Pole piece defect detection method and device and computer readable storage medium

Abstract

The invention provides a pole piece defect detection method and device and a computer readable storage medium, relates to the technical field of batteries, and can improve the accuracy of pole piece defect detection. The method comprises the steps that a plurality of first images and a plurality of second images of pole pieces are acquired, initial image labels of the first images indicate that no defect exists in the first images, and initial image labels of the second images indicate that defects exist in the second images; determining a third image and a fourth image in the plurality of first images based on a first preset model; the first preset model is an unsupervised model, the target image label of the third image indicates that no defect exists in the third image, and the target image label of the fourth image indicates that the defect exists in the fourth image; determining a fifth image and a sixth image in the plurality of second images based on a second preset model; the second preset model is a supervised model, the target image label of the fifth image indicates that no defect exists in the fifth image, and the target image label of the sixth image indicates that the defect exists in the sixth image.

Description

Technical Field

[0001] The present application relates to the field of battery technology, and in particular to a pole piece defect detection method, device and computer-readable storage medium. Background Art

[0002] The pole piece is an important part of the battery. In the process of battery production, in order to detect the defects of the pole piece, the existing solution is to take a picture of the pole piece to obtain an image of the pole piece, and then input the image into the defect recognition model. If the defect recognition model recognizes that there is no defect in the image, the image will be marked with an image label of a normal image. If the defect recognition model recognizes that there is a defect in the image, the image will be marked with an image label of a defective image. When multiple images of the pole piece are normal pictures, it means that the pole piece has no defects and is a qualified pole piece. When there is a defective image in multiple images of the pole piece, it means that the pole piece has defects and is an unqualified pole piece.

[0003] In some cases, over-detection and under-detection may occur during the process of detecting defects in the electrode. Over-detection means that there is actually no defect in the image, but the image is labeled as a defective image, and under-detection means that there is an abnormality in the image, but the image is labeled as a normal image. Therefore, the existing solutions have a low accuracy rate for detecting electrode defects. Summary of the invention

[0004] The present application provides a pole piece defect detection method, device and computer-readable storage medium, which can improve the accuracy of pole piece defect detection.

[0005] In order to achieve the above objectives, this application adopts the following technical solutions: In a first aspect, a method for detecting defects in an electrode piece is provided, the method comprising: acquiring a plurality of first images and a plurality of second images of the electrode piece, wherein an initial image label of the first image indicates that no defects exist in the first image, and an initial image label of the second image indicates that defects exist in the second image; determining a third image and a fourth image among the plurality of first images based on a first preset model; the first preset model is an unsupervised model, and a target image label of the third image indicates that no defects exist in the third image, and a target image label of the fourth image indicates that defects exist in the fourth image; determining a fifth image and a sixth image among the plurality of second images based on a second preset model; the second preset model is a supervised model, and a target image label of the fifth image indicates that no defects exist in the fifth image, and a target image label of the sixth image indicates that defects exist in the sixth image.

[0006] Based on this scheme, by obtaining a plurality of first images whose initial image labels indicate that there are no defects in the first image and a plurality of second images whose initial image labels indicate that there are defects in the second image, thereafter, based on a first preset model which is an unsupervised model, a third image whose target image labels indicate that there are no defects in the third image and a fourth image whose target image labels indicate that there are defects in the fourth image are determined in the plurality of first images; and based on a second preset model which is a supervised model, a fifth image whose target image labels indicate that there are no defects in the fifth image and a sixth image whose target image labels indicate that there are defects in the sixth image are determined in the plurality of second images. In this solution, the initial image label of the first image indicates that there are no defects in the first image, that is, the first image is an image that is determined to be defect-free after preliminary judgment. Due to the existence of missed detection, multiple first images may include images with defects. Since the first preset model is an unsupervised model, the first preset model can find outliers in multiple first images, and the outliers are images with defects, thereby reducing the missed detection rate; the initial image label of the second image indicates that there are defects in the second image, that is, the second image is an image that is determined to be defective after preliminary judgment. Due to the existence of over-killing, multiple second images may include images without defects. Since the second preset model is a supervised model, the second preset model can find images that clearly have defects in multiple second images, and the remaining images are images without defects, thereby reducing the over-killing rate. In summary, the accuracy of pole piece defect detection can be improved.

[0007] In combination with the first aspect, in certain embodiments of the first aspect, the method also includes: acquiring a first sample set; the first sample set includes multiple seventh images of the pole piece; there are no defects in the seventh image; and training the original first preset model based on the multiple seventh images to obtain the first preset model.

[0008] Based on this scheme, multiple seventh images of the pole piece are obtained, and then the original first preset model is trained based on the multiple seventh images to obtain the first preset model. Since there are no defects in the seventh image, the original first preset model, which is an unsupervised model, can be trained to obtain the first preset model, so that the pole piece can be subsequently defect detected based on the first preset model.

[0009] In combination with the first aspect, in certain embodiments of the first aspect, the method also includes: obtaining a second sample set; the second sample set includes multiple eighth images of the pole piece and defect labels of defects in each eighth image; there are defects in the eighth image; based on the multiple eighth images and the defect labels of defects in each eighth image, the original second preset model is trained to obtain a second preset model.

[0010] Based on this scheme, by obtaining multiple eighth images of the pole piece and the defect labels of the defects in each eighth image, the original second preset model is then trained based on the multiple eighth images and the defect labels of the defects in each eighth image to obtain the second preset model. Since there are no defects in the eighth image, the original second preset model, which is a supervised model, can be trained to obtain the second preset model, so that the pole piece can be subsequently defect detected based on the second preset model.

