CN118209561A

CN118209561A - Coil stock detection method and system

Info

Publication number: CN118209561A
Application number: CN202410630297.3A
Authority: CN
Inventors: 姜平; 彭灿福
Original assignee: Contemporary Amperex Technology Co Ltd
Current assignee: Contemporary Amperex Technology Co Ltd
Priority date: 2024-05-21
Filing date: 2024-05-21
Publication date: 2024-06-18

Abstract

The embodiment of the application discloses a detection method and a detection system for coiled materials, wherein the detection method for coiled materials is applied to a visual upper computer and comprises the following steps: respectively acquiring line scanning images acquired by at least two line scanning cameras on a detection station for the coil stock; at least two line scanning cameras have intersection areas in the acquisition visual field range of the coil stock, and the range of the union areas is larger than the width of the coil stock; and respectively carrying out defect detection on the line scanning images acquired by each line scanning camera, correspondingly obtaining a defect detection result of each line scanning image, and sending the defect detection result to the controller. In the embodiment of the application, the intersection area exists in the acquisition visual field range of the coil stock by at least two line scanning cameras on the detection station, and the range of the union area is larger than the width of the coil stock, so that the line scanning images acquired by the at least two line scanning cameras can be ensured to completely cover the coil stock, and the condition of missing shooting in partial areas of the coil stock can be avoided.

Description

Coil stock detection method and system

Technical Field

The application relates to the technical field of batteries, in particular to a detection method and a detection system for coil stock.

Background

For coil stock, a black and white line scan camera is generally used to photograph the coil stock after the light source is polished, so as to detect the coil stock. Due to uneven coating or missing coating of the base material and the coating material of the anode roll, it is difficult to distinguish the copper foil base and the film region from the line scan image, and thus it is difficult to detect defects of the roll.

Disclosure of Invention

In view of the above, the embodiment of the application provides a method and a system for detecting coil stock.

The technical scheme of the application is realized as follows:

In a first aspect, an embodiment of the present application provides a method for detecting a coil stock, which is applied to a visual host computer, where the method includes: respectively acquiring line scanning images acquired by at least two line scanning cameras on a detection station for the coil stock respectively; the at least two line scanning cameras have intersection areas for the acquisition visual field range of the coil stock, and the range of the union areas is larger than the width of the coil stock; and respectively carrying out defect detection on the line scanning images acquired by each line scanning camera, correspondingly obtaining a defect detection result of each line scanning image, and sending the defect detection result to a controller.

In the embodiment of the application, on one hand, the intersection area exists in the acquisition visual field range of the coil stock by at least two line scanning cameras on the detection station, and the range of the union area is larger than the width of the coil stock, so that the line scanning images acquired by the at least two line scanning cameras can be ensured to completely cover the coil stock, and the condition of missing shooting of part of the area of the coil stock can be avoided; on the other hand, the line scanning images collected by each line scanning camera are subjected to defect detection by the visual upper computer, so that the condition of missing detection of partial areas of the coiled material can be ensured.

In some embodiments, acquiring line scan images of the coil stock acquired by at least two line scan cameras on the detection station, respectively, includes: responding to a trigger signal, and controlling a light source on the detection station to polish the coil stock in a moving state; controlling at least two line scanning cameras on the detection station to simultaneously photograph the coiled material; and under the condition that the photographing times are larger than the preset times, controlling the at least two line scanning cameras to output line scanning images at the same time.

According to the embodiment of the application, on one hand, at least two line scanning cameras on the detection station are controlled by the visual host computer to simultaneously photograph the coil stock in a moving state, so that the condition of missing photographing of part of the area of the coil stock can be avoided; on the other hand, under the condition that the photographing times are larger than the preset times, at least two line scanning cameras are controlled by the vision upper computer to output line scanning images at the same time, and conditions are provided for the line scanning cameras to stop collecting line scanning images of the coil stock.

In some embodiments, each of the detection stations includes a line scan camera and a light source; wherein the long side of the visual field of the line scanning camera is parallel to the width of the coil stock; a first preset angle exists between the visual field direction of the line scanning camera and the normal line of the coil stock; a second preset angle exists between the polishing direction of the light source and the normal line of the coil stock.

In the embodiment of the application, the long side of the visual field of the line scanning camera is parallel to the width of the coil stock, so that the line scanning image acquired by the line scanning camera can be ensured to completely cover the coil stock.

In some embodiments, the coil stock is a lithium electrode sheet, and the detection stations include a front detection station and a back detection station; the front detection station is used for detecting the front of the lithium electrode slice; the back detection station is used for detecting the back of the lithium electrode slice; wherein, each detection station is provided with two brackets; two line scanning cameras are arranged on each support, a gap with a preset area exists between the two line scanning cameras, and the area of each line scanning camera is larger than that of the gap; four line scanning cameras on each detection station are arranged in a staggered mode.

In the embodiment of the application, on one hand, the front and the back of the lithium electrode slice are detected by adopting the front detection station and the back detection station respectively, so that the comprehensive detection of the lithium electrode slice is realized, and the performance and the safety of the battery are improved; on the other hand, through setting up four line scanning cameras on front detection station and the back detection station respectively and staggering each other and arranging, can ensure that the line that four line scanning cameras gathered sweeps the image and can cover the coil stock completely.

In some embodiments, the front side inspection station includes a first front side inspection station and a second front side inspection station; controlling at least two line scanning cameras on the detection station to simultaneously photograph the coil stock, comprising: controlling two line scanning cameras on the first front detection station to simultaneously photograph the coiled material; and controlling two line scanning cameras on the second front detection station to simultaneously photograph the coiled material.

In the embodiment of the application, the two line scanning cameras on the two front detection stations are respectively controlled by the visual upper computer to simultaneously photograph the coil stock, so that the condition of missing photographing of part of the area of the coil stock can be avoided.

In some embodiments, the defect detection includes a metal leakage detection of the lithium electrode sheet; performing defect detection on the line scan image acquired by each line scan camera respectively to correspondingly obtain a defect detection result of each line scan image, including: respectively carrying out binarization processing on the line scanning images acquired by each line scanning camera to correspondingly obtain a binary image of each line scanning image; determining a first communication area in which white pixels in each binary image are connected with each other; the white pixel points represent pixel points corresponding to a film region in the binary image; performing expansion operation on the first communication area corresponding to each binary image to correspondingly obtain a second communication area; determining the position of each second communication area in the corresponding line scan image; screening a third communication area from the second communication area; the third communication region characterizes a region possibly having a metal leakage defect; and determining a defect detection result of the third communication region corresponding to the area larger than the preset metal leakage area as metal leakage.

