CN110222679B - General battery polarity automatic detection method based on deep learning - Google Patents
General battery polarity automatic detection method based on deep learning Download PDFInfo
- Publication number
- CN110222679B CN110222679B CN201910386913.4A CN201910386913A CN110222679B CN 110222679 B CN110222679 B CN 110222679B CN 201910386913 A CN201910386913 A CN 201910386913A CN 110222679 B CN110222679 B CN 110222679B
- Authority
- CN
- China
- Prior art keywords
- battery
- layer
- polarity
- detected
- negative electrode
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/217—Validation; Performance evaluation; Active pattern learning techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- 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
- G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
- Y02E60/10—Energy storage using batteries
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of battery polarity detection, and particularly discloses a general battery polarity automatic detection method based on deep learning, which comprises the steps of obtaining a picture of a position to be detected; determining a battery polarity region to be detected; the battery polarity region to be detected is identified and classified by using a deep learning algorithm model to obtain a battery polarity classification result, and the battery polarity detection method is used for automatically detecting and classifying the battery polarity based on deep learning, and only all sample models are needed to be provided, and training samples can automatically learn classified characteristic expression, so that complex battery polarity characteristics can be well processed, battery polarity can be accurately identified, instability of a traditional detection method is effectively solved, and production efficiency and product quality are improved.
Description
Technical Field
The invention relates to the technical field of battery polarity detection, in particular to a general battery polarity automatic detection method based on deep learning.
Background
In the production of a battery pack for a power battery or other products, it is necessary to weld the positive and negative electrodes of a plurality of batteries in a certain order to form the battery pack. In order to prevent misplacement and misplacement, short circuit of products and resource waste, the positive and negative polarities of each battery area need to be accurately distinguished.
The positive electrode of the battery is stuck with highland barley paper, the middle of the positive electrode is provided with a silver round area with grooves, the negative electrode is provided with no highland barley paper, the middle of the battery is provided with a silver round area without grooves, the non-positive electrode comprises the positive electrode which is not stuck with highland barley paper, the negative electrode which is stuck with one layer of highland barley paper and the negative electrode which is stuck with two layers of highland barley paper, the non-negative electrode comprises the positive electrode which is not stuck with highland barley paper, the positive electrode which is stuck with two layers of highland barley paper, the negative electrode which is stuck with one layer of highland barley paper and the negative electrode which are stuck with two layers of highland barley paper, and although each polarity has certain distinguishing degree on the color of the appearance, the characteristics are extremely similar.
For battery polarity detection, most of the existing manual screening methods or traditional morphological analysis methods are adopted in the industry, but due to the diversity of battery polarity characteristics, the traditional two detection methods have great disadvantages, are difficult to stably distinguish, are extremely easy to detect and fail, and can be used for identifying battery polarity incorrectly.
Disclosure of Invention
Aiming at the technical problems, the invention provides a general battery polarity automatic detection method based on deep learning, which can well process the characteristics of complex battery polarities, accurately identify the battery polarities, effectively solve the instability of the traditional detection method and improve the production efficiency and the product quality.
In order to solve the technical problems, the invention provides the following specific scheme:
a general battery polarity automatic detection method based on deep learning, the method comprising:
acquiring a picture of a position to be detected;
determining a battery polarity region to be detected;
and identifying and classifying the battery polarity region to be detected by using a deep learning algorithm model to obtain a battery polarity classification result.
According to the invention, the battery polarity is automatically detected and classified based on deep learning, and the model can automatically learn the classified characteristic expression only by providing all sample models and training samples, so that the complex battery polarity characteristics can be well processed, the battery polarity can be accurately identified, the instability of the traditional detection method is effectively solved, and the production efficiency and the product quality are improved.
Optionally, the deep learning algorithm model includes positive, non-positive, negative and non-negative four-classification models.