[0011] In combination with the first aspect, in certain embodiments of the first aspect, obtaining multiple first images and multiple second images of the pole piece includes: sending indication information to a target storage device; the indication information is used to obtain multiple first images and multiple second images of the pole piece stored in the target storage device and determined online by a pole piece quality control device within a historical time period; receiving response information from the target storage device; the response information includes multiple first images and multiple second images.

[0012] Based on this solution, by acquiring multiple first images and multiple second images stored in the target storage device and determined online by the electrode quality control device within the historical time period, it is possible to subsequently detect whether the electrode has defects in an offline manner without the need to detect whether the electrode has defects in an online manner, thereby reducing the computing resources occupied by defect detection during electrode production and reducing the impact on electrode production.

[0013] In a second aspect, a pole piece defect detection device is provided for implementing the pole piece defect detection method of the first aspect. The pole piece defect detection device includes a module, unit, or means corresponding to the above method, and the module, unit, or means can be implemented by hardware, software, or by hardware executing corresponding software. The hardware or software includes one or more modules or units corresponding to the above functions.

[0014] In combination with the second aspect, in certain embodiments of the second aspect, the device includes: an acquisition module and a processing module; the acquisition module is used to acquire multiple first images and multiple second images of the pole piece, the initial image label of the first image indicates that there is no defect in the first image, and the initial image label of the second image indicates that there is a defect in the second image; the processing module is used to determine the third image and the fourth image among the multiple first images based on the first preset model; the first preset model is an unsupervised model, the target image label of the third image indicates that there is no defect in the third image, and the target image label of the fourth image indicates that there is a defect in the fourth image; the processing module is also used to determine the fifth image and the sixth image among the multiple second images based on the second preset model; the second preset model is a supervised model, the target image label of the fifth image indicates that there is no defect in the fifth image, and the target image label of the sixth image indicates that there is a defect in the sixth image.

[0015] In combination with the second aspect, in certain embodiments of the second aspect, the processing module is also used to: obtain a first sample set; the first sample set includes multiple seventh images of the pole piece; there are no defects in the seventh image; and train the original first preset model based on the multiple seventh images to obtain the first preset model.

[0016] In combination with the second aspect, in certain embodiments of the second aspect, the processing module is also used to: obtain a second sample set; the second sample set includes multiple eighth images of the pole piece and defect labels of defects in each eighth image; there are defects in the eighth image; based on the multiple eighth images and the defect labels of defects in each eighth image, the original second preset model is trained to obtain the second preset model.

[0017] In combination with the second aspect, in certain embodiments of the second aspect, the acquisition module is specifically used to: send indication information to the target storage device; the indication information is used to obtain multiple first images and multiple second images of the electrode determined online by the electrode quality control device within a historical time period stored in the target storage device; receive response information from the target storage device; the response information includes multiple first images and multiple second images.

[0018] In a third aspect, a pole piece defect detection device is provided, comprising: at least one processor and a memory for storing instructions executable by the processor; wherein the processor is configured to execute instructions to implement the method provided in the first aspect and any possible implementation manner thereof.

[0019] In a fourth aspect, a computer-readable storage medium is provided. When the instructions in the computer-readable storage medium are executed by a processor of a pole piece defect detection device, the pole piece defect detection device is enabled to execute the method provided in the first aspect and any possible implementation manner thereof.

[0020] In a fifth aspect, a computer program product comprising instructions is provided, which, when executed on a computer, enables the computer to execute the method provided in the first aspect and any possible implementation manner thereof.

[0021] Among them, the technical effects brought about by any implementation of the second to fifth aspects can refer to the technical effects brought about by different implementations of the first aspect mentioned above, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 A schematic diagram of a battery production process provided for this application; Figure 2 A schematic diagram of the architecture of a pole piece defect detection system provided in this application; Figure 3 A schematic diagram of a process flow of a pole piece defect detection method provided in this application; Figure 4 A schematic diagram of a process flow of a pole piece defect detection method provided in this application; Figure 5 A schematic diagram of a process flow of a pole piece defect detection method provided in this application; Figure 6 A schematic diagram of the structure of a pole piece defect detection device provided in this application; Figure 7 A schematic diagram of the structure of a pole piece defect detection device provided in this application. DETAILED DESCRIPTION

[0023] In the description of this application, unless otherwise specified, "plurality" means two or more than two. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b, or c can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple.

[0024] In addition, in order to clearly describe the technical solutions of the embodiments of the present application, in the embodiments of the present application, words such as "first" and "second" are used to distinguish the same items or similar items with substantially the same functions and effects. Those skilled in the art can understand that words such as "first" and "second" do not limit the quantity and execution order, and words such as "first" and "second" do not necessarily limit the difference.

[0025] Meanwhile, in the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a concrete manner for ease of understanding.

[0026] It is understood that the "embodiment" mentioned throughout the specification means that the specific features, structures or characteristics related to the embodiment are included in at least one embodiment of the present application. Therefore, the various embodiments in the entire specification do not necessarily refer to the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner. It is understood that in various embodiments of the present application, the size of the sequence number of each process does not mean the order of execution, and the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present application.

[0027] It can be understood that in the present application, "when", "if" and "if" all mean that corresponding processing will be carried out under certain objective circumstances, but do not limit the time, nor do they require judgment actions when implementing them, nor do they mean the existence of other limitations.

[0028] It can be understood that some optional features in the embodiments of the present application may be implemented independently in certain scenarios without relying on other features, such as the solution on which they are currently based, to solve corresponding technical problems and achieve corresponding effects, or may be combined with other features according to needs in certain scenarios. Accordingly, the devices provided in the embodiments of the present application may also realize these features or functions accordingly, which will not be elaborated here.

