Battery failure detection device, method, and system

Abstract

A battery defect detection device according to one embodiment disclosed in this document may include a communication module, a processor, and a memory for storing a first artificial intelligence model, a second artificial intelligence model, and instructions, and when the instructions are executed by the processor, the battery defect detection device may be configured to acquire an image of a target item using the communication module, input the image of the target item to the first artificial intelligence model to classify the target item, and if the target item is successfully classified, input the image of the target item to the second artificial intelligence model, and determine whether the image of the target item corresponds to first data among the learning data of the first artificial intelligence model for successfully classifying the target item.

Description

[Technical field]

[0001] The present invention claims the benefit of priority based on Korean Patent Application No. 10-2022-0027072 filed on March 2, 2022, and all contents disclosed in the documents of the Korean patent application are incorporated by reference into this specification.

[0002] SUMMARY OF THE DISCLOSURE The embodiments disclosed herein relate to battery fault detection apparatus, methods, and systems. [Background technology]

[0003] In recent years, the demand for portable electronic products such as laptops, video cameras, and mobile phones has increased dramatically, and the development of electric vehicles, energy storage batteries, robots, satellites, and other products has progressed in earnest. As a result, research into high-performance batteries that can be repeatedly charged and discharged is being actively conducted.

[0004] Batteries currently in commercial use include nickel-cadmium batteries, nickel-metal hydride batteries, nickel-zinc batteries, and lithium batteries. Of these, lithium batteries are attracting attention due to the advantages they have over nickel-based batteries, such as the almost complete absence of memory effect, which allows them to be charged and discharged freely, a very low self-discharge rate, and a high energy density. Summary of the Invention [Problem to be solved by the invention]

[0005] During the manufacturing and assembly process of a battery, the sealing portion of the battery's exterior material may become deformed, wrinkled, or damaged, which may result in problems with the appearance of the battery pouch (e.g., dents at the corners, dents, scratches, wrinkles, electrolyte contamination, etc.). If the appearance of the pouch becomes defective, this may affect the reliability and / or safety of the battery, so it is important to detect defects in the battery (e.g., pouch) before the battery is distributed.

[0006] A visual inspection machine that is generally used to detect visual defects of batteries (e.g., pouches) collects and learns images of both good and defective types, and inspects batteries that have undergone manufacturing and assembly processes to determine whether or not there is a visual defect. However, if a new type of defect occurs, rather than the conventional type that has been learned, the battery may be determined to be good. In addition, since the rate of new type defects is low in the manufacturing and assembly processes of pouches, it is difficult to collect suitable images of defective products for learning the visual inspection machine.

[0007] Therefore, the present invention aims to compensate for the uncertainty of detecting new defect types and prevent the outflow of defective products. Other objects and advantages of the present invention will be understood from the following description and will become more apparent from the embodiments of the present invention. Also, it will be easily understood that the objects and advantages of the present invention can be realized by the means and combinations thereof as set forth in the claims. [Means for solving the problem]

[0008] A battery defect detection device according to one embodiment disclosed in this document may include a communication module, a processor, and a memory for storing a first artificial intelligence model, a second artificial intelligence model, and instructions, and when the instructions are executed by the processor, the battery defect detection device may be configured to acquire an image of a target item using the communication module, input the image of the target item to the first artificial intelligence model to classify the target item, and if the target item is successfully classified, input the image of the target item to the second artificial intelligence model, and determine whether the image of the target item corresponds to first data among the learning data of the first artificial intelligence model for successfully classifying the target item.

[0009] According to one embodiment, when the instructions are executed by the processor, the battery defect detection device can be configured to reclassify the target item as defective if the image does not correspond to the first data.

[0010] According to one embodiment, when the instructions are executed by the processor, the battery defect detection device can be configured to judge the image with second data for classifying the target item as defective if the image does not correspond to the first data.

[0011] According to one embodiment, when the instructions are executed by the processor, the battery fault detection device may be configured to train the first artificial intelligence model based on the second data. According to one embodiment, the image of the object may be captured in multiple channels by an image capture device.

[0012] According to one embodiment, when the instructions are executed by the processor, the battery defect detection device can be configured to obtain a similarity between the image and the first data using the second artificial intelligence model, and reclassify the target item as defective or normal based on the similarity.

[0013] According to one embodiment, when the instructions are executed by the processor, the battery defect detection device can be configured to judge the image with second data for classifying the target item as defective if the similarity of the image is less than a pre-set reference value.

[0014] According to one embodiment, when the instructions are executed by the processor, the battery fault detection device may be configured to train the first artificial intelligence model based on the second data.

[0015] A battery defect detection method according to one embodiment disclosed in this document may include an operation of acquiring an image of a target item, an operation of inputting the image of the target item into the first artificial intelligence model to classify the target item, and, if the target item is classified correctly, an operation of inputting the image of the target item into the second artificial intelligence model and determining whether the image of the target item corresponds to first data among the learning data of the first artificial intelligence model for correctly classifying the target item.

[0016] According to one embodiment, the method for detecting a defective battery may further include reclassifying the item as defective if the image does not correspond to the first data.

[0017] According to one embodiment, the method for detecting a defective battery may further include judging the image with second data for classifying the target item as defective if the image does not correspond to the first data.

[0018] According to an embodiment, the battery fault detection method may further include an operation of training the first artificial intelligence model based on the second data. According to one embodiment, the images of the item of interest may be acquired in multiple channels.

[0019] According to one embodiment, the battery defect detection method may further include an operation of obtaining a similarity between the image and the first data using the second artificial intelligence model, and an operation of reclassifying the target item as defective or normal based on the similarity.

[0020] According to one embodiment, the battery defect detection method may further include an operation of judging the image with second data for classifying the target product as defective if the similarity of the image is less than a pre-set reference value. According to an embodiment, the battery fault detection method may further include an operation of training the first artificial intelligence model based on the second data.