In the embodiment of the application, on one hand, the accuracy of the first communication area can be improved by determining the first communication area by the visual upper computer according to the white pixel points in the binary image corresponding to the line scanning image; on the other hand, by employing the expansion operation, it is possible to enhance the image feature in the first communication area and enlarge the white highlight area in the image, so that the defect is more accurately located and identified.

In some embodiments, screening a third communication region from the second communication region includes: determining the number of pixel points in each second communication area and the pixel value corresponding to each pixel point; determining the pixel points with the pixel values larger than a second preset value in the second communication area as target pixel points under the condition that the number of the pixel points is larger than a first preset value; and determining the third communication area based on the target pixel point.

In the embodiment of the application, the second communication region which does not meet the conditions can be eliminated by combining the number of the pixel points and the pixel value to screen, so that the accuracy of the third communication region is improved.

In some embodiments, the defect detection comprises asperity detection of the lithium electrode sheet; performing defect detection on the line scanning images acquired by each line scanning camera respectively to correspondingly obtain defect detection results of each line scanning image, and further comprising: classifying each line scanning image through a pre-trained classifier to correspondingly obtain a fourth communication area; the fourth communication region represents a region where an concave-convex point possibly exists; and if the confidence coefficient of the fourth communication region is larger than a preset threshold value, determining that the defect detection result of the line scan image corresponding to the fourth communication region is concave-convex points.

In the embodiment of the application, on one hand, the visual upper computer classifies each line-scanned image through a pre-trained classifier to obtain the fourth communication area possibly provided with the concave-convex points, so that the effective classification of the line-scanned image of the area possibly provided with the concave-convex points is realized; on the other hand, the defect detection result of the line scan image corresponding to which the confidence coefficient is larger than the preset threshold value is determined as the concave-convex point through the visual upper computer, so that the accuracy and the efficiency of concave-convex point identification are improved.

In some embodiments, classifying, by a pre-trained classifier, each of the line scan images, correspondingly obtaining a fourth connected region, includes: carrying out semantic segmentation on each line scanning image to correspondingly obtain a segmentation area of each line scanning image; extracting features of the segmented regions of each line scan image, and correspondingly obtaining the features of each segmented region; classifying the characteristics of each segmented region through the pre-trained classifier to obtain a classification result; and determining a segmentation area corresponding to the classification result as the fourth communication area under the condition that the classification result represents that the lithium electrode plate has the concave-convex point defect.

According to the embodiment of the application, the concave-convex point defects on the lithium electrode slice can be more accurately identified and the range of the concave-convex point defects can be determined by combining the semantic segmentation with the pre-trained classifier.

In some embodiments, the defect detection results include a first result for characterizing the coil stock as acceptable and a second result for characterizing the coil stock as unacceptable, the detection method further comprising: sending the first result to a controller under the condition that the defect detection result indicates that the coil stock has no defect; and sending the second result to the controller when the defect detection result indicates that the coil stock has defects.

In the embodiment of the application, the visual upper computer sends the qualified or unqualified coil stock results to the controller, thereby realizing the quality control and tracing of the coil stock on an automatic production line.

In a second aspect, an embodiment of the present application provides a detection system for a coil stock, where the detection system includes at least two line scanning cameras and a vision upper computer; the at least two line scanning cameras are used for collecting line scanning images of the coil stock; the at least two line scanning cameras have intersection areas for the acquisition visual field range of the coil stock, and the range of the union areas is larger than the width of the coil stock; the visual upper computer is used for respectively acquiring line scanning images of the coil stock respectively acquired by at least two line scanning cameras on the detection station; and respectively carrying out defect detection on the line scanning images acquired by each line scanning camera, correspondingly obtaining a defect detection result of each line scanning image, and sending the defect detection result to a controller.

In some embodiments, the detection system further comprises a controller; the controller is used for marking a coil stock area where the defect is located in the line scanning image under the condition that the defect detection result of the line scanning image represents that the defect exists.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the application as claimed.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the application and together with the description, serve to explain the principles of the application.

FIG. 1 is a schematic diagram of an overall hardware layout of a coil stock inspection according to an embodiment of the present application;

Fig. 2 is a schematic implementation flow chart of a method for detecting coil stock according to an embodiment of the present application;

FIG. 3 is a schematic diagram of a hardware layout of a single inspection station according to an embodiment of the present application;

FIG. 4 is a top view of a line scan camera of a front vision inspection station according to an embodiment of the present application;

fig. 5 is a view field distribution schematic diagram of a front and a back camera of a visual inspection station coil stock according to an embodiment of the present application;

FIG. 6 is an overall plan view of a vision inspection station provided by an embodiment of the present application;

FIG. 7 is a schematic diagram of hardware configuration and installation structure of a front vision station for coil stock according to an embodiment of the present application;

Fig. 8 is a schematic implementation flow chart of another method for detecting coil stock according to an embodiment of the present application;

Fig. 9 is a schematic diagram of a composition structure of a coil stock detecting system according to an embodiment of the present application.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present application more apparent, the specific technical solutions of the present application will be described in further detail below with reference to the accompanying drawings in the embodiments of the present application. The following examples are illustrative of the application and are not intended to limit the scope of the application.

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of the application only and is not intended to be limiting of the application.

In the following description reference is made to "some embodiments," "this embodiment," "an embodiment of the application," and examples, etc., which describe a subset of all possible embodiments, but it is to be understood that "some embodiments" may be the same subset or different subsets of all possible embodiments and may be combined with one another without conflict.

If a similar description of "first/second" appears in the application document, the following description is added, in which the terms "first/second/third" are merely distinguishing between similar objects and not representing a particular ordering of the objects, it being understood that the "first/second/third" may be interchanged with a particular order or precedence, if allowed, so that embodiments of the application described herein may be practiced otherwise than as illustrated or described herein.

To facilitate understanding of the embodiments of the present application, an anode roll in the field of batteries will be described as an example.

(1) In the field of batteries, the anode roll is a rolled anode material that requires further processing to become an anode sheet. For example, the anode web is cut into anode sheets of a certain shape and size using a die cutting process.

(2) The laminated battery pole piece is an important link in the manufacturing process of the lithium battery, and mainly relates to the steps of dividing the coated positive pole piece and the coated negative pole piece according to the required size, and stacking the positive pole piece, the membrane, the negative pole piece and the membrane according to a specific sequence (usually the positive pole piece, the membrane, the negative pole piece and the membrane) to form a sandwich-like structure. A plurality of the sandwich structures are further overlapped to finally form the encapsulated battery cell.

(3) The coil stock unreeling is a key step in the manufacturing process of the lithium battery, particularly in the manufacturing process of the laminated battery pole piece, and mainly releases the coated anode and cathode materials from the winding drum for subsequent procedures such as punching, lamination and the like.