Optionally, the feature input of the positive electrode model comprises a positive electrode;
the characteristic inputs of the non-positive electrode model comprise positive electrode non-attached highland barley paper, negative electrode attached highland barley paper of one layer and negative electrode attached highland barley paper of two layers;
the characteristic input of the negative electrode model comprises a negative electrode;
the characteristic inputs of the non-negative electrode model comprise an anode, anode non-attached highland barley paper, anode attached two layers of highland barley paper, cathode attached one layer of highland barley paper and cathode attached two layers of highland barley paper.
Optionally, the battery polarity classification result includes four types of polarities of positive electrode, non-positive electrode, negative electrode and non-negative electrode.
Optionally, the method further comprises:
constructing a multi-layer convolutional neural network;
training the multi-layer convolutional neural network by adopting an error reverse conduction algorithm to obtain a battery polarity recognition model;
modifying the output layer node of the multi-layer convolutional network to be 4, and initializing the modified weight of the multi-layer convolutional neural network by using the weight of the trained battery polarity identification model;
and training the modified multi-layer convolutional neural network by using the positive electrode, non-positive electrode, negative electrode and non-negative electrode data sets to obtain a positive electrode, non-positive electrode, negative electrode and non-negative electrode four-classification model.
Optionally, the multi-layer convolutional neural network comprises an input layer, a hidden layer and an output layer;
the input layer is used for data input of the whole multi-layer convolutional neural network;
the hidden layer comprises four training models, namely an anode, a non-anode, a cathode and a non-cathode;
the output layer is used for outputting battery polarity classification results.
Optionally, the obtaining the picture of the position to be detected specifically includes:
manually placing the battery to be detected on the positioning module to fix the position;
and the industrial camera automatically photographs the battery to be detected on the positioning module to obtain a picture of the position to be detected.
Optionally, the automatic battery that waits to detect on the locating module of industry camera is photographed, specifically includes:
and the industrial camera determines the photographing times according to the number of the batteries to be detected on the positioning module and the focal length of the industrial camera, and if photographing for multiple times, the pictures obtained after photographing for multiple times are spliced into one picture to be detected.
Optionally, the determining the battery polarity area to be detected specifically includes:
and determining the detection position of each battery to be detected and the correct polarity to be placed at the position by adopting a region segmentation method, and finally determining the polarity region of the battery to be detected.
Compared with the prior art, the invention has the beneficial effects that: according to the invention, the battery polarity is automatically detected and classified based on deep learning, and the model can automatically learn the classified characteristic expression only by providing all sample models and training samples, so that the complex battery polarity characteristics can be well processed, the battery polarity can be accurately identified, the instability of the traditional detection method is effectively solved, and the production efficiency and the product quality are improved.
Drawings
Fig. 1 is a flowchart of a general battery polarity automatic detection method based on deep learning in an embodiment of the invention.
Fig. 2 is a flowchart of obtaining a deep learning algorithm model in an embodiment of the present invention.
Fig. 3 is a flowchart of acquiring a picture of a position to be detected in an embodiment of the present invention.
Detailed Description
In order to describe the technical solution of the present invention in detail, the following description will be made clearly and completely by referring to the drawings of the embodiments of the present invention. It will be apparent that the described embodiments are some, but not all, embodiments of the invention. All other embodiments, which can be made by a person skilled in the art without creative efforts, based on the described embodiments of the present invention fall within the protection scope of the present invention.
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 invention belongs. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and/or" as used herein includes any and all combinations of one or more of the associated listed items.
For example, a general battery polarity automatic detection method based on deep learning, the method comprising:
acquiring a picture of a position to be detected;
determining a battery polarity region to be detected;
and identifying and classifying the battery polarity region to be detected by using a deep learning algorithm model to obtain a battery polarity classification result.
According to the embodiment, the battery polarity is automatically detected and classified based on deep learning, and only all sample models are needed to be provided for training samples, so that the models can automatically learn classified characteristic expression, the characteristics of complex battery polarities can be well processed, the battery polarities can be accurately identified, the instability of a traditional detection method is effectively solved, and the production efficiency and the product quality are improved.
In some embodiments, as shown in fig. 1, a general battery polarity automatic detection method based on deep learning is provided, the method comprising:
s1, acquiring a picture of a position to be detected;
s2, determining a battery polarity region to be detected;
and S3, performing identification classification on the battery polarity region to be detected by using a deep learning algorithm model to obtain a battery polarity classification result.