[0021] A battery defect detection system according to one embodiment disclosed in this document may include a battery defect detection device that acquires an image of a target item and classifies the target item using a first artificial intelligence model that is trained using the image of the target item and already stored learning data, and a new defect detection device that, if the target item is classified correctly, inputs the image of the target item into a second artificial intelligence model and determines whether the image of the target item corresponds to first data among the learning data of the first artificial intelligence model for correctly classifying the target item. Effect of the Invention

[0022] The battery defect detection device disclosed in this document can detect new types of defective products that are not included in the learning data already stored, thereby preventing the leakage of defective products. The battery defect detection device disclosed in this document can improve the accuracy of defect detection by learning images of new types of defective products.

[0023] The battery defect detection system disclosed in this document can use a new defect detection device together with a conventional defect detection device, thereby increasing the convenience of introducing the new defect detection device.

[0024] The effects of the battery fault detection device, method, and system disclosed in this document are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the disclosure of this document. [Brief description of the drawings]

[0025] [Figure 1] FIG. 1 illustrates an embodiment of a battery failure detection device disclosed in this document. [Diagram 2] 1 illustrates an artificial intelligence model according to one embodiment disclosed herein. [Figure 3a] FIG. 2 is a diagram illustrating an example of first data for correctly classifying a target item among the training data according to an embodiment disclosed in this document. [Figure 3b]FIG. 2 is a diagram illustrating an example of second data for classifying a target product as defective among the learning data according to one embodiment disclosed in this document. [Figure 4] 1 illustrates an artificial intelligence model according to one embodiment disclosed herein. [Diagram 5] 1 illustrates an artificial intelligence model and a process for detecting good and / or bad items of interest using the artificial intelligence model according to one embodiment disclosed herein. [Figure 6] FIG. 1 illustrates an example of an image successfully classified by an artificial intelligence model. [Figure 7] FIG. 1 illustrates a method for detecting a faulty battery according to one embodiment disclosed herein. [Figure 8] FIG. 1 illustrates a method for detecting a faulty battery according to one embodiment disclosed herein. [Figure 9] FIG. 1 illustrates a method for detecting a faulty battery according to one embodiment disclosed herein. [Figure 10] FIG. 1 illustrates a method for detecting a faulty battery according to one embodiment disclosed herein. [Figure 11]

[0023] In connection with the description of the drawings, the same or similar reference numerals may be used to refer to the same or similar components. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0026] Various embodiments of the present invention will now be described with reference to the accompanying drawings, although it should be understood that this is not intended to limit the present invention to the particular embodiments, but rather includes various modifications, equivalents, and / or alternatives of the embodiments of the present invention.

[0027] The various embodiments and terms used in this document are not intended to limit the technical features described in this document to a specific embodiment, but should be understood to include various modifications, equivalents, or alternatives of the embodiment. In connection with the description of the drawings, similar reference numerals may be used for similar or related components. The singular form of a noun corresponding to an item may include one or more of the said item, unless the relevant context clearly dictates otherwise.

[0028] In this document, each phrase such as "A or B," "at least one of A and B," "at least one of A or B," "A, B, or C," "at least one of A, B, and C," and "at least one of A, B, or C" may include any one or all possible combinations of the items listed with that phrase. Terms such as "first," "second," "first," "second," "A," "B," "(a)," or "(b)" may be used merely to distinguish the component from other components and do not limit the component in other respects (e.g., importance or order) unless specifically stated to the contrary.

[0029] In this document, when a (e.g., first) component is referred to as being "coupled," "coupled," or "connected" to another (e.g., second) component, with or without the terms "functionally" or "communicatively," or when a reference is made to "coupled" or "connected," this means that the component may be coupled to the other component directly (e.g., by wire), wirelessly, or through a third component.

[0030] According to an embodiment, the method according to the various embodiments disclosed in this document may be provided in a computer program product. The computer program product may be traded between a seller and a buyer as a commodity. The computer program product may be distributed in the form of a machine-readable storage medium (e.g., a compact disc read only memory (CD-ROM)) or distributed online (e.g., downloaded or uploaded) via an application store or directly between two user devices. In case of online distribution, at least a part of the computer program product may be at least temporarily stored or temporarily generated in a machine-readable storage medium such as a memory of a manufacturer's server, an application store's server, or an intermediary server.

[0031] According to various embodiments, each of the components described above (e.g., modules or programs) may include one or more entities, and some of the entities may be located separately in other components. According to various embodiments, one or more of the components described above may be omitted, or one or more other components or operations may be added. Alternatively or additionally, multiple components (e.g., modules or programs) may be integrated into one component. In this case, the integrated component may perform one or more functions of each of the multiple components in the same or similar manner as those performed by the multiple components before the integration. According to various embodiments, operations performed by modules, programs, or other components may be performed sequentially, in parallel, iteratively, or heuristically, or one or more of the operations may be performed in a different order, omitted, or one or more other operations may be added.

[0032] FIG. 1 is a diagram illustrating a battery defect detection device according to one embodiment disclosed herein. Referring to FIG. 1, a battery fault detection device 101 according to an embodiment disclosed in this document can be connected to an image capture device 103 and a user terminal 105 via wired and / or wireless communication.

[0033] In one embodiment, the connection between the battery fault detection device 101 and the image capture device 103 may be a communication connection via a wired and / or wireless network. In one embodiment, the wired network may be based on a local area network (LAN) communication or a power line communication. In one embodiment, the wireless network may be based on a short-range communication network (e.g., Bluetooth, wireless fidelity (WiFi), or infrared data association (IrDA)), or a long-range communication network (cellular network, 4G network, 5G network).

