such as crossing hidden crack defects, wherein crossing points are used as defect source points rather than end points.
In some embodiments, the continuous defect region is labeled as the defect sample image is processed. In other embodiments, the defect type and defect source point may also be marked in the defect sample image.
In this embodiment of the present application, the first defect detection model may include a model component obtained by training in a machine learning manner, and may also be a combination of multiple models or calculation engines, which is not limited in this application. The machine learning mode may include a K-nearest neighbor algorithm, a perceptron algorithm, a support vector machine, a logistic-gaussian-based regression, a maximum entropy, and the like, and correspondingly, the generated model components may be naive bayes, hidden markov, and the like. Of course, in other embodiments, the machine learning manner may further include a deep learning manner, a reinforcement learning manner, and the like, and the generated model component may include a Convolutional Neural Network model Component (CNN), a Recurrent Neural Network model component (RNN), leNet, resNet, a Long-Short Term Memory Network model component (LSTM), a bidirectional Long-Short Term Memory Network model component (Bi-LSTM), and the like, which is not limited herein.
S205: and inputting the battery piece image to be detected into a second defect detection model to obtain a local defect, wherein the local defect comprises at least one defect source point, and the defect source point comprises a defect endpoint or a defect intersection point.
In the conventional battery piece defect detection technology, the sizes and forms of different defects of a battery piece are different, for example, the length of a subfissure defect and the form of a curve are different, and a local part of the defect is actually the defect. Because the data set for training the defect detection model is difficult to cover all defects, the final detection result is possibly positioned in the local region of the defects, so that the regression effect of the detection algorithm is poor, and the accuracy of the result of the defect detection is low. As shown in fig. 4, the actual defect of the battery piece is a region 11, a region 13 is a defect detection result of the detection model, the region 13 corresponds to a local defect region 12 of the actual defect, and the detection model does not detect the actual defect region 11. On the other hand, a local region of a certain type of defect may be another type of defect, for example, a local region of a cross hidden crack may be a cross hidden crack, and when a target region detected before the regression of the detection algorithm is located in the local region, the cross hidden crack is erroneously determined as a cross hidden crack, which results in a decrease in accuracy of defect detection. Moreover, the position of the defect in the battery piece is difficult to judge by the conventional battery piece defect detection model, and no matter the target detection algorithm or the semantic segmentation algorithm is difficult to have high regression accuracy, and as mentioned above, the training set is difficult to cover all the defects and the regression accuracy is also reduced, so that the defect detection model is also difficult to detect that the defect is located at the edge or inside of the battery piece.
Based on this, in the embodiment of the application, the image of the battery piece to be detected is input into the second defect detection model, so as to obtain the local defect. The second defect detection model can be used for identifying and matching the input local defects of the battery piece to be detected and outputting the local defects based on the characteristics of the local defects. The local defect includes at least one defect source point including a defect end point or a defect intersection point. The defect end point or the defect intersection point may include an inner point located inside the battery piece, and may also include an edge point located at an edge of the battery piece. In some embodiments, the local defect may further include a local continuous defect region including the defect source point, or a circumscribed rectangular frame of the local continuous defect region. In other embodiments, the local defect may further include a local defect type corresponding to the defect source point. The local defect types may include oblique subfissure, cross subfissure, and the like.
In the embodiment of the application, after the image of the battery piece to be detected is obtained, a training data set of a second defect detection model is established, and the second defect detection model is obtained based on the training of the training data set. In some embodiments, after the battery piece image to be detected is obtained, a qualified sample image and a local defect sample image can be further obtained, and data enhancement operations such as rotation, mirroring, turning, gray scale conversion and the like are performed on the sample image. In some embodiments, a local defect sample image containing at least one defect source point is screened out, and the defect source point is labeled, specifically, labeling the defect source point includes labeling a defect endpoint or a defect intersection point. In some embodiments, the labeling the defect source point further includes labeling a defect inner point or a defect edge point. In other embodiments, the corresponding defect type may also be labeled based on the defect source point. In other embodiments, a locally continuous defect region containing the defect source point may also be labeled.
In this embodiment of the application, the training mode of the second defect detection model may refer to the training mode of the first defect detection model in step S203, which is not described herein again.
S207: determining a target defect based on the primary defect and the local defect.
In the embodiment of the present application, the target defect may be determined based on the main defect and the local defect. The target defect includes the main defect and/or the local defect. Since the second detection model determines the local defect region or defect type based on the defect source point, the accuracy of the defect source point and defect type detection is higher than that of the first detection model. And comparing the local defects with the main defects output by the first detection model to obtain a comparison result, and determining target defects based on the comparison result. In some embodiments, the target defect comprises at least one of a target defect region, a target defect source point, and a target defect type, wherein the target defect region comprises a continuous defect region.