At present, in the previous procedure link of power battery production, for example, defects such as metal leakage, edge fold, pinhole breakage and the like of an anode coil are required to be detected before the anode coil is punched, on one hand, coating materials on the surface of an anode pole piece may fall off due to manual misoperation, production equipment failure or process design defects, and base metal materials leak out; on the other hand, as the manufacturing process of the laminated battery pole piece is different, the metal leakage area of the front side and the back side of the anode coil is continuously reduced along with the improvement of the unreeling speed of the anode coil, and the metal leakage area is difficult to identify by human eyes; on the other hand, due to the difference of nucleation energy barriers of the metal leakage area and the film area surface of the anode plate, the growth trend of sediment (such as lithium metal or other active substances) in the subsequent electrochemical reaction, the decomposition of electrolyte on the surfaces of the substrate and the film area, the consistency of the thickness of the micro-area of the cell and the like can be influenced, so that the overall stability and reliability of the battery are reduced. The existing optical scheme does not have the detection capability of the related detection items before punching, particularly the detection area precision of the metal leakage detection items is high, and the detection rate of the metal leakage detection items cannot be guaranteed in a hundred percent, so that the metal leakage problem of the anode coil stock is a key point and a difficult point in a previous procedure link of power battery production.

Based on the above, the embodiment of the application provides a method and a system for detecting a coil stock, wherein the method for detecting the coil stock is applied to a visual host computer, on one hand, through the fact that an intersection area exists in the acquisition visual field range of at least two line scanning cameras on a detection station and the range of the union area is larger than the width of the coil stock, line scanning images acquired by the at least two line scanning cameras can be ensured to completely cover the coil stock, so that the condition of missing shooting of part of the area of the coil stock can be avoided; on the other hand, the line scanning images collected by each line scanning camera are subjected to defect detection by the visual upper computer, so that the condition of missing detection of partial areas of the coiled material can be ensured.

The embodiment of the application provides an overall hardware layout for detecting coil stock, as shown in fig. 1, the overall hardware comprises a light source 11, a line scanning camera 12, an industrial personal computer 13, a central control display 14, a controller 15, an encoder 16 and a marking machine 17, wherein:

the size of the light source 11 can be selected according to the requirement, and the imaging consistency can be effectively ensured by selecting the light source with high brightness and good uniformity; the line scan camera 12 is used for exposing line by line according to the running speed and the beat of the coil stock 18 to form a complete line scan image meeting the size requirement.

The industrial personal computer 13 is used as an image processing terminal in the system, and a core computing model is arranged on the industrial personal computer 13 and can be adapted to various operating systems so as to meet the requirement of the operation of upper computer software in various development languages.

The central control display 14 is used for sorting and summarizing the processing results output by the industrial personal computer 13, arranging and displaying the processing results after graphic operation, linking various integrated functions, and interacting with user operation at the front end interface.

The controller 15 can interact with the industrial personal computer 13, for example, load the control instructions of the industrial personal computer into the memory at any time for storage and execution. The method is mainly responsible for carrying out distance calculation, and marks the coil stock through a certain distance control marking machine after receiving an alarm signal sent by an industrial personal computer. The controller 15 may be a programmable logic controller (Programmable Logic Controller, PLC).

The encoder 16 is generally pressed on the unreeling roller, and the rice pressing wheel is driven to rotate together in the advancing process of the coil stock, and pulse signals can be output when the encoder rotates for a certain angle, so that the line scanning camera is triggered to shoot one line of image data of the coil stock, and therefore the shooting frequency of the line scanning camera and the moving speed of the coil stock are unified, and the line scanning camera is ensured to stably output line scanning images at different speeds.

The marking machine 17 can be driven by a motor to paste the self-adhesive on the position where the coil stock has defects, so as to mark the defects, and prompt the following procedures to pay attention to avoiding and processing.

The embodiment of the application provides a coil stock detection method, which is applied to a visual host computer, as shown in fig. 2, and the coil stock detection method can comprise the following steps of S101 and S102, wherein:

Step S101, respectively acquiring line scanning images of at least two line scanning cameras on a detection station, which are respectively acquired by the coil stock; the at least two line scanning cameras have intersection areas for the acquisition visual field range of the coil stock, and the range of the union areas is larger than the width of the coil stock;

At least two line scanning cameras on the detection station are used for respectively acquiring line scanning images of the coil stock. The vision host computer is an important component of the machine vision system, typically a computer or dedicated device (e.g., an industrial personal computer) for controlling and managing the entire machine vision system, and may receive image data from an image acquisition device (e.g., a line scan camera), process and analyze the image data. The working mode of the line scanning camera is different from that of a common area array camera, the line scanning camera scans the coil stock in a moving state line by line, only one line of image data is collected at a time, and a complete line scanning image is obtained after the image data are spliced.

In some embodiments, at least two line scanning cameras on the detection station are used for photographing the coil stock, respective line scanning images are correspondingly obtained, and the visual upper computer is used for splicing the respective line scanning images of the at least two line scanning cameras to obtain an image which completely covers the coil stock, wherein the line scanning images corresponding to the two adjacent line scanning cameras have intersecting areas.

It should be noted that, in the embodiment of the present application, the at least two line scanning cameras may be two line scanning cameras, three line scanning cameras, four line scanning cameras, and the like. The number of line scan cameras can be determined by the width of the web. For example, if the width of the coil stock is 240 millimeters (mm), four line scan cameras with a total field of view range of 320mm (the field of view range of each line scan camera is 80 mm) are used to photograph the coil stock, so that the total field of view range of the four line scan cameras is larger than the width of the coil stock, the line scan images collected by the four line scan cameras can be ensured to completely cover the coil stock, and the condition of missing photographing in partial areas of the coil stock can be avoided.

Step S102, performing defect detection on the line scan images acquired by each line scan camera respectively, correspondingly obtaining a defect detection result of each line scan image, and sending the defect detection result to a controller.

Here, the PLC is a digital arithmetic operation electronic system which employs a programmable memory in which instructions for performing operations such as logical operations, sequence control, timing, counting, and arithmetic operations are stored, and controls various types of mechanical devices or production processes through digital or analog input and output. The working principle of the PLC is based on input, output and central processing, and works in a 'sequential scanning and continuous circulation' mode. The automatic production process monitoring, adjusting and optimizing device can realize the monitoring, adjusting and optimizing of the automatic production process through programming, so that the production efficiency and quality are improved.

The defect detection result of the line scan image can be that the coil stock is defective or not, and whether the line scan image is defective or not is detected by the visual upper computer, the defect detection result of the line scan image can be sent to the PLC. For example, if the visual upper computer detects that the line scan image has no defects, the visual upper computer sends defect detection results of no defects of the coil stock to the PLC; if the visual upper computer detects that the line scanning image has defects, the visual upper computer sends a defect detection result that the coil stock has defects to the PLC.

