CN112967245B - Battery detection method and device, electronic equipment and readable storage medium - Google Patents
Battery detection method and device, electronic equipment and readable storage medium Download PDFInfo
- Publication number
- CN112967245B CN112967245B CN202110228134.9A CN202110228134A CN112967245B CN 112967245 B CN112967245 B CN 112967245B CN 202110228134 A CN202110228134 A CN 202110228134A CN 112967245 B CN112967245 B CN 112967245B
- Authority
- CN
- China
- Prior art keywords
- cross
- battery
- section
- particle
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/30—Noise filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- General Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
- Secondary Cells (AREA)
Abstract
The application provides a battery detection method, a battery detection device, an electronic device and a readable storage medium. The method comprises the following steps: sectioning the first three-dimensional battery model at a specified interval from a first three-dimensional battery model obtained by transmission scanning of the battery to be tested to obtain a plurality of continuous sectional views, wherein the first three-dimensional battery model comprises a particle model obtained by transmission scanning of particles in the battery to be tested; and inputting the plurality of cross sections into the trained battery detection model to obtain a detection result output by the battery detection model for detecting each cross section in the plurality of cross sections. In this scheme, dissect through first three-dimensional battery model, then detect a plurality of cross-sectional views that obtain by battery test model to obtain the testing result to the battery that awaits measuring, so, need not artifical the detection, be favorable to improving detection efficiency, improve because of the artifical inefficiency that leads to battery test and the problem of making mistakes easily that detects.
Description
Technical Field
The present application relates to the field of artificial intelligence, and in particular, to a battery detection method, apparatus, electronic device, and readable storage medium.
Background
In the battery manufacturing industry, there is a need to detect the interior of a battery to ensure that the battery is qualified when it leaves the factory. When detecting the inside of the battery, the battery can be scanned by a ray with strong penetrating power (such as an X-ray) so as to obtain a three-dimensional battery model of the battery. If particles exist in the battery, the three-dimensional battery model comprises a particle model diagram obtained by scanning the particles in the battery. Then, the particulate matter in the three-dimensional battery model is manually detected. Wherein, when artifical the detection, there is visual fatigue easily, leads to the inefficiency of battery detection and makes mistakes easily.
Disclosure of Invention
An object of the embodiments of the present application is to provide a battery detection method, a battery detection device, an electronic device, and a readable storage medium, which can solve the problems of low efficiency and error susceptibility of battery detection.
In order to achieve the above object, the embodiments of the present application are implemented as follows:
in a first aspect, an embodiment of the present application provides a battery detection method, where the method includes:
sectioning and drawing the first three-dimensional battery model at a specified interval from a first three-dimensional battery model obtained by transmission scanning of a battery to be tested to obtain a plurality of continuous sectional views, wherein the first three-dimensional battery model comprises a particle model obtained by transmission scanning of particles in the battery to be tested;
inputting the plurality of cross sections into a trained battery detection model to obtain a detection result output by the battery detection model for detecting each cross section in the plurality of cross sections, wherein the detection result comprises the number of particles existing in each cross section, the contour size of the cross section of each particle and the position information of the cross section of each particle.
In the above embodiment, the first three-dimensional battery model is sectioned, and then the battery detection model detects the obtained plurality of sectional views, so that the detection result of the battery to be detected is obtained, and thus, manual detection is not needed, the detection efficiency is improved, and the problems that the efficiency of battery detection is low and errors are easy to occur due to manual detection are solved.
With reference to the first aspect, in some optional embodiments, the method further comprises:
judging whether target particles with the contour size larger than or equal to a first specified size exist according to the contour size of the cross section of each particle;
and when the target particles exist, sending prompt information for representing that the battery to be tested is abnormal.
In the above embodiment, by sending the prompt message, the manager can find the abnormal battery in time.
With reference to the first aspect, in some optional embodiments, the method further comprises:
determining a distance between a center point of a cross section of the target particle in the first cross section and a center point of a cross section of the target particle in the second cross section when the target particle exists in the detection results of the first cross section and the second cross section, wherein the first cross section and the second cross section are any group of adjacent cross sections;
when the distance between the central points is smaller than the section radius of the target particle in the first section view or the second section view, determining that the target particle in the first section view and the target particle in the second section view are the same target particle;
and determining the number of the particles in the battery to be tested according to the number of the repeated particles obtained by the same target particles and the number of the particles existing in each cross section.
In the above embodiment, the same particle with a larger particle size may appear in a plurality of adjacent cross-sectional views, and when the number of particles is calculated, the number of repeated particles is combined, which is beneficial to improving the accuracy of particle counting.
With reference to the first aspect, in some alternative embodiments, the specified spacing is less than or equal to
Multiple of the first specified size.
In the above embodiment, by defining the designated interval, it is advantageous to avoid missing detection of larger particles.
With reference to the first aspect, in some optional embodiments, before the cutting and drawing the first three-dimensional battery model at a specified interval in the first three-dimensional battery model obtained from the transmission scanning of the battery to be tested to obtain a plurality of continuous sectional views, the method further includes:
obtaining a training data set, wherein the training data set comprises a plurality of groups of cut-away views and a label corresponding to each cut-away view of the plurality of groups of cut-away views, and each group of cut-away views comprises a plurality of cut-away views of any one of a plurality of second three-dimensional battery models;
and training a deep learning model through the training data set to obtain the trained battery detection model.
In the above embodiment, the battery detection model is obtained by training the deep learning model, so that the accuracy and efficiency of the battery detection model for detecting the cross-sectional view of the battery can be improved.