In some embodiments, the target defect may be determined based on a degree of coincidence of the main defect region and the local defect region. At least one defect source point is included in the local defect output by the second defect model, and the at least one defect source point may be located in the same defect or different defects, so whether the local defect is consistent with the main defect or not can be determined based on the coincidence degree. If the local defects are consistent with the target defects, determining the local defects as target defects, and also determining the main defect area as a target defect area; and if the main defects are inconsistent, determining the main defects as the target defects.
According to the battery piece defect detection method provided by the embodiment of the application, the main defect and the local defect are obtained through the detection of the first detection model and the second detection model, and the target defect is determined based on the main defect and the local defect, so that on one hand, the internal relation between the characteristics of the main defect and the characteristics of the local defect can be fully considered, and the omission or error of the detection result of the main defect is complemented or corrected through the local defect; on the other hand, the local defects can more accurately express the positions of induced defects, important reference basis is provided for analyzing the causes of the defects of the battery piece, the links causing the defects in the production flow can be more quickly positioned, and further loss possibly caused in the generation of the battery piece is reduced. According to the battery piece defect detection method, the association between the main defect characteristics and the local defects is fully considered, so that the accuracy and the detection efficiency of the battery piece defect detection are effectively improved.
In this embodiment of the present application, in order to determine whether the main defect and the local defect both belong to a target defect, a method for determining the target defect is further provided, and in step S207, the determining the target defect based on the main defect and the local defect includes:
s301: determining a main defect region of the main defect and a local defect region of the local defect.
S303: determining the target defect based on the coincidence degree of the main defect region and the local defect region.
In the embodiment of the application, a plurality of defects may exist in the same battery piece to be detected, and a plurality of defect regions may be marked when the first defect detection model outputs the main defect. The second defect detection model may also mark a plurality of defect source points when outputting local defects. Therefore, the target defect can be determined by determining a main defect region of the main defect and a local defect region of the local defect, and then determining the target defect based on the coincidence degree of the main defect region and the local defect region, that is, determining the main defect and the local defect which belong to the target defect. In some embodiments, if the coincidence degree of the main defect region and the local defect region is greater than a preset threshold, determining a defect region of the main defect and a defect source point of the local defect as the target defect; and if the contact ratio is not greater than a preset threshold value, determining the main defect as the target defect. In other embodiments, the determining the target defect based on the coincidence-degree of the main defect region and the local defect region in step S303 includes:
s3031: and if the coincidence degree is larger than zero, determining the target defect based on the main defect area and the local defect.
S3033: and if the contact ratio is not greater than zero, determining the target defect based on the main defect.
In this embodiment, the contact ratio may be determined by one or more of calculation methods for calculating an IOU (Intersection-Intersection unit) or various derivative concepts thereof, such as a DIOU (Distance-Intersection-over unit, distance-based Intersection ratio) and the like. In some embodiments of the present application, determining the main defect region of the main defect includes determining a circumscribed rectangular frame of the main defect region as a, and determining the local defect region of the local defect includes determining a circumscribed rectangular frame of the local defect region as B, and then determining the degree of overlap as an IOU, where the IOU may be determined by equation (1):
Figure BDA0003872672380000101
in the formula (1), | a ≧ B | is an area of a rectangle where the rectangular frame a intersects the rectangular frame B, and | a ≧ B | is an area of a rectangle circumscribed by the rectangular frame a and the rectangular frame B.
If the IOU is larger than 0, determining a defect region of the main defect and a defect source point of the local defect as the target defect, wherein the target defect can also comprise a defect type corresponding to the local defect source point; and if the IOU is less than or equal to 0, determining the main defect as the target defect.
In the embodiment of the application, the target defect is determined through the main defect area and the local defect area, the main defect and the local defect included in the target defect can be determined, the overall distribution of the target defect can be determined through the main defect, the defect source point of the target defect can be accurately determined through the target defect, and the type of the target defect can also be determined, so that the accuracy of detecting the defects of the battery piece can be further improved.
In order to further improve the detection accuracy of the defect source point of the battery piece, in this embodiment of the application, the determining the target defect by the local defect in step S207 includes:
s401: and if the local defect is oblique hidden crack and the local defect comprises at least two oblique hidden crack source points, judging the position of the oblique hidden crack source points in the image of the cell to be detected.