In some embodiments, since the area of the coil stock collected by each line scanning camera is different, defect detection needs to be performed on the line scanning image corresponding to the coil stock area collected by each line scanning camera, so that the condition of missing detection in a part of the area of the coil stock can be avoided.

In some embodiments, after step S102, after the PLC receives the defect detection result that the line scan image sent by the vision host computer has a defect, the PLC determines the line scan image having the defect first, and then marks the coil stock area where the defect is located in the line scan image, thereby realizing quality control and tracing of the coil stock on the automated production line.

In some embodiments, the implementation of step S101 "to acquire the line scan images acquired by at least two line scan cameras on the detection station, respectively, on the web material may include the following steps S1011 to S1013, wherein:

step S1011, responding to a trigger signal, and controlling a light source on the detection station to polish the coil stock in a moving state;

Here, the light source includes a strip light source, an array light source, and the like. Taking a light emitting Diode (LIGHT EMITTING Diode) lamp as an example, in some embodiments, the strip light source may be arranged in a straight line by a plurality of LED lamps; the array light source can comprise a plurality of strip light sources which are connected end to end in sequence to form a square frame structure. The strip light source and the array light source can generate high-intensity light, and the illumination is uniformly distributed, so that the coil stock obtains clear and bright illumination effect. Such uniform illumination helps to reduce shadows and reflections, thereby improving image quality. It should be noted that, the light source in the embodiment of the present application may be any of the above light sources.

The trigger signal may be a pulse signal sent by the PLC or a pulse signal sent by the encoder. Next, a case where the trigger signal is a pulse signal from the PLC will be described.

In some embodiments, firstly, during the movement of the unreeling roller, an encoder on the shaft end of the unreeling roller is driven to synchronously move, and the encoder sends a pulse signal generated during the movement to a visual upper computer; then, after the vision upper computer receives the pulse signal, the light source on the detection station is controlled to polish the coil stock in a moving state (for example, the moving speed is 80 meters per minute (m/min)) in response to the pulse signal, and thus, the light source provides proper environment brightness for the line scanning camera to photograph the coil stock.

Step S1012, controlling at least two line scanning cameras on the detection station to simultaneously photograph the coil stock;

In some embodiments, at least two line scanning cameras are used for photographing the coil stock in a moving state, and each line scanning camera is arranged at a different position of the detection station, so that the photographed coil stock areas are different, and therefore, the at least two line scanning cameras on the detection station need to be controlled by the vision host computer to photograph the coil stock at the same time, and the condition that the part area of the coil stock is missed in photographing can be avoided.

Step S1013, controlling the at least two line scan cameras to output line scan images at the same time when the photographing times are greater than the preset times.

Here, the preset number of times may be a value preset according to the size of the roll, for example, 500 times, 600 times, 800 times, 1000 times, and the like. Under the condition that the size of the coil stock and the scanning width of the line scanning camera are given, the photographing times of the line scanning camera are in direct proportion to the length of the coil stock, so that under the condition that the photographing times of the line scanning camera are larger than the preset times, the line scanning camera can be ensured to output complete line scanning images of the coil stock.

In some embodiments, if the preset number of times is equal to 500, after each line scan camera collects 500 lines of image data, the vision host computer controls each line scan camera to simultaneously output a line scan image including 500 lines of image data.

The positional relationship between the line scan camera and the light source of the single detection station will be described with reference to fig. 3 and 4, wherein a first preset angle 25 exists between the view direction of the line scan camera 21 and the normal 24 of the coil 22 as shown in fig. 3; a second preset angle 26 exists between the shining direction of the light source 23 and the normal 24 of the coil 22, and the first preset angle 25 is smaller than the second preset angle 26. As shown in fig. 4, the long sides 30 of the field of view of the line scan cameras 31, 32, 33 and 34 are parallel to the width of the web 22.

It should be noted that, as shown in fig. 3, the first preset angle 25 between the view direction of the line scanning camera 21 and the normal 24 of the coil 22 is smaller than the second preset angle 26 between the light striking direction of the light source 23 and the normal 24 of the coil 22, and the angle between the light striking direction of the light source and the normal of the coil is larger, which is helpful to reduce the shadow generated on the surface of the coil, and the angle between the view direction of the line scanning camera and the normal of the coil is smaller, so that the illumination area in the view of the line scanning camera is more concentrated, and the quality of the line scanning image collected by the line scanning camera can be improved.

Here, the front detection station and the back detection station are respectively provided with two brackets, two line scanning cameras are placed on each bracket, and the area of a vacancy positioned between the two line scanning cameras is smaller than the area of any line scanning camera.

The following describes the positional relationship of four line scanning cameras on two brackets on the front surface detection station with reference to fig. 4, and as shown in fig. 4, the line scanning camera 31 and the line scanning camera 33 are located on one bracket of the front surface detection station, the line scanning camera 32 and the line scanning camera 34 are located on the other bracket of the front surface detection station, and the line scanning cameras 31 to 34 on the front surface detection station are staggered with each other. For example, there is a gap 35 between the line scan camera 31 and the line scan camera 33, a gap 36 between the line scan camera 32 and the line scan camera 34, and the area of the gap 35 or 36 is smaller than the area of either line scan camera on the front face detection station.

In some embodiments, the front side inspection station includes a first front side inspection station and a second front side inspection station; the implementation of step S1012 "controlling at least two line scanning cameras on the inspection station to simultaneously photograph the web material" may include the following steps S1121 and S1122, wherein:

Step S1121, controlling two line scanning cameras on the first front detection station to simultaneously photograph the coiled material;

In some embodiments, because the two line scanning cameras on the first front detection station are located on the same support and are arranged at intervals, the vision upper computer is required to control the two line scanning cameras on the first front detection station to photograph the coiled material at the same time, so that the condition that the part of the area of the coiled material is missed can be avoided.

And step S1122, controlling two line scanning cameras on the second front detection station to simultaneously photograph the coiled material.

In some embodiments, because the two line scanning cameras on the second front detection station are located on the same support and are arranged at intervals, the two line scanning cameras on the second front detection station need to be controlled by the vision upper computer to photograph the coiled material at the same time, so that the condition that the part of the area of the coiled material is missed can be avoided.

In some embodiments, the defect detection includes a metal leakage detection of the lithium electrode sheet; the implementation of the "performing defect detection on the line scan image acquired by each line scan camera in step S102, and correspondingly obtaining the defect detection result of each line scan image" may include the following steps S1021 to S1026, where:

Step S1021, performing binarization processing on the line scan image acquired by each line scan camera respectively to correspondingly obtain a binary image of each line scan image;

Here, the line scan image may be a Red Green Blue (RGB) color image. The binarization process is a process of setting the RGB value of each pixel point of a color image to 0 (black) or 255 (white), so that the original color range is changed from 256 colors to 2 black and white colors, and only black and white colors are formed in the binarized image.