With reference to the first aspect, in some optional embodiments, before training the deep learning model by the training data set, the method further comprises:
when the number of particles with contour sizes smaller than a second designated size in the plurality of groups of cross-sectional views is smaller than a first designated number, adding a cross-sectional view area with contour sizes smaller than the second designated size in a part or all of the plurality of groups of cross-sectional views to obtain a data-enhanced training data set, wherein the second designated size is smaller than the first designated size.
In the embodiment, the small-size particles with small quantity are filled with the samples, so that the data enhancement of the small-size particles can be realized, the detection effect of the deep learning model on the small-size particles is favorably improved, and the missing detection is avoided.
With reference to the first aspect, in some optional embodiments, before inputting the plurality of cross-sectional views into the trained battery testing model, the method further comprises:
and carrying out noise reduction treatment on the edge noise on the plurality of sectional views to obtain a plurality of sectional views after noise reduction, and inputting the plurality of sectional views into the trained battery detection model.
In the above embodiment, the noise reduction processing is performed on the cross-sectional view, which is beneficial to improving the definition of the contour line in the cross-sectional view, and is beneficial to improving the accuracy of particle detection.
In a second aspect, the present application also provides a battery testing apparatus, the apparatus comprising:
the device comprises a section unit, a data acquisition unit and a display unit, wherein the section unit is used for sectioning and drawing a first three-dimensional battery model at a specified interval from the first three-dimensional battery model obtained by transmission scanning of a battery to be tested to obtain a plurality of continuous section views, and the first three-dimensional battery model comprises a particle model obtained by transmission scanning of particles in the battery to be tested;
and the detection unit is used for inputting the plurality of cross sections into a trained battery detection model to obtain a detection result output by the battery detection model when the battery detection model detects each cross section in the plurality of cross sections, wherein the detection result comprises the number of particles existing in each cross section, the contour size of the cross section of each particle and the position information of the cross section of each particle.
In a third aspect, the present application further provides an electronic device, which includes a processor and a memory coupled to each other, wherein the memory stores a computer program, and when the computer program is executed by the processor, the electronic device is caused to perform the method described above.
In a fourth aspect, the present application further provides a computer-readable storage medium having stored thereon a computer program which, when run on a computer, causes the computer to perform the method described above.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments of the present application will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and that those skilled in the art can also obtain other related drawings based on the drawings without inventive efforts.
Fig. 1 is a schematic structural diagram of an electronic device according to an embodiment of the present application.
Fig. 2 is a schematic structural diagram of a three-dimensional battery model of a battery to be tested according to an embodiment of the present application.
Fig. 3 isbase:Sub>A sectional view of the sectionbase:Sub>A-base:Sub>A in fig. 2.
Fig. 4 is a sectional view of the section B-B in fig. 2.
Fig. 5 is a schematic flowchart of a battery detection method according to an embodiment of the present disclosure.
Fig. 6 is a block diagram of a battery detection apparatus according to an embodiment of the present application.
An icon: 10-an electronic device; 11-a processing module; 12-a storage module; 100-battery detection means; 110-a section unit; 120-detection unit.
Detailed Description
The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application. It should be noted that the terms "first," "second," and the like are used solely to distinguish one from another and are not to be construed as indicating or implying any relative importance. The embodiments described below and the features of the embodiments can be combined with each other without conflict.
Referring to fig. 1, an electronic device 10 for testing a battery under test to determine whether abnormal particles (particles) exist in the battery under test is provided according to an embodiment of the present disclosure. The electronic device 10 may improve the efficiency and accuracy of battery testing. The electronic device 10 may be a personal computer, a server, or the like, and is not particularly limited herein.
The electronic device 10 may include a processing module 11 and a memory module 12. The memory module 12 stores therein a computer program which, when executed by said processing module 11, enables the electronic device 10 to perform the steps of the method described below.
Of course, the electronic device 10 may also include other modules, for example, the electronic device 10 may also include a display screen, a battery detection apparatus 100 solidified in the storage module 12, and the like. The processing module 11, the storage module 12 and the components of the battery testing apparatus 100 are electrically connected directly or indirectly to realize data transmission or interaction. For example, the components may be electrically connected to each other via one or more communication buses or signal lines.
Referring to fig. 2, fig. 3 and fig. 4, in the present embodiment, the battery to be tested may be determined according to actual situations, and may be, but is not limited to, a lithium ion battery or a lead storage battery. In addition, the shape of the battery to be tested can be, but is not limited to, a cylinder and a cuboid. For example, the battery under test may be as shown in fig. 2. The battery under test shown in fig. 2 may be a schematic diagram of a three-dimensional battery model obtained by transmitting a ray having a high penetrating power (for example, X-ray) through the battery under test.
The schematic diagram of the three-dimensional battery model can be understood as a three-dimensional scanning image obtained by transmission scanning of the battery by a scanning device. For example, when the internal detection of the battery is performed, a user can scan and image the battery from top to bottom through an X-ray device, and the output of the X-ray device is a three-dimensional scan of the battery. If there are particles inside the cell, a corresponding scan of the particles can be visualized in the three-dimensional scan. The X-ray device and the manner of scanning to obtain a three-dimensional scan of the cell are well known to those skilled in the art and will not be described herein.
In this embodiment, the electronic device 10 may cut the three-dimensional battery model to obtain a corresponding cross-sectional view. For example, the cross-sectional views of the three-dimensional battery model shown in fig. 2 taken along thebase:Sub>A-base:Sub>A plane and the B-B plane are shown in fig. 3 and 4. It should be noted that the cross-sectional views shown in fig. 3 and 4 are schematic views for easy understanding, and in practical applications, the cross-sectional views may be determined according to practical situations.