S403: and if the oblique hidden crack source points are all internal end points, determining the main defect area, the oblique hidden crack and all the internal end points as the target defects.
S405: and if the oblique hidden crack source points comprise at least one boundary endpoint, determining the main defect area, the oblique hidden crack and all the boundary endpoints as the target defects.
In the embodiment of the application, in the local defects output by the second defect detection model, if the local defects are oblique hidden cracks and include at least two oblique hidden crack source points, the defect source points causing the oblique hidden cracks can be further determined by judging the positions of the oblique hidden crack source points in the image of the battery piece to be detected. In a specific embodiment, as shown in fig. 5, the local defect source points include an inner point shown as 22 and an edge point shown as 23, and the edge point 23 is preferentially determined as the defect source point of the target defect. In addition, if the local defect corresponding to the edge point 23 is oblique hidden crack and the main defect area is 21, determining 21, 23 and oblique hidden crack as the target defect. In another embodiment, as shown in fig. 6, the determined target defect includes the oblique hidden crack source points 17, 17 output by the second detection model as the cell boundary end points. The target defect also includes a primary defect region 18 output by the first inspection model.
If a hidden boundary crack source point and an internal hidden crack source point exist in the oblique hidden crack of the cell, the position causing the oblique hidden crack is usually the hidden boundary crack source point. In the embodiment of the application, when the local defects detected by the second detection model include oblique hidden cracking source points of the boundary end points and oblique hidden cracking source points of the internal end points, the boundary end points are preferentially determined as the source points of the target defects, the source point positions of induced defects can be accurately positioned in various hidden cracking source points, and the accuracy of the defect detection result of the cell is further improved.
In order to further improve the detection accuracy of the defect type of the battery piece, in the embodiment of the present application, the determining the target defect by the local defect in step S207 includes:
s501: and if the local defects comprise cross hidden cracks and oblique hidden cracks, determining the main defect region, the cross hidden cracks and all cross hidden crack source points as the target defects.
In this embodiment, if the local defect output by the second defect detection model includes multiple defect types, the actual defect type of the target defect may be further determined. Because the local cross hidden crack defect of the cell is represented as the oblique hidden crack, if the local defect output by the second defect detection model comprises a cross hidden crack source point and an oblique hidden crack source point, determining that the defect type of the target defect is the cross hidden crack. In a specific embodiment, as shown in fig. 7, the local defects output by the second defect detection model include oblique hidden crack source point edge points 31, oblique hidden crack source point edge points 32, oblique hidden crack source point interior points 33, oblique hidden crack source point interior points 34, oblique hidden crack source point interior points 35, and intersecting hidden crack source point intersection points 36, and the main defects output by the first defect detection model include main defect region intersecting hidden cracks 37. It will be appreciated that the source point inducing the defect in fig. 5 should be the intersection point 36, and therefore the other defect source points edge points and interior points detected are excluded, and the main defect region 37, the cross-hidden crack and all the cross-hidden crack source points 36 are determined to be the target defect.
If a hidden boundary crack source point and an internal hidden crack source point exist in the oblique hidden crack of the cell, the position causing the oblique hidden crack is usually the hidden boundary crack source point. In the embodiment of the application, when the local defects detected by the second detection model include oblique hidden cracking source points of the boundary end points and oblique hidden cracking source points of the internal end points, the boundary end points are preferentially determined as the source points of the target defects, the source point positions of induced defects can be accurately positioned in various hidden cracking source points, and the accuracy of the defect detection result of the cell is further improved.
In this embodiment of the application, after determining the target defect based on the main defect and the local defect in step S207, the method further includes:
s601: and determining the size ratio of the image of the battery piece to be detected to the battery piece to be detected.
S603: and determining the actual distribution of the target defects on the battery piece to be detected based on the position information of the target defects in the battery piece image to be detected and the size proportion.
In the embodiment of the application, mapping from pixel dimensions to physical dimensions can be determined by determining the size ratio of the image of the battery piece to be detected to the battery piece to be detected, and further, the actual distribution of the target defects on the battery piece to be detected is determined based on the position information of the target defects in the image of the battery piece to be detected and the size ratio, so that the distribution of the target defects on the corresponding battery piece to be detected is obtained. The distribution of the target defects on the battery piece to be detected, which is determined in one specific embodiment of the application, is shown in fig. 8, and from fig. 8, the defect concentration generated at certain positions can be determined, and the defect multiple positions are generally positions acted by a certain mechanism during the production of the battery piece. Therefore, the mechanism for generating the defects can be rapidly positioned based on the defect distribution diagram of the battery piece, and the loss possibly caused by the subsequent production of the battery piece can be timely reduced.
In this embodiment of the application, after determining the target defect based on the main defect and the local defect in step S207, the method further includes:
s701: and determining a comparison result of the target defect and a preset defect.
S703: determining unusable defective battery pieces based on the comparison result, and removing the defective battery pieces.
In the embodiment of the application, the preset defect may include a defect causing the battery piece to be unusable. In some embodiments, the preset defects may include oblique subfissure with a length greater than a preset threshold, cross subfissure with a total length greater than a preset threshold, black spot defects with a total area greater than a preset threshold, and the like. If the target defect is determined to be an unavailable defect after being compared with the preset defect, the battery piece to be detected containing the unavailable defect can be removed, the battery pieces output by the assembly line are all available battery pieces, the subsequent manual removing process of the unavailable battery pieces is saved, the labor cost is saved, and the selection efficiency of the available battery pieces is improved.
The overall process of cell inspection is described below with an embodiment. As shown in fig. 9, after the battery plate on the assembly line receives photo-induced luminescence caused by photo-electric triggering, the controller generates a trigger signal and sends the trigger signal to the PLC device, and the external encoder controls the line-scan camera to acquire an image of the battery plate to be detected. And after the battery piece to be detected is imaged, a training data set is established based on the qualified sample image and the defect sample image to train the target defect detection model. And after the training is finished, inputting the image of the battery piece to be detected into the target defect detection model, and obtaining a detection result image containing the target defect. Meanwhile, after receiving the trigger signal, the PLC equipment synchronously generates an identification ID of the battery piece to be detected, binds the identification ID with the detection result image to form an ID-result pair, and then sends the ID-result pair to the PLC equipment. And the PLC equipment rejects the battery piece to be detected containing the unavailable defect on the assembly line based on the ID-result pair and the preset defect.
It should be understood that, although the steps in the flowcharts related to the embodiments as described above are sequentially displayed as indicated by arrows, the steps are not necessarily performed sequentially as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least a part of the steps in the flowcharts related to the embodiments described above may include multiple steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, and the execution order of the steps or stages is not necessarily sequential, but may be rotated or alternated with other steps or at least a part of the steps or stages in other steps.
Based on the same inventive concept, the embodiment of the present application further provides a battery piece defect detection apparatus 900 for implementing the battery piece defect detection method. The implementation scheme for solving the problem provided by the apparatus is similar to the implementation scheme described in the above method, so specific limitations in one or more embodiments of the apparatus 900 for detecting defects of battery slices provided below can be referred to the limitations on the method for detecting defects of battery slices, and are not described herein again.
In one embodiment, as shown in fig. 10, there is provided a battery piece defect detecting apparatus 900, including:
the image acquisition module 901 is used for acquiring an image of a battery piece to be detected;
a first defect detection module 902, configured to input the image of the battery piece to be detected into a first defect detection model to obtain a main defect, where the main defect includes a continuous defect region;
a second defect detection module 903, configured to input the image of the battery cell to be detected into a second defect detection model to obtain a local defect, where the local defect includes at least one defect point, and the defect point includes a defect end point or a defect intersection point;
a defect determining module 904 for determining a target defect based on the primary defect and the local defect.
All or part of each module in the battery piece defect detection device can be realized through software, hardware and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
In one embodiment, a computer device is provided, which includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the battery piece defect detection method according to any one of the above items when executing the computer program.
In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, which when executed by a processor implements the steps of the battery sheet defect detection method of any one of the above.
It should be noted that, the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, presented data, etc.) referred to in the present application are information and data authorized by the user or sufficiently authorized by each party.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, database, or other medium used in the embodiments provided herein may include at least one of non-volatile and volatile memory. The nonvolatile Memory may include a Read-Only Memory (ROM), a magnetic tape, a floppy disk, a flash Memory, an optical Memory, a high-density embedded nonvolatile Memory, a resistive Random Access Memory (ReRAM), a Magnetic Random Access Memory (MRAM), a Ferroelectric Random Access Memory (FRAM), a Phase Change Memory (PCM), a graphene Memory, and the like. Volatile Memory can include Random Access Memory (RAM), external cache Memory, and the like. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM), among others. The databases referred to in various embodiments provided herein may include at least one of relational and non-relational databases. The non-relational database may include, but is not limited to, a block chain based distributed database, and the like. The processors referred to in the embodiments provided herein may be general purpose processors, central processing units, graphics processors, digital signal processors, programmable logic devices, quantum computing based data processing logic devices, etc., without limitation.
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the present application. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.