Specifically, this embodiment is mainly applied to when producing the battery pack of power battery or other products, need weld the positive pole and the negative pole of a plurality of batteries into the battery pack according to certain order, in order to prevent misplacing, leak and put, lead to the product short circuit and cause the wasting of resources, need accurate distinguishing the positive and negative polarities of each battery region, carry out automated inspection classification to battery polarity based on deep learning, provide all sample models, training the sample, make model automatic learn the characteristic expression of classification, the picture is obtained through the shooting of industrial camera, after obtaining the picture of waiting to detect the position, confirm the battery polarity region of waiting to detect on the picture of waiting to detect according to the characteristic of actual product, this product can be battery package or group battery, and the positive pole and the negative pole of a plurality of batteries weld according to certain order and form, after confirming the battery polarity region of waiting to detect, utilize the deep learning algorithm model that has trained in advance to carry out discernment classification to the battery polarity classification result, effectively solve the instability of traditional detection method, production operation efficiency and product quality of use are improved.
In some embodiments, the deep learning algorithm model includes positive, non-positive, negative, and non-negative four-classification models.
The polarity characteristics of the battery have diversity, for example, the positive electrode of the battery is stuck with highland barley paper, and the middle of the positive electrode is provided with a silver round area with grooves at the periphery; the negative electrode is free of highland barley paper, and the middle is a silver round area without grooves; the non-positive electrode comprises positive electrode non-attached highland barley paper, negative electrode attached highland barley paper of one layer and negative electrode attached highland barley paper of two layers; the non-negative electrode comprises a positive electrode, the positive electrode is not pasted with highland barley paper, the positive electrode is pasted with two layers of highland barley paper, the negative electrode is pasted with one layer of highland barley paper, and the negative electrode is pasted with two layers of highland barley paper. In order to achieve the purpose of well processing the characteristics of complicated battery polarity and accurately identifying the battery polarity effect, the deep learning algorithm model in the example comprises positive electrode, non-positive electrode, negative electrode and non-negative electrode four-classification models, and the positive electrode, non-positive electrode, negative electrode and non-negative electrode four-classification models are models obtained through training in advance, so that after a picture of a position to be detected is input and a battery polarity region to be detected is determined, each model can identify and classify the battery polarity region to be detected, and a battery polarity classification result is obtained.
In some embodiments, the feature input of the positive electrode model comprises a positive electrode.
The characteristic input of the non-positive electrode model comprises that the positive electrode is not pasted with highland barley paper, the negative electrode is pasted with highland barley paper on one layer, and the negative electrode is pasted with highland barley paper on two layers.
The feature input of the negative electrode model includes a negative electrode.
The characteristic inputs of the non-negative electrode model comprise an anode, anode non-attached highland barley paper, anode attached two layers of highland barley paper, cathode attached one layer of highland barley paper and cathode attached two layers of highland barley paper.
Specifically, in the training process of each model, the characteristic input is the characteristic of the polarity of the battery, for example, the characteristic input of the positive electrode model comprises a positive electrode; the characteristic input of the non-positive electrode model comprises that the positive electrode is not stuck with highland barley paper, the negative electrode is stuck with a layer of highland barley paper, and the negative electrode is stuck with two layers of highland barley paper; the characteristic input of the negative electrode model comprises a negative electrode; the characteristic inputs of the non-negative electrode model comprise positive electrode, positive electrode without highland barley paper, positive electrode with two layers of highland barley paper, negative electrode with one layer of highland barley paper and negative electrode with two layers of highland barley paper. The input of each battery polarity corresponds to a weight, and the input battery polarity can be output as the correct battery polarity by changing the weight and reconstructing the characteristic information of the battery polarity. The battery polarity classification result comprises four types of polarities, namely positive electrode, non-positive electrode, negative electrode and non-negative electrode.
