Weld seam defect detection method and apparatus, and electronic device and storage medium

Abstract

Provided are a weld seam defect detection method and apparatus, and an electronic device and a storage medium. The method comprises: when it is determined that an anomaly occurs during the welding of a cylindrical battery, acquiring a two-dimensional sealing-nail weld seam image of the cylindrical battery (S110), wherein the two-dimensional sealing-nail weld seam image is a two-dimensional image of a sealing-nail weld seam of the cylindrical battery, which image is collected by means of irradiating the sealing-nail weld seam by using a shadowless light source; inputting the two-dimensional sealing-nail weld seam image into a weld seam defect detection model, such that the weld seam defect detection model outputs, on the basis of the two-dimensional sealing-nail weld seam image, a corresponding weld seam defect detection result (S120), wherein the weld seam defect detection model is obtained by means of training a neural network on the basis of historical sealing-nail weld seam defect images; and receiving the weld seam defect detection result that is output by the weld seam defect detection model (S130). On the basis of a weld seam defect detection model, weld seam defect detection is performed on a two-dimensional sealing-nail weld seam image of a sealing-nail weld seam of a cylindrical battery, thereby improving the detection efficiency of performing weld seam defect detection on a sealing-nail weld seam of a cylindrical battery.

Description

Weld defect detection method, detection device, electronic equipment and storage medium

[0001] The present application claims priority to the Chinese patent application No. 202410414098.9, filed on April 8, 2024 in the China Patent Office, and entitled "Weld defect detection method, detection device, electronic equipment and storage medium", the whole content of which is incorporated herein by reference. TECHNICAL FIELD

[0002] The present application belongs to the technical field of defect detection, and in particular relates to a weld defect detection method, a detection device, an electronic equipment and a storage medium. BACKGROUND

[0003] In the production process of power lithium batteries, due to process and equipment reasons, there will be certain defects, which need to be detected by various detection means to improve the battery yield. In each link of the battery production line, the detection of the weld defects of the sealing pin is a crucial link, and the effectiveness of the detection result ensures the safety quality of the battery and the safety of the battery out of the factory.

[0004] At present, in the process of detecting the weld defects of the sealing pin of the battery, the weld defects of the sealing pin under the dome light irradiation are mainly judged by artificial, which leads to low detection efficiency of the weld defect detection of the sealing pin of the battery.

[0005] APPLICATION CONTENT

[0006] Therefore, the embodiments of the present application provide a weld defect detection method, a detection device, an electronic equipment and a storage medium to overcome the above problems of the prior art. TECHNICAL SOLUTION

[0007] The technical scheme adopted by the embodiments of the present application is:

[0008] In the first aspect, the embodiments of the present application provide a weld defect detection method, which comprises:

[0009] In the case that the welding process of the cylindrical battery is determined to be abnormal, a two-dimensional sealing pin weld image of the cylindrical battery is acquired, the two-dimensional sealing pin weld image being a two-dimensional image of the sealing pin weld of the cylindrical battery collected under the irradiation of the shadowless light source;

[0010] The two-dimensional sealing pin weld image is input to a weld defect detection model, so that the weld defect detection model outputs a corresponding weld defect detection result according to the two-dimensional sealing pin weld image, the weld defect detection model being obtained by training a neural network based on historical sealing pin weld defect images;

[0011] receive a weld defect detection result output by the weld defect detection model.

[0012] In some optional embodiments, the method further comprises:

[0013] In the case where it is determined that the welding process of the cylindrical battery is abnormal, the cylindrical battery is controlled to move to a first battery position.

[0014] It is determined whether the first battery position is within a two-dimensional image acquisition range of the two-dimensional camera.

[0015] In the case where it is determined that the first battery position is within the two-dimensional image acquisition range, the two-dimensional camera is controlled to acquire a two-dimensional image of the sealing pin weld, thereby obtaining a two-dimensional sealing pin weld image.

[0016] In some optional embodiments, the weld defect detection result includes a first detection result of a sealing pin weld defect type of the sealing pin weld, and the weld defect detection method further comprises:

[0017] In the case where the first detection result output by the weld defect detection model is received, a three-dimensional sealing pin weld image of the cylindrical battery is acquired.

[0018] According to the three-dimensional sealing pin weld image, a weld defect value of the sealing pin weld defect type is determined.

[0019] In some optional embodiments, the weld defect detection method further comprises:

[0020] In the case where the weld defect value is greater than a defect threshold value of the sealing pin weld defect type, it is determined that the weld defect detection of the sealing pin weld is unqualified.

[0021] In the case where the weld defect value is less than or equal to the defect threshold value, it is determined that the weld defect detection of the sealing pin weld is qualified.

[0022] In some optional embodiments, the weld defect detection method further comprises:

[0023] In the case where it is determined that the weld defect detection of the sealing pin weld is unqualified according to the weld defect value, a reminder information is generated and sent to a specified client.

[0024] In some optional embodiments, before the three-dimensional sealing pin weld image of the cylindrical battery is acquired, the weld defect detection method further comprises:

[0025] The cylindrical battery is controlled to move to a second battery position.

[0026] It is determined whether the second battery position is within a three-dimensional image acquisition range of the three-dimensional camera.

[0027] acquire a three-dimensional sealing nail weld image of the cylindrical battery, including:

[0028] In a case where it is determined that the second battery position is within the three-dimensional image acquisition range, control the three-dimensional camera to acquire a three-dimensional image of the sealing nail weld to obtain the three-dimensional sealing nail weld image.

[0029] In some optional embodiments, the weld defect detection result includes a second detection result for characterizing that the sealing nail weld does not have a sealing nail weld defect, and the weld defect detection method further includes:

[0030] In a case where the second detection result output by the weld defect detection model is received, it is determined that the sealing nail weld passes the weld defect detection.

[0031] In a second aspect, the embodiments of the present application provide a weld defect detection device, the weld defect detection device, including:

[0032] The two-dimensional image acquisition module is configured to acquire a two-dimensional sealing nail weld image of the cylindrical battery in a case where it is determined that the welding process of the cylindrical battery is abnormal, the two-dimensional sealing nail weld image being a two-dimensional image of the sealing nail weld of the cylindrical battery acquired under illumination of a shadowless lamp;

[0033] The two-dimensional image input module is configured to input the two-dimensional sealing nail weld image to the weld defect detection model, so that the weld defect detection model outputs a corresponding weld defect detection result according to the two-dimensional sealing nail weld image, the weld defect detection model being obtained based on training of a neural network by using historical sealing nail weld defect images;

[0034] The receiving module is configured to receive a weld defect detection result output by the weld defect detection model.

[0035] In a third aspect, the embodiments of the present application provide an electronic device, including:

[0036] a memory;

[0037] one or more processors coupled to the memory;

[0038] one or more application programs, wherein the one or more application programs are stored in the memory and configured to be executed by the one or more processors, and the one or more application programs are configured to execute the weld defect detection method provided in the first aspect.

[0039] In a fourth aspect, the embodiments of the present application provide a computer readable storage medium, the computer readable storage medium storing program codes, the program codes being executable by a processor to execute the weld defect detection method provided in the first aspect.