Claims (10)

1. A battery piece defect detection method is characterized by comprising the following steps:
acquiring an image of a battery piece to be detected;
inputting the battery piece image to be detected into a first defect detection model to obtain a main defect, wherein the main defect comprises a continuous defect area;
inputting the battery piece image to be detected into a second defect detection model to obtain a local defect, wherein the local defect comprises at least one defect point, and the defect point comprises a defect end point or a defect intersection point;
determining a target defect based on the primary defect and the local defect.
2. The method of claim 1, wherein the determining a target defect based on the primary defect and the local defect comprises:
determining a main defect area of the main defect and a local defect area of the local defect;
determining the target defect based on the coincidence degree of the main defect region and the local defect region.
3. The method of claim 2, wherein the determining the target defect based on a degree of coincidence of the primary defect region and the local defect region comprises:
if the contact ratio is larger than zero, determining the target defect based on the main defect area and the local defect;
and if the contact ratio is not greater than zero, determining the target defect based on the main defect.
4. The method of claim 3, wherein the determining the target defect based on the local defect comprises:
if the local defect is oblique hidden crack and the local defect comprises at least two oblique hidden crack source points, judging the position of the oblique hidden crack source points in the image of the battery piece to be detected;
if the oblique hidden crack source points are all internal end points, determining the main defect area, the oblique hidden crack and all the internal end points as the target defects;
and if the oblique hidden crack source points comprise at least one boundary endpoint, determining the main defect area, the oblique hidden crack and all the boundary endpoints as the target defects.
5. The method of claim 3, wherein the determining the target defect based on the local defect comprises:
and if the local defects comprise cross hidden cracks and oblique hidden cracks, determining the main defect region, the cross hidden cracks and all cross hidden crack source points as the target defects.
6. The method of claim 1, further comprising, after said determining a target defect based on the primary defect and the local defect:
determining the size ratio of the image of the battery piece to be detected and the battery piece to be detected;
and determining the actual distribution of the target defects on the battery piece to be detected based on the position information of the target defects in the battery piece image to be detected and the size proportion.
7. The method of claim 6, further comprising, after said determining a target defect based on the primary defect and the local defect:
determining a comparison result of the target defect and a preset defect;
determining unusable defective battery pieces based on the comparison result, and removing the defective battery pieces.
8. A device for detecting defects in a battery cell, the device comprising:
the image acquisition module is used for acquiring an image of the battery piece to be detected;
the first defect detection module is used for inputting the battery piece image to be detected into a first defect detection model to obtain a main defect, and the main defect comprises a continuous defect area;
the second defect detection module is used for inputting the image of the battery piece to be detected into a second defect detection model to obtain a local defect, wherein the local defect comprises at least one defect point, and the defect point comprises a defect end point or a defect intersection point;
a defect determination module to determine a target defect based on the primary defect and the local defect.
9. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the steps of the method of any of claims 1 to 7.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 7.
CN202211201816.1A 2022-09-29 2022-09-29 Battery piece defect detection method and device, computer equipment and readable storage medium Pending CN115526860A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211201816.1A CN115526860A (en) 2022-09-29 2022-09-29 Battery piece defect detection method and device, computer equipment and readable storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211201816.1A CN115526860A (en) 2022-09-29 2022-09-29 Battery piece defect detection method and device, computer equipment and readable storage medium