[0029] In this application, unless otherwise specified, the same or similar parts between the various embodiments can refer to each other. In the various embodiments in this application, and the various implementation methods in each embodiment, if there is no special description and logical conflict, the terms and / or descriptions between different embodiments and the various implementation methods in each embodiment are consistent and can be referenced to each other. The technical features in different embodiments and the various implementation methods in each embodiment can be combined to form new embodiments, implementation methods, implementation methods, or implementation methods according to their inherent logical relationships. The following implementation methods of this application do not constitute a limitation on the scope of protection of this application.

[0030] The electrode is an important part of the battery. Figure 1 A schematic diagram of a battery production process provided for this application, such as Figure 1 As shown, a solvent is added to the powder to form a slurry, and then the slurry is coated on a foil to form a pole roll, and then the pole roll is roll-die-cut to obtain a plurality of pole pieces, and then the plurality of pole pieces are stacked to obtain a battery cell, and then the plurality of battery cells are assembled into a shell to obtain a battery, and finally, the battery is injected with liquid for activation to obtain a usable battery.

[0031] Among the above-mentioned processes, roll-cutting, electrode stacking and shell assembly are important processes in battery production. During the above-mentioned processes, defects may occur in the electrode.

[0032] Pole sheet defects may include coating defects, rolling defects, cutting defects and other defects.

[0033] Among them, coating defects include: Exposed foil defect: Some areas of the electrode surface are not completely covered with slurry, resulting in exposed aluminum foil or copper foil.

[0034] Dark spot defect or light spot defect: Areas of uneven color appear on the surface of the pole piece, which may be caused by uneven mixing of the slurry or problems in the coating process.

[0035] Material falling defect: During the coating or subsequent processing of the electrode, part of the slurry falls off, forming voids or missing parts.

[0036] Scratches: Scratches caused by scraper or equipment problems during the coating process.

[0037] Rolling defects include: Pole piece wrinkling defect: The pole piece is subjected to uneven pressure during the rolling process, resulting in wrinkles on the surface of the pole piece.

[0038] Uneven thickness defect: The thickness of the electrode sheet is inconsistent after rolling, which may affect the performance and life of the battery.

[0039] Cutting defects include: Edge burr defect: irregular or sharp edge produced during the cutting process.

[0040] Size deviation defect: The size of the pole piece after cutting does not match the design size.

[0041] Other defects include Surface dust defect: Dust or impurities adhere to the surface of the electrode during the production process.

[0042] White spots after pressing: White spots appear on the surface of the pole piece during rolling or subsequent processing.

[0043] Large white spot defects on the surface: Large white areas appear on the surface of the pole piece, which may be due to slurry problems or uneven coating.

[0044] Surface streak defect: Strip-like marks appear on the surface of the pole piece, which may be caused by equipment problems or improper operation.

[0045] The above is a brief description of the possible defects of the pole piece. Of course, the pole piece may also have other unknown or known defects, and this application does not impose specific restrictions on this.

[0046] Defects in the electrode may have a negative impact on the battery's charge and discharge cycle, coulombic efficiency, rate performance, and cycle life. Therefore, it is necessary to detect defective electrodes to avoid adverse effects on the battery. In order to detect the defects of the electrode, the existing solution is to illuminate the electrode with a light source during the production process of the electrode, then take a picture of the electrode to obtain an image of the electrode, and then input the image into a defect recognition model. If the defect recognition model recognizes that there is no defect in the image, the image will be marked with an image label of a normal image. If the defect recognition model recognizes that there is a defect in the image, the image will be marked with an image label of a defective image. When multiple images of the electrode are normal pictures, it means that the electrode has no defects and is a qualified electrode. When there is a defective image in multiple images of the electrode, it means that there is a defect in the electrode and the electrode is an unqualified electrode.

[0047] Exemplary defect recognition models are charge coupled device (CCD) artificial intelligence (AI) models. CCD is a semiconductor sensor that can convert optical images into digital signals. Specifically, CCD captures light through a series of tiny photosensitive elements (usually photodiodes), each of which corresponds to a pixel in the image. When light reflected from the pole piece hits the CCD, the photosensitive element generates corresponding charges, which are then converted into digital signals to form a digital image.

[0048] During the production of the pole piece, the light source is used to illuminate the pole piece, and the CCD line scan camera takes images of the pole piece surface, which are then transmitted to the CCD AI model. The CCD AI model analyzes the images to detect whether there are defects on the pole piece surface.

[0049] In some cases, for example, unreasonable light source settings, low CCD AI model performance, and unknown defects in the pole piece, over-killing and under-killing may occur in the process of detecting pole piece defects. Over-killing means that there is actually no defect in the image, but the image is labeled as a defective image, and under-killing means that there is actually an abnormality in the image, but the image is labeled as a normal image. Therefore, the existing solutions have a low accuracy rate for pole piece defect detection.

[0050] In order to improve the accuracy of electrode defect detection, the common solution in the industry is to take pictures of the electrode using multiple light sources. However, the increase in the types of light sources will also lead to an increase in the number of images, increasing the subsequent workload, and there is no guarantee that there will be no overkill. Alternatively, the performance of the CCD AI model can be enhanced. However, due to the limitations of technological development and development costs, the performance of the CCD AI model cannot be greatly enhanced, and thus the accuracy of electrode defect detection cannot be greatly improved. Therefore, the common solutions in the industry also have various defects.

[0051] In order to improve the accuracy of pole piece defect detection, the present application provides a pole piece defect detection method, which is applied to a pole piece defect device, and the method includes: acquiring multiple first images and multiple second images of the pole piece, the initial image label of the first image indicates that there is no defect in the first image, and the initial image label of the second image indicates that there is a defect in the second image; determining a third image and a fourth image among the multiple first images based on a first preset model; the first preset model is an unsupervised model, the target image label of the third image indicates that there is no defect in the third image, and the target image label of the fourth image indicates that there is a defect in the fourth image; determining a fifth image and a sixth image among the multiple second images based on a second preset model; the second preset model is a supervised model, the target image label of the fifth image indicates that there is no defect in the fifth image, and the target image label of the sixth image indicates that there is a defect in the sixth image.