In some embodiments, before the binarization processing of the line scan image, the visual upper computer may perform gray scale processing on the line scan image to obtain a gray scale image; then, the pixel value corresponding to each pixel point in the gray image is scanned, the pixel value smaller than the preset threshold value is set to 0, and the pixel value larger than or equal to the preset threshold value is set to 255, so that a binary image of the line scan image is obtained.

Step S1022, determining a first connection area where white pixels in each binary image are connected to each other; the white pixel points represent pixel points corresponding to a film region in the binary image;

here, the first communication region may be an image region composed of white pixel points having the same pixel value and adjacent in position in the binary image.

Step S1023, performing expansion operation on the first communication area corresponding to each binary image to correspondingly obtain a second communication area;

Here, the expansion operation may be implemented by dilate functions. In image processing, the range of a white highlight region in an image can be enlarged by a dilation operation so as to more accurately locate and identify a defect.

In some embodiments, the visual upper computer performs expansion processing on the first communication area corresponding to each binary image, fills up a blank part in the first communication area, enhances image features in the first communication area, and enlarges a white highlight area in the image, so as to correspondingly obtain the second communication area.

Step S1024, determining the position of each second communication area in the corresponding line scan image;

In some embodiments, since the second communication region is a sub-region in the line scan image, it is necessary to convert the coordinates of the second communication region into a coordinate system of the line scan image by the vision host computer so that the second communication region can correctly correspond to the corresponding position in the line scan image.

Step S1025, screening a third communication area from the second communication area; the third communication region characterizes a region possibly having a metal leakage defect;

here, because the second communication area is an area corresponding to the film area of the coil stock, and the metal leakage defect may exist on the film area, the noise in the second communication area can be removed by the visual upper computer through corrosion operation, so that the third communication area which may have the metal leakage defect is more accurately screened, and whether the third communication area has the defect is more accurately positioned.

Step S1026, determining the defect detection result of the third communication area corresponding to the area larger than the preset metal leakage area as metal leakage.

In some embodiments, the visual host computer compares the area of the third communication region with the preset metal leakage area, and if the area of the third communication region is larger than the preset metal leakage area, it may be determined that the third communication region has a metal leakage defect.

In some embodiments, the implementation of step S1025 "screen the third communication area from the second communication area" may include steps S1251 to S1253, where:

step S1251, determining the number of pixel points in each second communication area and the pixel value corresponding to each pixel point;

Here, the number of pixel points in the second connected region and the pixel value corresponding to each pixel point may be determined by traversing all the pixel points in the second connected region.

Step S1252, determining, as a target pixel, a pixel having a pixel value greater than a second preset value in the second communication area, if the number of the pixel points is greater than the first preset value;

here, the first preset value may be any suitable value, for example, 25, 40, etc. The second preset value may be any suitable value, e.g., 205, 234, etc. By setting the first preset value, it is possible to exclude those second connected regions where the number of pixels is too small, possibly constituted by noise or extraneous details. By setting the second preset value, the pixel point with a higher pixel value can be screened out.

Step S1253, determining the third communication area based on the target pixel point.

Here, after the target pixel points are determined, the communication region may be formed based on the target pixel points.

In some embodiments, the defect detection comprises asperity detection of the lithium electrode sheet; the implementation of the step S102 of performing defect detection on the line scan image acquired by each line scan camera, and correspondingly obtaining the defect detection result of each line scan image, may further include the following steps S1121 and S1122, where:

Step S1121, classifying each line scan image by a pre-trained classifier, and correspondingly obtaining a fourth communication region; the fourth communication region represents a region where an concave-convex point possibly exists;

Here, the pre-trained classifier may be a deep learning model (e.g., convolutional neural network) for identifying potential regions of asperities. The pre-trained classifier may be obtained by training a classifier using images with the concave-convex points labeled.

In some embodiments, the implementation of step S1121 "categorize each of the line scan images by a pre-trained categorizer, which corresponds to obtaining a fourth connected region" may further include steps S1211 to S1214, where:

step S1211, performing semantic segmentation on each line scan image, so as to obtain a segmented region of each line scan image correspondingly;

Here, each line scan image may be semantically segmented by any one of a full convolution network (Fully Convolutional Networks, FCNs), a deep learning model (Deeplab), a Transformer-based semantic segmenter (SegFormer), corresponding to segmented regions resulting in line scan images, each line scan image may be divided into different regions, each region may correspond to a different portion of the lithium electrode sheet, e.g., a normal region, a region that may contain defects, etc.

Step S1212, extracting features of the segmented regions of each line scan image, and correspondingly obtaining features of each segmented region;

Here, in the detection of the lithium electrode sheet, it may be necessary to pay attention to specific features related to the concave-convex point defect, for example, local brightness variation, edge irregularities, and the like. Therefore, feature extraction needs to be performed on the segmented region of each line scan image, and features of each segmented region are correspondingly obtained.

The feature extraction method for the segmented region of the line scan image may include, but is not limited to, passing through a direction gradient histogram (Histogram of Oriented Gradient, HOG), a local binary pattern (Local Binary Pattern, LBP), and the like. For example, by HOG, the segmented region is divided into small connected regions, then the gradient or edge direction histograms of each pixel in the connected regions are collected, and finally these histograms are combined to obtain multiple features.

Step S1213, classifying the features of each of the segmented regions by the pre-trained classifier, to obtain a classification result;

Here, the pre-trained classifier may be a convolutional neural network (Convnext), a residual network (Resnet), or the like. In some embodiments, the extracted features of each segmented region are input into a pre-trained classifier, which classifies the features of each segmented region and outputs a classification result. The classification result may be a probability value or a class label indicating whether the segmented region contains a bump defect.

Step S1214, determining a segmentation region corresponding to the classification result as the fourth communication region when the classification result indicates that the lithium electrode sheet has the concave-convex point defect.

In step S1122, if the confidence coefficient of the fourth connected region is greater than the preset threshold, it is determined that the defect detection result of the line scan image corresponding to the fourth connected region is an uneven point.

Here, the confidence reflects the likelihood or confidence that the region is a concave-convex point. In some embodiments, the confidence of the fourth connected region related to the concave-convex point may be calculated by a confidence calculation formula.

In some embodiments, the defect detection results include a first result for characterizing the coil stock as acceptable and a second result for characterizing the coil stock as unacceptable, the detection method further including the following steps S111 and S112, wherein:

Step S111, when the defect detection result indicates that the coil stock has no defect, the first result is sent to a controller;

Step S112, when the defect detection result indicates that the coil stock has a defect, sending the second result to the controller.