[0034] In other embodiments, the connection between the battery fault diagnosis device 101 and the image acquisition device 103 may be via an inter-device communication method (e.g., a bus, a general purpose input and output (GPIO), a serial peripheral interface (SPI), or a mobile industry processor interface (MIPI)).

[0035] In one embodiment, the connection between the battery fault diagnosis apparatus 101 and the user terminal 105 may be a communication connection via a wired and / or wireless network.

[0036] In one embodiment, the image capture device 103 can capture an image of an object (e.g., the battery unit 115). According to an embodiment, the object (e.g., the battery unit 115) can include a secondary battery. According to an embodiment, the secondary battery can include a pouch-type secondary battery, a prismatic secondary battery, and / or a cylindrical secondary battery.

[0037] The image capture device 103 can capture an image of the target item by photographing the target item. According to one embodiment, the image capture device 103 can be realized by a camera. According to one embodiment, the image capture device 103 can capture an image of the target item in multiple channels.

[0038] In one embodiment, the user terminal 105 may be a mobile device (e.g., a mobile phone, a laptop computer, a smart phone, a smart pad) or a personal computer (PC). In one embodiment, the user terminal 105 may be a terminal used by an administrator of the battery failure detection device 101.

[0039] In one embodiment, the battery fault detection device 101 may include a communication circuit 120, a memory 140, and a processor 150. According to an embodiment, the battery fault detection device 101 shown in Fig. 1 may further include at least one component (e.g., a display, an input device, or an output device) other than the components shown in Fig. 1.

[0040] In one embodiment, the communication circuitry 120 can establish a wired and / or wireless communication channel between the battery fault detection device 101 and the image capture device 103 and / or the user terminal 105, and transmit and receive data with the image capture device 103 and / or the user terminal 105 via the established communication channel.

[0041] In one embodiment, memory 140 may include volatile memory and / or non-volatile memory. In one embodiment, the memory 140 may store data used by at least one component (e.g., the processor 150) of the battery fault detection device 101. For example, the data may include the program 130 (or instructions associated therewith), input data, or output data. In one embodiment, the instructions, when executed by the processor 150, may cause the battery fault detection device 101 to perform an operation defined by the instructions.

[0042] In one embodiment, the program 130 may include one or more pieces of software (eg, an artificial intelligence model training portion 141, an image acquisition portion 143, a diagnosis portion 145, and one or more artificial intelligence models 161, 165).

[0043] In one embodiment, the processor 150 may include a central processing unit, an application processor, a graphics processing unit, a neural processing unit (NPU), an image signal processor, a sensor hub processor, or a communication processor.

[0044] In one embodiment, the processor 150 can execute software (e.g., an artificial intelligence model learning unit 141, an image acquisition unit 143, a diagnosis unit 145, an abnormality processing unit 147, a pre-processing unit 149, and one or more artificial intelligence models 161, 165) to control at least one other component (e.g., a hardware or software component) of the battery defect detection device 101 coupled to the processor 150 and perform various data processing or calculations.

[0045] Below, with reference to Figures 2 to 5, we will explain how the battery defect detection device 101 diagnoses abnormalities in a target item (e.g., a battery unit 115) via an artificial intelligence model learning unit 141, an image acquisition unit 143, a diagnosis unit 145, and one or more artificial intelligence models 161, 165.

[0046] First, learning of the artificial intelligence model FIG 2 illustrates an artificial intelligence model 161 according to an embodiment disclosed herein. FIG 3a illustrates a first data example for classifying a target product as normal among the training data according to an embodiment disclosed herein. FIG 3b illustrates a second data example for classifying a target product as defective among the training data according to an embodiment disclosed herein.

[0047] In one embodiment, the artificial intelligence model learning unit 141 can train the artificial intelligence model 161 based on the image 201 and the image 205 of the target product that have already been acquired. Here, the image 201 and the image 205 may be training data that have already been acquired. The image 201 may represent first data for classifying the target product as normal among the training data. The image 205 may represent second data for classifying the target product as defective among the training data. In one embodiment, the image 201 and the image 205 may include an entire image of the target product or an image of each part of the target product. In one embodiment, when the image 201 and the image 205 are composed of an entire image, the image 201 may be composed of an image of the target product classified as a normal product, and the image 205 may be composed of an image of the target product classified as a defective product. In one embodiment, when image 201 and image 205 are composed of partial images, image 201 can be composed of images classified as normal among images obtained by dividing an entire image of the target item into inspection items (e.g., dents, indentations, printed text, welding), and image 205 can be composed of images classified as defective among images obtained by dividing an entire image of the target item into inspection items.

[0048] For example, referring to FIG. 3a, the first data 310 may include images classified as normal among images divided for each inspection item. For example, image 311 may be an image with normal printed text, and image 312 may be an image with a dent degree within a normal range. Similarly, referring to FIG. 3b, the second data 350 may include images classified as defective among images divided for each inspection item. For example, image 351 may be an image with abnormal printed text, and image 352 may be an image with a dent or indentation degree outside the normal range.

[0049] In one embodiment, the artificial intelligence model 161 may be a model capable of multi-class classification (e.g., a model based on a convolutional neural network (CNN)). In one embodiment, the artificial intelligence model 161 may be trained to classify whether an object is normal or defective based on images 201, 205 of the object. In one embodiment, the artificial intelligence model 161 may be trained to classify whether each of the inspection items of the object is normal or defective based on images 201, 205 of the object.