[0052] Based on this scheme, by obtaining a plurality of first images whose initial image labels indicate that there are no defects in the first image and a plurality of second images whose initial image labels indicate that there are defects in the second image, thereafter, based on a first preset model which is an unsupervised model, a third image whose target image labels indicate that there are no defects in the third image and a fourth image whose target image labels indicate that there are defects in the fourth image are determined in the plurality of first images; and based on a second preset model which is a supervised model, a fifth image whose target image labels indicate that there are no defects in the fifth image and a sixth image whose target image labels indicate that there are defects in the sixth image are determined in the plurality of second images. In this solution, the initial image label of the first image indicates that there are no defects in the first image, that is, the first image is an image that is determined to be defect-free after preliminary judgment. Due to the existence of missed detection, multiple first images may include images with defects. Since the first preset model is an unsupervised model, the first preset model can find outliers in multiple first images, and the outliers are images with defects, thereby reducing the missed detection rate; the initial image label of the second image indicates that there are defects in the second image, that is, the second image is an image that is determined to be defective after preliminary judgment. Due to the existence of over-killing, multiple second images may include images without defects. Since the second preset model is a supervised model, the second preset model can find images that clearly have defects in multiple second images, and the remaining images are images without defects, thereby reducing the over-killing rate. In summary, the accuracy of pole piece defect detection can be improved.

[0053] Figure 2 This is a schematic diagram of the architecture of a pole piece defect detection system provided in this application. The technical solution of the embodiment of this application can be applied to Figure 2 The pole piece defect detection system shown in FIG. Figure 2 As shown, the pole piece defect detection system 20 includes a pole piece defect detection device 21 and a target storage device 22 .

[0054] Among them, the pole piece defect detection device 21 is directly or indirectly connected to the target storage device 22. In this connection relationship, it can be connected in a wired manner or in a wireless manner, which is not limited in the embodiment of the present application.

[0055] In practical applications, the pole piece defect detection method provided in the embodiment of the present application can be applied to the pole piece defect detection device 21, and can also be applied to the device included in the pole piece defect detection device 21.

[0056] The target storage device 22 may be a solid state drive or a mechanical hard drive, and the present application does not impose any specific restrictions on the specific form of the target storage device 22. The target storage device 22 stores a plurality of first images and a plurality of second images determined online by the electrode quality control device.

[0057] The electrode quality control device can perform preliminary detection of electrode defects using existing technologies. The present application does not impose any specific restrictions on the specific detection method of the electrode quality control device.

[0058] For example, the pole piece quality control device performs online defect detection on the pole piece image through the existing CCD AI model to obtain multiple first images and multiple second images, and then stores the determined multiple first images and multiple second images to the target storage device 22.

[0059] Online defect detection refers to the real-time detection of pole piece defects during the pole piece production process.

[0060] Optionally, the target storage device 22 may also store an original image of the electrode, where the original image is an image of the electrode without an image label, that is, the original image is an image of the electrode after acquisition but without defect detection.

[0061] Optionally, the target storage device 22 may also store a seventh image of the pole piece, and there is no defect in the seventh image.

[0062] The seventh image may be an image confirmed by manual inspection to have no defects.

[0063] Optionally, the target storage device 22 may also store an eighth image of the pole piece, in which defects exist.

[0064] The eighth image may be an image confirmed to have defects by manual inspection.

[0065] Data can be exchanged between the electrode defect detection device 21 and the target storage device 22 .

[0066] It should be noted that the electrode defect detection device 21 and the target storage device 22 can be independent devices or integrated into the same device, and this application does not make any specific limitation on this.

[0067] When the electrode defect detection device 21 and the target storage device 22 are integrated into the same device, the communication mode between the electrode defect detection device 21 and the target storage device 22 is the communication between the internal modules of the device. In this case, the communication process between the two is the same as "the communication process between the electrode defect detection device 21 and the target storage device 22 when they are independent of each other".

[0068] In the following embodiments provided in the present application, the present application is explained by taking the example that the electrode defect detection device 21 and the target storage device 22 are independently configured.

[0069] The following describes the pole piece defect detection method provided in the embodiment of the present application by taking the pole piece defect detection method applied to the pole piece defect detection device 21 as an example in combination with the accompanying drawings.

[0070] Figure 3 A schematic diagram of a process flow of a pole piece defect detection method provided in this application, such as Figure 3 As shown, the method comprises the following steps: S301. A pole piece defect detection device obtains a plurality of first images and a plurality of second images of a pole piece.

[0071] The initial image label of the first image indicates that there is no defect in the first image, and the initial image label of the second image indicates that there is a defect in the second image.

[0072] As a possible implementation, combining Figure 1 , the electrode defect detection device sends a first request message to the target storage device, the first request message is used to obtain the original image of the electrode without the image label in the target storage device. Correspondingly, the target storage device receives the first request message from the electrode defect detection device.

[0073] The target storage device generates first reply information including the original image of the pole piece, and sends the first reply information to the pole piece defect detection device. Correspondingly, the pole piece defect detection device receives the first reply information from the target storage device.

[0074] After receiving the first reply information, the pole piece defect detection device performs preliminary defect detection on the original image of the pole piece through the CCD AI model to obtain multiple first images and multiple second images.

[0075] It should be noted that in this possible implementation, the specific scheme of performing defect detection on the original image of the pole piece through the CCD AI model can refer to the existing scheme, and this application will not explain it again.