The pole piece of the power battery is one of important components of the power battery, and the quality and the performance of the pole piece directly determine the quality and the safety of the power battery. The pole piece of the power battery is composed of a base material and a coating material, wherein the base material of the anode pole piece is copper foil with the thickness of about 8 micrometers (mum), the coating material is an active medium with the thickness of about 12 micrometers, and the anode coil stock is formed by the processes of stirring, coating, rolling, slitting and the like of slurry. During the production process of the process, the coil stock may have metal leakage, edge wrinkles, pinhole breakage, concave-convex points, wavy edges, excessive design of the negative electrode (Overhang) and other depressions. Therefore, the quality detection is carried out on the power battery pole piece in the production process, and the consistency is very important to ensure.

For the defect detection of the coil stock, a visual inspection person is initially arranged to conduct manual spot inspection, and can conduct spot inspection on the produced coil stock through sampling and spot inspection on two surface film areas of a section of coil stock through an optical microscope. However, with the improvement of production efficiency of the production line, the defects of low manual visual inspection efficiency and high omission factor appear, so that a visual inspection system is introduced, a black-and-white line scanning camera is adopted, and a line scanning light source is matched for polishing, so that the full inspection of the coiled material pole piece is completed.

The current visual detection system adopts a black-and-white line scanning camera to clean the light in cooperation with a line scanning light source, so that the full detection of the coil stock is completed, but the coating material in the film area of the coil stock is uneven or is not coated due to the process change of the coil stock, so that the general black-and-white line scanning camera cannot distinguish the metal leakage in the film area of the coil stock well to obtain the gray value under the condition of the line scanning light source dark field. In addition, with the improvement of the coil stock process, the precision of the detection requirement on the surface metal leakage of the coil stock is improved, the current optical detection scheme cannot meet the detection specification requirement, and the copper foil substrate and the film region can not be distinguished only through gray values.

The embodiment of the application is mainly used for solving the detection problem of leakage of the surface metal material of the anode plate film region caused by manual misoperation, production equipment failure or process design defects in the previous process link of power battery production. The embodiment of the application provides a visual detection route for detecting the surface characteristics of the front and back surfaces of a coil stock by multi-camera splicing based on an industrial visual detection technology, solves the problem that the existing optical visual scheme cannot effectively detect metal leakage on the surface of the coil stock and distinguishes a copper foil substrate from a film region, and further improves the quality and performance of a battery pole piece.

Aiming at the high-precision detection requirement of the metal leakage area of the power battery pole piece, the embodiment of the application provides an optical method for improving the visual detection precision by multi-camera splicing. A plurality of 16K (K represents the resolution of a line scanning camera) color TDI line scanning cameras (TIME DELAY Integration LINE SCAN CAMERA, a time delay Integration line scanning camera) are adopted at a metal leakage detection visual station at the unreeling position of the anode coil stock, and a high-resolution lens and a light source are matched for lighting and photographing. Because the surface of the film area material of the coil stock is smooth and reflective, the copper foil material of the substrate is rough, under the condition of dark field lighting, the light reflected by the film area does not enter the camera lens image of the line scanning camera and is represented as dark black in the film area, and the diffuse reflection of the copper foil of the substrate is represented as bright red in the image. And the width of the coil stock covered by the camera view field is ensured through multi-camera splicing, the single-pixel precision of the camera is improved, and the theoretical detection precision reaches 0.01 millimeter per pixel (mm/pix). According to the embodiment of the application, the visual field of the line scanning camera is ensured to cover the width of the anode coil stock through multi-camera splicing, and the single-pixel precision of the line scanning camera is improved. Visual detection stations are respectively arranged on the front side and the back side of the anode coil stock, a 16K color TDI line scanning camera is adopted according to the surface defect detection item characteristics, metal leakage detection precision, the width of the coil stock and the size of an installation space of the coil stock, lens focal lengths meeting the required visual field and precision of the camera are calculated, and the working distance meets the space arrangement and the number of the cameras on site.

As shown in fig. 5 and 6, four 16K color TDI line scan cameras are respectively arranged on the front and back sides of the anode roll, and a total of eight line scan cameras are arranged on the front surface 41 of the anode roll, wherein the line scan cameras 401 to 404 (not shown in fig. 5 and 6) and the light sources 411 and 412 are arranged on the front surface 41 of the anode roll, wherein the line scan cameras 401 and 403 are placed in the same line scan camera dust cover 415, and the line scan cameras 402 and 404 are placed in the same line scan camera dust cover 416; the opposite face 42 of the anode roll is provided with wire sweep cameras 405 to 408 (not shown in fig. 5 and 6) and light sources 413 and 414, wherein the wire sweep cameras 405 and 407 are placed in the same wire sweep camera dust cover 417 and the wire sweep cameras 406 and 408 are placed in the same wire sweep camera dust cover 418. Each line scanning camera shoots an anode coil material area; and performing image processing on the line scan image corresponding to each photographed coil stock area through an algorithm, identifying a film area of the line scan image, performing image processing on some columns such as binarization, and detecting a metal leakage area and an area of the film area.

Aiming at the anode pole pieces manufactured by different processes, two sets of polishing schemes (for example, front polishing of the anode coil stock and back polishing of the anode coil stock) are designed according to the embodiment of the application, so that the algorithm detection requirement is met, and the metal leakage area of the film area is accurately grasped and the area is detected. The embodiment of the application effectively solves the problems that the metal leakage of the anode coil stock is difficult to detect, the detection labor consumption is high and the like in the production process, thereby greatly reducing the leakage and killing of the metal leakage defects, and further achieving the purposes of reducing the product risk and improving the defect detection rate and the product quality.

The following describes an optical method for improving visual detection accuracy by multi-camera splicing by taking a front visual station of a coil stock as an example, and a back visual station is similar and will not be described in detail herein.

In the embodiment of the application, because the width of the coil stock is larger and the requirement on detection precision is high, the shooting vision of each line scanning camera needs to be distributed in consideration of the larger volume of the line scanning camera. As shown in fig. 4, the line scan cameras are respectively mounted on two brackets (not shown in fig. 4) in view, two line scan cameras are provided on each bracket, the line scan camera 31 and the line scan camera 33 are photographed on one bracket, and the line scan camera 32 and the line scan camera 34 are photographed on the other bracket. As shown in fig. 3, the included angle between the direction of the visual field photographed by each line scanning camera 21 and the normal 24 of the coil 22 is a first preset angle 25, the long side of the visual field of the line scanning camera is parallel to the width of the coil, the included angle between the light source 23 and the normal 24 of the coil 22 is a second preset angle 26, and the first preset angle 25 is smaller than the second preset angle 26. As shown in fig. 7, each vision station is provided with a dust-proof cover 61 of the line scanning camera, a dust-proof blowing mechanism 62 and a light source 63 (for example, a line scanning strip light source), wherein two line scanning cameras on each detection station are placed in the same dust-proof cover 61 of the line scanning camera, and the dust-proof blowing mechanism 62 is used for preventing dust from accumulating at a lens in the unreeling process of the roll material, so that the imaging effect of the line scanning camera is affected.