[0050] In one embodiment, the artificial intelligence model 161 may include an input layer 210, a hidden layer 220, and an output layer 230. Here, the input layer 210 may receive images 201 and 205 of the target item that have already been acquired. The hidden layer 220 may have a structure in which a plurality of layers are sequentially connected. The output layer 230 may be a layer for outputting a multi-class classification result 209 for each of the images 201 and 205. In one embodiment, the multi-class classification result 209 may include probability values ​​for classes 241, 243, 245, and 247 for the images 201 and 205 according to the inspection item. For example, the class 241 may indicate the probability that the printed wording is normal, and the class 243 may indicate the probability that the printed wording is defective. As another example, the class 245 may indicate the probability that the degree of the dent or indentation is within a normal range, and the class 247 may indicate the probability that the degree of the dent or indentation is outside the normal range. The sum of the probability values ​​for classes 241, 243, 245, 247 may be one.

[0051] In one embodiment, the artificial intelligence model learning unit 141 can input the image 201 and the image 205 into the artificial intelligence model 161, and adjust the parameters of the hidden layer 220 so that the obtained multi-classification result 209 matches or is equal to or smaller than a standard difference value with the respective pre-classification results of the image 201 and the image 205.

[0052] Learning the second artificial intelligence model FIG. 4 illustrates an artificial intelligence model 165 according to one embodiment disclosed herein.

[0053] In one embodiment, the artificial intelligence model learning unit 141 can train the artificial intelligence model 165 based on the image 201 of the target item that has already been acquired. Here, the image used to train the artificial intelligence model 165 may be the first data for correctly classifying the target item among the training data.

[0054] In one embodiment, the artificial intelligence model 165 may be a model capable of multi-class classification (e.g., a model based on CNN). In one embodiment, the artificial intelligence model 165 may be trained to output a similarity between the first data (i.e., image 201) of the target product and a normal image for each inspection item. Here, the similarity may mean a maximum value of a probability value according to the multi-class classification result obtained by the artificial intelligence model 165.

[0055] In one embodiment, the artificial intelligence model 165 can be trained to compare the similarity based on the first data (i.e., image 201) for the target product with a reference value to classify whether the image is in-distribution (i.e., corresponding to an image of an already learned good product) or out-of-distribution (i.e., corresponding to an image of a new type of defective product).

[0056] In one embodiment, the artificial intelligence model 165 may include an input layer 410, a hidden layer 420, and an output layer 430. Here, the input layer 410 may receive an image 201 of the target item that has already been acquired. The hidden layer 420 may have a structure in which a plurality of layers are sequentially connected. The output layer 430 may be a layer for outputting a multi-item classification result 409 for each image 201. In one embodiment, the multi-item classification result 409 may include a probability value for a class in which the image 201 corresponds to an inspection item. The classes set in the artificial intelligence model 165 may be fewer than the classes set in the artificial intelligence model 161. For example, the artificial intelligence model 165 may include only a class indicating normality. For example, the artificial intelligence model 165 may include a class indicating the probability that the printed wording is normal or a class indicating the probability that the printed wording is normal. The sum of the probability values ​​for the classes of the artificial intelligence model 165 may be 1.

[0057] In one embodiment, the artificial intelligence model learning unit 141 can adjust the parameters of the hidden layer 420 by inputting the image 201 into the artificial intelligence model 165 so that the maximum probability value 441 according to the multi-class classification result 409 obtained exceeds the reference value 449.

[0058] In one embodiment, the artificial intelligence model training unit 141 may test the artificial intelligence model 165 using the image 205. For example, the artificial intelligence model training unit 141 may check whether the maximum value 445 of the probability value according to the multi-class classification result 409 of the image 205 is equal to or less than the reference value 449. In one embodiment, if the maximum value 445 of the probability value according to the multi-class classification result 409 of the image 205 exceeds the reference value 449, the artificial intelligence model training unit 141 may adjust the reference value 449 of the artificial intelligence model 165. For example, the artificial intelligence model training unit 141 may adjust the reference value 449 based on a baseline algorithm or an out of distribution detector for neural networks (ODIN) algorithm. Here, the adjustment of the reference value 449 may be distinguished from the learning process for adjusting the parameters of the hidden layer 420.

[0059] Detecting defects in target products Figure 5 illustrates a process for detecting good and / or bad items among objects using artificial intelligence model 161 and artificial intelligence model 165 according to an embodiment disclosed herein. Figure 6 illustrates an example image successfully classified by the artificial intelligence model.

[0060] In one embodiment, the processor 120 can detect whether the target product is a good product and / or a defective product through the diagnosis unit 145. The processor 120 can classify the target product as normal or defective in order to detect whether the target product is a good product and / or a defective product through the diagnosis unit 145. The processor 120 can classify the target product as normal or defective based on already stored learning data through the diagnosis unit 145. Hereinafter, the operation of the processor 120 can be understood to be performed through the diagnosis unit 145 unless otherwise specified.

[0061] In one embodiment, the processor 120 can classify the object using the image 501 of the object and the already stored learning data. The processor 120 can compare the image 501 of the object captured by the image capture device 103 with the already stored learning data. Here, the image 501 of the object captured by the image capture device 103 can be captured by the processor 120 via the image capture unit 143. In one embodiment, the image capture unit 143 can include a network driver for controlling the communication circuit 120. According to an embodiment, the processor 120 can compare the image 501 of the object with the already stored learning data to classify the object. According to an embodiment, the processor 120 can classify the object based on the artificial intelligence model 161 to which the already stored learning data is applied.

[0062] According to an embodiment, the learning data may include at least one each of data for classifying the target item as normal and data for classifying the target item as defective. The first data may refer to data among the learning data for classifying the target item as normal. According to an embodiment, the first data may include an image of a non-defective item. According to an embodiment, the processor 120 may use the first data to learn a criterion for classifying the target item as normal.

[0063] The processor 120 can classify the target product as normal or defective based on the already stored learning data. The target product may refer to a target product that is correctly classified when the processor 120 classifies the target product based on the already stored learning data. According to an embodiment, the processor 120 may correctly classify a new type of defective product that has not been learned by the already stored learning data. In this case, the target product may include not only a good product but also a new type of defective product. For example, the image 610 may be an image of a conventional type of normal product, and the image 650 may be an image of a new type of defective product. In this case, when the image 610 is input to the artificial intelligence model 161, the target product for the image 610 may be classified as normal (i.e., a good product). As another example, when the image 650 of a new type of defective product is input to the artificial intelligence model 161, the target product for the image 650 may also be classified as normal (i.e., a good product).

[0064] The processor 120 may obtain a target image (i.e., images 610 and 650) corresponding to the target object. The processor 120 may determine characteristics of the target image using the first data and the target image. The processor 120 may detect a new type of defective product using the characteristics of the target image. According to an embodiment, the processor 120 may reclassify the target object as normal or defective using the characteristics of the target image. According to an embodiment, the processor 120 may classify the object based on the artificial intelligence model 165 to which the first data of the already stored learning data is applied. In one embodiment, the processor 120 may detect a new type of defective product by comparing a similarity obtained by inputting each of the target images (i.e., images 610 and 650) into the artificial intelligence model 165 with a reference value. Here, the similarity may mean a maximum value of a probability value according to a multi-item classification result obtained by the artificial intelligence model 165. For example, in the case of the image 610 having a similarity value exceeding a reference value among the images 610 and 650, the processor 120 may determine the target product as normal. As another example, if the similarity between the images 610 and 650 is less than a reference value, the processor 120 may determine that the target product is defective.

[0065] The target products that are reclassified as defective may correspond to a new type of defective product. The processor 120 may detect the new type of defective product by performing a reclassification process on the target products.

[0066] According to an embodiment, the processor 120 can compare whether the target image corresponds to the first data. If the target image does not correspond to the first data, the processor 120 can reclassify the target object as defective.

[0067] According to an embodiment, the processor 120 may determine a similarity between the target image and the first data. The processor 120 may determine a similarity between the target image and the first data by a previously set method. The processor 120 may reclassify the target object as defective or normal based on the similarity between the target image and the first data. According to an embodiment, the processor 120 may reclassify the target object as defective if the similarity is less than a previously set reference value, and may reclassify the target object as normal if the similarity is equal to or greater than the reference value. According to an embodiment disclosed herein, the processor 120 may determine a similarity between the target image and the first data and reclassify the target object as defective or normal based on the similarity, but is not limited thereto. According to another embodiment, the processor 120 may determine a dissimilarity between the target image and the first data and reclassify the target object as defective or normal based on the dissimilarity.

[0068] The processor 120 may determine a target image when the target object is reclassified as defective using the second data. The second data may correspond to an image for classifying the object as defective. The second data may include an image of a new type of defective product that is not included in the already stored learning data. According to an embodiment, the processor 120 may include the second data in the learning data. The processor 120 may learn a new type of defect by including the second data in the learning data. The processor 120 may retrain the artificial intelligence model 161 based on an image (e.g., image 650) newly included in the second data.

[0069] According to an embodiment, the processor 120 may determine the target image from the second data when the target image does not correspond to the first data. According to another embodiment, the processor 120 may determine the target image from the second data when the similarity of the target image is less than a pre-set reference value.

[0070] FIG. 7 illustrates a method for detecting a defective battery according to one embodiment disclosed herein. Referring to FIG. 7, a battery defect detection method according to one embodiment disclosed in this document may include a step of acquiring an image of a target item (S100), a step of classifying the target item using stored learning data (S110), a step of acquiring a target image corresponding to a target object (S120), and a step of determining characteristics of the target image using first data and the target image (S130).

[0071] Hereinafter, the method for detecting a defective battery will be described in detail with reference to FIG. 1. For convenience of explanation, the contents overlapping with those described above will be omitted or will be described briefly. In step S100, the battery defect detection device 101 can acquire an image of the target product. According to an embodiment, the target product can include a secondary battery. According to an embodiment, in step S100, the battery defect detection device 101 can acquire images of the target product through multiple channels. Step S100 can be performed by the image acquisition unit 143 of the battery defect detection device 101.

[0072] In step S110, the battery defect detection device 101 can classify the target item using the stored learning data. According to an embodiment, the battery defect detection device 101 can compare an image of the target item with already-stored learning data and classify the target item according to the comparison result. According to an embodiment, the battery defect detection device 101 can classify the target item based on the artificial intelligence model 161 to which the already-stored learning data is applied. Step S110 can be performed by the processor 120 of the battery defect detection device 101.

[0073] According to the embodiment, the learning data may include at least one data for classifying the target product as normal or defective. The first data may mean data for classifying the target product as normal among the learning data. According to the embodiment, the first data may include an image of a non-defective product. According to the embodiment, the battery defect detection device 101 may use the first data to learn a criterion for classifying the target product as normal.

[0074] In step S110, the battery defect detection device 101 can classify the target product as normal or defective based on the learning data. The target product correctly classified in step S710 can be referred to as a target product. According to the embodiment, the battery defect detection device 101 can classify the target product as normal or defective based on the learning data, and may correctly classify a new type of defective product that has not been learned by the learning data. In this case, the target product may include not only a good product but also a new type of defective product.

[0075] In step S120, the battery defect detection device 101 can acquire a target image corresponding to the target object. According to an embodiment, the battery defect detection device 101 can select a target image corresponding to the target object from images of the target objects already acquired. Step S120 can be performed by the processor 120 of the battery defect detection device 101.

[0076] In step S130, the battery defect detection device 101 may determine characteristics of the target image using the first data and the target image. The first data may mean data for correctly classifying the target product among the learning data. According to the embodiment, the first data may include an image of a non-defective product. According to the embodiment, the battery defect detection device 101 may determine characteristics of the target image using the first data and the target image based on the artificial intelligence model 165 to which the first data is applied. Step S130 may be performed by the processor 120 of the battery defect detection device 101.

[0077] In step S130, the battery defect detection device 101 can determine characteristics of the target image. According to an embodiment, the battery defect detection device 101 can determine whether the target image corresponds to the first data. According to another embodiment, the battery defect detection device 101 can determine a similarity between the target image and the first data.

[0078] The battery defect detection device 101 can reclassify the target object into normal or defective based on the characteristics of the target image determined in step S130. A specific method for reclassifying the target object into normal or defective based on the characteristics of the target image determined by the battery defect detection device 101 will be described later with reference to Figs. 8 and 9.

[0079] FIG. 8 illustrates a method for detecting a defective battery according to one embodiment disclosed herein. Referring to FIG. 8, a battery defect detection method according to one embodiment disclosed herein may further include a step of determining whether a target image corresponds to first data (S200), and reclassifying the target item as defective if the target image does not correspond to the first data (S210), a step of judging the target image based on second data (S220), and / or a step of including the second data in the learning data (S230).

[0080] Hereinafter, the battery defect detection method will be described in detail with reference to FIG. 1 and FIG. 7. For convenience of explanation, the contents overlapping with those described above will be omitted or will be described briefly. In step S200, the battery defect detection apparatus 101 may determine whether the target image corresponds to the first data. Step S200 may substantially correspond to a specific example of step S130 of FIG.

[0081] In step S200, if the battery defect detection device 101 determines that the target image does not correspond to the first data, the battery defect detection device 101 may reclassify the target object as defective (S210). According to the embodiment, the target object is an object normally classified by the battery defect detection device 101 based on the learning data, but the target object may also include a new type of defective product not included in the learning data. In step S200, the battery defect detection device 101 may detect a new type of defective product included in the target object by determining whether the target object corresponds to the first data. Step S200 may be performed by the processor 120 of the battery defect detection device 101.

[0082] In step S200, if it is determined that the target image corresponds to the first data, the battery defect detection device 101 can successfully reclassify the target object.

[0083] In step S200, if it is determined that the target image does not correspond to the first data, the battery defect detection device 101 may reclassify the target object as defective (S210).

[0084] In step S210, the target object product reclassified as defective by the battery defect detection device 101 may correspond to a new type of defective product not included in the learning data. Step S210 may be performed by the processor 120 of the battery defect detection device 101.

[0085] In step S220, the battery defect detection device 101 can determine a target image of the target object that the battery defect detection device 101 reclassified as defective in step S210 from the second data. According to an embodiment, the second data can include an image of a defective product of a new type to the battery defect detection device 101. Step S220 can be performed by the processor 120 of the battery defect detection device 101.

[0086] In step S230, the battery defect detection device 101 can include the second data in the learning data. According to an embodiment, the second data can correspond to an image for the battery defect detection device 101 to classify the target product as defective. According to an embodiment, the second data can include an image of a new type of defective product that is not included in the already stored learning data. In step S230, the battery defect detection device 101 can learn a new type of defect by including the second data in the learning data. Step S230 can be performed by the processor 120 of the battery defect detection device 101.

[0087] FIG. 9 illustrates a method for detecting a defective battery according to one embodiment disclosed herein. Referring to FIG. 9, the battery defect detection method according to one embodiment disclosed in this document may further include a step of determining a similarity between the target image and the first data (S300), and a step of reclassifying the target object as defective or normal based on the similarity (S310).

[0088] The battery defect detection method will be specifically described below with reference to FIG. 1 and FIGS. 7 to 8. For convenience of explanation, the contents that overlap with those described above will be omitted or described briefly.

[0089] In step S300, the battery defect detection apparatus 101 may determine a similarity between the target image and the first data. Step S300 may substantially correspond to a specific example of step S130 of Fig. 7. Step S300 may be performed by the processor 120 of the battery defect detection apparatus 101.

[0090] In step S300, the battery defect detection device 101 can determine the similarity between the target image and the first data by a previously set method. According to an embodiment, the battery defect detection device 101 can determine the similarity between the target image and the first data by converting it into a numerical value. According to the embodiment disclosed in this document, the battery defect detection device 101 determines the similarity between the target image and the first data, but is not limited thereto, and may determine the dissimilarity between the target image and the first data.

[0091] In step S310, the battery defect detection device 101 may reclassify the target object as defective or normal based on the similarity between the target image and the first data. The battery defect detection device 101 may reclassify the target object as defective or normal based on the degree to which the target image is similar to the first data. According to an embodiment, the battery defect detection device 101 may set a reference value for the similarity between the target image and the first data. According to an embodiment, the battery defect detection device 101 may reclassify the target object as defective or normal by comparing the similarity determined in step S300 with the reference value.

[0092] According to the embodiment, when the similarity between the target image and the first data is equal to or greater than a preset reference value, the battery defect detection device 101 can correctly reclassify the target object corresponding to the target image.

[0093] According to the embodiment, when the similarity between the target image and the first data is less than a preset reference value, the battery defect detection device 101 may reclassify the target object corresponding to the target image as defective. In this case, the target object reclassified as defective may correspond to a new type of defective product.

[0094] FIG. 10 illustrates a method for detecting a defective battery according to one embodiment disclosed herein. Referring to FIG. 10, a battery defect detection method according to one embodiment disclosed herein may further include a step of determining a similarity between a target image and first data (S400), a step of comparing the similarity between the target image and the first data with a previously set reference value (S410), a step of judging the target image with the second data if the similarity between the target image and the first data is less than the reference value (S420), and / or a step of including the second data in learning data (S430).

[0095] The battery defect detection method will be specifically described below with reference to FIG. 1 and FIGS. 7 to 9. For convenience of explanation, the contents that overlap with those described above will be omitted or described briefly.