In some embodiments, in the general battery polarity automatic detection method based on deep learning provided, S3, the battery polarity area to be detected is identified and classified by using a deep learning algorithm model, so as to obtain a battery polarity classification result, as shown in fig. 2, the obtaining process of the deep learning algorithm model includes:
s301, constructing a multilayer convolutional neural network.
The multi-layer convolutional neural network comprises a plurality of convolution and full-connection layers, the full-connection layers are changed into convolutional layers based on a deep learning method, convolutional layering is carried out on all sample characteristics, the input is an image, and the output is a battery polarity classification result.
S302, training the multi-layer convolutional neural network by adopting an error reverse conduction algorithm to obtain a battery polarity recognition model.
And training the multi-layer convolutional neural network by adopting an error reverse conduction algorithm so as to obtain a battery polarity recognition model, wherein an objective function used in the training process is the cross entropy of the type of the correct polarity of the battery and the type of the wrong polarity of the battery of the input image and the prediction result of the battery polarity recognition model respectively.
S303, modifying the output layer node of the multi-layer convolutional network to be 4, and initializing the modified weight of the multi-layer convolutional neural network by using the weight of the trained battery polarity identification model.
And S304, training the modified multi-layer convolutional neural network by using the positive electrode, non-positive electrode, negative electrode and non-negative electrode data sets to obtain a positive electrode, non-positive electrode, negative electrode and non-negative electrode four-classification model.
In some embodiments, the multi-layer convolutional neural network includes an input layer, a hidden layer, and an output layer;
the input layer is used for data input of the whole multi-layer convolutional neural network;
the hidden layer comprises four training models, namely an anode, a non-anode, a cathode and a non-cathode;
the output layer is used for outputting battery polarity classification results.
Specifically, in the construction of the multi-layer convolutional neural network, data is firstly acquired, neurons are constructed layer by layer, each layer can be regarded as a linear regression model, wherein the first layer is an input layer, data input of the whole multi-layer convolutional neural network is provided, each neuron of the input layer is not input, only 1 output is provided, in the example, a picture of a position to be detected is taken as the input layer and contains all characteristic information of the picture, the second layer is a hidden layer and comprises four training models of an anode, a non-anode, a cathode and a non-cathode, and the third layer is an output layer and is used for outputting a battery polarity classification result. And then adopting a layer-by-layer training mechanism to perform optimization on each layer by adopting a Wake-Sleep deep learning algorithm, and adjusting only one layer at a time and adjusting layer by layer, wherein the process can be regarded as a feature-learning process. The Wake stage is a cognitive process, wherein abstract expression codes of each layer are generated through Input features of a lower layer and upward cognitive encoding weights, then Reconstruction information Reconstruction is generated through current generated encoding weights, and Input features and Reconstruction information residual errors are calculated. In this example, the characteristic input is the characteristic of the battery polarity, including positive, positive non-stick highland barley paper, positive stick two layers highland barley paper, negative stick one layer highland barley paper and negative stick two layers highland barley paper, each battery polarity input will correspond to a weight, through changing the weight, rebuilding the characteristic information of the battery polarity, the input battery polarity can be output as the correct battery polarity, the output layer node of the multi-layer convolutional network is modified to 4, and the modified multi-layer convolutional neural network weight is initialized by the trained weight of the battery polarity identification model, and the modified multi-layer convolutional neural network is trained by the positive, non-positive, negative and non-negative data sets, so as to obtain the positive, non-positive, negative and non-negative four-classification model.
In some embodiments, as shown in fig. 3, the step S1 of obtaining a picture of the position to be detected specifically includes:
s101, manually placing a battery to be detected on a positioning module to fix the position;
s102, the industrial camera automatically photographs the battery to be detected on the positioning module to obtain a picture of the position to be detected.
S102, the industrial camera automatically photographs the battery to be detected on the positioning module, and the method specifically comprises the following steps:
and the industrial camera determines the photographing times according to the number of the batteries to be detected on the positioning module and the focal length of the industrial camera, and if photographing for multiple times, the pictures obtained after photographing for multiple times are spliced into one picture to be detected.