[0076] As another possible implementation method, combining Figure 1, the electrode defect detection device sends instruction information to the target storage device; the instruction information is used to obtain the multiple first images and multiple second images stored in the target storage device, which are determined online by the electrode quality control device within the historical time period. Correspondingly, the target storage device receives the instruction information of the electrode defect detection device.

[0077] The target storage device generates response information including a plurality of first images and a plurality of second images, and sends the response information to the pole piece defect detection device. Correspondingly, the pole piece defect detection device receives the response information from the target storage device, and obtains the plurality of first images and a plurality of second images from the response information.

[0078] Based on this possible implementation method, by acquiring multiple first images and multiple second images stored in the target storage device and determined online by the electrode quality control device within the historical time period, it is possible to subsequently detect whether the electrode has defects in an offline manner without the need to detect whether the electrode has defects in an online manner, thereby reducing the computing resources occupied by defect detection during electrode production and reducing the impact on electrode production.

[0079] S302. The pole piece defect detection device determines a third image and a fourth image among a plurality of first images based on a first preset model.

[0080] Among them, the first preset model is an unsupervised model, the target image label of the third image indicates that there is no defect in the third image, and the target image label of the fourth image indicates that there is a defect in the fourth image.

[0081] It should be noted that the first preset model can be a K-means clustering model, a hierarchical clustering model or a DBSCAN model. Of course, the first preset model can also be other unsupervised models, and this application does not impose any specific restrictions on this.

[0082] As a possible implementation manner, the pole piece defect detection device inputs the multiple first images into a first preset model, and after the first preset model processes the multiple first images, the multiple first images are divided into a third image and a fourth image.

[0083] As another possible implementation, the pole piece defect detection device inputs the multiple first images into a first preset model, and after the first preset model processes the multiple first images, the multiple first images are divided into an initial third image and an initial fourth image.

[0084] The pole piece defect detection device inputs the initial third image and the initial fourth image into a preset computer vision (CV) model to obtain a recognition result of each initial third image and a recognition result of each initial fourth image.

[0085] When the recognition result of the initial third image indicates that there is no defect in the initial third image, the initial third image is determined as the third image; when the recognition result of the initial third image indicates that there is a defect in the initial third image, the initial third image is determined as the pending image.

[0086] When the recognition result of the initial fourth image indicates that there are defects in the initial fourth image, the initial fourth image is determined as the fourth image; when the recognition result of the initial fourth image indicates that there are no defects in the initial fourth image, the initial fourth image is determined as the pending image.

[0087] Subsequently, it can be manually determined whether there are defects in the pending image.

[0088] It should be noted that the preset CV model can be a convolutional neural network, which is a widely used model for tasks such as image classification and target detection. It simulates the visual mechanism in organisms and uses multiple convolutional layers to extract features from images, thereby achieving efficient image recognition and processing. Of course, the preset CV model can also be other CV models, and this application does not make specific restrictions on this.

[0089] S303. The pole piece defect detection device determines a fifth image and a sixth image among the plurality of second images based on a second preset model.

[0090] Among them, the second preset model is a supervised model, the target image label of the fifth image indicates that there is no defect in the fifth image, and the target image label of the sixth image indicates that there is a defect in the sixth image.

[0091] It should be noted that the second preset model can be a naive Bayes classification model, a logistic regression model, a random forest model, a neural network model, a k-nearest neighbor model or a decision tree model. Of course, the second preset model can also be other supervised models, and this application does not impose any specific restrictions on this.

[0092] As a possible implementation manner, the pole piece defect detection device inputs the multiple second images into the second preset model, and after the second preset model processes the multiple second images, the multiple second images are divided into a fifth image and a sixth image.

[0093] As another possible implementation, the pole piece defect detection device inputs the multiple second images into the second preset model, and after the second preset model processes the multiple second images, the multiple second images are divided into an initial fifth image and an initial sixth image.

[0094] The pole piece defect detection device inputs the initial fifth image and the initial sixth image into a preset CV model to obtain a recognition result of each initial fifth image and a recognition result of each initial sixth image.

[0095] When the recognition result of the initial fifth image indicates that there is no defect in the initial fifth image, the initial fifth image is determined as the fifth image; when the recognition result of the initial fifth image indicates that there is a defect in the initial fifth image, the initial fifth image is determined as the pending image.

[0096] When the recognition result of the initial sixth image indicates that there are defects in the initial sixth image, the initial sixth image is determined as the sixth image; when the recognition result of the initial sixth image indicates that there are no defects in the initial sixth image, the initial sixth image is determined as the pending image.

[0097] Subsequently, it can be manually determined whether there are defects in the pending image.

[0098] Based on this scheme, by obtaining a plurality of first images of the pole piece whose initial image labels indicate that there are no defects and a plurality of second images whose initial image labels indicate that there are defects, thereafter, based on the first preset model of the unsupervised model, a third image whose target image labels indicate that there are no defects and a fourth image whose target image labels indicate that there are defects are determined among the plurality of first images; and based on the second preset model of the supervised model, a fifth image whose target image labels indicate that there are no defects and a sixth image whose target image labels indicate that there are defects are determined among the plurality of second images. In this solution, the initial image label of the first image indicates that there is no defect, that is, the first image is an image that is determined to be defect-free after preliminary judgment. Due to the existence of missed kills, multiple first images may include images with defects. Since the first preset model is an unsupervised model, the first preset model can find outliers in multiple first images, and the outliers are the fourth images with defects, thereby reducing the missed kill rate; the initial image label of the second image indicates that there is a defect, that is, the second image is an image that is determined to be defective after preliminary judgment. Due to the existence of overkill, multiple second images may include images without defects. Since the second preset model is a supervised model, the second preset model can find the sixth image that clearly has defects in multiple second images, and the remaining images are the fifth images without defects, thereby reducing the overkill rate. In summary, the accuracy of pole piece defect detection can be improved.