Referring to fig. 5, an embodiment of the present application further provides a battery detection method, which can be applied to the electronic device 10, and each step of the method is executed or implemented by the electronic device 10. The method may comprise the steps of:
step S210, sectioning and drawing the first three-dimensional battery model at a specified interval from a first three-dimensional battery model obtained by transmission scanning of a battery to be tested to obtain a plurality of continuous sectional views, wherein the first three-dimensional battery model comprises a particle model obtained by transmission scanning of particles in the battery to be tested;
step S220, inputting the plurality of cross-sectional views into a trained battery detection model, and obtaining a detection result output by the battery detection model when the battery detection model detects each cross-sectional view in the plurality of cross-sectional views, where the detection result includes the number of particles existing in each cross-sectional view, the contour size of the cross-section of each particle, and position information of the cross-section of each particle.
In the above embodiment, the first three-dimensional battery model is cut, and then the battery detection model detects the obtained plurality of cross-sectional views, so that the detection result of the battery to be detected is obtained, manual detection is not needed, the detection efficiency is improved, and the problems that the efficiency of battery detection is low and errors are easy to occur due to manual detection are solved.
The individual steps of the process are explained in detail below, as follows:
in step S210, the first three-dimensional battery model is a three-dimensional scan image obtained by transmission scanning of the battery to be tested by the scanning device. Understandably, if the particles exist in the battery to be tested, the particle model obtained by scanning the particles exists in the first three-dimensional battery model. The particle model is a three-dimensional scan of the particle.
In the present embodiment, when the particle size of the particles is greater than or equal to the first prescribed size, the size of the particles is considered abnormal, and it can be determined that there is an abnormality in the battery under test. The first designated size may be set according to actual conditions, and refers to a critical value of whether the size of the particles is abnormal or not. For example, the first specified dimension may be a dimension of 3 millimeters, 5 millimeters, or the like. The particle size of the particles is understood to mean the maximum diameter of the particles.
The designated spacing may be set according to actual conditions. To avoid leaving behind detection of particles of abnormal size due to excessive spacing, the specified spacing may be less than or equal to
Multiple of the first specified size. Through prescribing the interval and prescribing a limit to, be favorable to avoiding omitting to the detection of great granule to improve the reliability of detecting.
In step S220, the battery detection model is a detection model obtained by training a large number of cross-sectional views based on a deep learning model. And the deep learning model is the battery detection model after the training test is completed. Thereafter, the battery test model can directly perform identification test on one or more cross-sectional views, including detecting whether particles exist in the cross-sectional views and counting the outline size of the particles and the number of the particles when the particles exist.
Understandably, the battery detection model needs to perform a training test on the deep learning model before performing the first detection. The deep learning model can be selected according to actual conditions, including but not limited to AlexNet, vgg16, inclusion, and other models, and is well known to those skilled in the art.
As an optional implementation manner, before step S210, the method may further include:
obtaining a training data set, wherein the training data set comprises a plurality of groups of cross-sectional views and a label corresponding to each cross-sectional view of the plurality of groups of cross-sectional views, and each group of cross-sectional views comprises a plurality of cross-sectional views of any one of a plurality of second three-dimensional battery models;
and training a deep learning model through the training data set to obtain the trained battery detection model.
Understandably, the training data set is image data prepared in advance for the user. The number of groups of cross-sectional views included in the training data set may be set according to actual conditions, and may exceed thousands of groups, ten thousands of groups, and the like. In addition, the number of images in each sectional view may be determined according to actual conditions, and is not particularly limited herein. Wherein one set of cross-sectional views corresponds to a three-dimensional battery model of one battery and the different sets of cross-sectional views correspond to three-dimensional battery models of different batteries.
The second three-dimensional battery model is the three-dimensional battery model of the battery corresponding to the model training. For example, before model training, a plurality of batteries need to be transmitted and scanned, so as to obtain a plurality of three-dimensional battery models, and the obtained plurality of three-dimensional battery models are the second three-dimensional battery model. The second three-dimensional battery model may be the same as the first three-dimensional battery model, and may be different. Generally, the second three-dimensional battery model and the first three-dimensional battery model are three-dimensional battery models obtained by scanning different batteries.
In the training data set, the single cross-sectional view in each group of cross-sectional views is provided with a corresponding label, the label includes a label of whether the single cross-sectional view has particles, a label of the number of the particles, and the like, and the content of the label can be set according to actual conditions.
The user can input the training data set into the deep learning model, so that the deep learning model can perform feature extraction on the image, and convolution operation is performed, so that the image features of the image with particles and the image without particles are learned; and then testing the trained deep learning model by using the test data set so as to improve the accuracy and reliability of the detection of the deep learning model. The test data set may be obtained from actual conditions and may be a partial cutaway view of the training data set. After the test of the deep learning model is completed, the deep learning model is a battery detection model, and the section can be detected. The process of testing the deep learning model is well known to those skilled in the art, and is not described herein.
As an optional implementation, before training the deep learning model by the training data set, the method may further include:
when the number of particles with contour sizes smaller than a second designated size in the plurality of groups of cross-sectional views is smaller than a first designated number, adding a cross-sectional view area with contour sizes smaller than the second designated size in a part or all of the plurality of groups of cross-sectional views to obtain a data-enhanced training data set, wherein the second designated size is smaller than the first designated size.