Publications (1)

Publication Number Publication Date
CN115526860A true CN115526860A (en) 2022-12-27

Family

ID=84698875

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211201816.1A Pending CN115526860A (en) 2022-09-29 2022-09-29 Battery piece defect detection method and device, computer equipment and readable storage medium

Country Status (1)

Country Link
CN (1) CN115526860A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117973442A (en) * 2024-04-01 2024-05-03 青岛科技大学 Lithium ion battery SOC estimation method based on hybrid neural network

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117973442A (en) * 2024-04-01 2024-05-03 青岛科技大学 Lithium ion battery SOC estimation method based on hybrid neural network

Similar Documents

Publication Publication Date Title
JP7138753B2 (en) Image defect detection method, device, electronic device, storage medium and product
CN109977808B (en) Wafer surface defect mode detection and analysis method
US20230419472A1 (en) Defect detection method, device and system
TWI618161B (en) Methods, systems and non-transitory computer-readable medium for detecting repeating defects on semiconductor wafers using design data
Kim et al. Automatic fault recognition of photovoltaic modules based on statistical analysis of UAV thermography
CN104063873A (en) Shaft sleeve part surface defect on-line detection method based on compressed sensing
US20120243773A1 (en) Design-Based Inspection Using Repeating Structures
KR20210150970A (en) Detecting defects in semiconductor specimens using weak labeling
CN112381791A (en) Bolt looseness detection method based on 3D point cloud
JP2021515885A (en) Methods, devices, systems and programs for setting lighting conditions and storage media
Sovetkin et al. Automatic processing and solar cell detection in photovoltaic electroluminescence images
CN115526860A (en) Battery piece defect detection method and device, computer equipment and readable storage medium
Wu et al. An end-to-end learning method for industrial defect detection
CN110426395B (en) Method and device for detecting surface of solar EL battery silicon wafer
CN116542984A (en) Hardware defect detection method, device, computer equipment and storage medium
Bahreini et al. Point cloud semantic segmentation of concrete surface defects using dynamic graph CNN
JP2021143884A (en) Inspection device, inspection method, program, learning device, learning method, and trained dataset
Lang et al. Pavement cracking detection and classification based on 3d image using multiscale clustering model
CN113554645A (en) Industrial anomaly detection method and device based on WGAN
CN115937195A (en) Surface defect detection method and device and computer equipment
Ramlal et al. Toward automated utility pole condition monitoring: A deep learning approach
CN115690001A (en) Method for detecting defects in steel pipe welding digital radiographic image
Gao et al. Cylinder liner defect detection and classification based on deep learning
CN113177544A (en) Method for detecting damage and blood stain residue in tilapia processing process
Rahimzadeh et al. State-of-the-art deep learning anomaly detection method for analyzing electroluminescence images of solar cells

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