[0099] The above is a general description of the pole piece defect detection method provided in the present application. The pole piece defect detection method provided in the present application will be further described below in conjunction with the accompanying drawings.

[0100] In one design, Figure 4 A flow chart of another electrode defect detection method provided in this application is as follows: Figure 4 As shown, before S301, the electrode defect detection method provided by the present application may further include the following steps: S401. A pole piece defect detection device obtains a first sample set.

[0101] The first sample set includes a plurality of seventh images of the pole piece, and no defects exist in the seventh images.

[0102] As a possible implementation, combining Figure 1 , the electrode defect detection device sends a second request information to the target storage device, and the second request information is used to obtain a plurality of seventh images of the electrode. Correspondingly, the target storage device receives the second request information of the electrode defect detection device.

[0103] The target storage device generates second reply information including multiple seventh images of the pole piece, and sends the second reply information to the pole piece defect detection device. Correspondingly, the pole piece defect detection device receives the second reply information from the target storage device, and obtains multiple seventh images of the pole piece from the second reply information.

[0104] S402. The electrode defect detection device trains the original first preset model based on the multiple seventh images to obtain the first preset model.

[0105] As a possible implementation method, the pole piece defect detection device inputs multiple seventh images into the original first preset model, and obtains the first preset model through image reconstruction; or, extracts the initial features of multiple seventh images through multiple convolution kernels to obtain multiple initial features, and then, for the multiple initial features, focuses on the main features through downsampling to obtain the first preset model.

[0106] It should be noted that the above is a brief description of how to train the original first preset model to obtain the first preset model. The specific description of how to train the original first preset model to obtain the first preset model can refer to the existing scheme, and this application will not explain it again here.

[0107] Based on this scheme, multiple seventh images of the pole piece are obtained, and then the original first preset model is trained based on the multiple seventh images to obtain the first preset model. Since there are no defects in the seventh image, the original first preset model, which is an unsupervised model, can be trained to obtain the first preset model, so that the pole piece can be subsequently defect detected based on the first preset model.

[0108] In one design, Figure 5 A flow chart of another electrode defect detection method provided in this application is as follows: Figure 5 As shown, before S301, the electrode defect detection method provided by the present application may further include the following steps: S501. The electrode defect detection device obtains a second sample set.

[0109] The second sample set includes a plurality of eighth images of the pole piece and a defect label of a defect in each eighth image, and there is a defect in the eighth image.

[0110] It should be noted that the multiple eighth images may also be referred to as a second sample set.

[0111] The defect label information may be the type of defect.

[0112] As a possible implementation, combining Figure 1 , the pole piece defect detection device sends a third request information to the target storage device, and the third request information is used to obtain multiple eighth images of the pole piece and the defect label of the defect in each eighth image. Correspondingly, the target storage device receives the third request information of the pole piece defect detection device.

[0113] The target storage device generates a third reply message including a plurality of eighth images of the pole piece and a defect label of a defect in each eighth image, and sends the third reply message to the pole piece defect detection device. Correspondingly, the pole piece defect detection device receives the third reply message from the target storage device, and obtains a plurality of eighth images of the pole piece and a defect label of a defect in each eighth image from the third reply message.

[0114] S502. The pole piece defect detection device trains the original second preset model based on the multiple eighth images and the defect labels of the defects in each eighth image to obtain the second preset model.

[0115] As a possible implementation method, the pole piece defect detection device initializes the model parameters of the original second preset model, such as weights, biases, etc., and configures a preset loss function for measuring the difference between the model prediction result and the true label. Multiple eighth images and the defect labels of defects in each eighth image are input into the original second preset model, and the parameters of the original second model are continuously adjusted through iterative training. In each iteration, the gradient of the loss function with respect to the model parameters is calculated, and the parameters are updated using an optimization algorithm. The iterative process continues until the stop condition is met, and the second preset model is obtained.

[0116] It should be noted that the above is a brief description of how to train the original second preset model to obtain the second preset model. The specific description of how to train the original second preset model to obtain the second preset model can refer to the existing scheme, and this application will not explain it again here.

[0117] Based on this scheme, by obtaining multiple eighth images of the pole piece and the defect labels of the defects in each eighth image, the original second preset model is then trained based on the multiple eighth images and the defect labels of the defects in each eighth image to obtain the second preset model. Since there are no defects in the eighth image, the original second preset model, which is a supervised model, can be trained to obtain the second preset model, so that the pole piece can be subsequently defect detected based on the second preset model.

[0118] The above mainly introduces the solution provided by the embodiment of the present application from the perspective of the problem pole piece defect detection device executing the problem pole piece defect detection method. In order to realize the above functions, the problem pole piece defect detection device includes hardware structures and / or software modules corresponding to the execution of each function. Those skilled in the art should easily realize that, in combination with the units and algorithm steps of each example described in the embodiments disclosed herein, the embodiments of the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0119] The embodiment of the present application can divide the functional modules of the problem pole piece defect detection device according to the above method example. For example, each functional module can be divided according to each function, or two or more functions can be integrated into one processing module. The above integrated module can be implemented in the form of hardware or in the form of software functional modules. Optionally, the division of modules in the embodiment of the present application is schematic and is only a logical function division. There may be other division methods in actual implementation. In addition, the "module" here may refer to a specific application-specific integrated circuit (ASIC), a circuit, a processor and memory that executes one or more software or firmware programs, an integrated logic circuit, and / or other devices that can provide the above functions.