During the period of acquiring the training data set, if the number of the particles with smaller sizes is determined to be less, data enhancement can be performed on the particles with smaller sizes in the training data set, so that the particles with smaller sizes can be detected more easily, and missing detection is avoided. Understandably, a particle is a particle of lesser size if its overall size is smaller than the second designated size. If the number of smaller sized particles in the training data set is less than the first specified number, then the number of smaller sized particles is determined to be less. In the cross-sectional view, the contour dimension of the particle is the maximum diameter of the cross-sectional area of the particle. The first designated number, the first designated size and the second designated size can be set according to actual conditions. Illustratively, the first designated dimension may be 5 millimeters and the second designated dimension may be 1.5 millimeters.
The data enhancement mode can be as follows: copying a plurality of cross-sectional views with smaller particles by means of oversampling (so as to increase the number of cross-sectional areas of the smaller particles); or, training the cross-sectional view of the smaller particles for multiple times by using a deep learning model; alternatively, the cross-sectional regions of each type of smaller grain are extracted, then a plurality of copies are made, and the copied plurality of cross-sectional regions are added to regions of the training data set that are not grains, in some or all of the cross-sectional views, to increase the number of cross-sectional regions of the smaller grain.
Understandably, if the outline size of the cross-sectional region of the particle is small in the cross-sectional view, the area of the target (i.e., the cross-sectional region of the particle) in the cross-sectional view is small during model training, which results in fewer anchors (anchors) containing the target, and a smaller probability of detecting the small target. In this embodiment, the number of smaller sized particles in the training data set may be greater than or equal to the first specified number after the training data set has been enhanced by the above-described data. Based on the method, the number of regions for the particles with smaller sizes is increased in the sectional view, and the probability of being contained by the Anchor is improved, so that the problems that the deep learning model is insensitive and easy to miss detection due to the fact that the number of training samples of the particles with smaller sizes is smaller are solved.
As an optional implementation, the method may further include:
judging whether target particles with the contour size larger than or equal to a first specified size exist or not according to the contour size of the cross section of each particle;
and when the target particles exist, sending prompt information for representing that the battery to be tested is abnormal.
Understandably, the target particle is a particle having a contour size greater than or equal to a first prescribed size in a cross-sectional view, and the target particle is a particle having an abnormality in the above-mentioned size.
In this embodiment, the electronic device 10 may store a proportional relationship between the pixel distance in the cross-sectional view and the distance in the real object in advance. The proportional relation is a ratio of a distance between any two pixel points in the cross-sectional view to a distance between two corresponding points of the two pixel points in a real scene, and can be determined according to actual conditions.
In calculating the profile size of the cross-section of the particle in the cross-sectional view in the actual scene, the electronic device 10 may determine the pixel distance of the profile size of the particle in the cross-sectional view, and then, in combination with the proportional relationship, may determine the actual profile size of the cross-section of the particle in the cross-sectional view. The pixel distance for determining the contour size is well known to those skilled in the art and will not be described herein.
When the target particles exist, the prompt information is sent, so that managers can find abnormal batteries in time.
Since the same particle having a large particle size may appear in a plurality of adjacent sectional views, it is necessary to perform calculation in combination with the plurality of adjacent sectional views when performing the particle number statistics and calculating the particle size of the actual particle. As an optional implementation, the method may further include:
when the target particle exists in the detection results of the first cross section and the second cross section, determining the distance between the center point of the cross section of the target particle in the first cross section and the center point of the cross section of the target particle in the second cross section, wherein the first cross section and the second cross section are any group of adjacent cross sections;
when the distance between the central points is smaller than the section radius of the target particle in the first section view or the second section view, determining that the target particle in the first section view and the target particle in the second section view are the same target particle;
and determining the number of the particles in the battery to be tested according to the number of the repeated particles obtained by the same target particles and the number of the particles existing in each cross section.
The electronic device 10 may establish a spatial rectangular coordinate system 0-zyz for the three-dimensional battery model shown in fig. 2, and when obtaining the cross-sectional view, the cross-sectional view may be parallel to the plane where the x-axis and the y-axis are located, and the cross-sectional view is taken at the above specified interval until all the cross-sections of the three-dimensional battery model in the z-axis direction are completed. wherein,base:Sub>A partial section plane is shown asbase:Sub>A dotted line at the top of the three-dimensional battery model shown in fig. 2, and schematic diagrams of the section planes shown in fig. 3 and 4 can be obtained based on thebase:Sub>A-base:Sub>A section plane and the B-B section plane, respectively. In a three-dimensional battery model, each layer of sectional view can be provided with corresponding numbers, and the numbers of different sectional views are different so as to be convenient for distinguishing.
Illustratively, fig. 3 may be a first cross-sectional view, and fig. 4 may be a second cross-sectional view, with the two cross-sectional views positioned adjacent to one another. The electronic device 10 may determine the particle P1 in fig. 3 and the particle P2 in fig. 4 by a battery detection model. It is assumed that the determined contour sizes of the particles P1 and P2 are both larger than the first prescribed size, and at this time, the particles P1 and P2 are both target particles. The electronic device 10 may automatically determine the coordinates of the center point of the target particle in the coordinate system 0-xyz and then determine the center point distance between the center points of the particles P1, P2 based on the coordinates of the center points of the particles P1, P2. If the distance between the center points is smaller than the cross-sectional radius of any one of the particles P1 and P2 in the corresponding cross-sectional view, it means that the particles P1 and P2 are the same particle. If the distances between the center points are both greater than or equal to the cross-sectional radii of the particles P1 and P2 in the corresponding cross-sectional views, it means that the particles P1 and P2 are not the same particles. In this way, the electronic device 10 can perform preliminary judgment on the same particle, thereby being beneficial to improving the accuracy of detection. The cross-sectional radius is half of the size of the cross section of the particle P1 in the first cross-sectional view (or the particle P2 in the second cross-sectional view). In addition, the manner of determining the coordinates of the center point of the particle is well known to those skilled in the art and will not be described herein.