In some embodiments, the determining the battery polarity region to be detected S2 specifically includes:
and determining the detection position of each battery to be detected and the correct polarity to be placed at the position by adopting a region segmentation method, and finally determining the polarity region of the battery to be detected.
The technical features of the above-described embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above-described embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The foregoing examples illustrate only a few embodiments of the invention and are described in detail herein without thereby limiting the scope of the invention. It should be noted that it is possible for a person skilled in the art to make several variants and modifications without departing from the inventive concept, which fall within the scope of protection of the present invention, which is therefore subject to the appended claims.
Claims (6)
1. The utility model provides a general battery polarity automated inspection method based on degree of depth study, which is characterized in that the method includes:
acquiring a picture of a position to be detected;
determining a battery polarity region to be detected;
and identifying and classifying the battery polarity region to be detected by using a deep learning algorithm model of four classification models of positive electrode, non-positive electrode, negative electrode and non-negative electrode to obtain a battery polarity classification result, wherein the method comprises the following specific steps:
constructing a multi-layer convolutional neural network, adopting a layer-by-layer training mechanism, performing optimization on each layer by adopting a Wake-Sleep deep learning algorithm, adjusting only one layer each time, wherein the Wake stage is a cognitive process, generating abstract expression codes of each layer through Input features of a lower layer and upward cognitive encoding weights, generating Reconstruction information Reconstruction through current generated encoding weights, and calculating Input features and Reconstruction information residual errors;
training the multi-layer convolutional neural network by adopting an error reverse conduction algorithm to obtain a battery polarity recognition model;
modifying the output layer node of the multi-layer convolutional neural network to 4, and initializing the modified weight of the multi-layer convolutional neural network by using the weight of the trained battery polarity identification model;
training the modified multi-layer convolutional neural network by using positive electrode, non-positive electrode, negative electrode and non-negative electrode data sets to obtain positive electrode, non-positive electrode, negative electrode and non-negative electrode four-classification models;
the characteristic input of the positive electrode model comprises a positive electrode;
the characteristic inputs of the non-positive electrode model comprise positive electrode non-attached highland barley paper, negative electrode attached highland barley paper of one layer and negative electrode attached highland barley paper of two layers;
the characteristic input of the negative electrode model comprises a negative electrode;
the characteristic inputs of the non-negative electrode model comprise an anode, anode non-attached highland barley paper, anode attached two layers of highland barley paper, cathode attached one layer of highland barley paper and cathode attached two layers of highland barley paper.
2. The method for automatically detecting the polarity of a universal battery based on deep learning according to claim 1, wherein,
the battery polarity classification result comprises four types of polarities, namely positive electrode, non-positive electrode, negative electrode and non-negative electrode.
3. The method for automatically detecting the polarity of a universal battery based on deep learning according to claim 2, wherein,
the multi-layer convolutional neural network comprises an input layer, a hidden layer and an output layer;
the input layer is used for data input of the whole multi-layer convolutional neural network;
the hidden layer comprises four training models, namely an anode, a non-anode, a cathode and a non-cathode;
the output layer is used for outputting battery polarity classification results.
4. The method for automatically detecting the polarity of the universal battery based on deep learning according to claim 1, wherein the step of obtaining the picture of the position to be detected specifically comprises the steps of:
manually placing the battery to be detected on the positioning module to fix the position;
and the industrial camera automatically photographs the battery to be detected on the positioning module to obtain a picture of the position to be detected.
5. The automatic detection method for the polarity of the universal battery based on deep learning of claim 4, wherein the industrial camera automatically photographs the battery to be detected on the positioning module, specifically comprising:
and the industrial camera determines the photographing times according to the number of the batteries to be detected on the positioning module and the focal length of the industrial camera, and if photographing for multiple times, the pictures obtained after photographing for multiple times are spliced into one picture to be detected.