[0120] In the case of functional module division, Figure 6 FIG. 1 shows a schematic diagram of the structure of a defect detection device for a problematic electrode. Figure 6 As shown, the problematic pole piece defect detection device 60 includes an acquisition module 601 and a processing module 602 .

[0121] In some embodiments, the problem electrode defect detection device 60 may further include a storage module ( Figure 6 ), used to store program instructions and data.

[0122] Among them, the acquisition module 601 is used to acquire multiple first images and multiple second images of the pole piece, the initial image label of the first image indicates that there is no defect in the first image, and the initial image label of the second image indicates that there is a defect in the second image; the processing module 602 is used to determine the third image and the fourth image among the multiple first images based on the first preset model; the first preset model is an unsupervised model, the target image label of the third image indicates that there is no defect in the third image, and the target image label of the fourth image indicates that there is a defect in the fourth image; the processing module 602 is also used to determine the fifth image and the sixth image among the multiple second images based on the second preset model; the second preset model is a supervised model, the target image label of the fifth image indicates that there is no defect in the fifth image, and the target image label of the sixth image indicates that there is a defect in the sixth image.

[0123] In combination with the second aspect, in certain embodiments of the second aspect, the processing module 602 is also used to: obtain a first sample set; the first sample set includes multiple seventh images of the pole piece; there are no defects in the seventh image; and train the original first preset model based on the multiple seventh images to obtain the first preset model.

[0124] In combination with the second aspect, in certain embodiments of the second aspect, the processing module 602 is also used to: obtain a second sample set; the second sample set includes multiple eighth images of the pole piece and defect labels of defects in each eighth image; there are defects in the eighth image; based on the multiple eighth images and the defect labels of defects in each eighth image, the original second preset model is trained to obtain a second preset model.

[0125] In combination with the second aspect, in certain embodiments of the second aspect, the acquisition module 601 is specifically used to: send indication information to the target storage device; the indication information is used to obtain the target storage device, which stores multiple first images and multiple second images of the electrode determined online by the electrode quality control device within a historical time period; receive response information from the target storage device; the response information includes multiple first images and multiple second images.

[0126] All relevant contents of each step involved in the above method embodiment can be referred to the functional description of the corresponding functional module, and will not be repeated here.

[0127] When the functions of the above functional modules are implemented in the form of hardware, Figure 7 FIG. 2 shows a schematic diagram of the structure of another problematic electrode defect detection device. Figure 7 As shown, the problem electrode defect detection device 70 includes a processor 701, a memory 702 and a bus 703. The processor 701 and the memory 702 can be connected via the bus 703.

[0128] The processor 701 is the control center of the problem electrode defect detection device 70, which can be a processor or a general term for multiple processing elements. For example, the processor 701 can be a general-purpose central processing unit (CPU) or other general-purpose processors. Among them, the general-purpose processor can be a microprocessor or any conventional processor.

[0129] As an embodiment, the processor 701 may include one or more CPUs, such as Figure 7 CPU 0 and CPU 1 are shown in .

[0130] The memory 702 may be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, an electrically erasable programmable read-only memory (EEPROM), a disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto.

[0131] As a possible implementation, the memory 702 can exist independently of the processor 701, and the memory 702 can be connected to the processor 701 via a bus 703 to store instructions or program codes. When the processor 701 calls and executes the instructions or program codes stored in the memory 702, the problem electrode defect detection method provided in the embodiment of the present application can be implemented.

[0132] In another possible implementation, the memory 702 may also be integrated with the processor 701 .

[0133] The bus 703 may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 7 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0134] It should be pointed out that Figure 7 The structure shown does not constitute a limitation on the problematic electrode defect detection device 70. Figure 7 In addition to the components shown, the problematic pole piece defect detection device 70 may include more or fewer components than shown, or combine certain components, or arrange the components differently.

[0135] As an example, combining Figure 6 The functions implemented by the acquisition module 601 and the processing module 602 in the defect detection device 60 are similar to those Figure 7 The function of processor 701 in is the same.

[0136] Optional, such as Figure 7 As shown, the problematic pole piece defect detection device 70 provided in the embodiment of the present application may further include a communication interface 704 .

[0137] The communication interface 704 is used to connect with other devices through a communication network. The communication network may be Ethernet, wireless access network, wireless local area network (WLAN), etc. The communication interface 704 may include a receiving unit for receiving data and a sending unit for sending data.

[0138] In a possible implementation, in the problematic pole piece defect detection device 70 provided in the embodiment of the present application, the communication interface 704 may also be integrated in the processor 701, and the embodiment of the present application does not specifically limit this.

[0139] As a possible product form, the problematic pole piece defect detection device of the embodiment of the present application can also be implemented using the following: one or more field programmable gate arrays (FPGA), programmable logic devices (PLD), controllers, state machines, gate logic, discrete hardware components, any other suitable circuits, or any combination of circuits capable of performing the various functions described throughout the present application.

[0140] Through the description of the above implementation methods, those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above functional units is used as an example. In practical applications, the above functions can be assigned to different functional units as needed, that is, the internal structure of the device can be divided into different functional units to complete all or part of the functions described above. The specific working process of the system, device and unit described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0141] An embodiment of the present application also provides a computer-readable storage medium on which a computer program or instruction is stored. When the computer program or instruction is executed, the computer executes each step in the method flow shown in the above method embodiment.

[0142] An embodiment of the present application provides a computer program product including instructions. When the instructions are executed on a computer, the computer is caused to execute each step in the method flow shown in the above method embodiment.

[0143] An embodiment of the present application provides a chip system, including: a processor and an interface circuit; the interface circuit is used to receive a computer program or instruction and transmit it to the processor; the processor is used to execute the computer program or instruction so that the chip system executes each step in the method flow shown in the above method embodiment.