Of course, if the same larger particle appears in the cross-sectional view exceeding 2 layers, the electronic device 10 calculates the same manner as the above-mentioned manner of determining whether the particles are the same in the first cross-sectional view and the second cross-sectional view when detecting whether the target particles in the plurality of consecutive cross-sectional views are the same particle, and the description thereof is omitted here. If the same particle appears in N continuous sectional views, the repetition time of the same particle is N-1, and N is an integer greater than or equal to 2.
When calculating the number of particles of a single battery under test, the electronic device 10 may count the number of particles in each cross-sectional view based on the detection results of all cross-sectional views of the battery under test to obtain the number of particles in all cross-sectional views, and then subtract the repetition number of each same particle to obtain the final number of particles. The final particle number is the actual particle number of the battery to be measured or the particle number close to the actual particle number, so that the accuracy of particle counting is improved.
In calculating the particle size of the target particle, if the same particle appears in the N consecutive cross-sectional views, in this case, the electronic device 10 may use, as the particle size of the target particle, the diameter of a sphere that internally cuts the contour of the target particle in the cross-sectional views at both ends in the coordinate system 0-xyz based on the contour of the target particle in the cross-sectional views at both ends of the N consecutive cross-sectional views; or comparing the maximum outline size of the target particles in N continuous sectional views with N times of specified intervals, and selecting the maximum value as the final particle size of the target particles, so that the accuracy of particle size calculation can be improved.
As an alternative embodiment, before inputting the plurality of cross-sectional views into the trained battery test model, the method may further include:
and carrying out noise reduction treatment on the edge noise on the plurality of sectional views to obtain a plurality of sectional views after noise reduction, and inputting the plurality of sectional views into the trained battery detection model.
Understandably, the electronic device 10 may perform noise reduction processing on the cross-sectional view through a preset noise reduction filtering algorithm. The noise reduction process is used to remove edge noise of the particle profile in the cross-sectional view to make the particle profile clearer so as to facilitate the detection and identification of the particle profile by the electronic device 10. The filtering uses the average gray level of the neighborhood of the pixel instead of the pixel value, and is suitable for impulse noise because the gray level of impulse noise is generally not related to the gray level of the surrounding pixels and the brightness is much higher than that of other pixels (the higher value can be determined according to practical situations and is not limited in particular here). For example, a plane rectangular coordinate system is established for the cross-sectional view, so that the positions of the pixel points in the cross-sectional view can be represented by coordinates (i, j), and the noise reduction filtering algorithm can be the following formula:
in the above formula, a' (i, j) is a result obtained by filtering a pixel of the coordinate (i, j) (the result may include a gray value of the pixel after filtering); l refers to the radius of filtering; i. j respectively represents the horizontal and vertical coordinates of the starting point; k. l respectively indicate the positions of the horizontal and vertical coordinate shifts from the start points (i, j).
In another embodiment, the deep learning model may include labels of other defects in the cross-sectional view of the training image set during training, for example, the other defects may be cracks (Crack) in the battery, and thus, the battery inspection model may also be used to inspect other defects in the battery, such as cracks in the battery, so as to obtain inspection results such as the length, width, and number of cracks. The detection method of the crack is similar to the detection method of the particle, and is not described herein again.
Referring to fig. 6, an embodiment of the present application further provides a battery detection apparatus 100, which can be applied to the electronic device 10 described above for executing steps of the method. The battery detection apparatus 100 includes at least one software functional module which can be stored in the memory module 12 in the form of software or Firmware (Firmware) or solidified in an Operating System (OS) of the electronic device 10. The processing module 11 is used for executing executable modules stored in the storage module 12, such as software functional modules and computer programs included in the battery detection apparatus 100.
The battery test apparatus 100 may include a section unit 110 and a test unit 120, and may perform the following operations:
the section unit 110 is configured to perform section drawing on a first three-dimensional battery model obtained by transmission scanning of a battery to be tested at a specified interval to obtain a plurality of continuous section views, where the first three-dimensional battery model includes a particle model obtained by transmission scanning of particles in the battery to be tested;
the detecting unit 120 is configured to input the plurality of cross-sectional views into a trained battery detection model, and obtain a detection result output by the battery detection model when the battery detection model detects each of the plurality of cross-sectional views, where the detection result includes the number of particles present in each of the cross-sectional views, an outline size of a cross-section of each particle, and position information of the cross-section of each particle.
As an optional implementation manner, the battery detection apparatus 100 may further include a determination unit and a prompt unit. The judging unit is used for judging whether target particles with contour sizes larger than or equal to a first specified size exist according to the contour sizes of the cross sections of the particles; and the prompting unit is used for sending out prompting information for representing the abnormity of the battery to be tested when the target particles exist.
As an optional implementation, the battery detection apparatus 100 may further include a distance calculation unit, a particle determination unit, and a statistic unit, wherein:
a distance calculation unit configured to determine a distance between a center point of a cross section of a target particle in a first cross section and a center point of a cross section of a target particle in a second cross section when the target particle is present in both detection results of the first cross section and the second cross section, the first cross section and the second cross section being any set of adjacent cross sections;
a particle determining unit, configured to determine that the target particle in the first cross-sectional view and the target particle in the second cross-sectional view are the same target particle when the distance between the central points is smaller than the cross-sectional radius of the target particle in the first cross-sectional view or the second cross-sectional view;
and the counting unit is used for determining the number of the particles in the battery to be tested according to the number of the repeated particles obtained by the same target particles and the number of the particles existing in each cross section.