[0096] Step S400 may correspond to a step substantially similar to step S300 of Fig. 4. Steps S410, S420, and / or S430 may be performed simultaneously with step S310 of Fig. 4 or before or after step S310.

[0097] In step S410, the battery defect detection device 101 may determine whether the similarity between the target object and the first data is less than a previously set reference value. The reference value may correspond to a value set as a reference for the battery defect detection device 101 to reclassify the target object as normal or defective based on the similarity between the target image and the first data.

[0098] In step S410, the target image determined by the battery defect detection device 101 to have a similarity equal to or greater than the reference value may include an image of a non-defective product. According to an embodiment, the battery defect detection device 101 may not perform an additional operation on the target image determined to have a similarity equal to or greater than the reference value. According to another embodiment, the battery defect detection device 101 may add the target image determined to have a similarity equal to or greater than the reference value to the first data for inclusion. According to an embodiment disclosed herein, the battery defect detection device 101 may improve the accuracy of defect detection by adding the target image determined to have a similarity equal to or greater than the reference value to the first data for learning.

[0099] In step S420, the battery defect detection device 101 may determine, using the second data, a target image whose similarity is determined to be less than a reference value. According to an embodiment, the target image determined using the second data may include an image for a new type of defective product. According to an embodiment, in step S420, the battery defect detection device 101 may determine, using the second data, a target image corresponding to a new type of defective product.

[0100] In step S430, the battery defect detection device 101 can include the second data in the learning data. According to an embodiment, the second data can correspond to an image for the battery defect detection device 101 to classify the target product as defective. According to an embodiment, the second data can include an image of a new type of defective product that is not included in the already stored learning data. In step S430, the battery defect detection device 101 can learn the new type of defect by including the second data in the learning data.

[0101] FIG. 11 illustrates a battery failure detection system according to one embodiment disclosed herein. 11, a battery defect detection system 1000 according to an embodiment disclosed in the present document may include an inspection device 1100 and a novel defect detection device 1200. Here, the inspection device 1000 may be based on an artificial intelligence model 161, and the novel defect detection device 1100 may be based on an artificial intelligence model 165.

[0102] The battery defect detection system will be specifically described below with reference to Figs. 1 and 7 to 10, and for the sake of convenience, content that overlaps with the above content will be omitted or described briefly.

[0103] The inspection device 1100 can acquire an image of an object and classify the object using the image of the object and previously stored training data. According to an embodiment, the inspection device 1100 can include an image acquisition unit 1110 and a processor 1120.

[0104] The inspection device 1100 can acquire an image of the target item. According to an embodiment, the inspection device 1100 can acquire an image of the target item via the image acquisition unit 1110. According to an embodiment, the image acquisition unit 1110 can acquire an image from a remote image acquisition device that photographs the target item to acquire the image of the target item. According to an embodiment, the remote image acquisition device can be realized by a camera. According to an embodiment, the remote image acquisition device can acquire an image of the target item in multiple channels.

[0105] The inspection device 1100 may classify the object using an image of the object and previously stored learning data. According to an embodiment, the inspection device 1100 may classify the object via the processor 1120. According to an embodiment, the processor 1120 may compare the image of the object with previously stored learning data. According to an embodiment, the processor 1120 may classify the object based on a deep learning model to which the previously stored learning data is applied.

[0106] The processor 1120 can classify the target product as normal or defective based on the already stored learning data. The processor 1120 can classify the target product based on the already stored learning data, and set the normally classified target product as the target target product. According to an embodiment, the processor 1120 can normally classify a new type of defective product that is not included in the learning data among the target products, and the target target products can include both good products and new type of defective products.

[0107] The new defect detection device 1200 may acquire a target image corresponding to a target object correctly classified by the inspection device 1100, and determine characteristics of the target image using first data for correctly classifying the object from among the learning data and the target object image. According to an embodiment, the new defect detection device 1200 may include a target image acquisition unit 1210 and a processor 1220.

[0108] The new defect detection device 1200 may receive a target object from the inspection device 1100. The new defect detection device 1200 may acquire a target image corresponding to the target object. According to an embodiment, the new defect detection device 1200 may acquire the target image via a target image acquisition unit 1210.

[0109] According to an embodiment, the new defect detection device 1200 may receive a target image corresponding to a target object from the inspection device 1100. The target image acquisition unit 1210 may select a target image corresponding to the target object from among images of the object objects already acquired by the inspection device 1100. According to an embodiment, the target image acquisition unit 1210 may be integrated with the processor 1220 and implemented as one module.

[0110] According to another embodiment, the target image acquiring unit 1210 may acquire a target image by taking an image of a target object transferred from the inspection device 1100. In this case, the target image acquiring unit 1210 may be realized by a camera.

[0111] The new defect detection device 1200 may determine characteristics of the target image using the first data and the target image. The first data may mean data among the learning data for the inspection device 1100 to correctly classify the target product. According to an embodiment, the new defect detection device 1200 may share the learning data with the inspection device 1100. According to an embodiment, the new defect detection device 1200 may determine characteristics of the target image via the processor 1220.

[0112] In some embodiments, the processor 1220 may determine whether the target image corresponds to the first data. In other embodiments, the processor 1220 may determine a similarity between the target image and the first data.

[0113] The processor 1220 may determine the characteristics of the target image and detect a new type of defect based on the determined characteristics. According to an embodiment, the processor 1220 may determine the target image with the second data if the target image does not correspond to the first data. According to another embodiment, the processor 1220 may determine the target image with the second data if the similarity between the target image and the first data is less than a pre-set reference value. According to an embodiment disclosed herein, the target object corresponding to the target image determined with the second data may correspond to a new type of defective product. The processor 1220 may detect the new type of defective product by detecting the target object corresponding to the target image.