The embodiment of the application provides another detection method of coil stock, as shown in fig. 8, the detection method comprises the following steps:

step S710, the coil stock moves on the unreeling roller, and the line scanning camera is triggered to collect line scanning images of the coil stock;

here, the shaft end of the unreeling roller is provided with an encoder, and the encoder sends a generated pulse signal to the line scanning camera in the process that the reeling material moves on the unreeling roller, and the line scanning starts to collect line scanning images of the reeling material in response to the pulse signal.

Step S720, four line scanning cameras on the front visual station respectively shoot and take pictures of four areas on the front surface of the coil stock, and four line scanning cameras on the back visual station respectively shoot and take pictures of four areas on the back surface of the coil stock;

Step S730, detecting line scan images of 8 coil stock areas respectively by a software algorithm on a visual upper computer so as to judge whether the coil stock has metal leakage or concave-convex points; if so, the step S740 is performed, and if not, the step S760 is performed, if not, the step S is performed without the metal leakage or the concave-convex points;

Here, determining whether or not the coil stock has a metal leakage includes: firstly, carrying out image processing on line scan images of 8 areas of a coil stock respectively by an algorithm to identify a film area coating material and a copper foil substrate of the images, and secondly, carrying out binarization processing on the film area of the line scan image to detect the position and the area of a metal leakage area of the film area; then, comparing the metal leakage area with a preset area, and judging whether the metal leakage area exceeds the detection requirement; the step of judging whether the coil stock has concave-convex points comprises the following steps: and performing image processing on the line scan images of 8 areas of the coil stock by using a deep learning algorithm and a traditional algorithm so as to position and judge concave-convex points on the surface of the coil stock.

The metal leakage area of the detection film region includes: and respectively carrying out binarization processing on each line scan image to correspondingly obtain a binary image of each line scan image. An image area composed of white pixel points having the same pixel value and adjacent positions in each binary image is determined as a first communication area. And performing expansion processing on the first communication area corresponding to each binary image, filling up a blank part in the first communication area, enhancing image characteristics in the first communication area, expanding a white highlight area in the image, and correspondingly obtaining a second communication area. And determining the number of the pixel points in the second communication area and the pixel value corresponding to each pixel point by traversing all the pixel points in the second communication area. And determining the pixel point with the pixel value larger than the second preset value in the second communication area as a target pixel point under the condition that the number of the pixel points is larger than the first preset value. After the target pixels are determined, a connectivity detection algorithm may be used to find the third connected region formed by the target pixels. And determining a defect detection result of the third communication region corresponding to the area larger than the preset metal leakage area as metal leakage.

Step S740, the vision upper computer sends a coil stock OK result, namely a coil stock quality qualified result, to the PLC;

step S750, the PLC feeds back an OK result to the unreeling roller and controls the unreeling roller to normally operate;

step 760, the vision upper computer sends the coil stock NG result, namely the coil stock quality disqualification result, to the PLC.

If the coil stock has metal leakage or concave-convex points, the upper computer sends a coil stock NG result to the PLC, wherein the NG result comprises the defect type (metal leakage or concave-convex points) of the coil stock; if the defect type is metal leakage, the NG result also comprises the area and the position of a metal leakage area; if the defect type is concave-convex points, the NG result also contains the positions of the concave-convex points.

Compared with the prior art, the embodiment of the application has the following advantages:

1. According to the embodiment of the application, the 16K color TDI line scanning camera based on the multi-stage exposure accumulation type line scanning imaging principle is adopted on the front side and the back side of the coil stock, and the substrate and the film area of the coil stock are photographed, so that the color image of the surface characteristics of the coil stock is obtained. The color image obtained through the traditional algorithm processing can accurately distinguish the positions and the areas of the copper foil substrate, the film area and the metal leakage area of the film area, and the screened metal leakage area is compared with a preset area to judge whether the metal leakage area exceeds the detection requirement.

2. The visual detection station in the embodiment of the application can finish the identification of the surface defects of the coil stock, distinguish the copper foil substrate, the film region material and the metal leakage of the coil stock, and improve the detection capability of the metal leakage of the surface of the coil stock.

3. According to the embodiment of the application, deep learning is applied in the visual detection station, the obtained high-precision image is subjected to image processing by combining a traditional algorithm, the positions of the concave-convex points on the surface of the coil stock are rapidly positioned, and the positioning and judgment of the concave-convex points on the surface of the pole piece are realized.

4. The embodiment of the application can detect surface defects such as metal leakage and concave-convex points on the surface of the coil stock in real time under the condition of not influencing on-site production, obtain the area position and area information of the metal leakage and concave-convex points on the surface of the coil stock, and can interact corresponding NG signals or OK signals with unreeling equipment to discharge waste in a later process section.

The embodiment of the application provides a coil stock detection system, as shown in fig. 9, the coil stock detection system 800 comprises at least two line scanning cameras 810 and a visual upper computer 820; wherein the at least two line scan cameras 810 are configured to collect line scan images of the coil stock; the at least two line scanning cameras have intersection areas for the acquisition visual field range of the coil stock, and the range of the union areas is larger than the width of the coil stock; the visual upper computer 820 is configured to respectively acquire line scan images acquired by at least two line scan cameras on the detection station for the coil stock respectively; and respectively carrying out defect detection on the line scanning images acquired by each line scanning camera, correspondingly obtaining a defect detection result of each line scanning image, and sending the defect detection result to a controller.

In some embodiments, the coil stock detection system 800 further includes a controller; the controller is used for marking a coil stock area where the defect is located in the line scanning image under the condition that the defect detection result of the line scanning image represents that the defect exists.

It should be appreciated that reference throughout this specification to "one embodiment" or "an embodiment" means that a particular feature, structure or characteristic described in connection with the embodiment is included in at least one embodiment of the present application. Thus, the appearances of the phrases "in one embodiment" or "in an embodiment" in various places throughout this specification are not necessarily all referring to the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. It should be understood that, in various embodiments of the present application, the sequence number of each step/process described above does not mean that the execution sequence of each step/process should be determined by its functions and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application. The foregoing embodiment numbers of the present application are merely for the purpose of description, and do not represent the advantages or disadvantages of the embodiments.

It should be noted that, in the application, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.

In the several embodiments provided by the present application, it should be understood that the disclosed systems, devices, and methods may be implemented in other manners. The above described device embodiments are only illustrative, e.g. the division of the units is only one logical function division, and there may be other divisions in practice, such as: multiple units or components may be combined or may be integrated into another system, or some features may be omitted, or not performed. In addition, the various components shown or discussed may be coupled or directly coupled or communicatively coupled to each other via some interface, whether indirectly coupled or communicatively coupled to devices or units, whether electrically, mechanically, or otherwise.