As an alternative embodiment, the battery test apparatus 100 may further include an acquisition unit and a training unit. Before the section view unit 110 performs step S210, an acquisition unit configured to acquire a training data set including a plurality of sets of section views each including a plurality of section views of any one of a plurality of second three-dimensional battery models and a label corresponding to each of the plurality of sets of section views; and the training unit is used for training the deep learning model through the training data set to obtain the trained battery detection model.
As an alternative embodiment, the battery test apparatus 100 may further include a data enhancement unit. Before the training unit trains the deep learning model through the training data set, the data enhancement unit is used for adding section map areas of a second specified number of particles with contour sizes smaller than a second specified size to a part or all of the plurality of groups of section views to obtain the data enhanced training data set when the number of particles with contour sizes smaller than the second specified size is smaller than a first specified number in the plurality of groups of section views.
As an alternative embodiment, the battery detection apparatus 100 may further include an image noise reduction unit. Before the detection unit 120 inputs the plurality of cross-sectional views into the trained battery detection model, the image denoising unit is configured to perform denoising processing on the edge noise of the plurality of cross-sectional views to obtain a plurality of denoised cross-sectional views, and is configured to input the trained battery detection model.
In this embodiment, the processing module 11 may be an integrated circuit chip having signal processing capability. The processing module 11 may be a general-purpose processor. For example, the Processor may be a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), a Network Processor (NP), or the like; the method, the steps and the logic block diagram disclosed in the embodiments of the present Application may also be implemented or executed by a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other Programmable logic device, a discrete Gate or transistor logic device, or a discrete hardware component.
The memory module 12 may be, but is not limited to, a random access memory, a read only memory, a programmable read only memory, an erasable programmable read only memory, an electrically erasable programmable read only memory, and the like. In this embodiment, the storage module 12 may be used to store a three-dimensional battery model, a cross-sectional view, a first specified size, a second specified size, a specified pitch, and the like. Of course, the storage module 12 may also be used to store a program, and the processing module 11 executes the program after receiving the execution instruction.
It is understood that the configuration shown in fig. 1 is only a schematic configuration of the electronic device 10, and that the electronic device 10 may further include more components than those shown in fig. 1. The components shown in fig. 1 may be implemented in hardware, software, or a combination thereof.
It should be noted that, as will be clear to those skilled in the art, for convenience and brevity of description, the specific working process of the electronic device 10 described above may refer to the corresponding process of each step in the foregoing method, and will not be described in too much detail herein.
The embodiment of the application also provides a computer readable storage medium. The computer-readable storage medium has stored therein a computer program which, when run on a computer, causes the computer to execute the battery detection method as described in the above embodiments.
From the foregoing description of the embodiments, it is clear to those skilled in the art that the present application may be implemented by hardware or by software plus a necessary general hardware platform, and based on such understanding, the technical solution of the present application may be embodied in the form of a software product, which may be stored in a non-volatile storage medium (which may be a CD-ROM, a usb disk, a removable hard disk, or the like), and includes several instructions to enable a computer device (which may be a personal computer, a server, or a network device, or the like) to execute the method described in the various implementation scenarios of the present application.
In summary, the present application provides a battery detection method, a battery detection device, an electronic device, and a readable storage medium. The method comprises the following steps: sectioning the first three-dimensional battery model at a specified interval from a first three-dimensional battery model obtained by transmission scanning of the battery to be tested to obtain a plurality of continuous sectional views, wherein the first three-dimensional battery model comprises a particle model obtained by transmission scanning of particles in the battery to be tested; and inputting the plurality of cross sections into the trained battery detection model to obtain a detection result output by the battery detection model for detecting each cross section in the plurality of cross sections, wherein the detection result comprises the number of particles existing in each cross section, the contour size of the cross section of each particle and the position information of the cross section of each particle. In this scheme, through sectioning first three-dimensional battery model, then detect a plurality of cross-sectional views that obtain by the battery detection model to obtain the testing result to the battery that awaits measuring, so, need not artifical the detection, be favorable to improving detection efficiency, improve because of the artifical inefficiency that leads to the battery to detect and the problem of makeing mistakes easily that detects.
In the embodiments provided in the present application, it should be understood that the disclosed apparatus, system, and method may be implemented in other ways. The apparatus, system, and method embodiments described above are illustrative only, as the flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions. In addition, functional modules in the embodiments of the present application may be integrated together to form an independent part, or each module may exist alone, or two or more modules may be integrated to form an independent part.
The above description is only an example of the present application and is not intended to limit the scope of the present application, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present application shall be included in the protection scope of the present application.
Claims (8)
1. A battery testing method, the method comprising:
sectioning and drawing the first three-dimensional battery model at a specified interval from a first three-dimensional battery model obtained by transmission scanning of a battery to be tested to obtain a plurality of continuous sectional views, wherein the first three-dimensional battery model comprises a particle model obtained by transmission scanning of particles in the battery to be tested;
inputting the plurality of cross sections into a trained battery detection model to obtain a detection result output by the battery detection model when the battery detection model detects each cross section in the plurality of cross sections, wherein the detection result comprises the number of particles existing in each cross section, the contour size of the cross section of each particle and the position information of the cross section of each particle;
judging whether target particles with the contour size larger than or equal to a first specified size exist according to the contour size of the cross section of each particle;
when the target particles exist, sending prompt information for representing that the battery to be tested is abnormal;
determining a distance between a center point of a cross section of the target particle in the first cross section and a center point of a cross section of the target particle in the second cross section when the target particle exists in the detection results of the first cross section and the second cross section, wherein the first cross section and the second cross section are any group of adjacent cross sections;
when the distance between the central points is smaller than the section radius of the target particle in the first section view or the second section view, determining that the target particle in the first section view and the target particle in the second section view are the same target particle;
and determining the number of the particles in the battery to be tested according to the number of the repeated particles obtained by the same target particles and the number of the particles existing in each cross section.