6. The method for automatically detecting the polarity of the universal battery based on deep learning according to claim 1, wherein the determining the polarity region of the battery to be detected specifically comprises:
and determining the detection position of each battery to be detected and the correct polarity to be placed at the position by adopting a region segmentation method, and finally determining the polarity region of the battery to be detected.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910386913.4A CN110222679B (en) | 2019-05-10 | 2019-05-10 | General battery polarity automatic detection method based on deep learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910386913.4A CN110222679B (en) | 2019-05-10 | 2019-05-10 | General battery polarity automatic detection method based on deep learning |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110222679A CN110222679A (en) | 2019-09-10 |
CN110222679B true CN110222679B (en) | 2023-07-11 |
Family
ID=67820623
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910386913.4A Active CN110222679B (en) | 2019-05-10 | 2019-05-10 | General battery polarity automatic detection method based on deep learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110222679B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117372414A (en) * | 2023-11-09 | 2024-01-09 | 东莞市兆丰精密仪器有限公司 | Battery polarity detection method, device, equipment and storage medium |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103925893B (en) * | 2014-04-17 | 2017-01-11 | 广东正业科技股份有限公司 | Quality detection method of battery cells |
CN104091322B (en) * | 2014-05-06 | 2017-06-16 | 无锡日联科技有限公司 | The detection method of laminated lithium ion battery |
CN207123496U (en) * | 2017-06-22 | 2018-03-20 | 惠州市德赛电池有限公司 | A kind of battery plus-negative plate identification equipment |
CN109360186A (en) * | 2018-08-31 | 2019-02-19 | 广州超音速自动化科技股份有限公司 | Lithium battery diaphragm detection method, electronic equipment, storage medium and system |
CN109490320A (en) * | 2018-11-23 | 2019-03-19 | 奇瑞汽车股份有限公司 | Machine vision-based power battery pack positive and negative pole abnormity detection system and method |
CN109598721B (en) * | 2018-12-10 | 2021-08-31 | 广州市易鸿智能装备有限公司 | Defect detection method and device for battery pole piece, detection equipment and storage medium |
-
2019
- 2019-05-10 CN CN201910386913.4A patent/CN110222679B/en active Active
Also Published As
Publication number | Publication date |
---|---|
CN110222679A (en) | 2019-09-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109740657B (en) | Training method and device of neural network model for image data classification | |
CN110415215B (en) | Intelligent detection method based on graph neural network | |
CN104573669B (en) | Image object detection method | |
WO2017166137A1 (en) | Method for multi-task deep learning-based aesthetic quality assessment on natural image | |
CN111699499B (en) | Inspection, image recognition, recognizer generation system, and learning data generation device | |
CN111178120B (en) | Pest image detection method based on crop identification cascading technology | |
CN111257341A (en) | Underwater building crack detection method based on multi-scale features and stacked full convolution network | |
CN111758117B (en) | Inspection system, recognition system, and learning data generation device | |
JP2020060879A (en) | Learning device, image generator, method for learning, and learning program | |
CN117355038B (en) | X-shaped hole processing method and system for circuit board soft board | |
CN110222679B (en) | General battery polarity automatic detection method based on deep learning | |
CN114419029A (en) | Training method of surface defect detection model, surface defect detection method and device | |
CN110276405B (en) | Method and apparatus for outputting information | |
CN109886206B (en) | Three-dimensional object identification method and equipment | |
CN111652327A (en) | Model iteration method, system and computer equipment | |
CN116091836A (en) | Multi-mode visual language understanding and positioning method, device, terminal and medium | |
CN114580517A (en) | Method and device for determining image recognition model | |
CN113128522B (en) | Target identification method, device, computer equipment and storage medium | |
JP2020064364A (en) | Learning device, image generating device, learning method, and learning program | |
CN112949456A (en) | Video feature extraction model training method and device, and video feature extraction method and device | |
JP2019158684A (en) | Inspection system, identification system, and discriminator evaluation device | |
CN110595401A (en) | Detection method for detecting four corners of battery by using X-ray | |
CN114022698A (en) | Multi-tag behavior identification method and device based on binary tree structure | |
CN112396083B (en) | Image recognition, model training and construction and detection methods, systems and equipment | |
US11182646B2 (en) | User-generated visual guide for the classification of images |
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 |