The units described above as separate components may or may not be physically separate, and components shown as units may or may not be physical units; can be located in one place or distributed to a plurality of network units; some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment. In addition, each functional unit in each embodiment of the present application may be integrated in one processing unit, or each unit may be separately used as one unit, or two or more units may be integrated in one unit; the integrated units may be implemented in hardware or in hardware plus software functional units.

The foregoing is merely an embodiment of the present application, but the scope of the present application is not limited thereto, and any changes or substitutions easily contemplated by those skilled in the art within the technical scope of the present application should be included in the scope of the present application.

Claims

1. The detection method of the coil stock is characterized by being applied to a visual host computer, and comprises the following steps:

Respectively acquiring line scanning images acquired by at least two line scanning cameras on a detection station for the coil stock respectively; the at least two line scanning cameras have intersection areas for the acquisition visual field range of the coil stock, and the range of the union areas is larger than the width of the coil stock;

Performing defect detection on the line scanning images acquired by each line scanning camera respectively, correspondingly obtaining a defect detection result of each line scanning image, and sending the defect detection result to a controller;

the detection station comprises a front detection station and a back detection station;

the front detection station is used for detecting the front of the lithium electrode slice;

the back detection station is used for detecting the back of the lithium electrode slice;

wherein, each detection station is provided with two brackets;

Two line scanning cameras are arranged on each support, a gap with a preset area exists between the two line scanning cameras, and the area of each line scanning camera is larger than that of the gap;

Four line scanning cameras on each detection station are arranged in a staggered mode.

2. The method of claim 1, wherein the step of acquiring respective line scan images of the roll acquired by at least two line scan cameras at the inspection station, respectively, comprises:

Responding to a trigger signal, and controlling a light source on the detection station to polish the coil stock in a moving state;

controlling at least two line scanning cameras on the detection station to simultaneously photograph the coiled material;

and under the condition that the photographing times are larger than the preset times, controlling the at least two line scanning cameras to output line scanning images at the same time.

3. The method of claim 2, wherein each of the inspection stations includes a line scan camera and a light source;

Wherein the long side of the visual field of the line scanning camera is parallel to the width of the coil stock;

A first preset angle exists between the visual field direction of the line scanning camera and the normal line of the coil stock;

A second preset angle exists between the polishing direction of the light source and the normal line of the coil stock.

4. The method of claim 1, wherein the front side inspection station comprises a first front side inspection station and a second front side inspection station;

Controlling at least two line scanning cameras on the detection station to simultaneously photograph the coil stock, comprising:

Controlling two line scanning cameras on the first front detection station to simultaneously photograph the coiled material;

And controlling two line scanning cameras on the second front detection station to simultaneously photograph the coiled material.

5. The method of detecting a web according to claim 1, wherein the defect detection includes a metal leakage detection of the lithium electrode sheet;

Performing defect detection on the line scan image acquired by each line scan camera respectively to correspondingly obtain a defect detection result of each line scan image, including:

respectively carrying out binarization processing on the line scanning images acquired by each line scanning camera to correspondingly obtain a binary image of each line scanning image;

determining a first communication area in which white pixels in each binary image are connected with each other; the white pixel points represent pixel points corresponding to a film region in the binary image;

performing expansion operation on the first communication area corresponding to each binary image to correspondingly obtain a second communication area;

determining the position of each second communication area in the corresponding line scan image;

screening a third communication area from the second communication area; the third communication region characterizes a region possibly having a metal leakage defect;

And determining a defect detection result of the third communication region corresponding to the area larger than the preset metal leakage area as metal leakage.

6. The method of claim 5, wherein selecting a third communication area from the second communication areas, comprises:

Determining the number of pixel points in each second communication area and the pixel value corresponding to each pixel point;

Determining the pixel points with the pixel values larger than a second preset value in the second communication area as target pixel points under the condition that the number of the pixel points is larger than a first preset value;

And determining the third communication area based on the target pixel point.

7. The method of detecting a web material according to claim 1, wherein the defect detection includes a concave-convex point detection of the lithium electrode sheet;

performing defect detection on the line scanning images acquired by each line scanning camera respectively to correspondingly obtain defect detection results of each line scanning image, and further comprising:

classifying each line scanning image through a pre-trained classifier to correspondingly obtain a fourth communication area; the fourth communication region represents a region where an concave-convex point possibly exists;

And if the confidence coefficient of the fourth communication region is larger than a preset threshold value, determining that the defect detection result of the line scan image corresponding to the fourth communication region is concave-convex points.

8. The method of claim 7, wherein classifying each of the line scan images by a pre-trained classifier, the fourth connected region being obtained correspondingly, comprises:

Carrying out semantic segmentation on each line scanning image to correspondingly obtain a segmentation area of each line scanning image;

extracting features of the segmented regions of each line scan image, and correspondingly obtaining the features of each segmented region;

Classifying the characteristics of each segmented region through the pre-trained classifier to obtain a classification result;

And determining a segmentation area corresponding to the classification result as the fourth communication area under the condition that the classification result represents that the lithium electrode plate has the concave-convex point defect.

9. A method of inspecting a web according to any one of claims 2 to 8, wherein the defect inspection results include a first result for characterizing that the web is acceptable and a second result for characterizing that the web is unacceptable, the method further comprising:

sending the first result to a controller under the condition that the defect detection result indicates that the coil stock has no defect;

And sending the second result to the controller when the defect detection result indicates that the coil stock has defects.

10. The detection system for the coiled material is characterized by comprising at least two line scanning cameras and a visual upper computer;

The at least two line scanning cameras are used for collecting line scanning images of the coil stock; the at least two line scanning cameras have intersection areas for the acquisition visual field range of the coil stock, and the range of the union areas is larger than the width of the coil stock;

The visual upper computer is used for respectively acquiring line scanning images of the coil stock respectively acquired by at least two line scanning cameras on the detection station; performing defect detection on the line scanning images acquired by each line scanning camera respectively, correspondingly obtaining a defect detection result of each line scanning image, and sending the defect detection result to a controller; the detection station comprises a front detection station and a back detection station; the front detection station is used for detecting the front of the lithium electrode slice; the back detection station is used for detecting the back of the lithium electrode slice; wherein, each detection station is provided with two brackets; two line scanning cameras are arranged on each support, a gap with a preset area exists between the two line scanning cameras, and the area of each line scanning camera is larger than that of the gap; four line scanning cameras on each detection station are arranged in a staggered mode.

11. The coil stock inspection system of claim 10, further comprising a controller; the controller is used for marking a coil stock area where the defect is located in the line scanning image under the condition that the defect detection result of the line scanning image represents that the defect exists.