2. The method of claim 1, wherein the specified pitch is less than or equal to
Multiple of the first specified size.
3. The method of claim 1, wherein before the first three-dimensional cell model obtained from the transmission scanning of the cell under test is sectioned at a specified interval to obtain a plurality of consecutive sectional views, the method further comprises:
obtaining a training data set, wherein the training data set comprises a plurality of groups of cross-sectional views and a label corresponding to each cross-sectional view of the plurality of groups of cross-sectional views, and each group of cross-sectional views comprises a plurality of cross-sectional views of any one of a plurality of second three-dimensional battery models;
and training a deep learning model through the training data set to obtain the trained battery detection model.
4. The method of claim 3, wherein prior to training a deep learning model with the training data set, the method further comprises:
when the number of particles with contour sizes smaller than a second designated size in the plurality of groups of cross-sectional views is smaller than a first designated number, adding a cross-sectional view area with contour sizes smaller than the second designated size in a part or all of the plurality of groups of cross-sectional views to obtain a data-enhanced training data set, wherein the second designated size is smaller than the first designated size.
5. The method of claim 1, wherein prior to inputting the plurality of cross-sectional views into the trained battery inspection model, the method further comprises:
and carrying out noise reduction treatment on the edge noise on the plurality of sectional views to obtain a plurality of sectional views after noise reduction, and inputting the plurality of sectional views into the trained battery detection model.
6. A battery testing apparatus, the apparatus comprising:
the device comprises a section unit, a data acquisition unit and a display unit, wherein the section unit is used for sectioning and drawing a first three-dimensional battery model at a specified interval from the first three-dimensional battery model obtained by transmission scanning of a battery to be tested to obtain a plurality of continuous section views, and the first three-dimensional battery model comprises a particle model obtained by transmission scanning of particles in the battery to be tested;
the detection unit is used for inputting the plurality of cross sections into a trained battery detection model to obtain a detection result output by the battery detection model for detecting each cross section in the plurality of cross sections, wherein the detection result comprises the number of particles existing in each cross section, the profile size of the cross section of each particle and the position information of the cross section of each particle;
a judging unit for judging whether there is a target particle having a contour size larger than or equal to a first prescribed size, based on the contour size of the cross section of each particle;
the prompting unit is used for sending out prompting information for representing the abnormity of the battery to be tested when the target particles exist;
a distance calculation unit configured to determine a distance between a center point of a cross section of a target particle in a first cross section and a center point of a cross section of a target particle in a second cross section when the target particle is present in both detection results of the first cross section and the second cross section, the first cross section and the second cross section being any set of adjacent cross sections;
a particle determination unit, configured to determine that the target particle in the first cross-sectional view and the target particle in the second cross-sectional view are the same target particle when the center point distance is smaller than the cross-sectional radius of the target particle in the first cross-sectional view or the second cross-sectional view;
and the counting unit is used for determining the number of the particles in the battery to be tested according to the number of the repeated particles obtained by the same target particles and the number of the particles existing in each cross section.
7. An electronic device, characterized in that the electronic device comprises a processor and a memory coupled to each other, the memory storing a computer program which, when executed by the processor, causes the electronic device to perform the method according to any of claims 1-5.
8. A computer-readable storage medium, in which a computer program is stored which, when run on a computer, causes the computer to carry out the method according to any one of claims 1-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110228134.9A CN112967245B (en) | 2021-03-01 | 2021-03-01 | Battery detection method and device, electronic equipment and readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110228134.9A CN112967245B (en) | 2021-03-01 | 2021-03-01 | Battery detection method and device, electronic equipment and readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112967245A CN112967245A (en) | 2021-06-15 |
| CN112967245B true CN112967245B (en) | 2022-11-08 |
Family
ID=76276204
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110228134.9A Active CN112967245B (en) | 2021-03-01 | 2021-03-01 | Battery detection method and device, electronic equipment and readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112967245B (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102680903A (en) * | 2012-05-11 | 2012-09-19 | 山东轻工业学院 | Portable storage battery state detection system and method |
| CN105550452A (en) * | 2015-12-18 | 2016-05-04 | 中国科学院上海高等研究院 | Lithium ion battery P2D model parameter recognizing method based on heuristic algorithm |
| CN109239075A (en) * | 2018-08-27 | 2019-01-18 | 北京百度网讯科技有限公司 | Battery detection method and device |
| CN109949305A (en) * | 2019-03-29 | 2019-06-28 | 北京百度网讯科技有限公司 | Method for detecting surface defects of products, device and computer equipment |