[0040] In a fifth aspect, an embodiment of the present application provides a computer program product, which, when running on a computer device, causes the computer device to execute the welding seam defect detection method according to the first aspect. Advantages

[0041] The first aspect provided by the embodiment of the present application has the advantages that in the case where it is determined that the welding process of the cylindrical battery is abnormal, the two-dimensional sealing nail welding seam image of the sealing nail welding seam of the cylindrical battery is subjected to welding seam defect detection according to the welding seam defect detection model, so that the manual defect detection of the sealing nail welding seam is avoided, and the detection efficiency of the welding seam defect detection of the sealing nail welding seam of the cylindrical battery is improved. In the case where the sealing nail welding seam has defects, the welding process of the cylindrical battery is bound to be abnormal, so that the welding seam defect detection of the sealing nail welding seam is performed in the case where it is determined that the welding process of the cylindrical battery is abnormal, and the detection efficiency of the welding seam defect detection of the sealing nail welding seam of the cylindrical battery is further improved. Further, since the distance between the battery pole position and the sealing nail welding position of the cylindrical battery is small, the two-dimensional image of the sealing nail welding seam of the cylindrical battery under the illumination of the shadowless light source is collected to obtain the two-dimensional sealing nail welding seam image, so that the blackening of the inner circle of the two-dimensional sealing nail welding seam image collected under the illumination of the dome light is avoided, the detection accuracy of the welding seam defect detection is low, and the detection accuracy of the welding seam defect detection of the sealing nail welding seam of the cylindrical battery is improved.

[0042] It can be understood that the advantages of the second aspect to the fifth aspect of the present application can be referred to the related description of the first aspect of the present application, which will not be repeated here.

[0043] It should be understood that the foregoing general description and the following detailed description are only exemplary and explanatory, rather than limiting the technical solutions of the present application. BRIEF DESCRIPTION OF DRAWINGS

[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required to be used in the embodiments of the present application will be briefly introduced as follows. Obviously, the drawings described below only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative labor under the premise of the drawings.

[0045] FIG. 1 shows a scene schematic diagram of a welding seam defect detection system provided by an embodiment of the present application.

[0046] FIG. 2 shows a structure schematic diagram of a shadowless light source in a welding seam defect detection system provided by an embodiment of the present application.

[0047] FIG. 3 shows a flow schematic diagram of a welding seam defect detection method provided by an embodiment of the present application.

[0048] FIG. 4 shows another flow diagram of the method for detecting weld defects according to an embodiment of the present application.

[0049] FIG. 5 shows a structural diagram of the device for detecting weld defects according to an embodiment of the present application.

[0050] FIG. 6 shows a functional diagram of the electronic device according to an embodiment of the present application.

[0051] FIG. 7 shows a computer readable storage medium for storing or carrying program codes for implementing the method for detecting weld defects according to an embodiment of the present application.

[0052] FIG. 8 shows a computer program product for storing or carrying program codes for implementing the method for detecting weld defects according to an embodiment of the present application. DETAILED DESCRIPTION

[0053] In order to make the objectives, features and advantages of the present application more apparent and easy to understand, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the embodiments described below are only some of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0054] It should be understood that, when used in the specification and the appended claims, the term “comprising” indicates the presence of the described features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0055] It should also be understood that the terms used in the present application specification are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the present application specification and the appended claims, the singular forms “a”, “an” and “the” are intended to include the plural forms, unless the context clearly indicates otherwise.

[0056] It should be further understood that the term “and / or” used in the present application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes these combinations.

[0057] In addition, in the description of the present application, the terms “first”, “second”, “third” and the like are only used to distinguish descriptions, and cannot be understood as indicating or implying relative importance.

[0058] In the production process of power lithium battery, due to process and equipment reasons, there will be certain defects, which need to be detected by various detection means to improve the battery yield. In each link of the battery production line, the detection of the sealing pin weld defect is a crucial link, and the effectiveness of the detection result ensures the safety quality of the battery and the safety of the battery out of the factory.

[0059] At present, in the process of detecting the sealing pin weld defect of the battery, the sealing pin weld under the dome light irradiation is mainly judged for defects by manual operation, resulting in low detection efficiency of the sealing pin weld defect detection of the battery.

[0060] To solve the above problems, the welding defect detection method, detection device, electronic equipment and storage medium provided by the embodiments of the present application, by determining that the welding process of the cylindrical battery is abnormal, acquiring a two-dimensional sealing pin weld image of the cylindrical battery, the two-dimensional sealing pin weld image being a two-dimensional image of the sealing pin weld of the cylindrical battery collected under the irradiation of the shadowless light source, and inputting the two-dimensional sealing pin weld image into the welding defect detection model, so that the welding defect detection model outputs the corresponding welding defect detection result according to the two-dimensional sealing pin weld image, the welding defect detection model is trained based on the historical sealing pin weld defect image to obtain the neural network, and the welding defect detection result output by the welding defect detection model is received, realizing the welding defect detection of the two-dimensional sealing pin weld image of the sealing pin weld of the cylindrical battery according to the welding defect detection model under the condition that the welding process of the cylindrical battery is abnormal, avoiding manual defect detection of the sealing pin weld, and improving the detection efficiency of the welding defect detection of the sealing pin weld of the cylindrical battery.

[0061] And in the case of defects in the sealing pin weld, the welding process of the cylindrical battery must be abnormal, and the sealing pin weld is detected for welding defects under the condition that the welding process of the cylindrical battery is abnormal, further improving the detection efficiency of the welding defect detection of the sealing pin weld of the cylindrical battery.

[0062] Further, since the distance between the battery pole position and the sealing pin welding position of the cylindrical battery is small, the two-dimensional image of the sealing pin weld of the cylindrical battery under the irradiation of the shadowless light source is collected to obtain the two-dimensional sealing pin weld image, avoiding the blackening of the inner circle of the two-dimensional sealing pin weld image collected under the irradiation of the dome light, resulting in low detection accuracy of the welding defect detection, and improving the detection accuracy of the welding defect detection of the sealing pin weld of the cylindrical battery.

[0063] The technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application.

[0064] Please refer to FIG. 1, which shows a schematic diagram of an application scenario of a weld defect detection system provided by an embodiment of the present application, which can include a cylindrical battery 100, an unshadowed light source 200, a two-dimensional camera 300, a three-dimensional camera 400, and a processing device 500. The unshadowed light source 200 can be arranged relative to the two-dimensional camera 300 and the three-dimensional camera 400. The processing device 500 is communicatively connected to the two-dimensional camera 300 and the three-dimensional camera 400 and performs data interaction with the two-dimensional camera 300 and the three-dimensional camera 400.

[0065] The cylindrical battery 100 can include, but is not limited to, any one of a lithium storage battery, a lead-acid storage battery, a nickel-cadmium storage battery, a nickel-hydrogen storage battery, an iron-nickel storage battery, a sodium nickel chloride storage battery, a silver-zinc storage battery, a sodium-sulfur storage battery, an air storage battery, a fuel cell, or a solar storage battery.