[0144] Among them, the computer-readable storage medium, for example, can be but not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or devices, or any combination of the above. More specific examples of computer-readable storage media (non-exhaustive list) include: electrical connections with one or more wires, portable computer disks, hard disks. Random Access Memory (RAM), Read-Only Memory (ROM), Erasable Programmable Read Only Memory (EPROM), registers, hard disks, optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or the above people in a suitable combination, or any other form of computer-readable storage medium of numerical value in the art. An exemplary storage medium is coupled to a processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can be located in a specific purpose ASIC. In the embodiments of the present application, a computer-readable storage medium may be any tangible medium that contains or stores a program, which may be used by or in conjunction with an instruction execution system, apparatus, or device.

[0145] Since the problematic pole piece defect detection device, computer-readable storage medium, and computer program product provided in this embodiment can be applied to the problematic pole piece defect detection method provided in this embodiment, the technical effects that can be obtained can also refer to the above-mentioned method embodiments, and the embodiments of this application will not be repeated here.

[0146] Although the present application is described herein in conjunction with various embodiments, in the process of implementing the claimed application, those skilled in the art may understand and implement other changes to the disclosed embodiments by viewing the drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other components or steps, and "one" or "an" does not exclude multiple situations. A single processor or other unit may implement several functions listed in a claim. Certain measures are recorded in different dependent claims, but this does not mean that these measures cannot be combined to produce good results.

[0147] Although the present application has been described in conjunction with specific features and embodiments thereof, it is obvious that various modifications and combinations may be made thereto without departing from the spirit and scope of the present application. Accordingly, this specification and the drawings are merely exemplary illustrations of the present application as defined by the appended claims, and are deemed to have covered any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, those skilled in the art may make various modifications and variations to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. A method for detecting pole piece defects, characterized in that: The method comprises: Acquire a plurality of first images and a plurality of second images of the pole piece, wherein initial image labels of the first images indicate that there is no defect in the first images, and initial image labels of the second images indicate that there is a defect in the second images; Determine a third image and a fourth image among the plurality of first images based on a first preset model; the first preset model is an unsupervised model, the target image label of the third image indicates that there is no defect in the third image, and the target image label of the fourth image indicates that there is a defect in the fourth image; A fifth image and a sixth image among multiple second images are determined based on a second preset model; the second preset model is a supervised model, the target image label of the fifth image indicates that there is no defect in the fifth image, and the target image label of the sixth image indicates that there is a defect in the sixth image.

2. The method according to claim 1, characterized in that The method further comprises: Acquire a first sample set; the first sample set includes a plurality of seventh images of the pole piece; no defects exist in the seventh images; The original first preset model is trained based on the multiple seventh images to obtain the first preset model.

3. The method according to claim 1, characterized in that The method further comprises: Acquire a second sample set; the second sample set includes a plurality of eighth images of the pole piece and a defect label of a defect in each eighth image; there is a defect in the eighth image; The original second preset model is trained based on the multiple eighth images and the defect labels of the defects in each eighth image to obtain the second preset model.

4. The method according to any one of claims 1 to 3, characterized in that: The step of acquiring a plurality of first images and a plurality of second images of the pole piece comprises: Sending instruction information to the target storage device; the instruction information is used to obtain the plurality of first images and the plurality of second images of the electrode determined online by the electrode quality control device within a historical time period and stored in the target storage device; Receive response information from the target storage device; the response information includes the plurality of first images and the plurality of second images.

5. A pole piece defect detection device, characterized in that: The device comprises: an acquisition module and a processing module; The acquisition module is used to acquire a plurality of first images and a plurality of second images of the pole piece, wherein the initial image labels of the first images indicate that there are no defects in the first images, and the initial image labels of the second images indicate that there are defects in the second images; The processing module is used to determine a third image and a fourth image among the plurality of first images based on a first preset model; the first preset model is an unsupervised model, the target image label of the third image indicates that there is no defect in the third image, and the target image label of the fourth image indicates that there is a defect in the fourth image; The processing module is also used to determine the fifth image and the sixth image among multiple second images based on a second preset model; the second preset model is a supervised model, the target image label of the fifth image indicates that there is no defect in the fifth image, and the target image label of the sixth image indicates that there is a defect in the sixth image.

6. The device according to claim 5, characterized in that The processing module is further used for: Acquire a first sample set; the first sample set includes a plurality of seventh images of the pole piece; no defects exist in the seventh images; The original first preset model is trained based on the multiple seventh images to obtain the first preset model.

7. The device according to claim 5, characterized in that The processing module is further used for: Acquire a second sample set; the second sample set includes a plurality of eighth images of the pole piece and a defect label of a defect in each eighth image; there is a defect in the eighth image; The original second preset model is trained based on the multiple eighth images and the defect labels of the defects in each eighth image to obtain the second preset model.

8. The device according to any one of claims 5 to 7, characterized in that: The acquisition module is specifically used for: Sending instruction information to the target storage device; the instruction information is used to obtain the plurality of first images and the plurality of second images of the electrode determined online by the electrode quality control device within a historical time period and stored in the target storage device; Receiving response information from the target storage device; The response information includes the plurality of first images and the plurality of second images.

9. A pole piece defect detection device, characterized in that: The pole piece defect detection device comprises: a processor, the processor is coupled to a memory, the memory is used to store programs or instructions, when the program or instructions are executed by the processor, the device executes the method as described in any one of claims 1 to 4.

10. A computer-readable storage medium having a computer program or instruction stored thereon, characterized in that: When the computer program or instructions are executed, the computer is caused to perform the method according to any one of claims 1 to 4.

Images