[0114] According to the embodiment, the processor 1220 can include the second data in the learning data. The processor 1220 can learn a new type of defect by including the second data in the learning data. The new defect detection device 1200 can share the second data with the inspection device 1100. According to the embodiment, the inspection device 1100 can learn the second data, and by learning the second data, the accuracy of defect detection can be improved.

[0115] The terms "comprise", "constitute", or "have" described above, unless otherwise specified to the contrary, mean that the relevant element can be contained within, and should not be interpreted as excluding other elements, but as including other elements. All terms, including technical or scientific terms, have the same meaning as commonly understood by a person of ordinary skill in the art to which the embodiments disclosed in this document belong, unless otherwise defined. Commonly used terms, such as terms defined in a dictionary, should be interpreted to be consistent with the contextual meaning of the relevant art, and should not be interpreted in an idealized or overly formal sense, unless expressly defined in this document.

[0116] The above description is merely an illustrative example of the technical ideas disclosed in this document, and a person having ordinary knowledge in the technical field to which the embodiments disclosed in this document belong can make various modifications and variations within the scope of the essential characteristics of the embodiments disclosed in this document. Therefore, the embodiments disclosed in this document are intended to explain, not to limit, the technical ideas of the embodiments disclosed in this document, and such embodiments do not limit the scope of the technical ideas disclosed in this document. The scope of protection of the technical ideas disclosed in this document must be interpreted according to the claims below, and all technical ideas within the equivalent scope should be interpreted as being included in the scope of rights of this document. [Explanation of symbols]

[0117] 101 Battery defect detection device 103 Image acquisition device 105 User terminal 115 Battery Unit 120 Communication Circuit 130 Programs 140 Memory 141 Artificial Intelligence Model Learning Department 143 Image Acquisition Unit 145 Diagnostics Department 147 Abnormality Processing Unit 149 Pretreatment section 150 processors 161 Artificial Intelligence Model 165 Artificial Intelligence Model 201 images 205 images 209 Multinomial classification results 210 Input Layer 220 Layer 230 Output Layer 241 Class 243 Class 245 Class 247 Class 310 First Data 311 images 312 images 350 Second Data 351 images 352 images 409 Multinomial classification results 410 Input Layer 420 Layer 430 Output Layer 441 Maximum 445 Maximum 449 Reference Value 501 images 610 images 650 images 1000 Battery Defect Detection System 1100 Inspection Equipment 1200 New defect detection device 1110 Image acquisition unit 1120 Processor 1210 Target image acquisition unit 1220 Processor

Claims

1. A battery defect detection device, comprising: A communication module; A processor; a memory for storing a first artificial intelligence model, a second artificial intelligence model, and instructions; Including, The instructions, when executed by the processor, cause the battery fault detection device to: Acquiring an image of a target item using the communication module; inputting the image of the object into the first artificial intelligence model to classify the object; A battery defect detection device configured to, if the target item is classified correctly, input the image of the target item into the second artificial intelligence model and determine whether the image of the target item corresponds to first data among the learning data of the first artificial intelligence model for correctly classifying the target item.

2. The instructions, when executed by the processor, cause the battery fault detection device to: The battery defect detection device of claim 1 , configured to reclassify the target item as defective if the image does not correspond to the first data.

3. The instructions, when executed by the processor, cause the battery fault detection device to: The battery defect detection device of claim 1 , configured to judge the image with second data for classifying the target item as defective if the image does not correspond to the first data.

4. The instructions, when executed by the processor, cause the battery fault detection device to: The battery defect detection device of claim 3 , configured to train the first artificial intelligence model based on the second data.

5. The battery defect detection device according to claim 1 , wherein the image of the target item is acquired by an image acquisition device using multiple channels.

6. The instructions, when executed by the processor, cause the battery fault detection device to: Obtaining a similarity between the image and the first data by the second artificial intelligence model; The battery defect detection device according to claim 1 , configured to reclassify the target item as defective or normal based on the similarity.

7. The instructions, when executed by the processor, cause the battery fault detection device to: The battery defect detection device of claim 6, configured to judge the image with second data for classifying the target product as defective if the similarity of the image is less than a pre-set reference value.

8. The instructions, when executed by the processor, cause the battery fault detection device to: The battery defect detection device of claim 7 , configured to train the first artificial intelligence model based on the second data.

9. A method of operating a battery failure detection device, comprising: An operation of acquiring an image of a target item; inputting the image of the target item into the first artificial intelligence model to classify the target item; If the target item is classified correctly, inputting the image of the target item into the second artificial intelligence model and determining whether the image of the target item corresponds to first data among the learning data of the first artificial intelligence model for correctly classifying the target item; 4. A method of operation comprising:

10. The method of claim 9 , further comprising the act of reclassifying the item as defective if the image does not correspond to the first data.

11. 10. The method of claim 9, further comprising the act of determining, if the image does not correspond to the first data, a second data for classifying the item as defective.

12. The method of claim 11 further comprising the act of training the first artificial intelligence model based on the second data.

13. The method of claim 9 , wherein the images of the item are acquired in multiple channels.

14. An operation of obtaining a similarity between the image and the first data by the second artificial intelligence model; The method of claim 9 , further comprising: reclassifying the object as defective or normal based on the similarity.

15. The method of claim 14, further comprising judging the image with second data for classifying the target item as defective if the similarity of the image is less than a pre-established reference value.

16. The method of claim 15 further comprising the act of training the first artificial intelligence model based on the second data.

17. a battery defect detection device that acquires an image of a target product and classifies the target product using a first artificial intelligence model that is trained using the image of the target product and previously stored training data; a new defect detection device that, if the target product is classified correctly, inputs the image of the target product into the second artificial intelligence model and determines whether the image of the target product corresponds to first data for correctly classifying the target product among the learning data of the first artificial intelligence model; A battery failure detection system including:

Images