[0066] The unshadowed light source 200 is a planar coaxial light source. The unshadowed light source 200 can be used to unshadow the cylindrical battery 100. The unshadowed light source 200 can include, but is not limited to, any one of an unshadowed light emitting diode (LED) lamp, an unshadowed organic light emitting diode (OLED) lamp, an unshadowed fluorescent lamp, or an unshadowed halogen lamp.

[0067] As an example, as shown in FIG. 2, the light source size of the unshadowed light source 200 is 140 millimeters (mm) * 140 mm, and the light emitting window size is 110 mm * 110 mm.

[0068] The two-dimensional camera 300 can be used to collect a two-dimensional image of the sealing nail weld of the cylindrical battery 100 under the irradiation of the unshadowed light source 200, obtain a two-dimensional sealing nail weld image, and send the two-dimensional sealing nail weld image to the processing device 500.

[0069] The two-dimensional camera 300 can include, but is not limited to, any one of a two-dimensional white light camera, a two-dimensional infrared camera, or a two-dimensional laser camera.

[0070] The three-dimensional camera 400 can be used to collect a three-dimensional image of the sealing nail weld of the cylindrical battery 100 under the irradiation of the unshadowed light source 200, obtain a three-dimensional sealing nail weld image, and send the three-dimensional sealing nail weld image to the processing device 500.

[0071] The three-dimensional camera 400 can include, but is not limited to, any one of a three-dimensional white light camera, a three-dimensional infrared camera, or a three-dimensional laser camera.

[0072] The processing device 500 can be configured to receive the two-dimensional sealing nail weld seam image sent by the two-dimensional camera 300 and the three-dimensional sealing nail weld seam image sent by the three-dimensional camera 400, and perform weld defect detection on the sealing nail weld seam of the cylindrical battery 100 according to the two-dimensional sealing nail weld seam image and the three-dimensional sealing nail weld seam image. The processing device 500 can include, but is not limited to, any one of a server or a terminal device and the like.

[0073] The server can include, but is not limited to, any one of a stand-alone physical server, a server cluster composed of multiple physical servers, a distributed system, a cloud server and the like.

[0074] The terminal device can include, but is not limited to, any one of a mobile terminal device (for example, a mobile phone, a personal digital assistant (PDA), a tablet personal computer (Tablet PC), a notebook computer, a smart watch, a smart bracelet and the like) and a fixed terminal device (for example, a desktop computer, a smart panel, an all-in-one computer and the like).

[0075] Please refer to FIG. 3, which shows a flowchart of a weld defect detection method according to an embodiment of the present application. In specific embodiments, the weld defect detection method can be applied to the processing device 500 in the weld defect detection system shown in FIG. 1. In the following, the processing device 500 will be taken as an example to elaborate the flow shown in FIG. 3 in detail. The weld defect detection method can include the following steps S110 to S130.

[0076] Step S110: In a case where it is determined that an abnormality occurs in a welding process of welding the cylindrical battery, a two-dimensional sealing nail weld seam image of the cylindrical battery is acquired.

[0077] In the embodiments of the present application, in a case where a user needs to perform weld defect detection on the sealing nail weld seam, a detection instruction can be sent to the processing device. The processing device receives and responds to the detection instruction. In a case where it is determined that an abnormality occurs in a welding process of welding the cylindrical battery, a two-dimensional sealing nail weld seam image of the cylindrical battery is acquired.

[0078] The two-dimensional sealing nail weld seam image is a two-dimensional image of the sealing nail weld seam of the cylindrical battery collected under illumination of a shadowless light source.

[0079] Specifically, in the case that the user needs to perform weld defect detection on the sealing pin weld, a detection instruction can be sent to the processing device, the processing device receives and responds to the detection instruction, determines whether the welding process of the cylindrical battery is abnormal, and in the case that the welding process of the cylindrical battery is determined to be abnormal, sends a first image acquisition instruction to the two-dimensional camera, the two-dimensional camera receives and responds to the first image acquisition instruction, performs two-dimensional image acquisition on the sealing pin weld of the cylindrical battery under the irradiation of the shadowless light source, obtains a two-dimensional sealing pin weld image, and sends the two-dimensional sealing pin weld image to the processing device, and the processing device receives the two-dimensional sealing pin weld image returned by the two-dimensional camera.

[0080] The processing device can obtain the welding power of the cylindrical battery, and determine whether the welding process of the cylindrical battery is abnormal according to the welding power.

[0081] In the case that the welding power is included in the power threshold range, it is determined that the welding process of the cylindrical battery is abnormal; in the case that the welding power is not included in the power threshold range, it is determined that the welding process of the cylindrical battery is abnormal.

[0082] The power threshold range can be used to represent the welding power range in which the welding process of the cylindrical battery is normal, and the power threshold range can include but is not limited to the welding power range preset by the user, and the welding power range automatically generated by the processing device according to multiple weld defect detections on the sealing pin weld of the cylindrical battery.

[0083] In some embodiments, the processing device can be provided with an input panel, and in the case that the user needs to perform weld defect detection on the sealing pin weld, the detection instruction can be input on the input panel of the processing device, and the processing device receives the detection instruction through the input panel.

[0084] In some embodiments, the processing device can be provided with a voice recognition module, and in the case that the user needs to perform weld defect detection on the sealing pin weld, voice information can be sent within the voice acquisition range of the voice recognition module, the voice recognition module acquires the voice information issued by the user, performs voice recognition on the acquired voice information, and in the case that the recognition result of the voice recognition contains a keyword indicating weld defect detection on the sealing pin weld, such as the keyword “weld defect detection”, or the keywords “weld” and “defect detection”, etc., it is determined that the detection instruction for weld defect detection on the sealing pin weld is received.

[0085] As an example, the voice information issued by the user is: weld defect detection on the sealing pin weld, and the recognition result of the voice recognition contains the keyword “weld defect detection”, and it is determined that the detection instruction for weld defect detection on the sealing pin weld is received.

[0086] In some embodiments, the weld defect detection system can further comprise a client connected to the processing device via a network and exchanging data with the processing device via the network.

[0087] In the case that the user needs to perform weld defect detection on the sealing nail weld, a detection instruction can be sent to the client, the client receives and responds to the detection instruction, and forwards the detection instruction to the processing device via the network, and the processing device receives the detection instruction forwarded by the client.

[0088] The client can include, but is not limited to, any one of a mobile client (such as any one of a mobile phone client, a PDA client, a Tablet PC client, a notebook computer client, a smart watch client, a smart bracelet client, or a wearable client, etc.) or a fixed client (such as a desktop computer client, a smart panel client, etc.), etc.

[0089] The network can include, but is not limited to, any one of a ZigBee network, a Bluetooth (BT) network, a Wireless Fidelity (Wi-Fi) network, a Thread network, a Long Range Radio (LoRa) network, a Low-Power Wide-Area Network (LPWAN), an infrared network, a Narrow Band Internet of Things (NB-IoT), a Controller Area Network (CAN), a Digital Living Network Alliance (DLNA) network, a Wide Area Network (WAN), a Local Area Network (LAN), a Metropolitan Area Network (MAN), or a Wireless Personal Area Network (WPAN), etc.

[0090] In some embodiments, the weld defect detection system can further comprise a mobile mechanism connected to the processing device via a network and exchanging data with the processing device via the network. The cylindrical battery is loaded on the mobile mechanism and is moved with the movement of the mobile mechanism.

[0091] In the case that the user needs to detect the weld defects of the sealing nail weld, a detection instruction can be sent to the processing device, the processing device receives and responds to the detection instruction, in the case that it is determined that the welding process of the cylindrical battery is abnormal, a first movement instruction can be sent to the movement mechanism, the movement mechanism receives and responds to the first movement instruction, moves the cylindrical battery to a first battery position, and sends the first battery position to the processing device, the processing device receives the first battery position returned by the movement mechanism, and determines whether the first battery position is in the two-dimensional image acquisition range of the two-dimensional camera, and in the case that it is determined that the first battery position is in the two-dimensional image acquisition range, a first image acquisition instruction is sent to the two-dimensional camera, the two-dimensional camera receives and responds to the first image acquisition instruction, acquires a two-dimensional image of the sealing nail weld of the cylindrical battery under the irradiation of the shadowless light source, obtains a two-dimensional sealing nail weld image, and sends the two-dimensional sealing nail weld image to the processing device, the processing device receives the two-dimensional sealing nail weld image returned by the two-dimensional camera, realizes two-dimensional image acquisition of the sealing nail weld of the cylindrical battery in the case that it is determined that the cylindrical battery is moved to the two-dimensional image acquisition range of the two-dimensional camera, and improves the acquisition success rate of two-dimensional image acquisition of the sealing nail weld.

[0092] Step S120: input the two-dimensional sealing nail weld image to the weld defect detection model, so that the weld defect detection model outputs the corresponding weld defect detection result according to the two-dimensional sealing nail weld image.

[0093] In the embodiment of the present application, the processing device can input the two-dimensional sealing nail weld image to the weld defect detection model, the weld defect detection model receives and responds to the two-dimensional sealing nail weld image, detects the weld defects of the sealing nail weld according to the two-dimensional sealing nail weld image, obtains the weld defect detection result, and outputs the weld defect detection result to the processing device.

[0094] The weld defect detection model can be trained based on historical sealing nail weld defect images to obtain a neural network (NN). The neural network can include but is not limited to a feedforward neural network, a convolutional neural network, a recurrent neural network, a generative adversarial network, and an autoencoder.

[0095] Neural network is a research hotspot in the field of artificial intelligence since 1980s. Neural network abstracts the neuron network of human brain from the perspective of information processing, establishes a simple model, and forms different networks according to different connection modes. In engineering and academic circles, it is often directly referred to as neural network or neural network. Neural network is an operation model composed of a large number of nodes (or neurons) connected with each other. Each node represents a specific output function, called activation function. The connection between each two nodes represents a weight value of the signal passing through the connection, called weight, which is equivalent to the memory of artificial neural network. The output of the network is different due to different connection modes, weight values and activation functions. The network itself is usually an approximation of an algorithm or function in nature, or an expression of a logic strategy.

[0096] The weld defect detection result can include a first detection result for characterizing a seal pin weld defect type of the seal pin weld, and a second detection result for characterizing that the seal pin weld does not have a seal pin weld defect.

[0097] Step S130: receiving the weld defect detection result output by the weld defect detection model.

[0098] In the embodiment of the present application, the processing device can receive the weld defect detection result output by the weld defect detection model, realizing weld defect detection of the two-dimensional seal pin weld image of the seal pin weld of the cylindrical battery according to the weld defect detection model in the case that the welding process of the cylindrical battery is determined to be abnormal, avoiding manual defect detection of the seal pin weld, and improving the detection efficiency of weld defect detection of the seal pin weld of the cylindrical battery. In the case that the seal pin weld has a defect, the welding process of the cylindrical battery must be abnormal, and weld defect detection of the seal pin weld is performed in the case that the welding process of the cylindrical battery is determined to be abnormal, further improving the detection efficiency of weld defect detection of the seal pin weld of the cylindrical battery. Further, since the distance between the battery pole position and the seal pin welding position of the cylindrical battery is small, the two-dimensional image of the seal pin weld of the cylindrical battery under the illumination of the shadowless light source is collected to obtain the two-dimensional seal pin weld image, avoiding the low detection accuracy of weld defect detection caused by the black inner circle of the two-dimensional seal pin weld image collected under the dome light, and improving the detection accuracy of weld defect detection of the seal pin weld of the cylindrical battery.

[0099] In some embodiments, the processing device can acquire a three-dimensional sealing pin weld image of the cylindrical battery in a case where the first detection result output by the weld defect detection model is received, and determine a weld defect value of the sealing pin weld defect type according to the three-dimensional sealing pin weld image, so as to realize, in a case where the sealing pin weld defect type of the sealing pin weld is detected, detection of the weld defect value of the sealing pin weld according to the three-dimensional sealing pin weld image, so as to enable the user to dispose the cylindrical battery according to the weld defect value, and improve the detection experience of the user in weld defect detection of the sealing pin weld of the cylindrical battery.

[0100] Specifically, the processing device can send a second image acquisition instruction to the three-dimensional camera in a case where the first detection result output by the weld defect detection model is received, the three-dimensional camera receives and responds to the second image acquisition instruction, performs three-dimensional image acquisition on the sealing pin weld of the cylindrical battery under illumination of the shadowless light source, obtains a three-dimensional sealing pin weld image, and sends the three-dimensional sealing pin weld image to the processing device, the processing device receives the three-dimensional sealing pin weld image returned by the three-dimensional camera, and inputs the three-dimensional sealing pin weld image to a size measurement application program, the size measurement application program receives and responds to the three-dimensional sealing pin weld image, detects a weld defect value of the defective sealing pin weld, and sends the detected weld defect value to the processing device, and the processing device receives the weld defect value returned by the size measurement application program. The size measurement application program can be used for size measurement of a three-dimensional image.

[0101] In some embodiments, the processing device can determine whether the weld defect detection of the sealing pin weld is qualified according to the weld defect value, so as to realize, according to the weld defect value of the sealing pin weld, judgment of whether the weld defect detection of the sealing pin weld is qualified, so as to enable the user to remove the cylindrical battery with unqualified weld defect detection from the production line, and improve the qualified rate of the cylindrical battery.

[0102] In a case where the weld defect value is greater than a defect threshold value of the sealing pin weld defect type, the processing device determines that the weld defect detection of the sealing pin weld is unqualified; in a case where the weld defect value is less than or equal to the defect threshold value, the processing device determines that the weld defect detection of the sealing pin weld is qualified.

[0103] The defect threshold value can be used to represent the maximum weld defect value when the weld defect detection of the sealing pin weld is unqualified, and the defect threshold value can include but is not limited to a weld defect value pre-set by the user and a weld defect value automatically generated by the processing device according to multiple weld defect detection processes.

[0104] In some embodiments, in the case that the welding defect detection of the sealing nail welding seam of the cylindrical battery is determined to be unqualified according to the welding defect numerical value, a prompt information can be generated and sent to a specified client, so that the user can be reminded in time to dispose the cylindrical battery in the case that the welding defect detection of the sealing nail welding seam of the cylindrical battery is detected to be unqualified, and the production quality of the cylindrical battery is improved.

[0105] The prompt information can include at least one of sound prompt information, text prompt information, light prompt information, etc.

[0106] The specified client can be a client associated with the user, and can include at least one of a mobile client (such as any one of a mobile phone client, a PDA client, a Tablet PC client, a notebook computer client, a smart watch client, a smart bracelet client, or a wearable client, etc.), or a fixed client (such as a desktop computer client, a smart panel client, etc.), etc.

[0107] In some embodiments, in the case that the processing device receives the first detection result output by the welding defect detection model, the processing device can send a second movement instruction to the moving mechanism, the moving mechanism receives and responds to the second movement instruction, moves the cylindrical battery to a second battery position, and sends the second battery position to the processing device, the processing device receives the second battery position returned by the moving mechanism, determines whether the second battery position is within the three-dimensional image acquisition range of the three-dimensional camera, and in the case that the second battery position is within the three-dimensional image acquisition range, sends a second image acquisition instruction to the three-dimensional camera, the three-dimensional camera receives and responds to the second image acquisition instruction, performs three-dimensional image acquisition on the sealing nail welding seam of the cylindrical battery under the irradiation of the shadowless light source, obtains a three-dimensional sealing nail welding seam image, and sends the three-dimensional sealing nail welding seam image to the processing device, the processing device receives the three-dimensional sealing nail welding seam image returned by the three-dimensional camera, so that the sealing nail welding seam of the cylindrical battery can be acquired in three dimensions in the case that the cylindrical battery is moved to the three-dimensional image acquisition range of the three-dimensional camera, and the acquisition success rate of the three-dimensional image acquisition of the sealing nail welding seam is improved.

[0108] In some embodiments, in the case that the processing device receives the second detection result output by the welding defect detection model, the processing device can determine that the welding defect detection of the sealing nail welding seam is qualified, so that the welding defect detection of the sealing nail welding seam is determined to be qualified in the case that no sealing nail welding seam defect is detected according to the two-dimensional sealing nail welding seam image of the sealing nail welding seam, and the detection efficiency of the welding defect detection of the sealing nail welding seam is further improved.

[0109] The scheme provided in the application comprises the following steps: in the case that it is determined that the welding process of the cylindrical battery is abnormal, a two-dimensional sealing nail weld image of the cylindrical battery is acquired, the two-dimensional sealing nail weld image is a two-dimensional image of the sealing nail weld of the cylindrical battery collected under the irradiation of an shadowless light source, the two-dimensional sealing nail weld image is input into a weld defect detection model, the weld defect detection model outputs a corresponding weld defect detection result according to the two-dimensional sealing nail weld image, the weld defect detection model is trained based on historical sealing nail weld defect images to obtain a neural network, and the weld defect detection result output by the weld defect detection model is received, so that in the case that it is determined that the welding process of the cylindrical battery is abnormal, the two-dimensional sealing nail weld image of the sealing nail weld of the cylindrical battery is detected for weld defects according to the weld defect detection model, manual detection for weld defects of the sealing nail weld is avoided, and the detection efficiency of the weld defect detection of the sealing nail weld of the cylindrical battery is improved.

[0110] In addition, in the case that the sealing nail weld has defects, the welding process of the cylindrical battery is inevitably abnormal, and the sealing nail weld is detected for weld defects in the case that it is determined that the welding process of the cylindrical battery is abnormal, so that the detection efficiency of the weld defect detection of the sealing nail weld of the cylindrical battery is further improved.

[0111] Further, since the distance between the battery pole position and the sealing nail welding position of the cylindrical battery is small, the two-dimensional image of the sealing nail weld of the cylindrical battery under the irradiation of the shadowless light source is collected to obtain the two-dimensional sealing nail weld image, so that the inner circle of the two-dimensional sealing nail weld image collected under the irradiation of the dome light is blackened, the detection accuracy of the weld defect detection is low, and the detection accuracy of the weld defect detection of the sealing nail weld of the cylindrical battery is improved.

[0112] Please refer to FIG. 4, which shows a flowchart of a weld defect detection method provided in another embodiment of the application. In specific embodiments, the weld defect detection method can be applied to the processing device 500 in the weld defect detection system shown in FIG. 1, and the following will take the processing device 500 as an example to elaborate the flow shown in FIG. 4 in detail. The weld defect detection method can comprise the following steps S210 to S260.

[0113] Step S210: Acquire a historical sealing nail weld defect image.

[0114] In this embodiment, the processing device can acquire a historical sealing nail weld defect image, and the sealing nail weld corresponding to the historical sealing nail weld defect image has weld defects.

[0115] In some embodiments, the processing device has pre-stored historical sealing nail weld defect images, and the processing device can read the pre-stored historical sealing nail weld defect images.

[0116] In some embodiments, the weld defect detection system can further comprise a server, the server pre-storing historical seal pin weld defect images, the server being connected to the processing device through a network and performing data interaction with the processing device through the network.

[0117] The processing device can send an image acquisition instruction to the server through the network, the server receiving and responding to the image acquisition instruction, sending the pre-stored historical seal pin weld defect images to the processing device through the network, and the processing device receiving the historical seal pin weld defect images returned by the server.

[0118] Step S220: obtaining a sample set according to the historical seal pin weld defect images.

[0119] In this embodiment, the processing device can obtain a corresponding sample set according to the historical seal pin weld defect images. The sample set can include a training set and a test set.

[0120] Specifically, the processing device can label the historical seal pin weld defect images to obtain corresponding labeled images, and divide the labeled images according to a preset division rule to obtain the training set and the test set.

[0121] When labeling the historical seal pin weld defect images, the processing device mainly labels the weld defect type, weld defect position, weld defect value and the like in the historical seal pin weld defect images. For example, the processing device labels the weld defect type in the historical seal pin weld defect images with a label frame, labels the corner coordinates of the label frame, and labels the size of the label frame.

[0122] The preset division rule is a human division rule. For example, the preset division rule can be a human division rule of training set:test set=9:1. When the historical seal pin weld defect images are 20000, the training set is 18000 and the test set is 2000. The preset division rule can be a human division rule of training set:test set=8:1. When the historical seal pin weld defect images are 27000, the training set is 24000 and the test set is 3000.

[0123] Step S230: inputting the training set into a neural network for training to obtain a weld defect detection model.

[0124] In this embodiment, the processing device can input the training set into the neural network, the neural network receiving and responding to the training set, training according to the training set, and obtaining the weld defect detection model.

[0125] In some embodiments, the processing device can input the test set to the weld defect detection model, the weld defect detection model receives and responds to the test set, tests according to the test set, obtains the corresponding test result, and determines the test accuracy corresponding to the test result according to the test result and the test set, and determines whether the weld defect detection model converges according to the test accuracy.

[0126] When the test accuracy is greater than or equal to the preset accuracy threshold, it is determined that the weld defect detection model converges; when the test accuracy is less than the preset accuracy threshold, it is determined that the weld defect detection model does not converge.

[0127] Wherein, the preset accuracy threshold can be used to represent the minimum accuracy corresponding to the convergence of the weld defect detection model, and the preset accuracy threshold can include but is not limited to the accuracy preset by the user, the accuracy generated by the processing device according to multiple detection processes of the weld defect detection of the sealing nail weld, etc.

[0128] Step S240: In the case where it is determined that the weld defect detection model converges, and in the case where it is determined that the welding process of the cylindrical battery is abnormal, the two-dimensional sealing nail weld image of the cylindrical battery is obtained.

[0129] Step S250: Input the two-dimensional sealing nail weld image to the weld defect detection model, so that the weld defect detection model outputs the corresponding weld defect detection result according to the two-dimensional sealing nail weld image.

[0130] Step S260: Receive the weld defect detection result output by the weld defect detection model.

[0131] In this embodiment, steps S240, S250 and S260 can refer to the contents of the corresponding steps in the foregoing embodiments, which will not be repeated here.

[0132] The scheme provided by the embodiment comprises the following steps: acquiring historical sealing nail weld defect images, acquiring a sample set according to the historical sealing nail weld defect images, inputting the training set into a neural network for training to obtain a weld defect detection model, and when it is determined that the weld defect detection model converges and when it is determined that the welding process of the cylindrical battery is abnormal, acquiring a two-dimensional sealing nail weld image of the cylindrical battery, inputting the two-dimensional sealing nail weld image into the weld defect detection model, so that the weld defect detection model outputs a corresponding weld defect detection result according to the two-dimensional sealing nail weld image, and receiving the weld defect detection result output by the weld defect detection model. In the case where it is determined that the welding process of the cylindrical battery is abnormal, the two-dimensional sealing nail weld image of the sealing nail weld of the cylindrical battery is detected according to the weld defect detection model, which avoids manual detection of the sealing nail weld and improves the detection efficiency of the weld defect detection of the sealing nail weld of the cylindrical battery.

[0133] And in the case where the sealing nail weld has defects, the welding process of the cylindrical battery must be abnormal, so that the sealing nail weld is detected for weld defects in the case where it is determined that the welding process of the cylindrical battery is abnormal, further improving the detection efficiency of the weld defect detection of the sealing nail weld of the cylindrical battery.

[0134] Further, since the distance between the battery pole position and the sealing nail welding position of the cylindrical battery is small, the two-dimensional image of the sealing nail weld of the cylindrical battery under the illumination of the shadowless light source is collected to obtain a two-dimensional sealing nail weld image, which avoids the low detection accuracy of the weld defect detection caused by the blackening of the inner circle of the two-dimensional sealing nail weld image collected under the dome light, and improves the detection accuracy of the weld defect detection of the sealing nail weld of the cylindrical battery.

[0135] Further, when it is determined that the weld defect detection model converges, the sealing nail weld is detected for weld defects based on the weld defect detection model, which improves the stability of the weld defect detection model and improves the detection credibility of the weld defect detection of the sealing nail weld of the cylindrical battery.

[0136] Please refer to FIG. 5, which shows a weld defect detection device 600 provided by an embodiment of the present application. The weld defect detection device 600 can be applied to the processing device 500 in the weld defect detection system shown in FIG. 1. In the following, the weld defect detection device 600 shown in FIG. 5 will be described in detail by taking the processing device 500 as an example. The weld defect detection device 600 can comprise a two-dimensional image acquisition module 610, a two-dimensional image input module 620 and a receiving module 630.

[0137] The two-dimensional image acquisition module 610 can be configured to acquire a two-dimensional sealing pin weld seam image of the cylindrical battery in a case where it is determined that the welding process of the cylindrical battery is abnormal, the two-dimensional sealing pin weld seam image can be a two-dimensional image of the sealing pin weld seam of the cylindrical battery collected under illumination of a shadowless light source; the two-dimensional image input module 620 can be configured to input the two-dimensional sealing pin weld seam image to the weld seam defect detection model, so that the weld seam defect detection model outputs a corresponding weld seam defect detection result according to the two-dimensional sealing pin weld seam image, the weld seam defect detection model can be obtained by training a neural network based on historical sealing pin weld seam defect images; and the receiving module 630 can be configured to receive the weld seam defect detection result output by the weld seam defect detection model.

[0138] In some embodiments, the two-dimensional image acquisition module 610 can include a first control unit, a first determination unit, and a second control unit.

[0139] The first control unit can be configured to control the cylindrical battery to move to a first battery position in a case where it is determined that the welding process of the cylindrical battery is abnormal; the first determination unit can be configured to determine whether the first battery position is within a two-dimensional image acquisition range of a two-dimensional camera; and the second control unit can be configured to control the two-dimensional camera to acquire a two-dimensional image of the sealing pin weld seam to obtain a two-dimensional sealing pin weld seam image in a case where it is determined that the first battery position is within the two-dimensional image acquisition range.

[0140] In some embodiments, the weld seam defect detection result can include a first detection result for characterizing a sealing pin weld seam defect type of the sealing pin weld seam, and the weld seam defect detection device 600 can further include a three-dimensional image acquisition module and a first determination module.

[0141] The three-dimensional image acquisition module can be configured to acquire a three-dimensional sealing pin weld seam image of the cylindrical battery in a case where the first detection result output by the weld seam defect detection model is received; and the first determination module can be configured to determine a weld seam defect value of the sealing pin weld seam defect type according to the three-dimensional sealing pin weld seam image.

[0142] In some embodiments, the weld seam defect detection device 600 can further include a second determination module and a third determination module.

[0143] The second determination module can be configured to determine that the weld seam defect detection of the sealing pin weld seam is unqualified in a case where the weld seam defect value is greater than a defect threshold value of the sealing pin weld seam defect type; and the third determination module can be configured to determine that the weld seam defect detection of the sealing pin weld seam is qualified in a case where the weld seam defect value is less than or equal to the defect threshold value.

[0144] In some embodiments, the weld seam defect detection device 600 can further include a generation module.

[0145] The generating module can be configured to generate and send a prompt information to a specified client in a case that the welding defect detection of the sealing nail weld is not qualified according to the welding defect value.

[0146] In some embodiments, the welding defect detection device 600 can further include a control module and a fourth determination module.

[0147] The control module can be configured to control the cylindrical battery to move to a second battery position before the three-dimensional image acquisition module acquires the three-dimensional sealing nail weld image of the cylindrical battery; and the fourth determination module can be configured to determine whether the second battery position is in the three-dimensional image acquisition range of the three-dimensional camera.

[0148] In some embodiments, the three-dimensional image acquisition module can include a third control unit.

[0149] The third control unit can be configured to control the three-dimensional camera to acquire a three-dimensional image of the sealing nail weld to obtain the three-dimensional sealing nail weld image in a case that the second battery position is determined to be in the three-dimensional image acquisition range.

[0150] In some embodiments, the welding defect detection result can further include a second detection result for characterizing that the sealing nail weld does not have a sealing nail weld defect, and the welding defect detection device 600 can further include a fifth determination module.

[0151] The fifth determination module can be configured to determine that the welding defect detection of the sealing nail weld is qualified in a case that the second detection result output by the welding defect detection model is received.

[0152] The scheme provided in the embodiment, by determining that the welding process of the cylindrical battery appears abnormal, acquiring a two-dimensional sealing nail weld image of the cylindrical battery, the two-dimensional sealing nail weld image being a two-dimensional image of the sealing nail weld of the cylindrical battery collected under the illumination of the shadowless light source, inputting the two-dimensional sealing nail weld image into the welding defect detection model, causing the welding defect detection model to output a corresponding welding defect detection result according to the two-dimensional sealing nail weld image, the welding defect detection model being trained based on historical sealing nail weld defect images to obtain a neural network, and receiving the welding defect detection result output by the welding defect detection model, realizes the welding defect detection of the two-dimensional sealing nail weld image of the sealing nail weld of the cylindrical battery according to the welding defect detection model in a case that the welding process of the cylindrical battery appears abnormal, avoids manual defect detection of the sealing nail weld, and improves the detection efficiency of the welding defect detection of the sealing nail weld of the cylindrical battery.

[0153] And in the case of the existence of defects in the sealing nail weld, the welding process of the cylindrical battery will inevitably be abnormal, and the welding defect detection of the sealing nail weld is performed in the case of determining the welding process abnormality of the cylindrical battery, thereby further improving the detection efficiency of the welding defect detection of the sealing nail weld of the cylindrical battery.

[0154] Further, since the distance between the battery pole position of the cylindrical battery and the sealing nail welding position is small, the two-dimensional image of the sealing nail weld of the cylindrical battery under the irradiation of the shadowless light source is collected to obtain a two-dimensional sealing nail weld image, thereby avoiding the low detection accuracy of the welding defect detection caused by the blackening of the inner circle of the two-dimensional sealing nail weld image collected under the dome light irradiation, and improving the detection accuracy of the welding defect detection of the sealing nail weld of the cylindrical battery.

[0155] It should be noted that each embodiment in the specification adopts a progressive manner for description, and each embodiment focuses on the difference from other embodiments, and the same and similar parts between each embodiment can be referred to. For the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the part of the method embodiment. For any processing manner described in the method embodiment, it can be realized by a corresponding processing module in the device embodiment, and the device embodiment will not be described one by one.

[0156] In addition, each functional module in each embodiment of the present application can be integrated in one processing module, or each module can exist physically, or two or more modules can be integrated in one module. The integrated module can be realized in the form of hardware or in the form of a software functional module.

[0157] Please refer to FIG. 6, which shows the functional block diagram of the electronic device 700 provided by an embodiment of the present application. The electronic device 700 can include one or more of the following components: a memory 710, a processor 720, and one or more application programs, wherein the one or more application programs can be stored in the memory 710 and configured to be executed by the one or more processors 720, and the one or more application programs are configured to perform the method described in the foregoing method embodiments.

[0158] The memory 710 can include a random access memory (RAM) and can also include a read-only memory (ROM). The memory 710 can be used to store instructions, programs, codes, code sets, or instruction sets. The memory 710 can include a program storage area and a data storage area, where the program storage area can store instructions for implementing an operating system, instructions for implementing at least one function (such as determining that an abnormality occurs in a welding process, obtaining a two-dimensional sealing nail weld seam image, collecting a two-dimensional image, inputting a two-dimensional sealing nail weld seam image, outputting a weld defect detection result, training a neural network, receiving a weld defect detection result, moving a cylindrical battery to a first battery position, determining whether the first battery position is in a two-dimensional image collection range, determining that the first battery position is in the two-dimensional image collection range, controlling a two-dimensional camera, obtaining a two-dimensional sealing nail weld seam image, receiving a first detection result, outputting the first detection result, obtaining a three-dimensional sealing nail weld seam image, determining a weld defect value, determining whether a weld defect detection is qualified, determining that a sealing nail weld seam defect detection is unqualified, determining that a sealing nail weld seam defect detection is qualified, generating a reminder information, sending the reminder information, moving the cylindrical battery to a second battery position, determining whether the second battery position is in a three-dimensional image collection range, determining that the second battery position is in the three-dimensional image collection range, controlling a three-dimensional camera, performing three-dimensional image collection, obtaining a three-dimensional sealing nail weld seam image, receiving a second detection result, and outputting the second detection result, etc.), instructions for implementing each of the method embodiments described below, and the like. The data storage area can also store data created by the electronic device 700 in use (such as a cylindrical battery, a two-dimensional sealing nail weld seam image, a sealing nail weld seam, a shadowless light source, a two-dimensional image, a weld defect detection model, a weld defect detection result, a historical sealing nail weld seam defect image, a neural network, a sealing nail weld seam defect type, a first detection result, a second detection result, a first battery position, a two-dimensional camera, a two-dimensional image collection range, a three-dimensional sealing nail weld seam image, a weld defect value, a defect threshold value, a reminder information, a specified client, a second battery position, a three-dimensional camera, and a three-dimensional image collection range), and the like.

[0159] The processor 720 can include one or more processing cores. The processor 720 connects various parts within the entire electronic device 700 with various interfaces and lines, performs various functions of the electronic device 700 and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 710, and calling data stored in the memory 710. Optionally, the processor 720 can be implemented in at least one of a hardware form of a digital signal processing (DSP), a field-programmable gate array (FPGA), a programmable logic array (PLA). The processor 720 can integrate a combination of one or several of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. Among them, the CPU mainly processes an operating system, a user interface, and an application program; the GPU is responsible for rendering and drawing display content; and the modem is used for processing wireless communication. It can be understood that the above-mentioned modem can also not be integrated into the processor 720, but can be implemented by a separate communication chip.

[0160] Referring to FIG. 7, a structural block diagram of a computer readable storage medium provided by an embodiment of the present application is shown. The computer readable storage medium 800 stores program code 810, which can be called and executed by a processor to perform the methods described in the above method embodiments.

[0161] The computer readable storage medium 800 can be an electronic storage such as a flash memory, an EEPROM (electrically erasable programmable read-only memory), an EPROM, a hard disk, or a ROM. Optionally, the computer readable storage medium 800 includes a non-volatile computer readable medium. The computer readable storage medium 800 has a storage space for program code 810 to perform any of the above methods. These program codes can be read from or written to one or more computer program products. The program code 810 can be compressed in an appropriate form, for example.

[0162] Referring to FIG. 8, a structural block diagram of a computer program product 900 provided by the embodiment is shown. The computer program product 900 includes computer programs / instructions 910 stored in a computer readable storage medium of a computer device. When the computer program product 900 runs on the computer device, the processor of the computer device reads the computer programs / instructions 910 from the computer readable storage medium, and the processor executes the computer programs / instructions 910, so that the computer device executes the method described in the above method embodiments.

[0163] The scheme provided by the embodiment, by acquiring a two-dimensional sealing nail weld image of the cylindrical battery, the two-dimensional sealing nail weld image being a two-dimensional image of the sealing nail weld of the cylindrical battery collected under illumination of the shadowless light source, and inputting the two-dimensional sealing nail weld image to the weld defect detection model, so that the weld defect detection model outputs a corresponding weld defect detection result according to the two-dimensional sealing nail weld image, the weld defect detection model is trained based on historical sealing nail weld defect images to obtain a neural network, and the weld defect detection result output by the weld defect detection model is received, realizing weld defect detection of the two-dimensional sealing nail weld image of the sealing nail weld of the cylindrical battery according to the weld defect detection model, and improving the detection efficiency of weld defect detection of the sealing nail weld of the cylindrical battery.

[0164] Further, since the distance between the battery pole position and the sealing nail welding position of the cylindrical battery is small, the two-dimensional image of the sealing nail weld of the cylindrical battery under the illumination of the shadowless light source is collected to obtain the two-dimensional sealing nail weld image, which avoids the low detection accuracy of the weld defect detection caused by the blackening of the inner circle of the two-dimensional sealing nail weld image collected under the illumination of the dome light, and improves the detection accuracy of the weld defect detection of the sealing nail weld of the cylindrical battery.

[0165] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements do not drive the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for detecting weld defects, wherein: The weld defect detection method comprises: When it is determined that an abnormality occurs in the welding process of the cylindrical battery, a two-dimensional image of the sealing nail weld of the cylindrical battery is obtained, where the two-dimensional image of the sealing nail weld of the cylindrical battery is captured under illumination of a shadowless light source; Inputting the two-dimensional sealing nail weld image into a weld defect detection model, so that the weld defect detection model outputs a corresponding weld defect detection result according to the two-dimensional sealing nail weld image, wherein the weld defect detection model is obtained by training a neural network based on historical sealing nail weld defect images; Receive the weld defect detection result output by the weld defect detection model.

2. The weld defect detection method according to claim 1, wherein: When it is determined that the welding process of the cylindrical battery is abnormal, obtaining a two-dimensional sealing pin weld image of the cylindrical battery includes: When it is determined that an abnormality occurs in the welding process of the cylindrical battery, controlling the cylindrical battery to move to a first battery position; determining whether the first battery position is within a two-dimensional image acquisition range of a two-dimensional camera; When it is determined that the first battery position is within the two-dimensional image acquisition range, the two-dimensional camera is controlled to acquire a two-dimensional image of the sealing nail weld to obtain the two-dimensional sealing nail weld image.

3. The weld defect detection method according to claim 2, wherein: When it is determined that the welding process of the cylindrical battery is abnormal, controlling the cylindrical battery to move to the first battery position includes: When it is determined that an abnormality occurs in the welding process of the welded cylindrical battery, a first movement instruction is sent to the movement mechanism, so that the movement mechanism moves the cylindrical battery to the first battery position.

4. The weld defect detection method according to any one of claims 1 to 3, wherein: The weld defect detection result includes a first detection result for characterizing the sealing nail weld defect type of the sealing nail weld, and the weld defect detection method further includes: Upon receiving the first detection result output by the weld defect detection model, acquiring a three-dimensional sealing pin weld image of the cylindrical battery; Determine a weld defect value of the sealing nail weld defect type according to the three-dimensional sealing nail weld image.

5. The weld defect detection method according to claim 4, wherein: The weld defect detection method further includes: In a case where the weld defect value is greater than a defect threshold of the sealing pin weld defect type, determining that the weld defect detection of the sealing pin weld is unqualified; When the weld defect value is less than or equal to the defect threshold, it is determined that the weld defect detection of the sealing pin weld is qualified.

6. The weld defect detection method according to claim 5, wherein: The weld defect detection method further includes: When it is determined that the weld defect detection of the sealing pin weld is unqualified according to the weld defect value, a reminder message is generated and sent to a designated client.

7. The weld defect detection method according to any one of claims 4 to 6, wherein: Before obtaining the three-dimensional sealing nail weld image of the cylindrical battery, the weld defect detection method further includes: Controlling the cylindrical battery to move to a second battery position; determining whether the second battery position is within a three-dimensional image acquisition range of a three-dimensional camera; The obtaining of a three-dimensional sealing nail weld image of the cylindrical battery includes: When it is determined that the second battery position is within the three-dimensional image acquisition range, the three-dimensional camera is controlled to acquire a three-dimensional image of the sealing nail weld to obtain the three-dimensional sealing nail weld image.

8. The weld defect detection method according to claim 7, wherein: The controlling the cylindrical battery to move to the second battery position includes: A second movement instruction is sent to the movement mechanism, so that the movement mechanism moves the cylindrical battery to the second battery position according to the second movement instruction.

9. The weld defect detection method according to any one of claims 1 to 8, wherein the weld defect detection result includes a second detection result for indicating that the sealing nail weld does not have a sealing nail weld defect, wherein: The weld defect detection method further includes: Upon receiving the second detection result output by the weld defect detection model, it is determined that the weld defect detection of the sealing pin weld is qualified.

10. The weld defect detection method according to any one of claims 1 to 9, wherein: When it is determined that the welding process of the cylindrical battery is abnormal, before obtaining the two-dimensional sealing nail weld image of the cylindrical battery, the weld defect detection method further includes: Acquire the historical sealing pin weld defect image; The neural network is trained based on the historical sealing pin weld defect images to obtain the weld defect detection model.

11. The weld defect detection method according to claim 10, wherein: The neural network is trained according to the historical sealing nail weld defect images to obtain the weld defect detection model, including: Acquire a sample set based on the historical sealing pin weld defect image, wherein the sample set at least includes a training set; The training set is input into the neural network for training to obtain the weld defect detection model.

12. The weld defect detection method according to claim 11, wherein: The sample set also includes a test set, and when it is determined that the welding process of the cylindrical battery is abnormal, obtaining a two-dimensional sealing nail weld image of the cylindrical battery includes: When it is determined that the weld defect detection model converges according to the test set, and when it is determined that an abnormality occurs in the welding process of the welded cylindrical battery, the two-dimensional sealing nail welding image of the cylindrical battery is obtained.

13. A weld defect detection device, wherein: The weld defect detection device comprises: A two-dimensional image acquisition module is used to acquire a two-dimensional image of the sealing nail weld of the cylindrical battery when it is determined that the welding process of the cylindrical battery is abnormal, wherein the two-dimensional sealing nail weld image is a two-dimensional image of the sealing nail weld of the cylindrical battery captured under the illumination of a shadowless light source; a two-dimensional image input module, configured to input the two-dimensional sealing nail weld image into a weld defect detection model, so that the weld defect detection model outputs a corresponding weld defect detection result based on the two-dimensional sealing nail weld image, wherein the weld defect detection model is obtained by training a neural network based on historical sealing nail weld defect images; A receiving module is used to receive the weld defect detection result output by the weld defect detection model.

14. An electronic device, wherein: include: Memory; one or more processors coupled to the memory; One or more applications, wherein the one or more applications are stored in the memory and configured to be executed by one or more processors, and the one or more applications are configured to execute the weld defect detection method according to any one of claims 1 to 12.

15. A computer-readable storage medium, wherein: The computer-readable storage medium stores program code, and the program code can be called by a processor to execute the weld defect detection method according to any one of claims 1 to 12.

Images