| CN110180797A (en) * | 2019-05-10 | 2019-08-30 | 中国汽车技术研究中心有限公司 | Lithium battery alignment degree classification evaluation method for computed tomography detection |
| CN110824371A (en) * | 2019-11-08 | 2020-02-21 | 广东利元亨智能装备股份有限公司 | Battery detection method and device, electronic equipment and readable storage medium |
| CN111476307A (en) * | 2020-04-13 | 2020-07-31 | 河北工业大学 | Lithium battery surface defect detection method based on depth field adaptation |
| CN111524220A (en) * | 2020-03-18 | 2020-08-11 | 上海刻羽信息科技有限公司 | Method, system, electronic device, and medium for sectioning three-dimensional model |
| CN111598861A (en) * | 2020-05-13 | 2020-08-28 | 河北工业大学 | Improved Faster R-CNN model-based non-uniform texture small defect detection method |
| CN111768387A (en) * | 2020-06-30 | 2020-10-13 | 创新奇智(青岛)科技有限公司 | Flaw detection method, twin neural network training method and device and electronic equipment |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103643929A (en) * | 2013-12-03 | 2014-03-19 | 中国石油大学(北京) | Asphalt particle profile control effect predicting device and method for three-dimensional cementation model |
| CN104035048A (en) * | 2014-06-20 | 2014-09-10 | 上海出入境检验检疫局工业品与原材料检测技术中心 | Pyroelectric detection method and device for over-charged safety performance of lithium ion battery |
| CN105571571B (en) * | 2016-01-26 | 2017-11-17 | 中国科学院水利部成都山地灾害与环境研究所 | Accumulation section spatial structural form analysis method based on 3 D laser scanning |
| CN108595840A (en) * | 2018-04-25 | 2018-09-28 | 中国科学技术大学 | The method for establishing model of lithium ion battery, the optimization method of system and polar ear size |
| CN108646195A (en) * | 2018-07-03 | 2018-10-12 | 东莞百瑞自动化有限公司 | A kind of battery simulation simulation test machine and its control method |
| CN109447977B (en) * | 2018-11-02 | 2021-05-28 | 河北工业大学 | Visual defect detection method based on multispectral deep convolutional neural network |
-
2021
- 2021-03-01 CN CN202110228134.9A patent/CN112967245B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102680903A (en) * | 2012-05-11 | 2012-09-19 | 山东轻工业学院 | Portable storage battery state detection system and method |
| CN105550452A (en) * | 2015-12-18 | 2016-05-04 | 中国科学院上海高等研究院 | Lithium ion battery P2D model parameter recognizing method based on heuristic algorithm |
| CN109239075A (en) * | 2018-08-27 | 2019-01-18 | 北京百度网讯科技有限公司 | Battery detection method and device |
| CN109949305A (en) * | 2019-03-29 | 2019-06-28 | 北京百度网讯科技有限公司 | Method for detecting surface defects of products, device and computer equipment |
| CN110180797A (en) * | 2019-05-10 | 2019-08-30 | 中国汽车技术研究中心有限公司 | Lithium battery alignment degree classification evaluation method for computed tomography detection |
| CN110824371A (en) * | 2019-11-08 | 2020-02-21 | 广东利元亨智能装备股份有限公司 | Battery detection method and device, electronic equipment and readable storage medium |
| CN111524220A (en) * | 2020-03-18 | 2020-08-11 | 上海刻羽信息科技有限公司 | Method, system, electronic device, and medium for sectioning three-dimensional model |
| CN111476307A (en) * | 2020-04-13 | 2020-07-31 | 河北工业大学 | Lithium battery surface defect detection method based on depth field adaptation |
| CN111598861A (en) * | 2020-05-13 | 2020-08-28 | 河北工业大学 | Improved Faster R-CNN model-based non-uniform texture small defect detection method |
| CN111768387A (en) * | 2020-06-30 | 2020-10-13 | 创新奇智(青岛)科技有限公司 | Flaw detection method, twin neural network training method and device and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112967245A (en) | 2021-06-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20220036533A1 (en) | Image defect detection method and apparatus, electronic device, storage medium and product | |
| CN111340787B (en) | Method and device for detecting and identifying wire defects of power transmission line and computer equipment | |
| CN109741324B (en) | Detection method, detection device and terminal equipment | |
| CN111667448B (en) | Image processing method, device and equipment | |
| US20230298327A1 (en) | Information processing device, determination method, and information processing program | |
| CN108955901B (en) | Infrared temperature measurement method and system and terminal equipment | |
| CN109671082B (en) | Sealing ring detection method, device, equipment and computer readable storage medium | |
| CN114821274A (en) | Method and device for identifying state of split and combined indicator | |
| CN109102486B (en) | Surface defect detection method and device based on machine learning | |
| CN112967245B (en) | Battery detection method and device, electronic equipment and readable storage medium | |
| JP2002230562A (en) | Image processing method and device therefor | |
| CN115597494B (en) | Precision detection method and system for prefabricated part preformed hole based on point cloud | |
| CN107767372B (en) | Chip pin online visual detection system and method for layered parallel computing | |
| CN115272381A (en) | Metal wire segmentation method and device, electronic equipment and storage medium | |
| US8306311B2 (en) | Method and system for automated ball-grid array void quantification | |
| CN110634124A (en) | Method and equipment for area detection | |
| CN115187563A (en) | Perforation seal detection and identification method and system for electric power instrument inspection | |
| CN111504608B (en) | Brightness uniformity detection system and brightness uniformity detection method | |
| CN114998889A (en) | Intelligent identification method and system for immersive three-dimensional image | |
| CN112435224B (en) | Confidence evaluation method and device for stop line extraction | |
| CN112730427B (en) | Product surface defect detection method and system based on machine vision | |
| CN111524268B (en) | Method, device and equipment for detecting paper money adhesive substance | |
| JPH01112382A (en) | Cracked picture data processing method | |
| CN114387239A (en) | Compensation capacitor defect detection method and device | |
| Chen et al. | Circular object detection with mathematical morphology and geometric properties |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |